FF

A*L

R

AatBAscewttdr

na

ctarInshahma

iaEs

7

Biochemical and Biophysical Research Communications 282, 1161–1168 (2001)

doi:10.1006/bbrc.2001.4705, available online at http://www.idealibrary.com on

lavonoid B-Ring Chemistry and Antioxidant Activity:ast Reaction Kinetics

nanth Sekher Pannala,* Tom S. Chan,† Peter J. O’Brien,† and Catherine A. Rice-Evans*,1

Wolfson Centre for Age Related Diseases, GKT School of Biomedical Sciences, King’s College London, St. Thomas’s Street,ondon SE1 9RT, United Kingdom; and †Faculty of Pharmacy, University of Toronto, Toronto, Ontario M5S 2S2, Canada

eceived March 19, 2001

been demonstrated that the ABTS•1 formed reactsw8

dpbSvtte(haaicati

ctiiisiitddm

ttAsd

Rapid scavenging of the model stable radical cation,BTS•1, has been applied to screen for the antioxidantctivity of flavonoids. The reaction follows two dis-inct phases. For compounds with a monophenolic-ring there is a rapid initial phase of reduction ofBTS•1 within 0.1 s with no further change in theubsequent 2.9 s. In contrast, compounds with aatechol-containing B ring follow a fast initial scav-nging phase with a slow secondary phase. Flavonoidsith an unsubstituted B ring do not react within this

ime scale. The findings suggest that the structure ofhe B ring is the primary determinant of the antioxi-ant activity of flavonoids when studied through fasteaction kinetics. © 2001 Academic Press

Key Words: flavonoid; hydroxycinnamate; anthocya-idin; catechin; ABTS radical cation; structure-ctivity relationship; antioxidant activity; TEAC.

Long-lived free radicals with high extinction coeffi-ients are very useful in one-electron transfer reac-ions. 2,29-Azinobis-(3-ethylbenzthiazoline-6-sulfoniccid) (ABTS) can donate an electron to generate aelatively long-lived radical cation (ABTS•1) (Scheme) (1, 2) with absorption maxima at 645, 734, and 815m. Scavenging of ABTS•1 has been applied to thecreening of various compounds, both lipophilic andydrophilic, and food products for their antioxidantctivities (3–8). ABTS reacts at a rate of k . 108 withydroxyl, cysteinyl, glutathione, thiocyanate, and bro-ide radicals to yield the corresponding radical cation

s demonstrated by pulse radiolysis (1). It has also

Abbreviations used: TEAC, Trolox equivalent antioxidant capac-ty; ABTS, 2,29-azinobis-(3-ethylbenzthiazoline-6-sulfonic acid) di-mmonium salt; ECG, epicatechin gallate; EGC, epigallocatechin;GCG, epigallocatechin gallate; SAR, structure activity relation-hip.

1 To whom correspondence should be addressed. Fax: 1144 20848 6143. E-mail: catherine.rice-evans@kcl.ac.uk.

1161

ith ascorbic acid at neutral pH at a rate constant of3 106 M21 s21 and in acidic pH k , 105 M21 s21 (1).Flavonoids are a class of compounds that have been

emonstrated to be potent antioxidants based on theirhenolic hydroxyl groups. Antioxidant activity haseen attributed to their electron-donating ability.tructure-activity relationship (SAR) studies of fla-onoids have shown that the o-dihydroxy structure inhe B ring and the 2,3 double bond in conjugation withhe 4–oxo function in the C ring (as in flavones) aressential for effective free radical scavenging activity6, 9–12). The presence of a 3-hydroxyl group in theeterocyclic ring also increases the radical-scavengingctivity of flavonols, while additional hydroxyl groupst positions 5 and 7 of the A ring appear to be lessmportant. These structural features contribute to in-reasing the stability of the aroxyl radical, i.e., thentioxidant capacity of the parent compound. An addi-ional hydroxyl group on the B ring is reported toncrease the antioxidant activity (9).

