Mutation Research 523524 (2003) 920

Review
Methodological considerations for characterizing potential
antioxidant actions of bioactive components in plant foods
Okezie I. Aruoma

Department of Neuroinammation, Faculty of Medicine, Division of Neuroinammation and Psychological Medicine,

Imperial College London, Charing Cross Hospital Campus, Fulham Palace Road, London W6 8RF, UK
Received 5 April 2002; received in revised form 2 August 2002; accepted 7 August 2002
Abstract
The study of free radicals and antioxidants in biology is producing medical revolution that promises a newage of health and
disease management. Fromprevention of the oxidative reactions in foods, pharmaceuticals and cosmetics to the role of reactive
oxygen species (ROS) in chronic degenerative diseases including cancer, autoimmune, inammatory, cardiovascular and
neurodegenerative (e.g. Alzheimers disease, Parkinsons disease, multiple sclerosis, Downs syndrome) and aging challenges
continue to emerge from difculties associated with methods used in evaluating antioxidant actions in vivo. Our interest
presently is focused on development of neurodegeneration models based on the integrity of neuronal cells in the central
nervous system and how they are protected by antioxidants when challenged by neurotoxins as well as Fenton chemistry
models based on the prole of polyunsaturated fatty acids (PUFAs) for the assessment of antioxidant actions in vivo. Use
continues to be made of several in vitro analytical tools to characterise the antioxidant propensity of bioactive compounds
in plant foods and supplements. For example, the oxygen radical absorbance capacity (ORAC), ferric reducing antioxidant
power (FRAP), total oxidant scavenging capacity (TOSC), the deoxyribose assay, assays involving oxidative DNA damage,
assays involving reactive nitrogen intermediates (e.g. ONOO

), Trolox equivalent antioxidant capacity (TEAC) and the

2,2-diphenyl-1-picrylhydrazyl (DPPH) assay. There is need to agree governance on in vitro antioxidant methods based on
an understanding of the mechanisms involved. Because some of the assays are done in non-physiological pH values, it is
impossible to extrapolate the results to physiological environment. The consensus of opinion is that a mix of these tools should
be used in assessing the antioxidant activities in vitro. The proof of bio-efcacy must emanate from application of reliable
in vivo models where markers of baseline oxidative damage are examined from the standpoint of how they are affected by
changes in diet or by antioxidant supplements.
2002 Elsevier Science B.V. All rights reserved.
Keywords: Reactive oxygen species; Antioxidant; Degenerative diseases
1. Introduction
Several classes of dietary compounds have been
suggested to reduce the risk of some cancers, espe-
cially those of the gut, and there is some evidence

Tel.: +44-20-8846-7023; fax: +44-20-7635-9634.

E-mail address: o.aruoma@ic.ac.uk (O.I. Aruoma).
that consumption of certain foods leads to a reduc-
tion in biomarkers of oxidative damage. The active
principles in extracts from a variety of plant sources,
such as rosemary, sage, cocoa shells, oats, tea, olives,
garlic, ginger, red onion skin, grapes, apple cuticle,
wheat gliadin, korum rind, licorice, nutmeg, clove,
oregano, thyme, mustard leaf seed, chia seed, peanut
seed coat, birch bark, carob pod, tempeli, yam,
0027-5107/02/$ see front matter 2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0027-5107(02)00317-2
10 O.I. Aruoma / Mutation Research 523524 (2003) 920
Fig. 1. Flavonoid subclasses.
axseed, clover sprouts, mango, mangostum, vanilla
and a variety of plants and seeds from tropical and
sub-tropical regions of the world continue to be of
interest for use in complementary medicine supple-
ments. In the main, most of the extracts contain im-
portant vitamins, Vitamin E, Vitamin C, -carotene,
avonoids and other polyphenols [15]. Flavonoids
are polyphenolic compounds that occur ubiquitously
in plant foods and structurally have variations in the
C ring that characterizes the different types namely,
avonols, avones, isoavones, avonones, avanol
and anthocyanins (see Fig. 1). Quercetin, kaempferol
and quercetagetin are avonols. Myricetin, naringin
and naringenin are avonones. Phenolic compounds
are the major components of red wins and these
are responsible for their prominent characteristics
such as colour and astringency. The avan-3-ols,
catechin, () epicatechin, ()-epicatechin gallate,
()-epigallocatechin-3-gallate (EGCG), gallocatechin
are the major constitutive units of grape proantho-
cyanidins. Epigallocatechin-3-gallate is a potent an-
tiproliferative and cytostatic agent in human cancer
lines [6]. The genistein and diadzein are well-known
primary isoavones or phytoestrogens. Genestein in-
hibits the activities of topoisomerases I and II in vitro
[7] and is considered to be the primary anticancer
constituent of soy, soy-based products are a major
food source in Asia and South-East Asian popula-
tions (reviewed in [8]). The bioactive components of
plant and seed extracts are widely discussed as po-
tential antioxidant prophylactic agents for both health
and disease management [1,2,5,813]. By denition,
a substance which when present at low concentra-
tions compared with those of an oxidizable substrate
(lipid including polyunsarurated fatty acids, proteins,
carbohydrates or DNA signicantly delay or prevent
oxidation of the substrate) would be considered an
antioxidant. Acidic compounds (including phenols)
usable in foods, cosmetics and pharmaceutical prepa-
rations and which can readily donate an electron or
a hydrogen atom to a peroxyl or alkoxy radical to
terminate a lipid peroxidation chain reaction or to
regenerate a phenolic compound, or which can effec-
tively chelate a pro-oxidant transition metal can also
be classied as an antioxidant.
From prevention of the oxidative reactions in foods,
pharmaceuticals and cosmetics to the role of reactive
oxygen species (ROS) in chronic degenerative diseases
including diabetes, cancer, arthritis, cardiovascular
disease, macular degeneration, Alzheimers disease,
Parkinsons disease, multiple sclerosis, Downs syn-
drome and aging challenges continue to emerge from
difculties associated with methods used in evaluat-
ing antioxidant actions in vitro and in vivo. In order
to establish bio-efcacy of antioxidants, it is prudent
to measure markers of baseline oxidative damage in
vivo and examine how they are affected by changes in
diet or by antioxidant supplements. Lipid constituents
in foods and in living organisms derive from four
distinctive interfacial groups namely bulk food lipids
O.I. Aruoma / Mutation Research 523524 (2003) 920 11
(e.g. oils), dispersed food lipids (e.g. membranes and
emulsionssalad dressing), dispersed lipids in living
organisms (membranes and organelles) and watery
uids in organisms (e.g. cytoplasm, plasma) [1416].
The extent to which oxidation of fatty acids and their
esters occurs in foods depends on the chemical struc-
ture of the fatty acid, the nature of food processing
and the temperature at which the foods are stored
and/or cooked, and the minor constituent antioxidants
[14,15,1719]. Antioxidant efcacy may be demon-
strated in one system but to fail to protect, or even
sometimes to cause damage, in others. This is of par-
ticular importance as certain antioxidant inhibitors of
lipid peroxidation may not protect other targets (such
as DNA and protein) against damage, and some-
times can even aggravate such damage. Butylated
hydroxyanisole (BHA) is a powerful inhibitor of lipid
peroxidation but high doses of it can induce cancer
of the rat fore-stomach, a process that may involve
oxidative DNA damage [20]. The amount of BHA
that can be absorbed from foods may not pose imme-
diate problem to the human consumer. The chemical
nature of the food components, effects of compo-
nents within the food matrix, the mix of diet, and the
health status, all combine to affect the bioavailability
of the nutritive component to humans [1,2,16,21,22].
Food grade and oil-soluble antioxidants include
butylated hydroxyanisole, butylated hydroxytoluene
(BHT), tertiary butylhydroquinone (TBHQ), esters
of 3,4,5-tri-hydroxybenzoic acid (propyl, octyl and
dodecyl esters), ethoxyquin (6-ethoxy-1,2-dihydro
2,2,4-trimethyl quinoline) (used mostly in animal
feeds) and dl--tocopherols [2325].
2. Basics of antioxidant characterization
Antioxidant actions in vivo or in food may be
through inhibiting generation of ROS, or by direct
scavenging of free radicals. The levels of endogenous
antioxidants may also be up-regulated by increased
expression of the genes encoding the antioxidant en-
zymes superoxide dismutase, catalase or glutathione
peroxidase. Superoxide dismutases (SODs) [26] re-
move O
2

