45_69

ISSN 1330-9862 original scientific paper

(FTB-1574)

Effect of Flavonoids on Glutathione Level, Lipid

Peroxidation and Cytochrome P450 CYP1A1 Expression

in Human Laryngeal Carcinoma Cell Lines

Ksenija Durgo1*, Lidija Vukovi}2, Gordana Rusak3, Maja Osmak2

and Jasna Franeki} ^oli}11Faculty of Food Technology and Biotechnology, University of Zagreb, Pierottijeva 6,

HR-10 000 Zagreb, Croatia2Ru|er Bo{kovi} Institute, Bijeni~ka 54, HR-10 000 Zagreb, Croatia

3Faculty of Science, University of Zagreb, Rooseveltov trg 6, HR-10 000 Zagreb, Croatia

Received: November 4, 2005Accepted: June 29, 2006

Summary

Flavonoids are phytochemicals exhibiting a wide range of biological activities, amongwhich are antioxidant activity, the ability to modulate activity of several enzymes or cellreceptors and possibility to interfere with essential biochemical pathways. Using humanlaryngeal carcinoma HEp2 cells and their drug-resistant CK2 subline, we examined the ef-fect of five flavonoids, three structurally related flavons (quercetin, fisetin, and myricetin),one flavonol (luteolin) and one glycosilated flavanone (naringin) for: (i) their ability to in-hibit mitochondrial dehydrogenases as an indicator of cytotoxic effect, (ii) their influenceon glutathione level, (iii) antioxidant/prooxidant effects and influence on cell membranepermeability, and (iv) effect on expression of cytochrome CYP1A1. Cytotoxic action of theinvestigated flavonoids after 72 hours of treatment follows this order: luteolin>quercetin>fisetin>naringin>myricetin. Our results show that CK2 were more resistant to toxic con-centrations of flavonoids as compared to parental cells. Quercetin increased the total GSHlevel in both cell lines. CK2 cells are less perceptible to lipid peroxidation and damagecaused by free radicals. Quercetin showed prooxidant effect in both cell lines, luteolin onlyin HEp2 cells, whereas other tested flavonoids did not cause lipid peroxidation in thetested cell lines. These data suggest that the same compound, quercetin, can act as aprooxidant, but also, it may prevent damage in cells caused by free radicals, due to the in-duction of GSH, by forming less harmful complex. Quercetin treatment damaged cellmembranes in both cell lines. Fisetin caused higher cell membrane permeability only inHEp2 cells. However, these two compounds did not enhance the damage caused by hy-drogen peroxide. Quercetin, naringin, myricetin and fisetin increased the expression ofCYP1A1 in both cell lines, while luteolin decreased basal level of CYP1A1 only in HEp2cells. In conclusion, small differences in chemical structure of flavonoids led to drasticchange of their biological effects.

Key words: flavonoids, tumor cells, cytotoxicity, glutathione, lipid peroxidation, cytochrome 1A1

69K. DURGO et al.: Effect of Flavonoids in Laryngeal Carcinoma Cells, Food Technol. Biotechnol. 45 (1) 6979 (2007)

*Corresponding author; E-mail: [email protected] paper was presented at the 2nd Congress of Croatian Geneticists in Supetar, Island of Bra~, Croatia, September 24-27, 2005

Abbreviations: CYP1A1 cytochrome 1A1, F fisetin, HEp2 human laryngeal carcinoma cells, CK2 human laryngeal carcinoma cellsresistant to several cytostatics, GSH glutathione, DMSO dimethyl sulfoxide, C control, L luteolin, LDH lactate dehydrogenase, MDAmalondyaldehyde, MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, M myricetin, N naringin, ROS reactiveoxygen species, TBA thiobarbituric acid, Q quercetin

Introduction

Flavonoids are a group of low-molecular-mass po-lyphenolic substances, ubiquitous in all vascular plants(1). They occur naturally in a broad range of fruits, vege-tables and beverages such as green tea or red wine. Thedaily consumption of flavonoids is difficult to estimate,because the uptake of these specific compounds mayvary greatly depending on the food source. In plants,flavonoids are widely distributed in form of more polarglycosilated derivatives (mostly O-b-glycosides) (1,2) andtherefore these substances are considered to be nonab-sorbable (3,4). Nevertheless, microorganisms in the co-lon hydrolyze them into free flavonoids (aglycones), whichare expected to be able to pass through the gut wall.Also, some glycosilated flavonoids are metabolized in li-ver (3,5).

Flavonoids from 4-oxo-flavonoid family have beenreported to exert multiple biological effects, but biologi-cal outcome will depend on properties of each com-pound (2). Various experimental systems (e.g. membranesystems, cell cultures, animal models, and human clini-cal trials) are used to study the bioactivity of these plant--derived compounds. All of these systems have their ad-vantages and limitations. Cell cultures have a number ofadvantages over other experimental systems: ability toconduct mechanistic studies at molecular level, easycontrol of experimental environment and relatively lowcost, and finally, avoidance of ethical issues related toanimal or human studies. However, cell cultures cannotreflect precisely the conditions found in the body. In spiteof that, they are valuable tools in biomedical science.

In this work, we used two cell lines originated fromthe same parental cell line: human laryngeal carcinomaHEp2 cells and their drug-resistant CK2 subline. How-ever, CK2 cells acquired (during the development of re-sistance to cisplatin) many alterations that can influencetheir response to different cytotoxic agents. Both HEp2and CK2 cells have inducible glutathione, glutathione--S-transferases and cytochrome 1A1, but basal level ofthese enzymes is different. These properties make theselected pair of cell lines suitable for investigation of theeffect of flavonoids on the expression of the mentionedproteins. We determined the ability of five flavonoids to:(i) inhibit mitochondrial dehydrogenases as an indicatorof cytotoxic effect, (ii) determine the influence of flavo-noids on cellular level of glutathione, (iii) examine theirantioxidant/prooxidant effects (TBA-MDA formation) andinfluence on cell membrane permeability as an indicatorof secondary damage of cell (lactate dehydrogenase per-meability), and (iv) examine their influence on expres-sion of CYP1A1.