The ABTS assay initially developed for screening theompounds for their antioxidant activity was based onhe activation of metmyoglobin with hydrogen peroxiden the presence of ABTS to produce the radical cation,n the presence or absence of the antioxidants undernvestigation (3). More recently Re et al. (7) have de-cribed a modified ABTS•1 assay based on the decolor-zation of the preformed radical cation. This methodnvolved formation of the radical cation through oxida-ion by potassium persulfate, exposure of the antioxi-ant under investigation to the radical cation for aefined time period, and spectrophotometric measure-ent of the extent of the radical quenched.A method has been devised for measuring the rela-

ive activities of flavonoids and phenolic acids as elec-ron donors by measuring their rapid reaction withBTS•1 rather than their reactivity over longer time-cales, during which slower reacting functional hy-roxyl groups might also participate in the reaction.

0006-291X/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

Tiem

M

6l(ccvltgoL

adp–fpiaAarl

bStsads

absorbance change is triggered as soon as the cuvette is filled whichtmd

mTdttagat

Tiarboatf

rs0vtsosT

R

snFttiTcifbTFpn2p

attts

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

he purpose of the rapid assay described here is todentify the polyphenolic compounds which are moreffective reductants and the structural features of theolecules underlying these effects.

ATERIALS AND METHODS

Chemicals. Trolox, ABTS (2,29-azinobis-(3-ethylbenzthiazoline--sulfonic acid) diammonium salt), potassium persulphate, and gal-ic acid were purchased from Sigma-Aldrich Chemical CompanyPoole, Dorset). Hydroxycinnamates (p-, m-, and o-coumaric acids,affeic acid, chlorogenic acid, ferulic acid, and sinapic acid), antho-yanidins (malvidin, delphinidin, cyanidin, and pelargonidin), fla-onols (quercetin, kaempferol, and galangin), flavones (rutin, luteo-in, apigenin, chrysin, and disometin), flavanones (naringenin,axifolin, and hesperetin), and catechins (catechin, epicatechin, epi-allocatechin, epicatechin gallate, and epigallocatechin gallate) werebtained from Extrasynthese (Lyon, France). Rathburn Chemicalstd. (Walkerburn, Scotland) supplied HPLC grade ethanol.

ABTS decolorization assay. The assay was carried out by inter-cting the antioxidants with the ABTS radical cation prepared asescribed in Re et al. (7). A stock solution of 7 mM ABTS wasrepared in water. To this solution potassium persulphate (2.45 mMfinal concentration) was added and the solutions allowed to react

or a duration of 12 h at room temperature in the dark. ABTS andotassium persulphate react stoichiometrically at 1:0.5 leading to anncomplete oxidation to generate ABTS•1. The radical thus gener-ted is stable in the dark at room temperature for two days (7). TheBTS radical cation solution was diluted in ethanol to obtain anbsorbance of 0.70 6 0.02 at 734 nm. The final concentration of theadical cation was calculated to be 80 mM (« 5 16000 M21.cm21, at734) (7).The interaction between antioxidants and ABTS•1 was carried out

y stop-flow kinetics (SFA20, Hi-Tech Scientific, Salisbury, UK). TheFA20 is a two-syringe stop-flow kinetics system capable of mixingwo solutions in a spectrophotometer (Hewlett-Packard 8453). Oneyringe of the SFA20 is filled with the antioxidant/ethanol solutionnd the other syringe with ABTS•1 solution. Solutions are with-rawn into the driving syringes followed by rapid mixing of theolutions in the cuvette by pushing the base plate. Measurement of

SCHEME I. Formation of ABTS radical cation.

1162

riggers a signal to the spectrophotometer. The dead time-volume forixing is 8 ms. Absorbance changes are monitored every 0.1 s for a

uration of 3 s at 734 nm.Prior to the testing with the antioxidants, a baseline is obtained byonitoring the change in absorbance between ABTS•1 and ethanol.his reading is used as the basal value for calculating the antioxi-ant activity of the compounds. Subsequently three concentrations ofhe antioxidants (1, 5, and 10 mM, final concentration) are tested forheir antioxidant activity. Stock solutions of the antioxidants (1 mM)re prepared in ethanol and diluted subsequently with ethanol toive an initial concentration of 2, 10, and 20 mM. As the antioxidantsre mixed in the cuvette with the ABTS•1 solution, the final concen-rations are half the initial concentrations.