by greatly accelerating its conversion

to H
2
O
2
. Human cells have a SOD enzyme con-
taining manganese at its active site (MnSOD) in the
mitochondria. A SOD with copper and zinc at the
active site (Cu, Zn-SOD) is also present but largely
in the cytosol. The Cu/Zn SOD is often referred to
as SOD1 and has a strong link to the pathology of
amyotrophic lateral sclerosis where there is a muta-
tion in the gene encoding for this enzyme. Mutant
SOD1 exerts its deleterious effect by a toxic gain in
function rather than a loss in activity [2729]. Cata-
lases in the peroxisomes convert H
2
O
2
into water and
O
2
and help to dispose of H
2
O
2
generated by the ac-
tion of oxidase enzymes located in these organelles.
However, the most important H
2
O
2
-removing en-
zymes in human cells are glutathione peroxidases
(GSHPX); enzymes that require selenium (as seleno-
cysteine at the active site) for their action. GSHPX
enzymes remove H
2
O
2
by using it to oxidize reduced
glutathione (GSH) to oxidized glutathione (GSSG).
Glutathione reductase, an FAD-containing enzyme,
regenerates GSH from GSSG, with NADPH as a
source of reducing power. Antioxidant enzymes and
antioxidant molecules can inhibit free radical pro-
duction by chelating the transition metal catalysts,
breaking chain reactions, reducing concentrations of
ROS, and/or by scavenging initiating radicals. This
subject has been reviewed in many publications, e.g.
[1,3036] to which the reader is referred. Humans
contain a variety of radical-scavenging antioxidants,
including GSH, uric acid, -tocopherol (Vitamin E)
and ascorbic acid (Vitamin C). -Tocopherol delays
lipid peroxidation by reacting with chain-propagating
peroxyl radicals faster than these radicals can react
with proteins or fatty acid side-chains. In theory,
-carotene has remarkable antioxidant chemistry.
Chemically carotenoids are not chain-breaking an-
tioxidants, but their ability to scavenge singlet oxygen
suggests different reaction mechanisms. Mortensen
[31] has shown that carotenoids react with peroxyl
radicals (e.g. benzylperoxyl radicals) only slightly
more reactive than lipid peroxyl radicals neither by
electron transfer nor by hydrogen atom abstraction but
by adduct formation. Ubiquinol (reduced coenzyme
Q) might also regenerate -tocopherol in membranes
and lipoproteins (reviewed in [37]). It is also widely
believed that ascorbic acid (and possibly GSH) can
reduce the phenoxyl radical of tocopherol back to
-tocopherol. Ascorbate (Vitamin C) is often claimed
to be an important antioxidant in vivo. Its ability to
show antioxidant properties is related to the fact that
the dehydroascorbate radical is much less reactive
12 O.I. Aruoma / Mutation Research 523524 (2003) 920
than are many of the radicals that can be scavenged
by ascorbate. Enzymic systems exist in vivo to re-
duce this radical back to ascorbate using NADH (the
NADH-semidehydroascorbate reductase enzyme) or
GSH (the dehydroascorbate reductase enzyme) as
sources of reducing power. Unfortunately, these en-
zymes seem to be largely intracellular. It is therefore
not surprising that ascorbic acid is often rapidly de-
pleted in human extracellular uids under conditions
of oxidative stress.
3. Approaches to assessing dietary antioxidant
action in vitro
As antioxidants scavenge free radicals and oxidants,
the assumption is antioxidants in diet may therefore
prevent diseases. Simple experiments (Fig. 2) can
be performed to examine direct antioxidant ability
in vitro [1,16,3847]. Demonstrating antioxidant ef-
cacy in vivo requires use of valid in vivo models.
However, results from the in vitro methods can enable
assessment of potential in vivo efcacy. Measurements
of lipid peroxidation using rat liver or cardiac micro-
somes, ox-brain phospholipid liposomes, arachidonic
acid, and other lipid model systems (e.g. bulk oil and
emulsied oil systems) should be the rst line of tests
to establish the potential antioxidant action of dietary
antioxidant compounds. Antioxidant index based on
Table 1
Assessment of antioxidant actions using different in vitro methods that measure and/or predict index of antioxidant activities
Analytical method References
Trichrolomethyl peroxyl radical (CCl
3
O
2

) [66,67]
Alkoxyl peroxyl radical scavenging [68]
Deoxyribose assay for determination of rate constants of reactions with hydroxyl radicals [69,70]
Oxygen radical absorbance capacity assay (ORAC assay) [7173]
The photochemiluminescence assay (PCL assay) [74]
TEAC I (2,2