Glutathione, a tripeptide present in the majority ofcells, is responsible for hydrophilic xenobiotics conjuga-tion. Sulphydryl group of glutathione is essential for itsantioxidant activity against some forms of reactive oxy-gen species (ROS) in cells (6). There is evidence thatsome flavonoids can elevate intracellular basal level ofGSH, allowing better tolerance of free radicals (7). It hasbeen suggested that flavonoid activities depend heavilyon their antioxidant and chelating properties (1). There-fore, numerous studies concerning biological effects offlavonoids on the in vitro and in vivo free radical-medi-

ated processes have been carried out. Much effort hasbeen focused on the identification of flavonoids that ex-ert beneficial effects and mechanisms by which they in-hibit cellular injury and degradation (8,9). Free radicals,whose origin may be molecular oxygen that has accept-ed an electron from variety of sources, or activated formof foreign compound that has entered into a cell, can at-tack unsaturated lipids in a cell, resulting in the chainreaction of the formation of free radicals. This reaction isterminated by the production of lipid breakdown prod-ucts, lipid alcohols, aldehydes and malondialdehyde(MDA). Therefore, measurement of malondyaldehydeconcentration is a common method for determination ofprimary toxic effect caused by free radicals in experi-ments in vitro (9,10). It has been shown that flavonoidscan prevent autooxidation of linoleic acid in vitro, orperoxidation of phospholipid membranes, microsomalor mitochondrial peroxidation (1113).

Some flavonoids can stimulate lipid peroxidation, pro-bably depending on the number of hydroxyl substitu-ents in the B-ring of flavonoids (hydroxyl groups are sub-strates for cytochrome P450) (5). Among the proteinsthat interact with flavonoids, cytochromes P450 (CYPs)play a prominent role (5). Flavonoid compounds influ-ence these enzymes in several ways. They induce the ex-pression of several CYPs and inhibit or stimulate theirmetabolic activity. Some CYPs participate in the metabo-lism. Flavonoids enhance activation of carcinogens andinfluence the metabolism of drugs via induction of spe-cific CYPs. On the other hand, inhibition of CYPs in-volved in carcinogen activation and scavenging reactivespecies formed from carcinogens by CYP-mediated reac-tions can be a beneficial property of various flavonoids.Finally, cytochromes are producers of reactive oxygenspecies in the cells when pour coupling of the P450 cata-lytic cycle occurs (5,9).

Materials and Methods

Human cell lines

Human laryngeal carcinoma HEp2 cells were grownas monolayer cultures in DMEM medium supplementedwith 10 % of fetal bovine serum, 4500 mg/L glucose,pyridoxine and 1 % penicillin/streptomycin solution.Drug-resistant subline (CK2) was developed in Labora-tory for genotoxic agents, at Rudjer Bo{kovi} Institute,by the treatments with stepwise increased concentrationof cisplatin (14). CK2 cells became resistant to cisplatinand cross-resistant to several other anti-cancer drugs:vincristine, methotrexate, fluorouracil, mitomycin C andcarboplatin (15,16). This cross-resistance was accompa-nied with sensitivity to amphotericin B (17), more hy-perthermia-reduced intracellular cisplatin accumulation(18), increased levels of markers for invasion and metas-tasis, cathepsin D and urokinase plasminogen activator(19), and increased expression of integrins (20).

Chemicals3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium

bromide (MTT), malondyaldehyde (MDA), NADH, sodi-um pyruvate, thiobarbituric acid (TBA), Ellman reagent,NADPH, glutathione and glutathione reductase were

70 K. DURGO et al.: Effect of Flavonoids in Laryngeal Carcinoma Cells, Food Technol. Biotechnol. 45 (1) 6979 (2007)

purchased from Sigma Chemicals. Flavonoids (querce-tin, fisetin, myricetin, luteolin and naringin) were alsoobtained from Sigma Chemicals. Primary monoclonalantibody CYP1A1 was purchased from Santa Cruz Bio-technology. Secondary antimouse antibody was pur-chased from Amersham Pharmacia, USA. The chemicalsused in experiments were of analytical grade.

FlavonoidsQuercetin, myricetin and fisetin belong to a group

of flavonols. The difference in their structure is verysmall (Fig. 1). Myricetin has three hydroxyl groups atpositions 3, 4 and 5 at B ring and one hydroxyl groupat position 5 at A ring, quercetin lacks one hydroxylgroup at 5 position, and fisetin misses one hydroxylgroup at 3 position on B ring and one hydroxyl groupat position 5 at A ring. Luteolin belongs to a group offlavons, lacking hydroxyl group at 3 position at C ring.Naringin, belonging to a group of flavanons, is glyco-silated form of naringenin, which lacks hydroxyl groupsat positions 3 and 5 at B ring. The structure of the ex-amined compounds is presented in Fig. 1.

Flavonoids were dissolved in dimethyl sulphoxide(DMSO), in stock concentrations of 500 mM. Just beforeuse, stock solutions were dissolved in growth medium.

In all experiments final concentration of DMSO in themedium did not exceed 0.1 %.

Cytotoxicity assay

HEp2 and CK2 were seeded in 24-well plates inconcentration of 5.4104 cells, 1.8 mL/well. The next day,cells were treated with a range of concentrations (3.8500 mM; see Fig. 2) of each flavonoid. Following 72 hours(period sufficient for 3 cell divisions), growth mediumwas removed, and cells were treated with 10 % MTT so-lution (21). Viable cells have active mitochondrial dehy-drogenase enzymes which metabolize a yellow-colouredtetrazolium salt (MTT) into a blue formazan. Therefore,the number of surviving cells is directly proportional tothe level of the produced formazan. Formazan was dis-solved with 2-propanol, containing 0.01 M HCl. Intensityof absorbance was measured at 540 nm in Cecil spectro-photometer (Cecil Instruments Ltd, Cambridge TechnicalCentre, UK). Each experiment was repeated three times.

