Polyhydroxylated 2-styrylchromones as potent antioxidants

Article abstract of DOI:10.1016/j.bcp.2004.02.030

Four polyhydroxylated 2-styrylchromones, structurally related to flavones and cinnamic acid, have been studied. An SC derivative with OH groups only at positions 3′ and 4′ on the styryl moiety and another SC bearing an additional OH group at position 5 on

Full text of DOI:10.1016/j.bcp.2004.02.030

Biochemical Pharmacology 67 (2004) 2207–2218
Polyhydroxylated 2-styrylchromones as potent antioxidants
a,b
Paulo Filipe , Artur M.S. Silva , Patrice Morliere , Cristela M. Brito ,
c
b
c
`
Larry K. Patterson , Gordon L. Hug , Joa˜o N. Silva , Jose A.S. Cavaleiro ,
d,e
d
a,b
c
´
Jean-Claude Maziere , Joa˜o P. Freitas , Rene Santus
e
a
b,*
`
´
aCentro de Metabolismo e Endocrinologia da Faculdade de Medicina de Lisboa, Hospital de Santa Maria, 1699 Lisbon, Portugal
bINSERM U 532, Institut de Recherche sur la Peau, 1 Avenue Claude Vellefaux, 75475 Paris, Cedex 10, France
c
´
Department of Chemistry, University of Aveiro, Campus Universitario de Santiago, 3810-193 Aveiro, Portugal
dRadiation Laboratory, University of Notre Dame, Notre Dame, IN 46556, USA
e
ˆ
Service de Biochimie, Hopital Nord d’Amiens, 80054 Amiens, Cedex 01, France
Received 2 December 2003; accepted 12 February 2004
Abstract
Four polyhydroxylated 2-styrylchromones, structurally related to flavones and cinnamic acid, have been studied. An SC derivative with
OH groups only at positions 30 and 40 on the styryl moiety and another SC bearing an additional OH group at position 5 on the benzopyrone
ring were more potent inhibitors of the Cu2þ-induced peroxidation of LDL than the flavonoid quercetin. Fluorescence and absorption
spectroscopies suggested that one LDL particle may bind 40 SC molecules. A pulse radiolysis study in pH 7 buffered micellar solutions of
ꢀ
ꢀ
À
neutral TX100 and positively charged CTAB demonstrated that one-electron oxidation by Br2À, O2 and tryptophan radicals (ꢀTrp)
depends strongly on the micellar microenvironment. All SCs were readily oxidized by ꢀO2 in CTAB micelles À(rate constants:
À
6–18 Â 106 MÀ1 sÀ1). In TX100 micelles only the SC derivative with OH groups in position 30 and 40 reacted with ꢀO2 (rate constant:
1:1 Â 106 MÀ1 sÀ1). In CTAB, electron transfer to Trp radicals was observed for all SCs with rate constants ꢁ3:2 Â 107 MÀ1 sÀ1
In TX100 micelles, this reaction occurred solely with the derivative bearing OH groups only at positions 30 and 40.
# 2004 Elsevier Inc. All rights reserved.
.
ꢀ
Keywords: LDL oxidation; Fast kinetics; Micelles; Pulse radiolysis; Superoxide radical; Tryptophan; Redox reactions
1. Introduction
to good health and are found in large quantities in fruits and
vegetables [5,6]. Literature data suggest that antioxidants
often work by the combination of several mechanisms
such as scavenging of lipid alkoxyl and peroxyl radicals
through hydrogen donation, regeneration of a-tocopherol
by a-tocopheroxyl radical reduction and chelation of
transition metal ions [7]. More recently, the repair of
oxidative damage to proteins or lipoproteins has been
shown to also contribute to the flavonoid antioxidant
activity [8,9].
In practice, the protection afforded by antioxidants
against oxidative stress and its consequences to human
health are limited. Thus, antioxidants may exhibit pro-
oxidant properties under appropriate conditions. Another
important factor that limits the efficacy of antioxidants is
their bioavailability [10,11]. Consequently, it is of interest
to search out new molecules with improved antioxidant
properties to help circumvent some limitations of presently
available antioxidants.
Reactive oxygen species (ROS) such as ꢀO2À, ꢀOH, and
H2O2 generated by cellular metabolism or arising from
environmental sources (chemicals, UV light) are recog-
nized as factors contributing to major diseases such as
atherosclerosis [1], cancer [2], and central nervous system
degeneration [3] and are implicated in the general aging
process [4]. In addition to the constitutive cell antioxidant
defense system, there exist various small molecules,
vitamins (ascorbic acid and a-tocopherol), carotenoids,
flavonoids and simpler phenolic coumponds, which exhibit
potent antioxidant properties and which are excellent
reactive oxygen species scavengers. These are essential
Abbreviations: CTAB, cetyltrimethylammonium bromide; LDL,
low-density lipoprotein; ROS, reactive oxygen species; Q, quercetin; SC,
2-styrylchromone; TX100, Triton X100
* Corresponding author. Tel.: þ33-1-40793726; fax: þ33-1-40793716.
E-mail address: santus@mnhn.fr (R. Santus).
0006-2952/$ – see front matter # 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.bcp.2004.02.030

2208
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
The present study concerns polyhydroxylated SCs.
readily monitored [22]. Since oxidized Trp residues in
proteins can be repaired by antioxidants such as ascorbate
[23] and flavonoids [8,9], repair of ꢀTrp was also chosen as
a probe of the antioxidant capacity of the SCs studied.
These compounds are structurally related to cinnamic acids
and flavones, i.e. two families of molecules which exhibit
potent antioxidant properties. To date, only two natural
SCs are known [12,13]. However, synthetic molecules of
this class of compounds exhibit antiallergic, antitumor,
and anticancer properties [12–15]. More specifically,
some polyhydroxylated SCs are potent inhibitors of
xanthine oxidase [16] and potent hepatoprotectors against
tert-butylhydroperoxide [17]. Little is known about the
biochemical mechanisms responsible for their biological
effects.
This study, is therefore, intended to explore possible
antioxidant properties of four synthetic SCs substituted
with an increasing number of hydroxyl groups on both the
chromone and styryl moieties. To this end, we have studied
their inhibitory effect on Cu2þ-induced oxidation of iso-
lated human serum low-density lipoproteins (LDLs), a
well-established in vitro model of lipid peroxidation via
radical chain reaction [18]. The formation of conjugated
dienes and the consumption of carotenoids were chosen as
markers of LDL lipid peroxidation [19]. Their inhibition
by polyhydroxylated SCs was compared to that observed in
the presence of the flavone quercetin since the latter is
known to be an effective inhibitor of the LDL lipid
peroxidation [10].
2. Materials and methods
2.1. Chemicals and routine equipment
All routine chemicals were of analytical grade and were
used as received from the suppliers. Quercetin, D,L-trypto-
phan, CTAB, and TX100 were purchased from Sigma.
Dimethyl sulfoxide, absolute ethanol, and methanol were
supplied by Merck and were of spectroscopic grade. The
phosphate buffer (pH 7) was prepared in pure water
obtained with a reverse osmosis system from Ser-A-Pure
Co. This water exhibits a resistivity of >18 MO cmÀ1 and a
total organic content of <10 ppb. Solutions were saturated
with pure N2O or O2. Absorption spectrophotometry was
carried out with an Uvikon 943 spectrophotometer whereas
fluorescence spectra were recorded with a Shimadzu
RF5000 spectrofluorometer.
The polyhydroxylated 2-styrylchromones SC2–SC4
(Scheme 1) were synthesized as described in [24]. 30,40-
Dihydroxy-2-styrylchromone SC1 was prepared as
detailed below and as summarized in Scheme 2.
This result prompted us to undertake a more detailed
mechanistic study of their antioxidant and redox properties
by means of fast kinetics spectroscopy. This subsequent
investigation was carried out in neutral Triton X100
(TX100) and cationic cetyltrimethylammonium bromide
(CTAB) pH 7 buffered micellar solutions. These two kinds
of micelles are related to two important microenviron-
ments found in the LDLs. These are the hydrophobic
apolar neutral lipid core and the heavily positively charged
apolipoprotein B containing 353 Lys and 152 Arg residues
which play a key role in the recognition of the so-called
Apo-B/E cell receptor [20]. The apolipoprotein B interacts
with the outer layer of phospholipids and cholesterol in
contact with the aqueous phase and with the lipidic core of
the LDLs. In the TX100 and CTAB micelles, the phenoxyl
radicals resulting from the onꢀe-electron oxidation of the
2.2. Synthesis of 30,40-dihydroxy-2-styrylchromone (SC1)
2.2.1. 20-(3,4-Dibenzyloxycinnamoyloxy)acetophenone
(1)
3,4-Dibenzyloxycinnamic acid (6.8 g, 18.9 mmol) and
phosphorus oxychloride (8.0 ml, 86 mmol) were added
to a solution of the 20-hydroxyacetophenone (2.0 ml,
16.6 mmol) in dry pyridine (300 ml). The solution was
stirred at 60 8C for 4 h. Subsequently, the solution was
poured into ice and water (200 ml), and the pH adjusted to
4 with hydrochloric acid. The mixture was then extracted
with chloroform (2 Â 100 ml) and the organic layer washed
with water. The solvent was evaporated and the residue was
purified by silica gel column chromatography using a (3:7)
À
SCs, by the mildly oxidizing Br2 radical anions, were
characterized spectroscopically by pulse radiolysis. Using
the spectral data obtained, we demonstrated that polyhy-
ꢀ
À
droxylated SCs also react very effectively with the O2
radical anion in these aqueous micellar systems.
Amino acid residues of proteins are also important
biological targets of oxidative damage by ROS [21].
The neutral indolyl radical (ꢀTrp) is the primary transient
species produced by oxidation of the aromatic amino acid
Trp, which is among the residues most energetically
vulnerable to radical attack. ꢀTrp can be produced by
ꢀ
strongly oxidizing OH radicalsꢀ or, more seleꢀctively, by
ꢀ
milder oxidants such as Br2À, ðSCNÞ2À, or N3. Since
ꢀTrp absorbs in the UV–Vis, its kinetic behavior can be
Scheme 1.