-azino-bis (3-ethylbenzthiazoline-6-sulfonic acid radical ABTS

+
/metmyoglobin) [75,76]
Trolox equivalent antioxidant capacity assay. TEAC II (ABTS
+
with manganese dioxide MnO
2
) [77,78]
Novel methods involving oxidative DNA damage [79,80]
TEAC III (ABTS
+
with potassium persulfate KO
8
S
2
) [81]
Oil stability index method [82]
Total oxidant scavenging assay (TOSCA) [83]
Tocopheroxyl radical attenuating ability (TRAA) [84]
Diphenyl-1-picrylhydrazyly assay (DPPH assay) [85]
N, N-Dimethyl p-phenylendiamine assay (DMPD assay) [86]
Total radical-trapping antioxidant parameter (TRAP assay) [8789]
The ferric reducing ability of plasma assay (FRAP assay) [9092]
ability to scavenge peroxyl radicals provides support
for antioxidant efcacy in vitro (see Table 1). Es-
sentially the methods either involve interaction with
radicals or interaction with metal ions. The interfacial
characteristics of the antioxidants (in part determined
by their partition coefcients) are equally critical
[3841]. There is need to agree governance on in vitro
antioxidant methods based on an understanding of the
mechanisms involved. Because some of the assays are
done in non-physiological pH values, it is impossible
to extrapolate the results to physiological environ-
ment. For this reason, knowledge-based antioxidant
efcacy must derive from in vivo data where effects
on surrogate biomarkers are considered. Table 1 gives
some examples of several analytical methods, which
are routinely used to characterize antioxidants in
vitro. Comments follow on some of the methods.
The spectrophotometric technique, the Trolox equ-
ivalent antioxidant capacity involves the generation
of the long-lived specic radical cation chromophore
of 2,2

-azinobis-(3-ethylbenzothiazoline-6-sulphonic
acid) (ABTS) by controlled chemical oxidation.
The ABTS
+
radical cation has absorption max-
ima in the near-infrared region at 645, 734 and
815 nm. The TEAC reects the ability of hydrogen-
or electron-donating antioxidants to scavenge the
ABTS
+
radical cation compared with that of Trolox.
The antioxidant suppresses the A
734
to an extent
and on a time scale dependent on the antioxidant
O
.
I
.
A
r
u
o
m
a
/
M
u
t
a
t
i
o
n
R
e
s
e
a
r
c
h
5
2
3

5
2
4
(
2
0
0
3
)
9

2
0
1
3
Fig. 2. Antioxidant strategies showing prole of various analytical methods.
14 O.I. Aruoma / Mutation Research 523524 (2003) 920
activity. Other modications of the TEAC assay
exist. TEAC I allows for the measurement of hy-
drophilic antioxidants. Carotenoids, tocopherols and
other lipophilic antioxidants can be measured us-
ing TEAC II whilst TEAC III may be applied to
both. Trolox C has a value one in all types but as
would be expected the antioxidant index for other
compounds are varied. Table 2 shows TEAC values,
which are compared with the antioxidant index from
other methods where available. Brand-Williams et al.
described a method involving use of the free radical
2,2-diphenyl-1-picrylhydrazyl (DPPH

), where an-
tioxidants are allowed to react with the stable radical
in a methanol solution. The reduction in the concen-
tration of the DPPH

is followed by monitoring the

decrease in its absorbance at a characteristic wave-
length during the reaction. In its radical form, DPPH

absorbs at 515 nm, but upon reduction by an antiox-

idant (AH) or a radical species (R

), the absorption
disappears. The antioxidant activities are determined
using DPPH as a free radical. Antioxidant solution in
methanol is added to a methanol DPPH solution. The
absorbance at 515 nm is measured until the reaction
reached a plateau. The exact initial DPPH

concen-
tration (CDPPH) in the reaction medium is calculated
from a calibration curve determined by linear regres-
Table 2
Antioxidant index using different analytical methods for a variety of naturally occurring avonoids, phenylpropanoids, carotenoids and
vitamins
Antioxidant ORAC Values CCl
3
O
2

(M
1
s
1
) TEAC (Mm) DPPH values (antiradical efciencies)
Hydroxytyrosol Not available 8.37 10
6
Not available Not available
Vitamin C 0.95 0.02 1.3 10
8
1.0 0.02 3.70
Vitamin E (-tocopherol) 0.5 0.02 4.9 10
8
1.0 0.03 Not available
Vitamin E (-tocopherol) 1.36 0.14 Not available Not available 4
Vanillin Not available 3.97 10
6
Not available 0.05
Vanillic acid Not available 9.54 10
7
Not available 0.17
6-Gingerol Not available 4.67 10
6
Not available Not available
Zingerone Not available 5.63 10
6
Not available 3.7
Ferulic acid Not available 7.5 10
6
1.9 0.02 2.33
Quercetin 6.47 0.29 3.9 10
7
4.7 0.10 Not available
Trolox C 1.0 2.23 10
8
1 Not available
Chrysin Not available 9.86 10
7
1.4 0.07 Not available
Apigenin Not available Not available 1.5 0.08 Not available
Fisetin Not available 4.09 10
6
Not available Not available
Epicatechin Not available 7.3 10
6
2.4 0.02 Not available
Ergothioneine Not available 6.3 10
8
0.60 0.03 Not available
The reader may refer to the references in Table 1 from which the values quoted have been extracted. This is illustrative of the current
problem in several methods yielding values that are not comparable.
sion. Antiradical activity was dened as the amount
of antioxidant necessary to decrease the initial DPPH

concentration by 50% (efcient concentration = EC

50
((mol l
1
)AO/(mol l
1
)DPPH

). The larger the anti-

radical power (ARP) the more efcient the antioxidant
action.
A rapid and convenient method for screening poten-
tial inhibitors of the initiation of low-density lipopro-
tein lipid peroxidation. The so-called -tocopheroxyl
radical attenuating ability (TRAA) method is based
on the capacity of an antioxidant to attenuate
-tocopheroxyl radicals generated by irradiation of
-tocopherol-containing micelles with UV light, and
measured directly by electron spin resonance spec-
troscopy. TRAA makes use of the positively and nega-
tively charged micelles with added -tocopherol. Like
the method employing the trichloromethylperoxyl rad-
ical (CCl
3
O
2

), the TRAA method can be expensive,

requiring an ESR machine. The method involving the
CCl
3
O
2

radicals require a pulse radiolysis facility.