Measurement of total glutathione

A volume of 13.5 mL of cell suspension (4.5105 cells)was seeded in 10-centimetre Petri dishes. After attach-ment, the cells were treated with highest nontoxic con-


quercetin

OO OH

OO

HH

OO

HH

A C

B

OO

HH

OO

OO

HH

OO

1

1'

2'

3'

4'

5'

6'

2

3

45

6

7

8 HH

myricetin

OOH

OH

OOH

OH

luteolin

OOH

OH

O

OH

OH

fisetin

O

OH

O

O

OH

*

naringin*7-O-neohesperidosid

Fig. 1. Chemical structure of investigated flavonoids

centration of the examined flavonoids. After 72 hours ofincubation, the cells were scraped, washed twice withcold PBS (containing Ca2+ and Mg2+ salts), lysed andcentrifuged at 15 000 rpm. Supernatant was examinedfor the level of total glutathione by the Tietzes method(22). Briefly, cell supernatant (10 or 20 L) was added

into quivette containing 0.7 mL of 100 mM Na2HPO4/5M EDTA, 200 L of Ellman reagent (3 mM) and 100 Lof NADPH (2 mM). Prior to measurement, 10 L of glu-tathione reductase (activity 12 units/mL) was added intothe mixture. The absorbance was determined at 412 nmevery 15 seconds for 2 min. Concentration of total gluta-


0

10

20

30

40

50

60

70

80

90

100

110

4.12 41.20 412

Concentration of myricetin/ M

Surv

ival

/%

HEp2

CK2

0

10

20

30

40

50

60

70

80

90

100

110

3.81 38.1 381

Concentration of naringin/ M

HEp2

CK2

0

10

20

30

40

50

60

70

80

90

100

110

120

5.46 54.6 546

Concentration of quercetin/ M

HEp2

CK2

0

10

20

30

40

50

60

70

80

90

100

110

5 50 500

Concentration of fisetin/ M

HEp2

CK2

0

10

20

30

40

50

60

70

80

90

100

110

4.12 41.2 412

Concentration of luteolin/ M

HEp2

CK2

Surv

ival

/%

Surv

ival

/%

Surv

ival

/%

Surv

ival

/%

Fig. 2. Survival curves obtained after the 72-hour treatment of HEp2 (triangle) and CK2 (circle) cells with five flavonoids. Percentagesof survival are expressed in comparison with negative control. Pooled data from three experiments (the mean at the point s.d.)

thione was calculated from standard curve. Total amountof proteins in supernatants was determined using Brad-ford assay (23). Each experiment was carried out in trip-licate.

Measurements of lipid peroxidation and lactatedehydrogenase

A volume of 13.5 mL of cell suspension (4.5105 cells)was seeded in 10-centimetre Petri dishes. After attach-ment, the cells were incubated with highest nontoxicconcentration of the examined flavonoids or the flavo-noid and hydrogen peroxide for 72 hours. Then the cellswere scraped, washed twice with PBS (containing Ca2+

and Mg2+ salts) and lysed in 520 L of potassium chlo-ride (1.15 %) for 30 min. The mixture was centrifuged at5000 rpm. A volume of 500 L of cell supernatants wasincubated with 2 mL of TBA (100 g/L) for 15 min at 100C. The mixture was cooled with tap water, centrifugedat 1000 rpm/10 min; then 2.5 mL of supernatant weremixed with 1 mL of TBA (0.8 %) and incubated for 15min at 100 C. After cooling with tap water, the absor-bance of the samples was examined at 532 nm and thenat 600 nm (value of nonspecific absorbance). Concentra-tion of MDA-TBA complex as an indicator of lipid per-oxidation was calculated from standard curve (24,25).Concentration of cellular proteins was determined ac-cording to Bradford method (23). Each experiment wascarried out in triplicate.

Changes in the membrane permeability can be de-tected in vitro as the leakage of enzymes such as lactatedehydrogenase. The measurements were done in themedium where cells were grown and treated with dif-ferent flavonoids or with flavonoids and hydrogen per-oxide mixture. One unit of LDH activity presents thequantity of an enzyme needed for the formation of 1mol of NAD+ in one minute (26). Growth medium wascentrifuged for 5 min at 3000 rpm to remove detachedcells. In 470 mL of buffered water (5 mM) and 430 mL ofMOPS buffer (500 mM), 30 L of NADH and 30 L ofsodium pyruvate were added. Prior to measurement, 40L of sample were added. Absorbance was measured at340 nm for 2 min, every 15 seconds. Activity of lactatedehydrogenase was expressed as a rate of absorbancechange in unit of time. Concentration of cellular pro-teins was determined according to Bradford method(23). Each experiment was carried out in triplicate.

Western blot analysis

A volume of 13.5 mL (4.5105 cells) was seeded in10-centimetre Petri dishes. After attachment, the cellswere incubated with highest nontoxic concentration ofthe examined flavonoids for 72 hours. After that, thecells were scraped, washed twice with cold PBS (con-taining Ca2+ and Mg2+ salts), lysed and centrifuged at10 000 rpm/15 min at 4 C. The supernatant was col-lected and placed on ice. Protein concentration was de-termined according to Lowry method (27). An amountof 25 g of proteins was separated by SDS polyacril-amide electrophoresis (2 h/80 mA). The proteins weretransferred (1 h/400 mA) on nitro-cellulose membranes(Hybond C, Amersham Pharmacia, USA). Membraneswere blocked in 5 % non-fat milk in TTBS (50 mM

Tris-HCl, 200 mM NaCl and 0.05 % Tween 20) and incu-bated with monoclonal primary mouse antibody raisedagainst human CYP1A1 (dilution 1:1000; Santa Cruz).Detection of CYP1A1 was enabled by incubation of mem-branes with anti-mouse secondary antibody (AmershamPharmacia, USA). CYP1A1 was determined by enhancedchemiluminescence (ECL) reagent (Amersham Pharma-cia, USA). Relative intensities of the bands were estima-ted from densitometric analysis of the blot with an Im-age Master VDS software (Amersham Pharmacia Biotech).Each experiment was repeated three times.

Statistical analsyes

Statistical analyses were performed with SPSS ver-sion 8.0 (SPSS Inc., Chicago, IL, USA). Each experimentis the result of three independently performed experi-ments. The one-way analysis of variance (ANOVA) wasemployed to determine whether the means of differentgroups were significantly different. The Dunnett post--hoc test was used to determine the significance. A proba-bility level of pquer-cetin>fisetin>naringin>myricetin. The highest nontoxicconcentrations of flavonoids were established as follows:luteolin (4.12 mM), quercetin (5.46 mM), fisetin (5 mM),naringin (38.1 mM) and myricetin (41.2 mM). These con-centrations were used in further experiments. SolventDMSO in the highest applied concentration did notcause any biological effects in control experiments. Gen-erally, CK2 cells were more resistant to toxic effects ofthe examined flavonoids than parental cells.