P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
2209
OBn
dichloromethane (20 ml) and purified by silica gel column
chromatography using dichloromethane as eluent. This
solvent was evaporated to dryness and the residue was
then crystallized from ethanol giving 1-(2-hydroxyphenyl)-
5-(3,4-dibenzyloxyphenyl)-3-hydroxy-2,4-pentadiene-1-
one (2) (2.0 g, yield: 84%). m.p. 133–136 8C (recrystallized
from ethanol). 1H NMR: d ¼ 5:21 (s, 2H, 300-OCH2C6H5),
5.22 (s, 2H, 400-OCH2C6H5), 6.28 (s, 1H, H-2), 6.40 (d, J
15.7 Hz, 1H, H-4), 6.90 (ddd, J 8.5, 7.6 and 1.0 Hz, 1H,
H-50), 6.94 (d, J 8.4 Hz, 1H, H-500), 6.99 (dd, J 8.4 and
1.0 Hz, 1H, H-30), 7.12 (dd, J 8.4 and 1.9 Hz, 1H, H-600),
7.16 (d, J 1.9 Hz, 1H, H-200), 7.30–7.49 (m, 11H, H-40 and
H-2,3,4,5,6 of 300,400-OCH2C6H5), 7.55 (d, J 15.7 Hz, 1H,
H-5), 7.69 (dd, J 8.5 and 1.5 Hz, 1H, H-60), 12.27 (s, 1H, 20-
OH), 14.72 (s, 1H, 3-OH). 13C NMR: d ¼ 70:9 (400-
OCH2C6H5), 71.4 (300-OCH2C6H5), 96.5 (C-2), 113.7
(C-200), 114.3 (C-500), 118.7 (C-30), 119.0 (C-50), 119.1
(C-10), 120.1 (C-4), 122.9 (C-600), 127.2, 127.3, 127.5,
128.0 and 128.4 (C-2,3,4,5,6 of 300,400-OCH2C6H5), 128.4
(C-60), 135.7 (C-40), 136.7 and 136.9 (C-100 and C-1 of 300,400-
OCH2C6H5), 139.8 (C-5), 149.0 (C-300), 151.0 (C-400), 162,5
(C-20), 174.9 (C-3), 195.6 (C-1). EI-MS: m/z (%) ¼ 478 (8)
(Mꢀþ), 387(3), 369(2), 181(4), 163(3), 121(18), 92(11),
91(100), 77(2), 65(10), 63(2), 51(2). C31H26O5 (478.5):
Calc. C, 77.81; H, 5.48. Found C, 77.94; H, 5.52.
O
OBn
CO2H
BnO
O
+
A
OH
O
BnO
O
1
OBn
OBn
B
5’
5"
2"
OBn
6"
O
4’
3’
6’
4"
C
4
2
5
3
1
3"
OBn
2’
OH
O
HO
O
3
2
D
5’
4’ OH
6’
3’
β
OH
2’
8
5
A: POCl3, dry pyridine
B: KOH, DMSO
9
O
4
α
7
6
2
C: p-Toluenesulfonic acid, DMSO
D: HBr/Acetic acid
3
10
O
SC1
Scheme 2.
mixture of light petroleum:dichloromethane as eluent.
This solvent was then evaporated to dryness and the residue
was crystallized from ethanol to give 20-(3,4-dibenzylox-
ycinnamoyloxy)acetophenone (1) (4.6 g, yield: 58%). m.p.
108–109 8C (recrystallized from ethanol). 1H NMR:
d ¼ 2:56 (s, 3H, 2-CH3), 5.21 (s, 2H, 300-OCH2C6H5),
5.23 (s, 2H, 400-OCH2C6H5), 6.46 (d, J 15.8 Hz, 1H, H-
a), 6.94 (d, J 8.4 Hz, 1H, H-500), 7.14 (dd, J 8.4 and 2.0 Hz,
1H, H-600), 7.18 (dd, J 7.4 and 1.5 Hz, 1H, H-30), 7.19 (d, J
2.0 Hz, 1H, H-200), 7.26–7.49 (m, 11H, H-50 and H-2,3,4,5,6
of 300,400-OCH2C6H5), 7.56 (dt, J 7.4 and 1.6 Hz, 1H,
H-40), 7.78 (d, J 15.8 Hz, 1H, H-b), 7.83 (dd, J 7.7 and
1.6 Hz, 1H, H-60). 13C NMR: d ¼ 29:9 (C-2), 70.9 (400-
OCH2C6H5), 71.3 (300-OCH2C6H5), 113.8 (C-200), 114.1
(C-500), 114.5 (C-a), 123.5 (C-600), 123.8 (C-30), 126.0
(C-50), 127.1, 127.3, 127.4, and 128.0 (C-2,3,4,5,6 of
300,400-OCH2C6H5), 130.1 (C-60), 131.4 (C-10), 133.3 (C-
40), 136.6 and 136.8 (C-100 and C-1 of 300,400-OCH2C6H5),
147.2 (C-b), 148.9 (C-20), 149.2 (C-300), 151.5 (C-400),
165.4 (C=O), 197.9 (C-1). EI-MS: m/z (%) ¼ 478(7)
(Mꢀþ), 344(9), 343(34), 253(5), 251(3), 181(8), 163(2),
136(3), 121(6), 105(2), 92(11), 91(100), 77(4), 65(9),
63(3), 51(3). C31H26O5 (478.5): Calc. C, 77.81; H,
5.48. Found C, 77.81; H, 5.40.
2.2.3. 30,40-Dibenzyloxy-2-styrylchromone (3)
p-Toluenesulfonic acid monohydrate (0.38 g, 2.0 mmol)
was added to a solution of 1-(2-hydroxyphenyl)-5-(3,4-
dibenzyloxyphenyl)-3-hydroxy-2,4-pentadiene-1-one (2)
(1.9 g, 4 mmol) in dimethyl sulfoxide (30 ml). This
solution was heated, under nitrogen, at 90 8C, until the
complete consumption of the starting material. After that
period, the solution was poured into ice and water (100 ml)
and the solid obtained was removed by filtration. This solid
was dissolved in dichloromethane (50 ml) and the organic
layer washed with water (2 Â 100 ml). After evaporation of
the solvent to dryness, the residue was crystallized from
ethanol giving 30,40-dibenzyloxy-2-styrylchromone (3)
(1.6 g, yield: 85%). m.p. 161–163 8C (recrystallized from
1
ethanol). H NMR: d ¼ 5:22 (s, 2H, 40-OCH2C6H5), 5.23
(s, 2H, 30-OCH2C6H5), 6.29 (s, 1H, H-3), 6.58 (d, J
16.0 Hz, 1H, H-a), 6.96 (d, J 8.4 Hz, 1H, H-50), 7.14
(dd, J 8.4 and 2.0 Hz, 1H, H-60), 7.19 (d, J 2.0 Hz,
1H, H-20), 7.33–7.43 (m, 10H, H-2,3,4,5,6 of 30,40-
OCH2C6H5), 7.48 (dd, J 8.3 and 7.6 Hz, 1H, H-6), 7.51
(d, J 16.0 Hz, 1H, H-b), 7.52 (d, J 8.2 Hz, 1H, H-8), 7.68
(ddd, J 8.2, 7.6 and 1.6 Hz, 1H, H-7), 8.20 (dd, J 8.3 and
1.6 Hz, 1H, H-5). 13C NMR: d ¼ 71:0 (40-OCH2C6H5),
71.4 (30-OCH2C6H5), 110.1 (C-3), 113.4 (C-20), 114.4 (C-
50), 117.8 (C-8), 118.3 (C-a), 122.4 (C-60), 124.1 (C-10),
124.9 (C-6), 125.6 (C-5), 127.2, 127.3, 128.0 and 128.5 (C-
2,3,4,5,6 of 30,40-OCH2C6H5), 133.6 (C-7), 136.7 (C-b),
136.9 (C-10), 136.9 (C-1 of 30,40-OCH2C6H5), 149.1 (C-30),
150.7 (C-40), 156.0 (C-9), 162.0 (C-2), 178.4 (C-4). EI-MS:
m/z (%) ¼ 460 (22) (Mꢀþ), 370(11), 369(23), 352(4),
2.2.2. 1-(2-Hydroxyphenyl)-5-(3,4-dibenzyloxyphenyl)-
3-hydroxy-2,4-pentadiene-1-one (2)
Potassium hydroxide (powder, 1.0 g, 25 mmol) was
added to a solution of 20-(3,4-dibenzyloxycinnamoylox-
yacetophenone (1) (2.4 g, 5 mmol) in dimethyl sulfoxide
(30 ml). The solution was stirred, under nitrogen, at room
temperature for 2 h. Following that period the solution
was poured into ice and water (200 ml), and the pH
was adjusted to 4 with diluted hydrochloric acid. The
solid obtained was removed by filtration, dissolved in