Results are expressed in terms of the stoichiometric factor androlox equivalent antioxidant capacity (TEAC). Stoichiometric factor

s calculated on the basis of extent of ABTS•1 scavenged by thentioxidant. Results are calculated at 0.1 s and 3 s. The extent ofadical cation present at 0.1 s and 3 s is calculated from the absor-ance recorded and the extinction coefficient (7). The concentrationf the ABTS•1 reacted is then plotted against the concentration of thentioxidant applied. The ratio of the concentration of the antioxidanto the concentration of the ABTS•1 is expressed as the stoichiometricactor.

TEAC is calculated based on the percentage scavenging of theadical cation by the flavonoid relative to that by Trolox. Percentagecavenging of ABTS•1 is calculated from the absorbance values at.1 s and 3 s compared to the base value at 0 s. To calculate TEACalue, percentage scavenging of the radical cation is plotted againsthe concentration of the flavonoid, which exhibits a linear relation-hip. A similar plot is also obtained with Trolox. The ratio of the slopebtained from the flavonoid percentage inhibition graph with thelope obtained from Trolox is the antioxidant activity expressed asEAC.

ESULTS

The structures of the families of phenolic compoundstudied, flavonol, flavone, flavanone, hydroxycin-amate, anthocyanidin, and flavanol are shown inigs. 1a–1d. The different families show close struc-ural similarities with major variations in the nature ofhe C ring, and individual family members have vary-ng numbers of hydroxyl groups in the B and C rings.he rapid rates of reaction with the ABTS radicalation are assessed by the spectrophotometric monitor-ng of the change in absorbance at 734 nm every 0.1 sor a total duration of 3 s. Figure 2a shows the reactionetween ABTS•1 and increasing concentrations ofrolox (1, 5, and 10 mM) relative to the ABTS•1 control.or all concentrations the reaction appears to be com-lete within the first measured time point (0.1 s) witho further reaction taking place in the subsequent.9 s. The extent of scavenging, or reduction of ABTS•1, isroportional to the concentration of Trolox used.Interaction of flavonoids with the radical cation was

ssessed at 1, 5, and 10 mM concentrations as a func-ion of time. The mode of inhibition exhibited two dis-inct types of reaction depending on the structure ofhe B ring (Figs. 2b and 2c). For compounds with aingle hydroxyl group in the B ring, the reaction ap-

p2flatahcrs

Ttram

psct

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ears to be rapid and complete within the 0.1 s (Fig.b). This is particularly the case for naringenin—theavanone with a single 49-hydroxyl group in the B ring,pigenin—a flavone with a single 49-hydroxyl group inhe B ring, p-coumaric acid (4-hydroxycinnamic acid)nd pelargonidin—the anthocyanidin with a single 49-ydroxyl group in the B ring, as well as the standardompound for comparison, Trolox. Other compoundseact very rapidly in the initial 0.1 s followed by slowecondary phase in the subsequent 2.9 s (Fig. 2c).

FIG. 1. Chemical s

1163

hese include catechol-containing compounds (querce-in, luteolin, catechin, and epicatechin), trihydroxy Bing structures such as delphinidin, EGC and EGCGnd hindered phenols such as ferulic acid (4-hydroxy,3-ethoxy cinnamic acid).The extent of scavenging of the radical cation by the

henolics relative to that scavenged by Trolox, repre-ented by the TEAC value, is reported in Table 1. Theatechol-containing B ring structures, especially therihydroxy compounds such as EGC, delphinidin, and

ctures of flavonoids.

tru

the flavanol gallate esters are the most potent electron-dtpcTsEaec

isafAwcaTspnqcdtsa3romt

ct0it(pAttcatp(pst

cR0nTp2E

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

1164

onating compounds according to the TEAC values ofhe various flavonoid classes tested. This class of com-ounds shows a reactivity with ABTS•1 greater than orlose to that of Trolox after 3 s of reaction. As shown inable 1, most of the catechol-rich structures demon-trate their reactivity essentially within 0.1 s, namely,GC, EGCG, luteolin, quercetin, caffeic acid, and galliccid. Other catechol-containing phenolics continue anxtended reaction to 3 s such as taxifolin, ECG, cate-hin, epicatechin, delphinidin, and cyanidin.The potency of the antioxidant activity is also shown