It is known that small changes in chemical structureof flavonoids cause significant changes in their biologi-cal activities. Therefore, it is not surprising that narin-gin, the only glycosilated flavonoid among the tested


substances, has shown lower toxic effect on both celllines compared to other flavonoids. This could be due tohydrophilic nature of this compound and its differentbioavailability in cells.

Similar to our results, Rusak et al. (28,29) also foundthat quercetin is the most cytotoxic compound amongthe tested flavonoids after 72 hours of incubation. Theyalso pointed out that C2C3 double bond is an impor-tant structural requirement for the cytotoxic effects of fla-vonoids. All flavonoids used in our study possess C2C3double bond and they showed cytotoxic effects. Our dataare in accordance with the data of Hodek et al. (5), whoshowed that 7-hydroxyl group of the ring A is an im-portaint factor for cytotoxic effect of flavonoids. Rusaket al. (28,29) pointed out that higher number of hydroxylgroups at ring B enhanced cytotoxic effects of flavonoids.We showed that myricetin, the flavonoid with 3 hydro-xyl groups at ring B, was the least toxic compound ofthe tested flavonoids, suggesting that cytotoxic effects offlavonoids are cell specific. The mechanism of cytotoxicactivity of quercetin and luteolin was proposed by Hi-rano et al. (30). They supposed that quercetin and luteo-lin inhibit human promyelocytic leukemia cell growth pos-sibly via the cessation of DNA, RNA and protein synthe-sis of leukemic cells. Scambia et al. (31) estimated quer-cetin as a powerful antiproliferative agent against somehuman cancer cell lines, such as primary colorectal, ova-rian, lymphoblastoid and breast cancer cells.

Glutathione levels

Probably the most important protective mechanismfor free radical scavenging and inhibition of electrophilicxenobiotics attack on cellular macromolecules involvestripeptide glutathione (6). Due to nucleophilic thiol group,it can detoxify substances in one of three ways: (i) conju-

gation catalyzed by glutathione-S-transferases (GST); (ii)chemical reaction with a reactive metabolite to form aconjugate; and (iii) donation of proton or hydrogen atomto reactive metabolites or free radicals. Reactive interme-diates can react with GSH either by a direct chemical re-action or by a GST-mediated reaction preventing possi-ble cell death. In this work we have investigated theinfluence of flavonoids on intracellular glutathione level.These results are presented in Fig. 3.

CK2 cells have higher basal level of total GSH thanparental HEp2 cells; 7.8 mM/mg of protein for HEp2cells, 9.5 mM/mg of protein for CK2 cells, and, for com-parison, 8.125 mM/mg of protein for normal human lungfibroblasts. However, according to our previous results,GSH was not involved in the resistance of CK2 cells tocisplatin (32). In HEp2 cell line, myricetin and quercetinelevated the level of total GSH, but this increase wassignificant only for quercetin. In CK2 cell line, with in-creased basal level of GSH, fisetin, luteolin and querce-tin elevated GSH level, but again, only for quercetin thisincrease was statistically significant. This implicates thatcells treated with quercetin can tolerate exposure tohigher concentrations of different electrophilic xenobio-tics, as well as compounds that cause oxidative damageand oxidative stress.

Similarly, protective in vivo effect of quercetin waspublished by Devi Priya and Shyamala Devi (33). Theyinvestigated the influence of quercetin on rat kidneyswhere rats were simultaneously treated with cisplatinand quercetin. Quercetin administration per se preventedlipid peroxidation and cell membrane damage. This sug-gests that quercetin has the same effect on GSH induc-tion in both in vivo and in vitro systems as well as thatthis effect is not cell specific. Beneficial effects of quer-cetin were also observed by Pattipati et al. (34). They in-


luteolin

(4.12 M)

fisetin

(5 M)

quercetin*/**

(5.46 M)

naringin

(38.1 M)

myricetin

(41.2 M)control

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

c(G

SH

)/m

(pro

tein

)/(M

/mg

)

HEp2

CK2

()

Fig. 3. Concentration of total glutathione in HEp2 (white columns) and CK2 (gray columns) cell lines after 72 hours of treatmentwith the highest nontoxic concentrations of flavonoids indicated as data labels. Pooled data from three experiments (the mean at thepoint s.d.). Statistically significant difference compared to control: * in HEp2 and ** in CK2 cell line

vestigated the prevention of occurrence of dyskinesia, aserious neurological syndrome associated with long-termadministration of neuroleptics to experimental animals.Repeated treatment with neuroleptics decreased the levelof GSH, but treatment with quercetin significantly re-duced neuroleptically induced changes in the brain ofrats. Further, administration of quercetin restored intra-cellular GSH levels, the decrease of which was causedby chronic neuroleptic treatment.

Lipid peroxidationFree radicals, whose origin may be molecular oxy-

gen which has accepted an electron from a variety ofsources or activated form of foreign compound that hasentered into a cell, can attack unsaturated lipids in thecell, resulting in chain reaction of free radical formation.This reaction is terminated by the production of lipidbreakdown products; lipid alcohols, aldehydes and ma-londialdehyde. Therefore, measurement of MDA concen-tration is a commonly used method to assess lipid per-oxidation in cultures in vitro.

It is believed that flavonoids provide only beneficialeffects on human health, which is associated with an im-proved antioxidant response in vivo. This presumption isbased on their low absorption, extensive metabolismand their ability to bind to proteins. But, numerous liter-ature data show that some flavonoids could behave asboth antioxidants and prooxidants, depending on theconcentration and free radical source (8,35).

In this work, we have examined the influence offive flavonoids on the formation of MDA after pro-longed exposure of HEp2 and CK2 cells to nontoxic con-centrations of flavonoids, as well as to the mixture offlavonoids and hydrogen peroxide. Because of low con-centration of free radicals induced by highest nontoxicconcentrations of flavonoids (see section Cytotoxicity) andhydrogen peroxide (8 mM), incubation time was pro-longed to three days in order to be able to determinesome biological effects.