The CCl
3
O
2

radical is a reactive peroxyl radical that

is frequently used as a model to assess the ability of
a compound to react with peroxyl radicals. CCl
3
O
2

is a lipid-soluble radical generated by radiolysis of an

aqueous mixture of propan-2-ol and CCl
4
. The abili-
ties of compounds to react with the radical can then
O.I. Aruoma / Mutation Research 523524 (2003) 920 15
be assessed. Peroxyl radicals generated by thermal
homolysis of 2,2

-azobis-amidinopropane (ABAP)
can cause the oxidation of -keto--methiolbutyric
acid (KMBA) to ethylene. The formation of ethylene
can be monitored by gas chromatographic analysis
of headspace from the reaction vessel. Antioxidants
compete with KMBA for the oxidants thus inhibiting
the formation of ethylene. The area under the ki-
netic curve is calculated and total oxidant scavenging
capacity (TOSC) values are quantied by compar-
ing the areas of control and sample reaction. This
is the basis of the total oxidant scavenging capac-
ity assay (TOSC assay). In use also it is the ferric
reducing antioxidant power (FRAP) assay, which
has found increased applications in the evaluation
of the antioxidant components of dietary polyphe-
nols. The FRAP assay depends on the reduction of a
ferric tripyridyltriazine [Fe(III)-TPTZ)
2
] complex to
the ferrous tripyridyltriazine [Fe(II)-TPTZ)
2
] by an
antioxidant, usually a non-physiological conditions
with low pH of about 3.6. The FRAP and TEAC
assays measure the reducing capabilities (albeit only
partially). Ou et al. [72] described an improved oxy-
gen radical absorbance capacity (ORAC) assay using
uorescein (3

-dihydrospiro[isobenzofuran-1[3H],
9

[9H]-xanthe]-3-one as a uoresecent probe instead

of the original -phycoerythrin. The -phycoerythrin
(a protein isolated from P. cruentum) probe has
distinct excitation and emission wavelengths, high
uorescence yield, and sensitivity to ROS and water
solubility. The newer ORAC assay is reported to allow
a direct measurement of hydrophilic chain-breaking
antioxidants capacity against peroxyl radicals (see
Huang et al. [73] in Table 1). Proteggente et al. [46]
assessed the antioxidant actions of extracts from reg-
ularly consumed fruit and vegetables using TEAC,
FRAP and ORAC and found that the indices corre-
lated well with total phenolic and Vitamin C contents.
Schlesier et al. [47] applied the TEAC IIII, TRAP,
DPPH, DMPD and PCL to tea and juices and found
that blackcurrant juice showed the highest antioxidant
activity in all the test systems when compared with
tea, apple juice and tomato juice. The tools all give
different values and use of a mixture of these must
now be the adopted procedure for in vitro antioxidant
screening.
The deoxyribose assay allows determination of rate
constants of reactions with OH

radicals, assessment
of abilities to exert pro-oxidant action, and assess-
ment of abilities to chelate metal iron. The positive
pro-oxidant actions in the deoxyribose system rely on
the ability of the compounds to promote reduction of
Fe
3+
to Fe
2+
chelates and hence OH formation in the
presence of H
2
O
2
. In tests, for the ability of a com-
pound to exert pro-oxidant action in vitro using the
deoxyribose assay, it is the ability of the compound to
mediate a reaction similar to that of ascorbate that con-
stitutes the basis of the evaluation. The Fe(III)EDTA
complex has a tested propensity to be reduced by the
pro-oxidant if the redox potential of other metal com-
plex is favourable. In the absence of EDTA, iron ions
are equally available to both the deoxyribose and the
compound under test. Thus, compounds that are able
to preferentially chelate iron and present the result-
ing metal complex in a less redox-active form com-
pared with EDTAmetal complex will protect deoxyri-
bose against damage in the presence of ascorbate and
H
2
O
2
. Substances that inhibit in the assay are also
those that are able to bind iron ions strongly enough
to remove them from deoxyribose. Thus, EDTA re-
moves iron ions from deoxyribose, but ironEDTA
chelates are very effective in generating OH

so that
the deoxyribose is still degradedthis time by OH

in free solution, rather than by OH

formed on the
deoxyribose molecule. Studies using the deoxyribose
assay can provide useful information on the likelihood
that molecules could chelate iron ions in a way that
prevents them from catalyzing OH

formation. Thus,
when iron is added to the assay mixture as ferric chlo-
ride instead of as ferricEDTA, some of the Fe
3+
ions
bind to deoxyribose, and damage to the sugar becomes
site-specic such that the OH

formed by bound iron

ions immediately attacks the deoxyribose. The ability
of a substance to inhibit deoxyribose degradation un-
der these reaction conditions is a measure of its ability
to interfere with site-specic Fenton chemistry. When
ascorbate is omitted from the deoxyribose reaction
mixture, the ability of added compounds to reduce the
Fe
3+
EDTA complex can be tested. This gives an in-
dex of pro-oxidant action.
Aside from lipids, DNA is also a major cellular
component and is prone to oxidative attack. DNA
damage is often measured as single-strand breaks,
double-strand breaks or chromosomal aberrations.
Mechanisms involving the Fenton system, ionizing
radiation and nuclease activation been suggested to
16 O.I. Aruoma / Mutation Research 523524 (2003) 920
account for much of the DNA damage that occurs
in biological systems [48]. Incubation of calf-thymus
DNA with a system producing OH

radicals gives
rise to extensive chemical modication of the DNA
bases in a way that appears to be diagnostic for OH

.
Incubation of DNA with Fe(III)EDTA, ascorbate
and H
2
O
2
led to signicant rises in the amounts of
several oxidatively-modied bases; this is character-
istic of attack by OH

[49]. Thus dietary antioxidants

will modulate the DNA base modications if they are
able to scavenge OH

. Omission of ascorbate from

the reaction mixture greatly decreased the DNA base
modication but dietary antioxidants that can act as
pro-oxidants, would be able to promote oxidative
DNA damage in this system. The method discussed
by Aruoma et al. [79] and Sakakibara et al. [80] (see
citations in Table 1) would also allow for antioxi-
dant assessment based on the modulation of oxidative
DNA damage.
4. The bleomyciniron-dependent DNA
damage
Bleomycin, an antitumor antibiotic, binds to DNA
by using its bithiazole and terminal amine residues,
and it also complexes with metals (such as iron)
using the -amino-alaninepyrimidine--hydroxy his-
tidine portion of the molecule. The bleomycin assay
was rst described to measure non-transferrin bound
iron in biological samples but has now been adapted
as a method for assessing the pro-oxidant action
of food additives and/or nutrient components [1].
Bleomycin binds iron ions and the bleomyciniron
complex will degrade DNA in the presence of O
2
and a reducing agent such as ascorbic acid. The
reaction occurs by attack of a ferric bleomycin per-
oxide (BLM-Fe(III)-O
2
H