n terms of stoichiometric ratio (Table 1), which repre-ents the concentration of ABTS•1 reacted per unitntioxidant concentration (mM). The stoichiometricactor is calculated by comparing the concentration ofBTS•1 reacted (based on the extinction coefficient)ith the concentration of the antioxidant applied. The

oncentration of the radical cation scavenged is plottedgainst the concentration of the antioxidant applied.he slope represents the stoichiometric factor, ashown in Fig. 3 for Trolox and three representativehenolics: sinapic acid (4-hydroxy,3,5-dimethoxy cin-amic acid)—a hindered phenol, chlorogenic acid (theuinic acid ester of 3,4-dihydroxy cinnamic acid)—aatechol compound, and pelargonidin—an anthocyani-in with a monohydroxyl group in the 49-position ofhe B ring. The ratios of the stoichiometric factor ob-erved for each compound relative to that for Trolox,re given in Table 1 obtained for reaction at 0.1 s ands. Hence, by definition, the stoichiometric factor is

elative to and is consistent with the TEAC valuesbtained. For the majority of compounds the stoichio-etric value is between 1 and 2 at 0.1 s, depending on

he reactivity of the B ring.However, stoichiometric factors do not take into ac-

ount the kinetics of the reaction. Since the majority ofhe reaction of the flavonoids occurs within the first.1 s as indicated in Table 1, an alternative approachncorporating the reaction time was used to calculatehe antioxidant activity, as described by Lebeau et al.13). The kinetics of the reaction were calculated bylotting the reciprocal of the concentration of theBTS•1 reacted against time at 0.1 s for each concen-

ration of the compound. The slope of this graph isaken as the rate constant of the reaction. The rateonstant obtained from this graph was then plottedgainst the ratio of the concentration of antioxidant tohat of ABTS•1 reacted. The slope (r 2 . 0.95) gave thearameter Z defined as radical scavenging activity13). The ratio of the Z value obtained for each com-ound to that obtained for Trolox, the relative radicalcavenging activity, is given in Table 1. The Z value isime-dependent.

FIG. 2. (a) Scavenging of ABTS radical cation by Trolox. },ontrol; h, 1 mM Trolox; ‚, 5 mM Trolox; E, 10 mM Trolox. (b)eaction of compounds (10 mM) exhibiting a fast initial phase within.1 s with no significant change in subsequent 2.9 s. }, control; ‚,aringenin; Œ, apigenin; h, p-coumaric acid; ■, pelargonidin; F,rolox. (c) Reaction of compounds (10 mM) exhibiting a fast initialhase within 0.1 s and a slow secondary phase in the subsequent.9 s. }, control; ‚, ferulic acid; Œ, epicatechin; h, delphinidin; F,CG; E, EGCG.

D

cdtd

aapiacitatq

tdtte(oecba(

torfCa

TABLE 1

T

C

3

H

M

U

T

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

ISCUSSION

The three structural categories of the B ring:atechol-containing compounds (including the trihy-roxy compounds), hindered phenols containing me-hoxy groups, and the monophenolic B ring structuresisplay different radical scavenging properties.The catechol group in the B ring of flavonoids is

mong the major structural considerations underlyingntioxidant activity due to the favourable reductionotential (12, 14, 15). Furthermore, phenolics contain-ng three adjacent hydroxyl groups such as delphinidinnd EGC are more effective than their dihydroxylounterparts, cyanidin, and epicatechin, respectively,llustrating the greater oxidizability of these specificrihydroxy structures. The predominant mechanism ofction is probably via donation of a single electron tohe radical cation resulting in the formation of a semi-uinone, which can donate a further electron to form

TEAC Values and Relative Stoic

Compounds

TEAC

0.1 s 3 s

rihydroxy compoundsEGCG 2.07 2.24EGC 1.77 1.88ECG 1.64 2.02Delphinidin 1.16 1.62Gallic acid 1.21 1.24atecholsLuteolin 1.04 1.01Quercetin 1.01 1.21Taxifolin 0.65 0.95Catechin 0.96 1.4Epicatechin 0.83 1.44Cyanidin 0.69 0.94Caffeic acid 0.82 0.92Chlorogenic acid 0.8 0.98

,49-dihydroxy compoundKaempferol 0.61 0.83indered phenolsHesperetin 0.18 0.45Ferulic acid 0.62 1.18Sinapic acid 0.95 1.12Malvidin 0.63 0.79onophenolic compoundsp-Coumaric acid 0.48 0.53Pelargonidin 0.46 0.51Apigenin 0.16 0.16Naringenin 0.11 0.12nsubstituted B ringChrysin 0.08 0.09

rolox 1 1

a From reference 9.b From reference 15.