Basal level of MDA-TBA complex in HEp2 cells was10.64 mM/mg of protein, in CK2 cells 5.12 mM/mg ofprotein. For comparison, in normal lung fibroblasts ba-sal level of MDA-TBA was 12.7 mM/mg of protein. Asshown in Fig. 4, fisetin (5 mM), quercetin (5.46 mM) andluteolin (4.12 mM) induced lipid peroxidation in HEp2cell line. Naringin (38.1 mM) reduced basal level of per-oxidation but also inhibited lipid peroxidation that wasinduced with 8 mM H2O2. Luteolin and fisetin enhancedprimary toxic effect of hydrogen peroxide. These two com-pounds behave as prooxidants showing synergistic ef-fect with free radicals of different origin.

Contrary to our results, Shimoi et al. (36) showedthat luteolin in vivo protected serum lipids against oxi-dation induced with 2,2-azobis(2-aminopropane)-dihydro-chloride (AAPH). This discrepancy could be explainedby the hypothesis that luteolin in vivo has passed throughmetabolic changes (glucuronidation, sulphatation, me-thylation) prior to entering into a blood system. We sup-posed that in in vitro system luteolin is not transformedto a more electrophilic form by enzymes involved inphase I of detoxification (this phase could be responsiblefor free radical formation). Kumar et al. (37) showed that

naringin protects hemoglobin from nitrite-induced oxi-dation to methemoglobin in vitro. The protection wasnot observed when naringin was added after autocata-lytic stage of the oxidation of hemoglobin by nitrite. Theability of naringin to scavenge oxygen free radicals maybe responsible for this phenomenon, because superoxide,hydroxyl and other free radicals are implicated in pro-moting the autocatalytic stage of oxidation of hemoglo-bin by nitrite.

Fig. 5 presents levels of lipid peroxidation in CK2cells following the treatment with flavonoids. It shouldbe pointed out that basal level of MDA in CK2 cells is


HE

p2

Q*

(5.4

6M

)m

HE

p2

M+

HO

(8M

)2

2m

HE

p2

M(4

.12

M)

m

HE

p2

F(5

M)

m

HE

p2

N(3

8.1

M)

m

HE

p2

L*

(4.1

2M

)m

HE

p2

C

HE

p2

F+

HO

(8M

)*2

2m

HE

p2

L+

HO

(8M

)*2

2m

80

MH

O*

m2

2

8M

HO

m2

2

HE

p2

Q+

HO

(8M

)2

2m

HE

p2

N+

HO

(8M

)2

2m

((T

BA

-MD

Aco

mp

lex

)/(p

rote

in))

/(M

/mg

)c

mm

Fig. 4. Concentration of TBA-MDA complex in HEp2 cells afterthe treatment for 72 hours with the highest nontoxic concentra-tions of flavonoids (white columns) and flavonoids and hydro-gen peroxide (gray columns). The highest nontoxic concentra-tions of flavonoids are presented in brackets. M=myricetin,F=fisetin, Q=quercetin, N=naringin; L=luteolin. *Statisticallysignificant difference compared to nontreated control (C)

CK

2C

CK

2M

(41.2

M)

m

CK

2M

+H

O(8

M)

22

m

CK

2F

(5M

)m

CK

2F

+H

O(8

M)

22

m

CK

2Q

(5.4

6M

)*m

CK

2Q

+H

O(8

M)*

22

m

CK

2N

(38.1

M)

m

CK

2N

+H

O(8

M)

22

m

CK

2L

(41.2

M)

m

CK

2L

+H

O(8

M)

22

m

80

MH

O*

m

22

8M

HO

m2

2

((T

BA

-MD

Aco

mple

x)/

(pro

tein

))/(

M/m

g)

cm

m

Fig. 5. Concentration of TBA-MDA complex in CK2 cells afterthe treatment for 72 hours with the highest nontoxic concentra-tions of flavonoids (white columns), and flavonoids and hydro-gen peroxide (gray columns). M=myricetin, F=fisetin, Q=quer-cetin, N=naringin; L=luteolin. *Statistically significant differencecompared to nontreated control (C)

significantly lower (5.12 mM/mg protein) than in paren-tal cells (10.64 mM/mg of protein in HEp2 and 12.7 mM/mg of protein in normal lung fibroblasts). Furthermore,it is obvious that drug-resistant CK2 cells are less sus-ceptible to lipid peroxidation than HEp2 cells. CK2 cellsare also less sensitive to lipid peroxidation and damagecaused by free radicals. Quercetin caused significant in-crease in MDA concentration, but not synergistic effectwith 8 mM H2O2. Fisetin showed slight (but not statisti-cally significant) prooxidant nature when it was addedto cells treated with 8 M H2O2.

Quercetin caused significant MDA-TBA complexformation in hydrogen peroxide-treated cells as well asin quercetin-treated cells in both cell lines, indicating itsprooxidative activity. Young et al. (38) investigated theinfluence of quercetin present in juice on healthy volun-teers. They analyzed the presence of MDA in bloodsamples and urine. Quercetin was excreted in urine(0.290.47 %), and its plasma concentration was verylow. Total plasma MDA level decreased with time, indi-cating that reduction of lipid oxidation had occurred.2-adipic semialdehyde, a biomarker of protein oxida-tion, appeared, indicating a prooxidant effect of thejuice. This suggests that even within the plasma, thereare several subcompartments which may respond differ-ently to a dietary challenge. These results also contradicta general pro- or antioxidant state in blood, but also adifferential protection or damage to specific structures,depending on their interaction with the dietary compo-nents that reach them. The differences between our re-sults and those obtained by Young et al. (38) in in vivosystem could be explained by the synergistic effects ofdifferent components present in juice. Therefore, theseeffects cannot be attributed solely to quercetin.

Lactate dehydrogenase activity

Changes in the membrane permeability can be de-tected in vitro as leakage of enzymes such as lactate de-hydrogenase. In the present study we have examined cellmembrane permeability following the treatment of cellswith the highest nontoxic concentrations of flavonoidsand hydrogen peroxide. One unit of LDH activity pres-ents the quantity of an enzyme needed for the formationof 1 mol of NAD+ in one minute (25).