2210
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
221(2), 181(3), 121(2), 92(14), 91(100), 65(11), 63(3),
51(2). C31H24O4 (460.5): Calc. C, 80.85; H, 5.25. Found
C, 80.84; H, 5.26.
2.4. Conjugated diene determination
Conjugated diene formation was monitored by second
derivative absorption spectroscopy (220–300 nm) based on
an earlier methodology described in the literature [27].
With protein solutions, second derivative absorption spec-
troscopy is generally preferred to direct absorption spectro-
scopy because it is much less sensitive to scattered light.
Before recording spectra, 80 ml of the sample were diluted
10-fold with 10 mM phosphate buffer (pH 7.4). The second
derivative spectrum was subtracted from the second deri-
vative spectrum of the matching control sample without
Cu2þ. The increase in conjugated dienes expressed in
relative units was obtained from the amplitude of the
254 nm peak.
2.2.4. 30,40-Dihydroxy-2-styrylchromone (SC1)
30,40-Dibenzyloxy-2-styrylchromone (3) (1.2 g, 2.55
mmol) was added to a solution of hydrogen bromide in
acetic acid (33%, 50 ml). The mixture was refluxed for 2 h.
The solution was then carefully poured into ice and water
and the solid obtained was removed by filtration. The solid
was dissolved in ethyl acetate (20 ml) and washed with
water several times. After evaporation of the solvent, this
residue was purified by silica gel column chromatography
using a (8:2) mixture of chloroform:acetone as eluent. The
solvent was evaporated to dryness, and the residue was
precipitated from ethanol with water giving 30,40-dihy-
droxy-2-styrylchromone (SC1) (286 mg, yield: 40%).
m.p. 176–177 8C.1H NMR (DMSO-d6): d ¼ 6:42 (s, 1H,
H-3), 6.79 (d, J 8.2 Hz, 1H, H-50), 6.88 (d, J 16.0 Hz, 1H,
H-a), 7.03 (dd, J 8.2 and 2.0 Hz, 1H, H-60), 7.11 (d, J
2.0 Hz, 1H, H-20), 7.46 (ddd, J 7.7, 7.3 and 0.8 Hz, 1H, H-
6), 7.55 (d, J 16.0 Hz, 1H, H-b), 7.71 (d, J 8.2 Hz, 1H, H-
8), 7.80 (ddd, J 8.2, 7.3 and 1.6 Hz, 1H, H-7), 8.00 (dd, J
7.7 and 1.6 Hz, 1H, H-5) 13C NMR (DMSO-d6): d ¼ 109:0
(C-3), 114.3 (C-20), 115.9 (C-50), 116.7 (C-a), 118.2 (C-8),
120.9 (C-60), 123.5 (C-10), 124.8 (C-5), 125.1 (C-6),
126.6 (C-10), 134.1 (C-7), 137.2 (C-b), 145.7 (C-30),
148.0 (C-40), 155.5 (C-9), 162.4 (C-2), 176.9 (C-4). EI-
MS: m/z (%) ¼ 280 (100) (Mꢀþ), 263(29), 251(5), 234(10),
233(11), 223(8), 205(9), 187(13), 160(29), 131(7), 121(47),
114(16), 103(9), 92(13), 84(17), 77(9), 76(9), 66(20),
63(12), 51(7). C17H12O4 (280.3): Calc. C; 72.85; H,
4.32. Found C, 72.82; H 4.33.
2.5. Consumption of carotenoids
The basal carotenoid content of LDL preparations was
spectrophotometrically determined [28]. The concentra-
tion of total carotenoids can be determined using an
average extinction coefficient of 140,000 MÀ1 cmÀ1 at
448 nm. This value is based on a calculation of contribu-
tions from the four main carotenes in human plasma,
b-carotene, b-carotene, b-cryptoxanthin, and lycopene
[29]. Changes in carotenoid concentration during LDL
oxidation were monitored by second derivative absorption
spectroscopy (400–550 nm) through measurement of the
second derivative spectrum amplitude between 489 and
516 nm.
2.6. Pulse radiolysis
Pulse radiolysis measurements were carried out with the
Notre Dame Radiation Laboratory 8-MeV linear accel-
erator, which provides 5 ns pulses of up to 30 Gy. In
general, the doses used in this work were approximately
2–20 Gy. The principles of the detection system have been
previously described [30,31]. A Corning O-51 optical filter,
which removes all wavelengths shorter than 350 nm, was
placed in the analyzing light beam preceding the sample
cell.
2.3. Preparation and treatment of LDL
Serum samples were obtained from healthy volunteers.
The LDLs (d ¼ 1:024À1:050 g mlÀ1) were prepared by
sequential ultracentrifugation according to Havel et al. [25].
Protein content was determined by the technique of Peter-
son [26]. The LDL samples were used within 2–3 weeks.
Just before experimentation, the LDLs were dialyzed twice
for 8 and 16 h against 1 l of 10 mM phosphate buffer (pH
7.4) to remove EDTA. Then, the LDLs were diluted to a
final concentration of 0.15 mg mlÀ1 (300 nM). To obtain
the desired concentration, the appropriate volume of
0.5 mM 2-styrylchromone (SC) or quercetin (Q) stock
solutions in ethanol as well as 150 ml of buffer were added
800 ml of the diluted LDLs. Blank LDL solutions without
SC or Q were also prepared. The LDL solutions loaded with
SC or Q and the blank LDL solutions were then incubated at
37 8C for 15 min. Lipid peroxidation was triggered by
adding 50 ml of CuCl2 to achieve a Cu2þ concentration
of 5 mM. After Cu2þ addition, conjugated diene formation
was measured periodically during continuous incubation
at 37 8C.
Radical concentrations calculated froÀm transient
ꢀ
absorption data are referenced to ðSCNÞ2 dosimetry.
ꢀ
À
The extinction coefficient for ðSCNÞ2 is taken to be
7580 Æ 60 MÀ1 cmÀ1 at 472 nm, and the G value for ꢀOH
in N2O-saturated solution has been measured as
6:13 Æ 0:09 [32]. The G value or radiolytic yield is the
number of radicals generated per 100 eV of absorbed
energy. Such numbers may be recast as radical concentra-
tions per unit radiation (e.g. a G value of 6.1 corresponds to
a concentration of ꢂ0.63 mM GyÀ1).
Solutions for pulse radiolysis were prepared in 10 mM
phosphate buffer solutions (pH 7) of CTAB and TX100.
The concentrations used were 10 and 3 mM, respectively,
well above the critical micellar concentration for each