) on the DNA. The fer-

ric peroxide can be formed by direct reaction of
ferricbleomycin with hydrogen peroxide or from a
BLM-Fe(III)-O
2
complex. It is possible that under
certain conditions the BLM-Fe(III)-O
2

might de-
compose to yield O
2

and BLM-Fe(III)-O
2
H

to
release OH

. Hydroxyl radical is not necessarily the

major DNA damaging species in the bleomycin sys-
tem. The bleomyciniron(III) complex by itself is
inactive in inducing damage in DNA. Oxygen and
a reducing agent or hydrogen peroxide are required
for the damage of DNA to occur. DNA cleavage by
bleomycin releases some free bases and base prope-
nals in amounts that are stoichiometric with strand
cleavage. When heated with thiobarbituric acid (TBA)
at low pH, base propenals rapidly decompose to give
malondialdehyde (MDA) which combines with TBA
to form a pink (TBA)
2
MDA adduct. A positive test
is obtained when the compound is able to reduce
bleomycinFe
3+
DNA complex in the absence of
ascorbate to the more active bleomycinFe
2+
DNA
complex (in the presence of oxygen) which would
result in DNA damage [43].
5. DNA damage using copper-1,10-
phenanthroline complex
The original copper-phenanthroline assay was de-
veloped to measure copper ions in biological uids,
but has been adapted as a method for assessing the
pro-oxidant action of food additives and/or nutrient
components [1]. Hydrogen peroxide is implicated in
the mechanism of the DNA damage by the copper-
phenanthroline system. Food additives and/or nutri-
ent components which exert pro-oxidant action in the
copper-phenanthroline system do not react with hy-
drogen peroxide to the extent that would affect the
outcome of the assay. Hydroxyl radicals are in-
volved in the damage to DNA caused by the copper-
phenanthroline system. Unlike the bleomyciniron
mediated damage to DNA, damage in the copper-
phenanthroline system is conned mainly to the
DNA bases. The small amount of DNA sugar dam-
age is what the copper-phenanthroline assay mea-
sures. When a reducing agent is omitted from the
reaction mixture, no damage to deoxysugar in DNA
occurs. Ascorbate and mercaptoethanol or other re-
ducing agents lead to increased deoxysugar damage
if added. Assays involving DNA rely on ability to re-
duce either the ironbleomycinDNA or copper-1,10-
phenanthrolineDNA complex. Fortunately, organic
solvents do not affect the outcome of DNA-dependent
assays. Thus, where the deoxyribose assay can-
not be performed due to solubility restriction, the
copper-phenanthroline assay would sufce. Circum-
venting potential pro-oxidant action could contribute
to increased protective ability of dietary antioxidants
towards susceptible substrates. For example, proteins
O.I. Aruoma / Mutation Research 523524 (2003) 920 17
protect DNA against the pro-oxidant actions of some
avonoids and polyphenolic compounds in in vitro
systems. The work with plant extracts Herbor 025
and Spice Cocktail Provenal, illustrates the concepts
[43].
6. The prole of polyunsaturated fatty acids
as biomarkers of oxidative stress
The iron complex of the chelating agent nitrilotri-
acetic acid (NTA) is nephrotoxic. Intraperitoneal in-
jection of ferric nitrilotriacetate (Fe-NTA) induces re-
nal proximal tubular damage associated with oxida-
tive damage that eventually leads to a high incidence
of renal cell carcinoma in rodents after repeated ad-
ministration [5052]. In the kidney, Fe-NTA can be
ltered through the glomeruli into the lumen of the
renal proximal tubule where Fenton chemistry me-
diated oxidative damage occurs. Deiana et al. [53]
have described an in vivo biomarker model based on
the prole of polyunsaturated fatty acids (PUFAs).
Fe-NTA induced a time-dependent reduction in the
levels of polyunsaturated fatty acid, together with an
increase of conjugated dienes value and a decrease
of cellular antioxidants -tocopherol and glutathione.
A strong reduction of PUFA concentrations, up to
3545% of the control levels, were observed in the
kidney and liver of rats at the time of the maximum
detectable oxidation (3 h) after the injection of a sub-
lethal dose of Fe-NTA. The working model is shown in
Fig. 3.
Fig. 3. The Fe-NTA model based on polyunsaturated fatty acids (PUFA) proles. The peak of the levels of the PUFAs at 3 h following
the injection of acute dose of Fe-NTA may serve as a sampling point if this model is to be used for the assessment of the biopotency of
antioxidants (full details in [53]).
7. Method for assessing antioxidant capacity of
nutrient components based on reactions with
peroxynitrite
The reaction between nitric oxide (NO) and su-
peroxide radical (O
2

) produce the oxidant perox-

ynitrite (ONOO

) and occurs with a rate constant

of 6.7 10
9
M
1
s
1
[54]. Peroxynitrite can induce
peroxidation of lipids, oxidizes methionine and -SH
residues in proteins, depletes antioxidants and causes
DNA damage. Addition of peroxynitrite to biolog-
ical uids leads to nitration of tyrosine residues,
and the presence of these appear to be a marker
of peroxynitrite-dependent damage in vivo [55,56].
Indeed, increased levels of 3-nitrotyrosine have
been detected in numerous human diseases such as
rheumatoid arthritis, Parkinsons disease, Alzheimers
disease and asthma. Scavenging of ONOO

by
antioxidants based on assays involving tyrosine nitra-
tion and inactivation of
1
-antiproteinase serves as
useful research tool providing in vitro information on
the antioxidant prole of proposed neuroprotectants
[1,5760]. Methodological considerations on the de-
tection of 3-nitrotysrosine in cardiovascular systems
are reviewed in [6164], the keys issues that may
affect the detection limits for the various methods
(e.g. acidication of Tyr and Tyr-protein contain-
ing biological samples, protein denaturation, basal
plasma levels of nitrotyrosines on the amino acid of
within protein material to be measured, sensitivity to
immunohistochemical methods, and the application
of GC/MS, GC/MS/MS, LC/MS of LC/MS/MS) are
18 O.I. Aruoma / Mutation Research 523524 (2003) 920
applicable across several disease groups where the
level of 3-nitrotyrosine may be considered as a surro-
gate biomarker.
The understanding of the chemistry of free radical
damage to DNA and of the mechanisms of DNA repair
processes is of fundamental importance. The level of
DNA damage at a particular time point will reect the
rate of DNA damage and repair. Oxidative damage to
DNA is important as it not only may represent an early
stage of carcinogensis but could provide a valuable
biomarker of overall oxidative stress. Human neuronal
hybridoma cells (N-18-RE-105) were challenged with
ONOO