Basal LDH activity in HEp2 cells was 0.25 nmol/(minmg protein), in CK2 it was 0.27 nmol/(minmgprotein). In normal lung fibroblasts, Hef cells, basal ac-tivity of LDH was 0.1 nmol/(minmg protein). Nontoxicconcentrations of fisetin and quercetin increased cellmembrane permeability. LDH activity was not increasedwhen the cells were simultaneously treated with fisetinand quercetin in the presence of 8 M H2O2, suggestingthat these two compounds did not enhance the toxic ef-fect of hydrogen peroxide (Fig. 6).

CK2 cells were more sensitive to the toxic concen-tration of hydrogen peroxide (80 M), possibly due tohigher permeability of cell membrane. In CK2 cells, onlyquercetin caused significant leakage of LDH. There wasno difference between cells treated with quercetin andthose treated with quercetin and H2O2 mixture, suggest-ing that quercetin did not participate in LDH leakagefrom the cells treated with hydrogen peroxide (Fig. 7).

Expression of cytochrome CYP1A1

Cytochrome P450 are heme containing mixed-func-tion oxidases that play a key role in the metabolism ofhydrophobic endogenic substances (sterols, prostaglan-dins, and fatty acids) and ingested foreign compoundssuch as drugs, pesticides and pollutants (5). These pro-teins are involved in interactions with flavonoid com-pounds in three ways: flavonoids can induce biosynthe-


HE

p2

Q(5

.46

M)

m

HE

p2

M+

HO

(8M

)2

2m

HE

p2

M(4

1.2

M)

m HE

p2

F(5

M)

m

n(L

DH

)

tm

(pro

tein

)(n

mo

l/m

inm

g)

HE

p2

N(3

8.1

M)

m

HE

p2

L(4

.12

M)

m

HE

p2

C

HE

p2

F+

HO

(8M

)2

2m

HE

p2

L+

HO

(8M

)2

2m

80

MH

O*

m2

2

8M

HO

m2

2

HE

p2

Q+

HO

(8M

)2

2m

HE

p2

N+

HO

(8M

)2

2m

Fig. 6. Lactate dehydrogenase activity measured in growth me-dium as a measure of cell membrane permeability in HEp2cells after 72 hours of treatment with the highest nontoxic con-centrations of flavonoids (white columns) and flavonoids andhydrogen peroxide (gray columns). Pooled data from three ex-periments (the mean at the point s.d.). *Statistically signifi-cant difference compared to control

n(L

DH

)

tm

(pro

tein

)(n

mol/

min

mg

)

CK

2C

CK

2M

(41.2

M)

m

CK

2M

+H

O(8

M)

22

m

CK

2F

(5M

)m

CK

2F

+H

O(8

M)

22

m

CK

2Q

(5.4

6M

)m

CK

2Q

+H

O(8

M)

22

m

CK

2N

(38.1

M)

m

CK

2N

+H

O(8

M)

22

m

CK

2L

(4.1

2M

)m

CK

2L

+H

O(8

M)

22

m

80

MH

O*

m

22

8M

HO

m2

2

Fig. 7. Lactate dehydrogenase activity measured in growth me-dium as a measure of cell membrane permeability in CK2 cellsafter 72 hours of treatment with the highest nontoxic concentra-tions of flavonoids (white columns) and flavonoids and hydro-gen peroxide (gray columns). Pooled data from three experi-ments (the mean at the point s.d.). *Statistically significantdifference compared to control

sis of certain CYPs, they can modulate enzymatic activityof CYPs and, finally, flavonoids can be metabolized byseveral CYPs.

In the present study, we examined the ability offlavonoids to induce biosynthesis of cytochrome P450,CYP1A1. Induction of CYP1A1 is regulated by bindingof a certain compound to aryl hydrocarbon receptor(AhR), a ligand-activated transcription factor (5). Bind-ing affinities of these xenobiotics to AhR appear to bedependent on structural constrains; planar aromatic com-pounds with bulky substituent groups are preferred (5).Depending on flavonoid concentrations, flavonoids ex-hibit different biological activities. At lower concentra-tions, flavonoids act as AhR antagonists, binding to thereceptor without activation of a transcription factor, whileat higher concentrations the same flavonoids can act asAhR agonists and modulate gene expression of CYPs in-duced by benz(a)pyrene, for example (5).

In this study, we determined the effects of nontoxicconcentrations of investigated flavonoids on CYP1A1expression. Basal expression of CYP1A1 in HEp2 cellswas higher than in CK2 cell line (Fig. 8). Human lungfibroblasts have low basal expression of CYP1A1. Asshown in Fig. 8, quercetin did not induce CYP1A1 inHEp2 cell line. Naringin, fisetin and myricetin slightlyincreased expression of CYP1A1 compared to control(from 22 to 27 %). Luteolin reduced basal CYP1A1 ex-pression for 40 %. In CK2 cell line, quercetin, naringin,myricetin, fisetin and luteolin induced expression ofCYP1A1 from 43 to 60 %, as compared to basal level ofCYP1A1 in CK2 cell line (Fig. 8). Inhibition of CYP1A1protein expression in HEp2 cells treated with luteolincould play an important role in cancer chemoprotection.In cells with high basal CYP1A1 expression, luteolin isprobably metabolized through this system and is trans-formed to reactive intermediate which binds as an an-tagonist to AhR, inhibiting consequently polycyclic aro-matic hydrocarbon adduct formation (5,36). Our resultsindicate that the same flavonoid could induce, but alsosuppress CYP1A1 expression in different cell lines, sug-gesting cell type-specific effect of flavonoids on expres-sion of CYP1A1.

Connection between the chemical structure of theinvestigated flavonoids and their influence on CYP1A1expression is difficult to establish. All examined flavo-noids have 7-hydroxyl group at the ring A and 4-ketogroup at the ring C. They differ in number of hydroxylgroups at 3, 4 and 5 positions at the ring B, absence of

hydroxyl group at 5 position at A ring (fisetin) or ab-sence of hydroxyl group at 3 position of C ring(luteolin). Flavonoids with more hydroxyl groups at-tached at B ring have stronger influence on CYP1A1 ex-pression. Naringin, a flavonoid with one hydroxylgroup at ring B, showed the weakest effect on CYP1A1 ex-pression. Quercetin and luteolin, flavonoids with twohydroxyl groups at ring B, showed more potent effectson CYP1A1 expression, which depended on cellular sur-rounding.