P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
2211
surfactant involved. This allows one to assume that most
No Cu
Cu
Cu + Q
Cu + SC1
Cu + SC2
Cu + SC3
Cu + SC4
detergent is in micellar form. All solutions also contained
0.1 M BrÀ unless otherwise stated. Triton X100 could be
used in this work since it was preÀviously shown by one of
us that it does not react with ꢀBr2 radical anions [33]. To
minimize the volumes of SC solutions required, a microcell
(optical path: 1 cm, volume: 120 ml and 2 mm i.d. Teflon1
tubing) was used for transient recording. Where necessary,
differences in radical G value, arising from the dependence
of spur scavenging on BrÀ concentration, are taken into
account [34]. Additionally, production of superoxide radi-
cal by oxygen scꢀavenging of Hꢀ is included in calculating
the G value for O2À. Numerical integrations, applied to
the analysis of rate data, were carried out using the
Scientist software from Micromath Scientific Software.
0.04
0.03
0.02
0.01
0.00
0
50
100
150
Time (min)
(A)
0.04
0.03
0.02
0.01
0.00
3. Results and discussion
3.1. Inhibition of conjugated diene formation by SCs
during Cu2þ-induced LDL lipid peroxidation
It is common knowledge that LDLs are natural carriers
of important antioxidants such as Vitamin E and carote-
noids. Because antioxidants compete with LDL lipids for
the propagation of the radical chain reactions, a lag time,
due to the antioxidant consumption, is observed between
the start of the oxidation by Cu2þ ions and the appearance
of lipid peroxidation products. The duration of this induc-
tion period depends on the constitutive antioxidant content
of the LDLs which may vary among blood donors [35].
Fig. 1A shows the time course of the conjugated diene
formation after addition of Cu2þ ions to native LDL and
after loading the LDL with 3 mM SC or Q. It may be seen
that under these concentration conditions, SC1, SC2, and Q
totally inhibit the formation of conjugated dienes for a
period of at least 3 h of incubation at 37 8C. On the other
hand, SC4 only partially retards the propagation phase, and
inclusion of SC3 exhibits only a minimal effect on the
kinetics. Thus, as reflected in Cu2þ-induced LDL oxida-
tion, only SC1 and SC2 can be considered as antioxidants
competitive with Q, one of the most potent antioxidant
polyphenols [7,10,36].
To further assess the antioxidant potential of SC1 and
SC2 in comparison to that of Q, the concentration threshold
was determined at which SC1, SC2, and Q can totally
inhibit conjugated diene formation or carotenoid consump-
tion over a standard 3 h period. Fig. 1B demonstrates that
SC1 at a concentration of 0.75 mM is the most powerful
antioxidant in this model system, whereas SC2 by itself is
more potent than Q. A similar order of effectiveness was
observed for the inhibition of the carotenoid consumption
by the SC derivatives (Fig. 1C).
0
50
100
150
(B)
Time (min)
100
80
60
40
20
0
0
50
100
150
(C)
Time (min)
Fig. 1. (A) Kinetics of conjugated diene formation during LDL oxidation
induced by 5 mM Cu2þ. LDL solutions at 0.12 mg mlÀ1 (240 nM) in
10 mM phosphate buffer (pH 7) were incubated for 15 min at 37 8C with
3 mM SC or Q before Cu2þ addition. Note that time zero denotes
measurements made at about 1 min after Cu2þ addition. Data are the
mean Æ S:D: of three independent experiments. (B) Same as (A) but the
concentrations of SC and Q were 0.75 mM. (C) Kinetics of carotenoid
consumption during LDL oxidation induced by 5 mM Cu2þ. LDL solutions
at 0.12 mg mlÀ1 (240 nM) in 10 mM phosphate buffer (pH 7) were
incubated for 15 min at 37 8C with 1.5 mM SC or Q before Cu2þ addition.
Data are the mean Æ S:D: of three independent experiments.
regarding the antioxidant capacity of the SCs investigated.
The SC1 derivative with only two OH groups on the styryl
moiety is the most effective antioxidant in Cu2þ-induced
LDL oxidation. Addition of another hydroxyl at position 5
on the benzopyrone ring (SC2 derivative; Scheme 1) leads
to less efficient antioxidant behavior. Alternatively, when
It should be noted that the present experimental system
defines an overall apparent antioxidant capacity. However,
these data suggest several structure–activity relationships