. Ergothioneine protected against oxidative

DNA damage in the cells. The levels of products,
4,6-diamino-5-formamidopyrimidine, xanthine, hypo-
xanthine, 5-hydroxyuracil, 2,6-diamino-5-formamido-
pyrimidine, 8-hydroxyguanine, 5-hydroxycytosine
and 8-hydroxyadenine in the treated cells were sig-
nicantly lower than those of untreated cells. The
antioxidant ergothioneine was effective even at con-
centrations lower than that of ONOO

. The fate and

action of ONOO

in biological systems would be

dependent on the biological environment in which
the oxidant is present. Ergothioneine preferentially
scavenged ONOO

[65].
8. Conclusion
The implication of redox mechanisms in the patho-
genesis of human diseases and in the process of aging
has led to the suggestion that antioxidants in particu-
lar, plant diet-derived antioxidants, might have health
benets as prophylactic agents. The mechanism of an-
tioxidant action in vitro may involve direct inhibition
of the generation of reactive oxygen species, or the
scavenging of free radicals. From the foregoing dis-
cussions, it is clear that not a single method can give a
comprehensive prediction of antioxidant efcacy. So
use of more than one method is recommended and
there should be greater caution in extrapolating the in
vitro data to in vivo situations. For in vivo consid-
erations, the question of bioavailability and the fate
of metabolites of the antioxidant components must be
addressed. Thus, if one is interested in the therapeutic
strategies to prevent progressive neuronal loss based
on antioxidant activity, the antioxidant must be able
to cross the blood brain barrier and occur at the re-
spective brain region for neuroprotection. Current re-
search directed towards understanding the role of free
radicals, plant extracts, plant-derived antioxidants in
foods and in nutrition and in human health is being
complemented with the development and validation of
biological markers with which scientists could begin
to delineate the efcacy of dietary antioxidants. First
we have to agree governance on in vitro antioxidant
methods based on an understanding of the mechanisms
involved.
References
[1] O.I. Aruoma, Nutrition and health aspects of free radicals
and antioxidants, Food Chem. Toxicol. 32 (1994) 671683.
[2] P.C.H. Hollman, M.B. Katan, Dietary avonoids: intake,
health effects and bioavailability, Food Chem. Toxicol. 37
(1999) 937942.
[3] K. Hermann, Flavonols and avones in food plants: a review
(1976), J. Food Technol. 11 (1976) 433448.
[4] S. Kuhnau, The avonoids: a class of semi-essential food
components: their role in human nutrition, World Rev. Nutr.
Diet 24 (1976) 117191.
[5] W.A. Pryor, Vitamin E and heart disease: basic science to
clinical intervention trials, Free Radical Biol. Med. 28 (2000)
141164.
[6] Y. Asano, S. Okamura, T. Ogo, T. Eto, T. Otsuka, Y. Niho,
Effect of () epigallocatechin gallate on leukemic blast cells
from patients with acute myeloblastic leukaemia, Life Sci.
60 (1997) 135142.
[7] Effect of genistein on topoisomerase activity and on the
growth of (Val 12) Ha-ras-transformed NIH 3T3 cells,
Biochem. Biophys. Res. Commun. 157 (1988) 183189.
[8] Y.-J. Surh, K.-S. Chun, H.-H. Cha, Y.-S. Keum, K.-K. Park,
S.S. Lee, Molecular mechanisms underlying chemopreventive
activities of anti-inammatory phytochemicals: down-regu-
lation of COX-2 and iNOS through suppression of NF-B
activation, Mutat. Res. 480-481 (2001) 243268.
[9] A.-N.T. Kong, R. Yu, V. Hebbar, C. Chen, E. Owuor, R. Hu,
R. Ee, S. Mandlekar, Signal transduction events elicited by
cancer prevention compounds, Mutat. Res. 480-481 (2001)
231241.
[10] C.F. Skibola, M.T. Smith, Potential health impacts of
excessive avonoid intake, Free Radical Biol. Med. 29 (2000)
375385.
[11] O.I. Aruoma, Free radicals, J. Am. Oil Chem. 75 (1998)
199212.
[12] H.C. Greenspan, O.I. Aruoma, Oxidative stress and apoptosis
in HIV infection: a role for plant-derived metabolites with
synergistic antioxidant activity, Immunol. Today 15 (1994)
209213.
[13] I. Cantuti-Castelvetri, B. Shukitt-Hale, J.A. Joseph, Neuro-
behavioural aspects of antioxidants in aging, Int. J. Dev.
Neurosci. 18 (2000) 367381.
O.I. Aruoma / Mutation Research 523524 (2003) 920 19
[14] W.L. Porter, Paradoxical behaviour of antioxidants in food
and biological systems, Toxicol. Ind. Health 9 (1993) 93122.
[15] E.N. Frankel, Lipid oxidation (1980), Prog. Lipid Res. 19
(1980) 122.
[16] O.I. Aruoma, Assessment of potential pro-oxidant and antio-
xidant actions, J. Am. Oil Chem. Soc. 73 (1996) 16171625.
[17] N.A. Porter, Mechanism for the autoxidation of
polyunsaturated lipids, Acc. Chem. Res. 19 (1986) 262268.
[18] M. Namiki, Antioxidants/antimutagens in foods, Crit. Rev.
Food Sci. Nutr. 29 (1990) 273300.
[19] J. Lliger, The use of antioxidants in food, in: O.I. Aruoma,
B. Halliwell (Eds.), Free Radicals and Food Additives, Taylor
& Francis, London, 1991, pp. 121150.