Care should be taken if flavonoids are added in dietas food supplements when they are consumed withother medicaments, because their presence and chronicintake can drastically change metabolism of other drugsand their plasma level. Intensive degradation of drugsor their retention in cells due to altered expression ofcytochrome P450 could lead to serious consequencesand unwanted effects. It would be interesting to investi-gate synergistic and/or antagonistic effects of variousflavonoids using the same experimental approach estab-lished in this work.

Conclusions

Small differences in chemical structure of flavonoidsled to drastic change of their biological effects, and werecell specific. CK2 cell resistant to anti-cancer drugs sur-vived the treatment with flavonoids better than parentallaryngeal carcinoma HEp2 cells. CK2 cells were less per-ceptible to lipid peroxidation and damage caused byfree radicals, as compared to HEp2 cells. Fisetin andquercetin showed prooxidant effect in both cell lines.Quercetin increased glutathione level in both cell lines,pointing out that the same compound can act as pro-oxidant, but also prevent severe damage of the cell byinduction of GSH, which reacts with a free radical andforms harmless complex. Fisetin and quercetin damagedcell membrane, but did not enhance the damage causedby hydrogen peroxide. Quercetin, naringin, myricetin,fisetin and luteolin induced CYP1A1 protein expressionin CK2 cell line, while luteolin decreased basal level ofCYP1A1 in HEp2 cell line. Although the in vitro systemprovides sufficient conditions for determination of mo-lecular mechanisms of biological effects of flavonoids,this system cannot precisely reflect conditions in the sys-tem in vivo. Extensive studies on structure-function rela-tionship of flavonoids in different test systems couldprovide rational approach to drug and chemopreventiveagent design.


HEp2 cell line CK2 cell line

Fig. 8. Western blot analysis of cytochrome CYP1A1 expression in HEp2 and CK2 cells following 72 hours of treatment with thehighest nontoxic concentrations of flavonoids in HEp2 and CK2 cell line. b-actin was used as equal loading control. Values in brack-ets are results of densitometry measurements. C=control, Q=quercetin, N=naringin, F=fisetin, M=myricetin, L=luteolin

Acknowledgements

Ministry of Science, Education and Sports of the Re-public of Croatia (project No 058013 and No 0098076) isgratefully acknowledged for financial support.

References

1. L.G. Korkina, I.B. Afanasev, Antioxidant and chelating pro-perties of flavonoids, Adv. Pharmacol. 38 (1997) 151163.

2. S.A. Aherne, N.M. OBrien, Dietary flavonols: Chemistry,food content and metabolism, Nutrition, 18 (2002) 7581.

3. A.J. Day, M.R.A. Morgan, M.J.C. Rhodes, G. Williamson:The First Step in the Metabolism of Flavonoid Glycosides.In: Dietary Anticarcinogens and Antimutagens: Chemical andBiological Aspects, I.T. Johnson, G.R. Fenwick (Eds.), RoyalChemistry Society, Cambridge, UK (2000) p. 55.

4. J.B. Harborne: Nature, Distribution, and Function of PlantFlavonoids. In: Plant Flavonoids in Biology and Medicine: Bio-chemical, Pharmacological and Structure-Activity Relationships,V. Cody, E. Middleton, J.B. Harborne (Eds.), Alan R. Liss,New York, USA (1986) p. 15.

5. P. Hodek, P. Trefil, M. Stiborova, Flavonoids Potent and ver-satile biologically active compounds interacting with cyto-chromes P450, Chem. Biol. Interact. 139 (2002) 121.

6. N.H.P. Cnubben, I.M.C.M. Rietjens, H. Wortelboer, J. vanZanden, P.J. van Bladeren, The interplay of glutathione-re-lated processes in antioxidant defense, Environ. Toxicol. Phar-macol. 10 (2001) 141152.

7. L. Cipak, E. Berceliova, H. Paulikova, Effects of flavonoidson glutathione and glutathione-related enzymes in cyspla-tin-treated L1210 leukemia cells, Neoplasma, 50 (2003) 443446.

8. A.W. Boots, J.M. Balik, A. Bast, G.R.M.M. Haenen, The re-versibility of the glutathionyl-quercetin adduct spreads oxy-dised quercetin-induced toxicity, Biochem. Biophys. Res. Com-mun. 338 (2005) 923929.

9. N.M. OBrien, J.A. Woods, S.A. Aherne, Y.C. OCallaghan,Cytotoxicity, genotoxicity and oxidative reactions in cellculture models: Modulatory effects of phytochemicals, Bio-chem. Soc. Trans. 28 (2000) 2226.

10. J. Timbrell: Primary Events: Lipid Peroxidation. In: Princi-ples of Biochemical Toxicology, J. Timbrell (Ed.), Taylor andFrancis, London, UK (2000) pp. 204213.

11. J. Torel, J. Cillard, Antioxidant activity of flavonoids andreactivity with peroxy radical, Phytochemistry, 25 (1986)383387.

12. J. Terao, M. Piskula, Q. Yao, Protective effect of epicate-chin, epicatechin gallate and quercetin on lipid peroxida-tion in phospholipid bilayers, Arch. Biochem. Biophys. 308(1994) 278284.

13. M.R. Cholbi, M. Paya, M.J. Alcaraz, Inhibitory effects ofphenolic compounds on CCl4 induced microsomal lipid per-oxidation, Experientia, 47 (1991) 195199.

14. M. Osmak, Collateral resistance or sensitivity of human la-rynx carcinoma HEp2 cells resistant to cis-dichlorodiam-mineplatinum(II) or vincristine sulfate, Neoplasma, 39 (1992)197202.

15. M. Osmak, L. Beketi}-Ore{kovi}, M. Matuli}, J. Sori}, Re-sistance of human larynx carcinoma cells to cisplatin, gam-ma irradiation and methotrexate do not involve overex-pression of c-myc or c-Ki-ras oncogenes, Mutat. Res. 303(1993) 113120.