2212
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
there is a single OH group on the benzopyrone ring at
position 7, there is a marked decrease in the antioxidant
capacity (SC3 derivative). Finally, the behavior of SC4,
with two OH substitutions on the benzopyrone ring at
positions 5 and 7, is comparable to that of SC3 in anti-
oxidant capacity. Thus, in these systems, increasing the
number of hydroxyl groups seems to decrease the antiox-
idant potential of the SC derivatives. Interestingly, SC4, a
rather inactive molecule in this system, has been found to
be the most effective xanthine oxidase inhibitor with a 50%
inhibition concentration as low as 0.5 mM [16].
SC1
LDL
LDL + SC1
0.4
0.3
0.2
0.1
0
300
350
400
450
500
550
600
650
600
This large variation observed in antioxidant effective-
ness may be due to different partitioning of the hydro-
phobic SCs into LDL as a function of the number of their
hydroxyl groups present. The interaction of the SCs with
the LDL, was therefore, investigated by absorption and
fluorescence spectroscopies. Absorption spectroscopy sug-
gests that, like Q [9], the hydrophobic SCs do bind to LDL.
Fig. 2A shows that in LDL solutions with the same [LDL]/
[SC] ratio as in Fig. 1A, the maximum for SC1 absorbance
is shifted to 379 nm—a value approaching those measured
in ethanol (380 nm) and methanol (377 nm) (Fig. 2C)—as
compared to the value of 373 nm in buffer. Similar data
were obtained with the other SCs (data not shown).
Fluorescence studies demonstrate that SC1 is the only
fluorescent SC in this study. Fig. 2B displays the fluores-
cence spectra of 2.5 mM SC1 in ethanol (fluorescence
maximum: 520 nm), in pH 7.4, 10 mM phosphate buffer
(fluorescence maximum at 527 nm) and in the presence of
100 nM LDL in buffer. Thus, in changing solvent from
ethanol to buffer, the SC1 fluorescence is strongly
quenched and the maximum is shifted to the red by
7 nm. Addition of 100 nM LDL to the buffer partially
restores the SC1 fluorescence and causes significant
changes to the spectral shape. The fluorescence maximum
is shifted to 475 nm, and a less intense band is observed at
about 520 nm, close to the fluorescence maximum found in
ethanol. These fluorescence data, along with the absorption
spectra, suggest that SC1 does interact with the LDL
particle possibly at two different sites.
(A)
(B)
(C)
Wavelength (nm)
Ethanol
Buffer
LDL
2.5
2
1.5
1
0.5
0
400
450
500
550
600
Wavelength (nm)
TX100, pH 7
CTAB, pH 6
CTAB, pH 7
Methanol
1.5
1
Water, pH 11
0.5
0
300
350
400
450
500
550
Wavelength (nm)
Fig. 2. (A) Absorbance spectra of solutions containing 10 mM SC1 (SC1),
800 nM LDL (LDL), or 800 nM LDL þ 10 mM SC1 (LDL þ SC1) in
10 mM phosphate buffer (pH 7). Spectra were recorded at 20 8C in 1 cm
light-path quartz cells. (B) Fluorescence spectra of solutions of 2.5 mM
SC1 in ethanol (ethanol), in 10 mM phosphate buffer, pH 7.4 (buffer) or in
the presence of 100 nM LDL in buffer (LDL). Spectra were recorded at
20 8C in 1 cm light-path quartz cells. The excitation wavelength was
375 nm in all cases. Sensitivity has been multiplied by 10 for spectra
recorded in buffer and LDL. (C) Absorbance spectra of solutions
containing 50 mM SC1 in buffered 1% TX100 containing 10 mM
phosphate buffer (TX100, pH 7), in aqueous 10 mM CTAB (CTAB, pH
6), in buffered 10 mM CTAB containing 10 mM phosphate buffer (CTAB,
pH 7), in 1 mM aqueous NaOH (water, pH 11) or in methanol (methanol).
All spectra were recorded at 25 8C in 1 cm light-path quartz cells.
While the above data point to a possible role for specific
LDL-interaction(s) in the antioxidant effectiveness of the
SCs, it does not mean that the intrinsic antioxidant poten-
tial of these molecules are different. A study of the
molecular redox mechanisms underlying the antioxidant
activity of the SCs may also contribute to an understanding
of the above results. This investigation, presented below,
was carried out by pulse radiolysis.
3.2. One-electron oxidation of SCs by pulse radiolysis
The very low solubility of the SCs (approximately
50 mM) in pH 7 phosphate buffer requires the use of
solubilizing agents for the study of their redox reactions
by pulse radiolysis. Because antioxidants encounter multi-
ple microenvironments under in vivo conditions, micelles
of differing charge provide good primary model systems
for determining environmental effects on radical reactions.
Hence, redox reactions of SCs were investigated in solu-
tions containing positively charged CTAB and neutral

P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
2213
ꢀ
À
TX100 micelles. Furthermore, as indicated in the intro-
duction, they mimic two important microenvironments
related to the biological properties of LDLs. For interpre-
tation of radical behavior discussed below, it is first
necessary to establish the levels of micellar interaction
with SCs.
immediate trapping of Br2 by these positively charged
micelles, the SCs are readily oxidized by the ꢀBr2À radical
anions
ꢀBr2À þ SC ! ꢀSC þ 2 BrÀ
in competition with
(4)
À
ꢀBr2À þ ꢀBr2 ! products
(5)
3.2.1. Influence of bulk pH and of micelle solubilization
on spectra of SCs
In analogy to other phenolic compounds, the hydroxyl
groups are probably the electron donating sites for the one-
electron oxidation of the SCs. Fig. 3A–D show transient
absorbance spectra in the range from 450 to 700 nm
obtained at 2 ms after pulse radiolysis of 100 mM SC in
pH 7 buffered CTAB and TX100 micellar solutions. These
transients all absorb in the 450–700 nm region and all
spectra exhibit well-defined peaks. Their spectral shapes
remain constant over the 0–200 ms time range (data not
shown). The second-order recombination reactions
As suggested by the absorbance measurements carried
out in LDL solutions, see above, the optical spectra of all
these compounds are sensitive to solvent and microenvir-
onmental factors. Furthermore, the presence of hydroxyl
groups leads to pH-dependent absorbance changes. Thus,
Fig. 2C reports the absorbance spectra of 50 mM SC1, the
simplest polyhydroxylated SC under investigation (see
Scheme 1), in methanol, 1% TX100 (pH 7), 10 mM CTAB
in water (pH 6), 10 mM CTAB in pH 7 buffer, and 1 mM
aqueous NaOH. In these solvents, SC1 as well as SC2,
SC3, and SC4 (data not shown) exhibit rather high molar
absorbances (approximately about 20,000 MÀ1 cmÀ1). In
TX100 solution, the SC1 absorbance is similar to that
found in methanol which is consistent with the incorpora-
tion of SC1 into the polar but uncharged interfacial micel-
lar region. The 59 nm red-shift in the absorbance
maximum of SC1 in going from pH 7 (Fig. 2A) to pH
11 (Fig. 2C) reflects the ionization of the hydroxyl groups
at the higher pH. As a result, the pH-dependence of the
complicated transient spectrum of SC1 in CTAB is con-
sistent with the incorporation of SC1 into the interfacial
micellar region where, due to surface electrostatic
charges, the pH is raised by about three pH units as
compared to the pH of the bulk solution [37]. Fig. 2C
shows that the proportion of ionized SC1 molecules
increases 5.8 times when the pH of the bulk is raised
by one unit. It can be estimated that one of the two OH
groups of SC1 has a pKa of about 9.5. However, the
comparison of the spectrum obtained in CTAB with those
obtained in methanol and in alkaline aqueous solution
demonstrates that other factors specific to the microenvir-
onment such as the polarizability and/or partial solvation
in the micellar interior may contribute to the electronic
energy of SC1 in CTAB.
ꢀSC þ ꢀSC ! productðsÞ
(6)
occur with rate constants 2k6 in the range ð1:1À3:9Þ Â
107 MÀ1 sÀ1 in CTAB.
The neutral and ionized forms of the SCs at equilibrium
ꢀ
À
in the CTAB micelles should react differently with Br2
radical anions since phenolates are generally more oxidiz-
able than neutral phenols [38]. As a result, it may be
expected that in CTAB both the neutral and ionized species
contribute to the overall transient spectra. The rate con-
stants for the formation of the ꢀSC radicals (k4) were
measured by following the growth kinetics of these radicals
at the corresponding absorption maxima. These rate con-
stants are all approximately ð6:5À10:6Þ Â 108 MÀ1 sÀ1
.
As can be seen in Fig. 3A–D, the transient spectra,
recorded in neutral TX100 miꢀcelles after the one-electron
oxidation of 100 mM SC by Br2À, are characterized by
much weaker transient absorbances than those of the
corresponding spectra in CTAB. Additionally, the values
of k4 are smaller (k4 ¼ 1:7 Â 107 MÀ1 sÀ1 for SC1). As
indicated in Fig. 2C, the neutral SC species is the one
predominantly found in TX100 micelles. ThisÀmay be
reflected both in lower reactivity toward ꢀBr2 and in
À
weaker absorbance. Under these conditions, ꢀBr2 mainly
disappears via reaction (5).
3.2.2. Transient absorbance spectra of semi-oxidized SC
in micellar media
Pulse radiolysis of a N2O-saturated buffered aqueous
solution containing 0.1 M Br , generates Br2 radical
anions by the sequence of reactions
3.2.3. One-electron oxidation of SCs by the
ꢀO2 radical anion
À
In biological systems, the presence of O2 with the
attendant formation of O2À, via reactions such as (7)
À
ꢀ
À
ꢀ
and (8) below, adds complexity to the radical kinetics.
Antioxidants such as flavonoids are known to react slowly
H2O ! eaqÀ; Hꢀ; OH; H2O2; H3Oþ
eaqÀ þ N2O þ H2O ! ꢀOH þ N2 þ OHÀ
ꢀOH þ 2 BrÀ ! ꢀBr2À þ OHÀ
(1)
(2)
(3)
ꢀ
À
with ꢀO2 in homogeneous aqueous solutions [39]. Given
that SC1 and SC2 are more effective than Q in the
protection of LDL against Cu2þ-induced oxidation and
given that ꢀO2 might be Àinvolved as an intermediate
À
The ꢀBr2À radical anion is produced with a radiolytic yield
of 6.2 (i.e. 0.64 mM GyÀ1). In CTAB micelles, after
[39,40], the presence of ꢀO2 must be taken into account.
ꢀ
À
Here O2 reactivity with the SCs in CTAB and TX100