1165

he quinone (Scheme II). For example, it has beenemonstrated by EPR spectroscopy that the spin dis-ribution during oxidation of quercetin remains en-irely on the B ring favouring the donation of twolectrons leading to the formation of an ortho-quinone16). This has also been related to the total conjugationf the quercetin molecule over the B and C rings. How-ver, in the case of taxifolin which lacks this extendedonjugation, the spin distribution was also reported toe on ring B emphasising the ready electron-donatingbility of the catechol group leading to its oxidation16).

Compounds containing a 49-monohydroxyl group onhe B ring are less potent antioxidants, the mechanismf action probably being via the formation of a phenoxyladical (17, 18) (Scheme III). This is especially the caseor phenolics in which there is no conjugation with the-ring (i.e., the C-ring is either saturated or unsatur-ted at the 2,3 position but lacking a 4-carbonyl group),

metric Factors at 0.1 s and 3 s

Ratio stoichiometricactor relative to Trolox Z value ratio

E7 (V)0.1 s 3 s 0.1 s

2.46 2.77 2.03 0.43a

1.98 2.19 1.67 0.42a

1.84 2.39 1.53 —1.16 1.69 1.15 —1.31 1.42 1.04 —

1.19 1.19 1.06 0.60a

1.05 1.35 0.97 0.33a

0.68 1.02 0.68 0.50a

1.04 1.65 0.77 0.57a

0.96 1.72 0.66 0.57a

0.72 1.02 0.72 —0.96 1.10 0.66 0.54b

0.65 0.96 0.65 —

1.04 1.26 0.90 0.75a

0.21 0.47 0.21 0.72a

0.77 1.46 0.47 —0.96 1.18 0.95 —0.64 0.84 0.63 —

0.50 0.59 0.25 —0.76 0.84 0.67 —0.17 0.19 0.17 .1.00a

0.13 0.15 0.13 .1.00a

0.11 0.12 0.11 —1 1 1

hio

f

o3m

mwgddtodaTatdtggigCraatd

g on

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

r for flavones lacking a hydroxyl group in the-position. Thus, apigenin and naringenin have mini-al reactivity possibly due to the relatively slow for-

FIG. 3. Calculation of stoichiometric factor dependin

SCHEME II. Potential oxidation products of catechols

1166

ation of the phenoxyl radical compared to the caseith hindered phenols. The position of the hydroxylroup in hydroxycinnamates plays an important role inetermining the antioxidant activity (19). A para hy-roxyl group enhances the antioxidant activity, whilehere is little or no effect when present in the meta orrtho positions. Hence, as predicted, m-coumaric acidemonstrates no activity while p-coumaric acid is rel-tively more reactive, with about 50% of the value ofrolox at 100 ms and 3 s. Kaempferol, a flavonol withsingle 49-hydroxyl group in the B ring, has a rela-

ively high activity, compared with other monohy-roxyl compounds studied which is most likely due tohe potential for conjugation between the 49-hydroxylroup and the 3-hydroxyl group through the conju-ated C ring. The phenoxyl radical that is formed couldn theory abstract an electron from the radical cation toenerate the di-cation and the phenolate. However,han et al. (17), based on their observation of the

eaction between monophenolics and NADH to gener-te NAD•, have suggested that the predominant mech-nism of action for these compounds is via the forma-ion of a phenoxyl radical. It has also recently beenemonstrated that the phenoxyl radicals formed from

the concentration of ABTS radical cation scavenged.

attr

teFmdmrnma

ptas(tpbtwohads

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

pigenin and naringenin are capable of oxidising GSHo GS•. However, flavonoids with a catechol group inhe B-ring, quercetin, and luteolin, form semi-quinoneadicals which are not capable of oxidising GSH (20).