16. L. Beketi}-Ore{kovi}, M. Osmak, Human larynx carcinomacells resistant to cis-diamminedichloroplatinum(II): Cross--resistance pattern, Neoplasma, 41 (1994) 171176.

17. L. Beketi}-Ore{kovi}, M. Osmak, Modulation of resistanceto cis-diamminedichloroplatinum(II) by amphotericin B andaphidicolin in human larynx carcinoma cells, Cancer Che-mother. Pharm. 35 (1995) 327333.

18. L. Beketi}-Ore{kovi}, M. Jak{i}, S. Ore{kovi}, M. Osmak,Hyperthermic modulation of resistance to cis-diammine-dichloroplatinum(II) in human larynx carcinoma cells, Int.J. Hyperth. 13 (1997) 205214.

19. M. Osmak, D. Nik{i}, A. Brozovi}, A. Ambriovi}-Ristov, I.Vrhovec, J. [krk, Drug resistant tumor cells have increasedlevels of tumor markers for invasion and metastasis, Anti-cancer Res. 19 (1999) 31933198.

20. A. Ambriovi}-Ristov, J. Gabrilovac, T. ^imbora-Zovko, M.Osmak, Increased adenoviral transduction efficacy in hu-man laryngeal carcioma cells resistant to cisplatin is asso-ciated with increased expression of integrin avb3 and cox-sackie adenovirus receptor, Int. J. Cancer, 110 (2004) 660667.

21. G. Mickish, S. Fajta, G. Keilhauer, E. Schlick, R. Tschada, P.Alken, Chemosensitivity testing of primary human renalcell carcinoma by tetrazolium based microculture assay(MTT), Urol. Res. 18 (1990) 131136.

22. F. Tietze, Enzymatic method for quantitative determina-tion of nanogram amounts of total and oxidized glutathi-one: Applications to mammalian blood and other tissues,Anal. Biochem. 27 (1969) 502522.

23. M.M. Bradford, A rapid and sensitive method for the quan-tification of microgram quantities of protein utilising theprinciple of protein-dye binding, Anal. Biochem. 72 (1976)248254.

24. C. Cereser, S. Boget, P. Parvaz, A. Revol, Thiram-inducedcytotoxicity is accompanied by a rapid and drastic oxida-tion of reduced glutathione with consecutive lipid peroxi-dation and cell death, Toxicology, 163 (2001) 153162.

25. S. McCarthy, M. Somayajulu, M. Sikorska, H. Borowy-Bo-rowski, S. Pandey, Paraquat induces oxidative stress andneuronal cell death; Neuroprotection by water soluble co-enzyme Q-10, Toxicol. Appl. Pharm. 201 (2004) 2131.

26. E. Hurna, S. Hurna, Influence of zinc on cadmium inducedcytotoxicity in cultured rat liver cells, J. Trace MicroprobeTech. 16 (1998) 437444.

27. O.H. Lowry, N.J. Rosenbrough, A.L. Farr, R.J. Randall, Pro-tein measurement with the Folin phenol reagent, J. Biol.Chem. 193 (1951) 265275.

28. G. Rusak, H.O. Gutzeit, J. Ludwig Mller, Structurally re-lated flavonoids with antioxidative properties differen-tially affect cell cycle progression and apoptosis of humanacute leukaemia cells, Nutr. Res. 25 (2005) 141153.

29. G. Rusak, H.O. Gutzeit, J. Ludwig-Mller, Effects of struc-turally related flavonoids on hsp gene expression in hu-man promyeloid leukaemia cells, Food Technol. Biotechnol.40 (2002) 267273.

30. T. Hirano, M. Gotoh, K. Oka, Natural flavonoids and lig-nans are potent cytostatic agents against human leukemicHL-60 cells, Life. Sci. 55 (1994) 10611069.

31. G. Scambia, F. Ranelletti, P.B. Panici, M. Piantelli, G. Bo-nanno, R. De Vincenzo, G. Ferrandina, N. Maggiano, N.Capelli, S. Mancuso, Inhibitory effect of quercetin on pri-mary ovarian and endometrial cancers and synergistic an-tiproliferative activity with cis-diamminedichloroplatinum(II), Gynecol. Oncol. 45 (1992) 1319.

32. L. Beketi}-Ore{kovi}, M. Osmak, M. Jak{i}, Human larynxcarcinoma cells resistant to cis-diamminedichloroplati-num(II): Mechanisms involved in resistance, Neoplasma, 41(1994) 163169.

33. S. Devi Priya, C.S. Shyamala Devi, Protective effect of quer-cetin in cisplatin-induced cell injury in the rat kidney, In-dian J. Pharmacol. 31 (1999) 422426.


34. S.N. Pattipati, A. Singh, S.K. Kulkarni, Reversal of reser-pine-induced orofacial dyskinesia and cognitive disfunc-tion by quercetin, Pharmacology, 70 (2004) 5967.

35. N. Cotelle, Role of flavonoids in oxidative stress, Curr. Top.Med. Chem. 1 (2001) 569590.

36. K. Shimoi, H. Okada, J. Kaneko, M. Furugori, T. Goda, S.Takase, M. Suzuki, Y. Hara, N. Kinae: Bioavailability andAntioxidant Properties of Luteolin. In: Natural Antioxidantsand Anticarcinogens in Nutrition, Health and Disease, J.T. Kum-pulainen, J.T. Salonen (Eds.), Royal Society of Chemistry,Finland (1998) pp. 166174.

37. M.S. Kumar, M.K. Unnikrishnan, S. Patra, K. Murthy, K.K.Srinivasan, Naringin and naringenin inhibit nitrite-inucedmethemoglobin formation, Pharmazie, 58 (2003) 564566.

38. J.F. Young, S.E. Nielsen, J. Haraldsdotir, B. Daneshvar, S.T.Lauridsen, P. Knuthsen, A. Crozier, B. Sandstrm, L.O.Dragsted: Effect of Fruit Intake on Urinary Quercetin Ex-cretion and Biomarkers of Antioxidative Status. In: NaturalAntioxidants and Anticarcinogens in Nutrition, Health andDisease, J.T. Kumpulainen, J.T. Salonen (Eds.), Royal Soci-ety of Chemistry, Finland (1998) pp. 193203.


Documents

45_69