2214
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
0.030
0.020
0.010
0.000
-0.010
0.030
0.020
0.010
0.000
-0.010
-0.020
-0.030
SC2 in CTAB
SC2 in TX100
SC1 in CTAB
-0.020
SC1 in TX100
-0.030
400
500
600
700
400
500
600
700
(A)
(B)
Wavelength (nm)
Wavelength (nm)
0.030
0.020
0.010
0.000
-0.010
-0.020
-0.030
0.030
0.020
0.010
0.000
-0.010
-0.020
-0.030
SC3 in CTAB
SC3 in TX100
600
SC4 in CTAB
SC4 in TX100
600 700
400
500
700
400
500
(C)
(D)
Wavelength (nm)
Wavelength (nm)
Fig. 3. Transient absorption spectra obtained 2 ms after the radiolytic pulse in N2O-saturated, 10 mM phosphate buffer (pH 7) containing 0.1 M KBr and
either 10 mM CTAB (*) or 3 mM TX100 (&) in the presence of 100 mM SC. (A) SC1, (B) SC2, (C) SC3, and (D) SC4. Note the bleaching below 450 nm
due to SC depletion. Dose was 8 Gy.
micelles was measured by pulse radiolysis of O2-saturated
micellar solutions containing 10 mM phosphate buffer
(pH 7), 500 mM SC and 0.1 M HCOOÀ. Under these
conditions, all radiÀcals produced by water radiolysis are
not unexpected in TX100 micelles in view of the found
limited reactivity of the SCs with ꢀBr2À, a stronger oxidant
than O2À. The transient spectra obtained following the
ꢀ
ꢀ
À
oxidation of SC1, SC2, SC3, and SC4 by O2 in CTAB
are shown on Fig. 4A. The spectral shapes and positions of
absorbance maxima of these transients are similar to those
shown in Fig. 3A–D. However, differences in the ratio of
absorbances at wavelengths of maximum absorbance in the
blue and red regions are sometimes observed in the spectra
taken in CTAB compared to those in TX100. This behavior
converted into O2 radical anions with G ðꢀO2ÀÞ ¼ 6:3
ꢀ
(0.65 mM GyÀ1) by the reactions
À
eaqÀ þ O2 ! ꢀO2
(7)
(8)
Hꢀ þ O2 ! ꢀO2À þ Hþ
ꢀOH þ HCOOÀ ! ꢀCO2À þ H2O
ꢀCO2À þ O2 ! ꢀO2À þ CO2
(9)
ꢀ
À
mayꢀreflÀect differences in the relative reactivity of Br2
and O2 with the SC species whose ionic equilibrium is
governeÀd by the micellar microenvironment. Assuming
full ꢀO2 scavenging by the SCs, an estimate of the molar
absorbances of the ꢀSC radicals has been determined
(Table 1).
(10)
In CTAB micelles all the SCs are readily oxidized by
ꢀO2À. On the other hand, in the TX1ꢀ00 micelles, only SC1
showed measurable reactivity with O2À. This result was
Table 1
À
À
Molar absorbance (e) of ꢀSC radicals obtained by one-electron oxidation of SC by ꢀO2 and reaction rate constants (k11) for the reaction of SC with ꢀO2
Compounds
Conditions
Wavelength (nm)
e (MÀ1 cmÀ1
)
k11 (Â107 MÀ1 sÀ1
)
À
SC1
SC2
SC3
SC4
SC1
CTAB, ꢀO2À
CTAB, ꢀO2À
CTAB, ꢀO2À
CTAB, ꢀO2
600
630
550
560
460
6100
7400
4500
6400
1.4
1.1
0.6
1.8
0.11
À
TX100, ꢀO2
ND (see text)
ND: not determined. Temperature was 25 8C. e and k11 values were determined in 10 mM phosphate buffer (pH 7). SC concentration was 250 mM. Dose was
about 6 Gy.

P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
2215
ꢀ
3.2.4. Repair of the Trp radical by SC
0.060
0.040
0.020
0.000
-0.020
Pulse radiolysis of 5 mM Trp in N2O-saturated micꢀellar
solutions of CTAB at pH 7, generates the neutral Trp
radical by the one-electron oxidation reaction
ꢀBr2À þ Trp ! ꢀTrp þ Hþ þ 2BrÀ
In the CTAB micellar solution, the stoichiometric reaction
(12)
ꢀ
À
of the Br2 radical with a large excess of Trp occurs
within a few milliseconds with a rate constant k12
SC1
SC2
SC3
SC4
¼
8:6 Â 108 MÀ1 sÀ1. This reaction generates the character-
ꢀ
istic neutral Trp radical transient absorption with a max-
400
500
600
700
imum at 520 nm (emax ¼ 1750 MÀ1 cmÀ1) and negligible
(A)
Wavelength (nm)
ꢀ
absorption beyond 580 nm [22]. The Trp radical is pro-
duced with a G value of 6.2, corresponding to full scaven-
ging of the ꢀBr2À radicals by Trp [8]. As shown in Fig. 5A,
the ꢀTrp transient subsequently undergoes bimolecular
0.040
0.030
0.020
0.010
0.000
CTAB
TX100
0.010
0.005
0.000
0
0.0005
0.001
0.0015
(B)
Time (ms)
Fig. 4. (A) Transient absorption spectra obtained 2 ms after the radiolytic
pulse of O2-saturated, 10 mM phosphate buffer (pH 7) containing 0.1 M
HCOOÀ and 10 mM CTAB in the presence of (*) 125 mM SC1, (~)
500 mM SC2, (&) 250 mM SC3, or (5) 70 mM SC4. Dose was 10 Gy. Note
the bleaching of SC below 450 nm. (B) Growth of ꢀSC1 radicals in O2-
saturated 10 mM phosphate buffered micellar solutions (pH 7) containing
0.1 M sodium formate and 250 mM SC1 in the presence of either 10 mM
CTAB (*) or 3 mM TX100 (&). Transient growth was recorded at
460 nm. Solid line: fit of the growths assuming pseudo-first-order reaction
0
0.4
0.8
1.2
1.6
(A)
Time (ms)
À
for SC1 reacting with ꢀO2 (see text). Dose was about 10 Gy. For TX100,
absorbances were multiplied by 10 and the time scale by 0.1.
12.5
The reaction rate constants, k11, for the oxidation of SC
ꢀ
0.010
0.005
0.000
À
by O2 depicted by the reaction
10.0
7.5
5.0
2.5
0.0
ꢀO2À þ SC ! ꢀSC
(11)
were calcuꢀlated from the pseudo-first-order growth
kinetics of SC radicals in CTAB and TX100 micelles
(Fig. 4B) and are included in Table 1. The k11 values found
here in CTAB micelles are comparable to the value mea-
0
50
100
[SC1] (µM)
150
200
250
ꢀ
sured for Q formation under similar experimental condi-
tionÀs [8]. The rate constant for the reaction of SC1 and
ꢀO2 is 13 times higher in CTAB than in TX100 micelles
(Fig. 4B, Table 1). As with otherÀspecies bearing negative
charge such as ꢀBr2À, the ꢀO2 radical is most likely
localized in the positively charged headgroup region of
CTAB, thereby increasing the local concentration of reac-
tants. No effect of this nature is expected or observed with
TX100. Furthermore, as discussed above, it is likely that
the ionized hydroxyl groups of the SCs in the CTAB
micellar solution are much more oxidizable than the
un-ionized SCs found in TX100.
0
0.2
0.4
0.6
0.8
Time (ms)
(B)
Fig. 5. (A) Decay of the ꢀTrp radical at 520 nm in absence of SC after
pulse radiolysis of 5 mM Trp in N2O-saturated, 10 mM phosphate buffer
(pH 7) containing 0.1 M KBr and 10 mM CTAB at 25 8C. Dose was about
12 Gy. Solid line: fit of the ꢀTrp decay to a bimolecular second-order
kinetics. (B) Growth of the ꢀSC1 radical at 600 nm with the same
constituents as in (A) but in the presence of 125 mM SC1. Solid line: fit of
the growth assuming a pseudo-first-order reaction between ꢀTrp and SC1
competing with the ꢀTrp recombination (see text for details). Insert: plot of
the pseudo-first-order rate constant (kapp) measured from growths recorded
at 600 nm as a function of [SC1]. Dose was about 6 Gy.