The presence of a hindered phenol on the B ring viahe presence of a methoxyl group greatly enhances thelectron-donating properties in the 4- or 49-position.or example, the hydroxycinnamates, ferulic (3-ethoxy, 4-hydroxycinnamic) acid, and sinapic (3,5-

imethoxy, 4-hydroxycinnamic) acid are approxi-ately twice as effective in scavenging the ABTS

adical cation, in relation to the equivalent monophe-olic structures (4-hydroxycinnamic acid), at 3 s. Theore hindered the phenol, the more rapid the reaction,

s shown by sinapic acid and by malvidin, in which the

SCHEME III. Possible mechanism

SCHEME IV. Reaction of hinder

1167

henolic group is hindered by the presence of two me-hoxyl groups. This increases the rate of the reactionnd stabilises the formation of the phenoxyl radicaluch that the reaction is almost complete at 100 msScheme IV). In contrast, ferulic acid shows a biphasicrend in scavenging the radical cation, a fast initialhase followed by a continued and extensive reactionetween 100 ms and 3 s (Fig. 2c). It is of interest to notehat ferulic acid dimer is a far more potent antioxidanthen compared to the monomer (21). The low activityf hesperetin is accounted for by the location of theydroxyl group in the B-ring at the 39-position akin tom-hydroxyphenolic structure, with little electron-

onating potential. In contrast, the 49-position wouldhow more rapid formation of the phenoxyl radical as

action of monophenolic compounds.

henols with ABTS radical cation.

of

ed p

demonstrated by the fast reaction at 0.1 s. Under thesecmA

tvgUttgfl

frtactmtr

A

vt

R

7. Re, R., Pellegrini, N., Proteggente, A., Pannala, A. S., Yang, M.,

1

1

1

1

1

1

1

1

1

1

2

2

Vol. 282, No. 5, 2001 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

onditions, there is a possibility that hindered phenolsight regenerate the radical cation by interacting withBTS itself.For the flavonoids studied, hydroxyl groups at posi-

ions 5 and 7 on the A ring are common features, withariations in the number and position of hydroxylroups in the B ring and in the 3-position on the C ring.nder the time-scale of these fast reaction conditions,

here is negligible contribution to the antioxidant ac-ivity from the slower acting A-ring meta-hydroxylroups as demonstrated from the reaction of chrysin, aavone with an unsubstituted B-ring.The method of analysis described here using a model

ree radical in a chemical system can provide a fast andeproducible assay to screen compounds and plant ex-racts to give an indication of their relative abilities toct as electron-donating antioxidants, based on thehemistry of the B-ring. (It is of course, conceivablehat the reaction occurs more rapidly than the firsteasured value at 0.1 s). Indeed, the antioxidant ac-

ivity observed for each compound is reflected in theeported reduction potentials.

CKNOWLEDGMENT

The authors wish to thank Dr. Paul Talalay (Johns Hopkins Uni-ersity, Baltimore, MD, USA) for his valuable contribution towardshe discussion.

EFERENCES

1. Wolfenden, B. S., and Willson, R. L. (1982) Radical-cations asreference chromogens in kinetic studies of one electron trans-fer reactions: Pulse radiolysis studies of 2,29-azinobis-(3-ethyl-benzthiazoline-6-sulphonate). J. Chem. Soc. Perkin Trans. 2,805–812.

2. Maruthamuthu, P., Venkatsubramanian, L., and Dharmal-ingam, P. (1987) A fast kinetic study of formation and decay of2,29-azinobis(3-ethyl-benzthiazoline-6-sulphonate) radical cationin aqueous solution. Bull. Chem. Soc. Jpn. 60, 1113–1117.

3. Wang, H., Cao, G., and Prior, R. L. (1996) Total antioxidantcapacity of fruits. J. Agric. Food Chem. 44, 701–705.

4. Miller, N. J., and Rice-Evans, C. A. (1996) Spectrophotometricdetermination of antioxidant activity. Redox Report 2, 161–171.