[20] P.A.E.L. Schildermann, F. Ten-Hoor, J.C.S. Kleinjas,
Induction of oxidative DNA damage and early lesions in
rat gastro-intestinal epithelium in relation to prostaglandin H
synthase-mediated metabolism of butylated hydroxyanisole,
Food Chem. Toxicol. 33 (1995) 99109.
[21] P.C.H. Hollman, J.M.P. van Trijp, M.N.C.P. Buysman, M.S.
van der Gaag, M.J.B. Mengelers, J.H.M. de Vries, M.B.
Katan, Relative bioavailability of the antioxidant quercetin
from various foods in man, FEBS Lett. 418 (1997) 152156.
[22] F. Toms-Barbern, Are dietary polyphenols available for
bioactivity, Polyphenols Actul. 21 (2001) 2526.
[23] D.E. Pratt, Antioxidants indigenous to foods, Toxicol. Ind.
Health 9 (1993) 6375.
[24] H. Domnguez, J. Torres, M.J. Nnez, Antioxidant phenolics
as food additives from agricultural wastes, Polyphenols Actul.
21 (2001) 2732.
[25] F. Shahidi, J.P.K.P.D. Wanasundara, Phenolic antioxidants,
Crit. Rev. Food Sci. Nutr. 32 (1992) 67103.
[26] I. Fridovich, Superoxide radical: an endogenous toxicant,
Ann. Rev. Pharmacol. Toxicol. 23 (1983) 239257.
[27] F.M. Menzies, P.G. Ince, P.J. Shaw, Mitochondrial involve-
ment in amyotrophic lateral slerosis, Neurochem. Int. 40
(2002) 543551.
[28] D.R. Rosen, T. Siddique, T. Patterson, D.A. Figlewicz,
Mutations in Cu/Zn superoxide dismutase gene are associated
with familial amyotrophic lateral sclerosis, Nature 362 (1993)
5962.
[29] J.B. Crow, J.B. Sampson, Y. Zhuang, J.A. Thompson, J.S.
Beckman, Decreased zinc afnity of amyotrophic lateral
sclerosis associated mutants leads to enhanced catalysis of
tyrosine nitration by peroxynitrite, J. Neurochem. 69 (1997)
19361944.
[30] B. Halliwell, J.M.C. Gutteridge, Free Radicals in Biology and
Medicine, third ed., Oxford University Press, Oxford, 1999.
[31] A. Mortensen, Scavenging of benzyperoxyl radicals by
carotenoids, Free Radical Res. 36 (2002) 211216.
[32] D. Fuchs, O.I. Aruoma, Pharmacology of HIV infection and
AIDS, OICA International, London, 2000.
[33] J. Chaudire, R. Ferrari-Iliou, Intracellular antioxidants: from
chemical to biochemical mechanisms, Food Chem. Toxicol.
37 (1999) 949962.
[34] M.L. Sagrista, A.F. Garcia, M.A. De Madariaga, M.
Mora, Antioxidant and pro-oxidant effect of the thiolic
compounds N-acetyl-l-cysteine and glutathione against free
radical-induced lipid peroxidation, Free Radical Res. 36
(2002) 329340.
[35] D.W. Lamson, M.W. Brignal, Antioxidants and cancer III:
quercetin, Altern. Med. Rev. 5 (2000) 196208.
[36] L.M. Milchak, J.D. Bricker, The effects of glutathione and
Vitamin E on iron toxicity in isolated rat hepatocytes, Toxicol.
Lett. 126 (2002) 169177.
[37] T. Yoshikawa, S. Toyokuni, Y. Yamamoto, Y Naito, Free
Radicals in Chemistry and Medicine, OICA International,
London 2000.
[38] E.N. Frankel, In search of better methods to evaluate natural
antioxidants and oxidative stability in food lipids, Trends
Food Sci. Technol. 4 (1993) 220225.
[39] E.N. Frenkel, A.S. Meyer, The problem of using
one-dimensional method to evaluate multifunctional food and
biological antioxidants, J. Agric. Food Chem. 80 (2000) 1925
1941.
[40] E.N. Frankel, S. W Huang, J. Kanner, J.B. German, Interfacial
phenomena in the evaluation of antioxidants: bulk oils vs
emulsions, J. Agric. Food Chem. 42 (1994) 10541059.
[41] A.I. Hopia, S.W. Huang, K. Schwarz, J.B. German, N.N.
Frankel, Effect of different lipid systems on antioxidant
activity of rosemary constituents carnosol and carnosic acid
with and without -tocopherol, J. Agric. Food Chem. 44
(1996) 20302036.
[42] O.I. Aruoma, B. Halliwell, R. Aeschbach, J. Lliger,
Antioxidant and pro-oxidant properties of active rosemary
constituents: carnosol and carnosic acid, Xenobiotica 22
(1992) 257268.
[43] O.I. Aruoma, J.P.E. Spencer, P. Rossi, R. Aeschbach, A.
Khan, N. Mahmood, A. Munoz, A. Murcia, J. Butler, B.
Halliwell, An evaluation of the antioxidant and antiviral action
of extracts of rosemary and Provenal herbs, Food Chem.
Toxicol. 34 (1996) 449456.
[44] B.C. McGorum, S.C. Fry, G. Wallace, K. Coenen, J. Robb,
G. Williamson, O.I. Aruoma, Properties of herbage in
relation to equine dysautonomia: Biochemical composition
and antioxidant and pro-oxidant actions, J. Agric. Food Chem.
48 (2000) 23462352.
[45] P.K. Vayalil, Antioxidant and antimutagenic properties of
aqueous extract of date fruit (Phoenix dactylifera L.
Arecaceae), J. Agric. Food Chem. 50 (2002) 610617.
[46] A.R. Proteggente, A.S. Pannala, G. Paganga, L. van Buren,
E. Wagner, S. Wiseman, F. van de Put, C. Dacombe, C.A.
Rice-Evans, The antioxidant activity of regularly consumed
fruit and vegetables reects their phenolic and Vitamin C
composition, Free Radical Res. 36 (2002) 217233.
[47] K. Schlesier, M. Harwat, V. Bhm, R. Bitsch, Assessment of
antioxidant activity by using different in vitro methods, Free
Radical Res. 36 (2002) 177187.
[48] O.I. Aruoma, B. Halliwell, DNA and Free radicals.
Techniques, principles and applications, OICA International,