2216
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
decay (2k13 ¼ 6:0 Â 108 MÀ1 sÀ1) [9] according to the
Values of kET greater than 1 Â 107 MÀ1 sÀ1 for all the
SCs are obtained in a similarly straightforward manner and
are compiled in Table 2. It can be seen that these values are
rather similar to those previously obtained with Q in CTAB
micellar solutions [8]. The repair reaction is also observed
with SC1 and ꢀTrp in TX100 micelles confirming that the
SCs under study are antioxidants at least potentially as
good as Q, the reference flavonoid antioxidant.
reaction
2ꢀTrp ! products
(13)
Upon addition of 250 mM SC1 to a CTAB micellar solution
containing 5 mM Trp, strong transient absorbance changes
are observed on a time scale of several hundred micro-
seconds yielding the transient spectrum attributed to the
ꢀSC1 radical shown in Fig. 3A. Fig. 5B illustrates the slow
transient growth at 600 nm, the wavelength maximum of
the ꢀSC1 radical in the red region. Similar long term radical
formation is observed with all SCs (data not shown). The
formation of the ꢀSC1 radical over such a lonÀg time scale
4. Conclusion
Both the SC1 derivative, with two OH groups on the
styryl moiety, and SC2, which has an additional hydroxyl
group at position 5 on the benzopyrone ring, compare
favorably to quercetin in the protection of LDL lipids from
peroxidation in Cu2þ-induced oxidation. The study of the
electron donating properties of the four SCs by pulse
radiolysis in micellar solutions demonstrates that they
are equally capable of reacting with ꢀO2À and of repairing
ꢀTrp radicals in CTAB micelles. In such micelles, bearing
positively charged head groups, the effectiveness of elec-
tron transfer reactions strongly depends on microenviron-
mental factors such as local pH and electrostatic
interactions which modify both the redox parameters
and the accessibility (local concentration) of radicals to
these hydrophobic SCs. Such factors are also commonly
encountered in the control of structure and function in
proteins. The superiority of SC1 over the other three SCs is
made clear when electron transfer reactions are carried out
in neutral TX100 micelles in which SC1 is the sole SC
derivative exhibiting measurable reactivity. This observa-
tion may be related to the better protection afforded
specifically by this antioxidant in the oxidation of LDL
by Cu2þ ions at pH 7. In this regard, the repair of ꢀTrp by
SCs suggests that it would be interesting to study the
protection brought by SCs against the oxidation of key
residues of LDL apolipoprotein B. As a matter of fact, the
binding of LDL to their cell receptors requires the integrity
of apolipoprotein B. However, other factors must also be
taken into account when considering the antioxidant capa-
city of SCs. Among them, chelation of transition metal ions
by SCs should not be ignored since it is believed that the
chelating properties of related antioxidants such as flavo-
noids contribute to their antioxidant activity. Thus, Brown
et al. demonstrated that the phenolics that had the greatest
propensity for Cu2þ ion chelation were the most effective
in inhibiting LDL oxidation in the model of copper-
mediated LDL oxidation. Their study also showed a sig-
nificant correlation between partition of phenolics and the
inhibition of propagation rate due to redox reactions
generated by the hydrophilic Cu2þ [41]. Similar factors
could play a role in the antioxidant activity of the SCs.
Finally, it may also be deduced from this study that the
antioxidant potential of the 2-styrylchromones is not
necessarily correlated to the number of OH substitutions.
ꢀ
cannot be due to theꢀdireÀct reaction of Br2 with SC1
(reaction 4) since all Br2 radical anions react with Trp
within a few microseconds. This behavior may, however,
be explained by the electron transfer reaction
ꢀTrp þ SC ! ꢀSC þ Trp
occurring with the rate constants k14 or kET. The electron
(14)
transfer yield from SC, measured 500 ms after the pulse,
may be seen to approach the G value for the Trp radical
ꢀ
production. Tryptophan itself is soluble (ꢂ10 mM) in the
buffer, and therefore, predominantly in the aqueous phase
of micellar solutions. The inset in Fig. 5B, showing the
growth rate of the 600 nm transient as a function of the
SC1 concentration, demonstrates the pseudo-first-order
ꢀ
nature of the SC1 growth which is in agreement with
ꢀ
the large excess of SC1 compared to the Trp concentra-
tion. The excellent fit of the transient growth at 600 nm
(Fig. 5B) hasꢀbeen obtained assuming that the reaction of
SC with the Trp radical competes with the bimolecular
ꢀTrp reaction (reaction 13) illustrated in Fig. 5A. The
associated kinetics of ꢀTrp and ꢀSC can be represented by
the equations
d½ꢀTrpꢃ
2
À
¼ 2k13½ꢀTrpꢃ þ kET½SCꢃ½ꢀTrpꢃ
dt
d½ꢀSCꢃ
dt
¼ kET½SCꢃ½ꢀTrpꢃ
where 2k13 ¼ 6:0 Â 108 MÀ1 sÀ1 and [SC] was varied
between 62 and 250 mM.
Table 2
Rate constant (kET) for the electron transfer from SC to ꢀTrp radicals in
N2O-saturated pH 7 buffered micellar solutions
Compounds
Conditions
kET (Â107 MÀ1 sÀ1
)
SC1
SC2
SC3
SC4
SC1
CTAB
CTAB
CTAB
CTAB
TX100
5.2
4.5
3.2
3.6
2.2
Temperature was 25 8C. Pseudo-first-order decays (see text and insert of
Fig. 5B) were determined for successive dilutions of a micellar solution
containing 250 mM SC and 5 mM Trp with a 5 mM Trp stock micellar
solution.