5. Paganga, G., Miller, N., and Rice-Evans, C. A. (1999) The poly-phenolic content of fruit and vegetables and their antioxidantactivities. What does a serving constitute? Free Radic. Res. 30,153–162.

6. Rice-Evans, C. A., Miller, N. J., and Paganga, G. (1996) Struc-ture antioxidant-activity relationships of flavonoids and phenolicacids. Free Radic. Biol. Med. 20, 933–956.

1168

and Rice-Evans, C. A. (1999) Antioxidant activity applying animproved ABTS radical cation decolorization assay. Free Radic.Biol. Med. 26, 1231–1237.

8. van den Berg, R., Haenen, G. R. M. M., van den Berg, H., andBast, A. (1999) Applicability of an improved Trolox equivalentantioxidant capacity (TEAC) assay for evaluation of antioxidantcapacity measurements of mixtures. Food Chem. 66, 511–517.

9. Pietta, P-G. (2000) Flavonoids as antioxidants. J. Nat. Prod. 63,1035–1042.

0. Lien, E. J., Ren, S., Bui, H. H., and Wang, R. (1999) Quantitativestructure activity relationship analysis of phenolic antioxidants.Free Radic. Biol. Med. 26, 285–294.

1. Bors, W., and Saran, M. (1987) Radical scavenging by flavonoidantioxidants. Free Radic. Res. Commun. 2, 289–294.

2. Jovanovic, S. V., Steenken, S., Hara, Y., and Simic, M. G. (1996)Reduction potentials of flavonoid and model phenoxyl radicals.Which ring in flavonoids is responsible for antioxidant activity.J. Chem. Soc. Perkin Trans. 2, 2497–2504.

3. Lebeau, J., Furman, C., Bernierm, J-L., Duriez, P., Teissier, E.,and Cotelle, N. (2000) Antioxidant properties of di-tert-butylhydroxylated flavonoids. Free Radic. Biol. Med. 29, 900–912.

4. Jovanovic, S. V., Steenken, S., Tosic, M., Marjanovic, B., andSimic, M. G. (1994) Flavonoids as antioxidants. J. Am. Chem.Soc. 116, 4846–4851.

5. Jovanovic, S. V., Steenken, S., Simic, M. G., and Hara, Y. (1998)Antioxidant properties of flavonoids: Reduction potentials andelectron transfer reactions of flavonoid radicals. In Flavonoids inhealth and disease (Rice-Evans, C. A., and Packer, L., Eds.), pp.137–161, Marcel-Dekker Inc., New York, NY.

6. van Acker, S. A. B. E., de Groot, M. J., van den Berg, D-J.,Tromp, M. N. J. L., den Kelder, G. D-O., van der Vijgh, W. J. F.,and Bast, A. (1996) A quantum chemical explanation of theantioxidant activity of flavonoids. Chem. Res. Tox. 9, 1305–1312.

7. Chan, T., Galati, G., and O’Brien, P. J. (1999) Oxygen activationduring peroxidase catalysed metabolism of flavones or fla-vanones. Chemico-Biol. Interac. 222, 15–25.

8. Galati, G., Chan, T., Wu, B., and O’Brien, P. J. (1999)Glutathione-dependent generation of reactive oxygen species bythe peroxidase-catalysed redox cycling of flavonoids. Chem. Res.Tox. 12, 521–525.

9. Pannala, A. S., Razaq, R., Halliwell, B., Singh, S., and Rice-Evans, C. A. (1998) Inhibition of peroxynitrite dependent ty-rosine nitration by hydroxycinnamates: Nitration or electrondonation? Free Radic. Biol. Med. 24, 594–606.

0. Galati, G., Moridani, M. Y., Chan, T. S., and O’Brien, P. J. (2001)Peroxidative metabolism of apigenin and naringenin versus lu-teolin and quercetin: Glutathione oxidation and conjugation.Free Radic. Biol. Med. 30, 370–382.

1. Garcia-Conesa, M. T., Plumb, G. W., Waldron, K. W., Ralph, J.,and Williamson, G. (1997) Ferulic acid dehydrodimers fromwheat bran: Isolation, purification, and antioxidant properties of8-O-4-diferulic acid. Redox Report 3, 319–323.