London, 1998.
[49] O.I. Aruoma, B. Halliwell, E. Gajewski, M. Dizdaroglu,
Damage to the bases in DNA induced by hydrogen peroxide
and ferric iron chelates, DNA. J. Biol. Chem. 264 (1989)
2050920512.
[50] Y. Ebina, S. Okada, S. Hamazaki, F. Ogino, J.-L. Li, O.
Midorikawa, Nephrotoxicity and renal cell carcinoma after
20 O.I. Aruoma / Mutation Research 523524 (2003) 920
use of iron and aluminum-nitrilotriacetate complexes in rats,
J. Natl. Cancer Inst. 76 (1986) 107113.
[51] K. Ikeda, F. Sun, K. Tanaka, S. Tokumaru, S. Kojo, Increase
of lipid hydroperoxides in the rat liver and kidney after
administering ferric nitrilotriacetate, Biosci., Biotechnol.,
Biochem. 62 (1998) 14381439.
[52] S. Toyokuni, Iron induced carcinogenesis: the role of redox
regulation, Free Radical Biol. Med. 20 (1996) 553566.
[53] M. Deiana, O.I. Aruoma, A. Rosa, V. Crobu, V. Casu, R.
Piga, M.A. Dess`, The effect of ferric-nitrilotriacetic acid on
the prole of polyunsaturated fatty acids in the kidney and
liver of rats, Toxicol. Lett 123 (2001) 123125.
[54] R.E. Huie, S. Padmaja, The reaction of NO with superoxide,
Free Radical Res. Commun. 18 (1993) 195199.
[55] J.S. Beckman, Oxidative damage and tyrosine nitration from
peroxynitrite, Chem. Res. Toxicol. 9 (1996) 836844.
[56] J. Lancaster, The Biological Chemistry of Nitric Oxide,
Academic Press, New York, 1995.
[57] O.I. Aruoma, M. Whiteman, T.G. England, B. Halliwell,
Antioxidant action of ergothioneine: assessment of its ability
to scavenge peroxynitrite, Biochem. Biophys. Res. Commun.
231 (1997) 389391.
[58] M. Whiteman, H. Kaur, B. Halliwell, Protection against
peroxynitrite dependent tyrosine nitration and
1
-antipro-
teinase inactivation by some anti-inammatory drugs and by
the antibiotic tetracycline, Ann. Rheu. Dis. 55 (1996) 383
387.
[59] M.S. Ramezanian, S. Padmaja, W.H. Koppenol, Nitration and
hydroxylation of phenolic compounds by peroxynitrite, Chem.
Res. Toxicol. 9 (1996) 232240.
[60] M. Deiana, O.I. Aruoma, P. Bianchi, J.P.E. Spencer, H.
Kaur, B. Halliwell, R. Aeschbach, S. Banni, M.A. Dessi, F.P.
Corongiu, Inhibition of peroxynitrite dependent DNA base
modication and tyrosine nitration by the extra virgin olive
oil-derived antioxidant hydroxytyrosol, Free Radical Biol.
Med. 26 (1999) 762769.
[61] I. Fridovich, M.M. Tarpey, Methods of detection of vascular
reactive species. Nitric oxide, superoxide, hydrogen peroxide
and peroxynitrite, Circ. Res. 89 (2001) 224236.
[62] D. Tsikas, E. Schwedhelm, J.C. Frlich, Methodological
considerations on the detection of 3-nitrotyrosine in the
cardiovascular system, Circ. Res. 90 (2002) (Letter to the
Editor).
[63] E. Schwedhelm, D. Tsikas, F.-M. Gutzki, J.C. Frlich, Gas
chromatographic-tandem mass spectrometric quantication of
free 3-nitrotyrosine in human plasma at the basal state, Anal.
Biochem. 276 (1999) 195203.
[64] M.T. Frost, B. Halliwell, K.P. Moore, Analysis of free
and protein-bound nitrotyrosine in human plasma by a
gas chromatography/mass spectrometry method that avoids
nitration artefacts, Biochem. J. 345 (2000) 453458.
[65] O.I. Aruoma, J.P.E. Spencer, N. Mahmood, Protection against
oxidative damage and cell death by the natural antioxidant
ergothioneine, Food Chem. Toxicol. 37 (1999) 10431053.
[66] O.I. Aruoma, et al., Free Radical Res. 22 (1995) 187190.
[67] T.J. Hill, et al., J. Am. Chem. Soc. 117 (1995) 83228326.
[68] T. Akaike, et al., J. Agric. Food Chem. 43 (1995) 18641870.
[69] B. Halliwell, et al., Anal. Biochem. 165 (1987) 215219.
[70] O.I. Aruoma, et al., Methods Enzymol. 233 (1994) 5766.
[71] G. Cao, et al., Free Radical Biol. Med. 14 (1993) 303311.
[72] B. Ou, et al., J. Agric. Food Chem. 49 (2001) 46194626.
[73] D. Huang, et al., J. Agric. Food Chem. 50 (2002) 18151821.
[74] I. Popov, G. Lewin, Meth. Enzymol. 300 (1999) 437456.
[75] N.J. Miller, et al., Clin. Sci. 84 (1993) 407412.
[76] R. van den Berg, et al., Food Chem. 66 (1999) 511517.
[77] N.J. Miller, et al., FEBS Lett. 384 (1996) 240242.
[78] V. Bohm, et al., J. Agric. Food Chem. 50 (2002) 221226.
[79] O.I. Aruoma, et al., Free Radical Res. 30 (1999) 419427.
[80] H. Sakakibara, et al., Free Radical Res. 36 (2002) 307316.
[81] R. Re, et al., Free Radical Biol. Med. 26 (1999) 12311237.
[82] K. Miura, et al., J. Agric. Food Chem. 50 (2002) 18451851.
[83] G.W. Winston, et al., Free Radical Biol. Med. 24 (1998)
480493.
[84] P.K. Witting, et al., J. Lipid Res. 37 (1997) 853867.
[85] W. Brand-Williams, et al., Lebensm.-Wiss. Technol. 28 (1995)
2530.
[86] V. Fogliano, et al., 47 (1999) 10351040.
[87] D.D.M. Wayner, et al., FEBS Lett. 187 (1985) 3337.
[88] A. Ghiselli, et al., Free Radical Biol. Med. 18 (1995) 29
36.
[89] E. Lissi, et al., Free Radical Biol. Med. 18 (1995) 153
158.
[90] I.F.F. Benzie, J.J. Strain, et al., Anal. Biochem. 239 (1996)
7076.
[91] A. Bub, et al., J. Nutr. 130 (2000) 22002206.
[92] R. Pulido, et al., J. Agric. Food. Chem. 48 (2000) 33963402.