P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
2217
[16] Fernandes E, Carvalho M, Silva AMS, Santos CMM, Pinto DCGA,
Cavaleiro JAS, et al. 2-Styrylchromones as novel inhibitors of
xanthine oxidase. A structure–activity study. J Enzyme Inhib Med
Chem 2002;17:45–8.
Acknowledgments
This work was supported by an INSERM/GRICES
Franco-Portuguese exchange program. J.N.S. thanks the
Gulbenkian Foundation for a travel grant and C.M.B.
thanks the Organic Chemistry Research Unit and the
University of Aveiro for a research grant. We wish to thank
Mrs. J. Haigle for her skillful technical assistance. Notre
Dame Radiation Laboratory is supported by the Office of
Basic Energy Sciences of the US Department of Energy.
This is document NDRL-4480 from the Notre Dame
Radiation Laboratory.
[17] Fernandes E, Carvalho M, Carvalho F, Silva AMS, Santos CMM,
Pinto DCGA, et al. Hepatoprotective activity of polyhydroxylated 2-
styrylchromones against tert-butylhydroperoxide induced toxicity in
freshly isolated rat hepatocytes. Arch Toxicol 2003;77:500–5.
[18] Rice-Evans CA, Leake D, Bruckdorfer R, Diplock AT. Practical
approaches to low-density lipoprotein oxidation: whys, wherefores
and pitfalls. Free Radic Res 1996;25:285–311.
[19] Patel RP, Darley-Usmar VM. Molecular mechanisms of the copper
dependent oxidation of low-density lipoproteins. Free Radic Res
1999;30:1–9.
[20] Cladaras C, Hadzopoulou-Cladaras M, Nolte RT, Atkinson D, Zannis
VI. The complete sequence and structural analysis of human apoli-
poprotein B-100: relationship between apoB-100 and apoB-48 forms.
EMBO J 1986;5:3495–507.
References
[21] Davies KJA. Protein damage and degradation by oxygen radicals. J
Biol Chem 1987;262:9895–901.
[1] Steinbrecher UP, Zhang H, Lougheed M. Role of oxidatively modified
LDL in atherosclerosis. Free Radic Biol Med 1990;9:155–68.
[2] Trush MA, Kensler TW. Role of free radicals in carcinogen activation.
In: Sies H, editor. Oxidative stress. Oxidants and antioxidants. San
Diego: Academic Press; 1991. p. 277–318.
[22] Redpath JL, Santus R, Ovadia J, Grossweiner LI. The oxidation of
tryptophan by radical anions. Int J Radiat Biol 1975;27:201–4.
[23] Jovanovic SV, Simic MG. Repair of tryptophan radicals by antiox-
idants. J Free Radic Biol Med 1985;1:125–9.
[3] Mattson MP, Duan W, Pederson WA, Culmsee C. Aging alters the
apoptotic response to genotoxic stress. Apoptosis 2001;6:69–81.
[4] Melow S, Ravenscroft J, Malik S, Gill MS, Walker DW, Clayton PE,
et al. Extension of life span with superoxide dismutase. Science
2000;289:1567–9.
[24] Santos CMM, Silva AMS, Cavaleiro JAS. Synthesis of new hydroxy-
2-styrylchromones. Eur J Org Chem 20034575-4585.
[25] Havel JR, Edre HA, Bragdon J. The distribution and chemical
composition of ultracentrifugally-separated lipoproteins in human
serum. J Clin Invest 1955;34:1345–53.
[5] Halliwell B. Antioxidant defence mechanisms: from the beginning to
the end (of the beginning). Free Radic Res 1999;31:261–72.
[6] Hermann K. Occurrence and content of hydroxycinnamic acid and
hydroxybenzoic compounds in foods. Crit Rev Food Sci Nutr
1989;28:315–47.
[26] Peterson GL. Simplification of the protein assay method of Lowry et al.
which is more generally applicable. Anal Biochem 1977;83:346–56.
[27] Corongiu FP, Banni S, Dessi MA. Conjugated dienes detected in tissue
lipid extracts by second derivative spectrophotometry. Free Radic Biol
Med 1989;7:183–6.
[7] Rice-Evans CA, Miller J, Paganga G. Structure–antioxidant activity
relationships of flavonoids and phenolic acids. Free Radic Biol Med
1996;20:933–56.
`
[28] Behrens WA, Thompson JN, Madere R. Distribution of alpha-toco-
pherol in human plasma lipoproteins. Am J Clin Nutr 1982;35:691–6.
[29] Thurnham DI, Smith E, Flora PS. Concurrent liquid-chromatography
assay of retinol, a-tocopherol, b-carotene, a-carotene, lycopene and b-
cryptoxanthin in plasma, with tocopherol acetate as internal standard.
Clin Chem 1988;34:377–81.
`
`
`
[8] Filipe P, Morliere P, Patterson LK, Hug GL, Maziere J-C, Maziere C,
et al. Mechanisms offlavonoid repair reactions with amino acid radicals
in models of biological systems: a pulse radiolysis study in micelles and
human serum albumin. Biochim Biophys Acta 2002;1572:150–62.
[30] Patterson LK, Lilie JA. Computer-controlled pulse radiolysis system.
Int J Radiat Phys Chem 1974;6:129–41.
`
`
`
[9] Filipe P, Morliere P, Patterson LK, Hug GL, Maziere JC, Maziere C,
et al. Repair of amino acid radicals of apolipoprotein B100 of low-
density lipoproteins by flavonoids. A pulse radiolysis study with
quercetin and rutin. Biochemistry 2002;41:11057–64.
[31] Hug GL, Wang Y, Schoneich C, Jiang PY, Fessenden RW. Multiple
time scale in pulse radiolysis: application to bromide solutions and
dipeptides. Radiat Phys Chem 1999;54:559–66.
´
´ ´
[10] Manach C, Morand C, Crespy V, Demigne C, Texier O, Regerat F, et al.
Quercetin is recovered in human plasma as conjugated derivatives
which retain antioxidant properties. FEBS Lett 1998;426:331–6.
[11] Peng HW, Cheng FC, Huang YT, Chen CF, Tsai TH. Determination of
naringenin and its glucuronide conjugate in rat plasma and brain tissue
[32] Schuler RH, Patterson LK, Janata E. Yield for the scavenging of OH
radicals in the radiolysis of N2O-saturated aqueous solutions. J Phys
Chem 1980;84:2088–9.
[33] Chauvet JP, Viovy R, Land EJ, Santus R, Truscott TG. One-electron
oxidation of carotene and electron transfers involving carotene
cations and chlorophyll pigments in micelles. J Phys Chem 1983;
87:592–601.
by high-performance liquid chromatography.
J Chromatogr B
1998;714:369–74.
[12] Gerkick WH. 6-Desmethoxyhormothamnione, a new cytotoxic styr-
ylchromone from marine cryptophyte chrysophaeum tailori. J Nat
Prod 1989;22:252–6.
[34] La Verne JA, Pimblott SM. Scavenger and time dependence of radicals
and molecular products in the electron radiolysis of water: examina-
tion of experiments and models. J Phys Chem 1991;95:3196–206.
[35] Esterbauer H, Striegl G, Puhl H, Rotheneder M. Continuous monitor-
ing of in vivo oxidation of human low density lipoprotein. Free Rad
Res Commun 1989;6:67–75.
[13] Gerwick WH, Lopez A, Van Duyne V, Clardy J, Ortiz W, Baez A.
Hormothamnione, a novel cytotoxic styrylchromone from the marine
cyanophyte Hormothamnion enteromorphoides Grunow. Tetrahedron
Lett 1986;27:1979–87.
[36] Bors W, Heller W, Michel C, Saran M. Flavonoids as antioxidants:
determination of radical-scavenging efficiencies. Methods Enzymol
1990;186:343–55.
[14] Brion JD, Le Baut G, Zammattio F, Pierre A, Atassi G, Belachmi L.
Preparation of 2-styryl-4-chromanones as anti-cancer agents. Eur-
opean Patent Applied EP 1991,454,587 [Chem Abstr 1992;116:755
(106092K)].
´
[37] Fernandez MS, Fromherz P. Lipoid pH indicators as probes of
electrical potential and polarity in micelles. J Phys Chem 1977;81:
1755–61.
[15] Doria G, Romeo C, Forgione A, Sberze P, Tibolla N, Cormo ML, et al.
Antiallergic agents. III. Substituted trans-2-ethenyl-4-oxo-4H-1-benzo-
pyran-6-carboxylic acids. Eur J Med Chem—Chim Ther 1979;27:
347–51.
[38] Jovanovic SV, Steenken S, Simic M, Hara Y. Antioxidant properties of
flavonoids: reduction potentials and electron transfer reactions of

2218
P. Filipe et al. / Biochemical Pharmacology 67 (2004) 2207–2218
flavonoid radicals. In: Rice-Evans CA, Packer L, editors. Flavonoids
in health and disease. New York: Marcel Dekker; 1998. p. 137–61.
[40] Pecci L, Montefoschi G, Cavallini D. Some new details on the copper–
hydrogen peroxide interaction. Biochem Biophys Res Commun
1997;235:264–7.
[39] Perugini C, Seccia M, Bagnati M, Cau C, Albano E, Bellomo G.
Different mechanisms are progressively recruited to promote Cu(II)
reduction by isolated human low density lipoprotein undergoing
oxidation. Free Radic Biol Med 1998;25:519–28.
[41] Brown JE, Khodr H, Hider RC, Rice-Evans CA. Structural depen-
dence of flavonoid interactions with Cu2þ ions: implications for their
antioxidant properties. Biochem J 1998;330:1173–8.