Protective effects of 4-hydroxycinnamic ethyl ester derivatives and related dehydrodimers against oxidation of LDL: Radical scavengers or metal chelators?

Article abstract of DOI:10.1021/jf052923p

4-Hydroxycinnamate derivatives are known to be potent protectors against oxidation of low-density lipoproteins (LDL), via a combination of free radical scavenging and transition metal chelation. Through a series of 4-hydroxycinnamic ethyl ester derivative

Full text of DOI:10.1021/jf052923p

1898 J. Agric. Food Chem. 2006, 54, 1898−1905
Protective Effects of 4-Hydroxycinnamic Ethyl Ester Derivatives
and Related Dehydrodimers against Oxidation of LDL: Radical
Scavengers or Metal Chelators?
ANNE NEUDO¨ RFFER,† JEAN-PIERRE DESVERGNE,§
DOMINIQUE BONNEFONT-ROUSSELOT,‡ ALAIN LEGRAND,‡
MAURICE-BERNARD FLEURY,† AND MARTINE LARGERON*,†
Unite´ Mixte de Recherche CNRS-Universite´ Paris 5 (UMR 8638, Synthe`se et Structure de Mole´cules
d’Inte´reˆt Pharmacologique), Laboratoire de Biochimie Me´tabolique et Clinique (EA 3617), Faculte´ des
Sciences Pharmaceutiques et Biologiques, 4 Avenue de l’Observatoire, 75270 Paris Cedex 06, France,
and Unite´ Mixte de Recherche CNRS-Universite´ Bordeaux 1 (UMR 5802, Laboratoire de Chimie
Organique et Organome´tallique), 351 Cours de la Libe´ration, 33405 Talence Cedex, France
4-Hydroxycinnamate derivatives are known to be potent protectors against oxidation of low-density
lipoproteins (LDL), via a combination of free radical scavenging and transition metal chelation. Through
a series of 4-hydroxycinnamic ethyl ester derivatives and related 8-8 dehydrodimers, we have tried
to bring out the structural requirements for radical scavenging and cupric ion chelation. We found
that the monomeric compounds, except for highly lipophilic tert-butyl derivative 3, exhibited rather
low radical scavenging properties. Furthermore, they did not chelate copper but, in contrast, reduced
cupric ion to cuprous ion, affording the related 8-8 dehydrodimers, for which they could be considered
as precursors in vitro. In the copper-dependent human LDL oxidation in vitro, the cyclic 8-8
dehydrodimer forms behaved essentially as efficient copper chelators, while related noncyclic 8-8
forms, which were found to be the best protectors, mainly acted as radical scavengers.
KEYWORDS: 4-Hydroxycinnamic acid derivatives; dehydrodimers; LDL oxidation; radical scavenging;
lipophilicity; metal chelation
INTRODUCTION
Table 1. Distribution of Ferulic Acid Dehydrodimers Released from
Insoluble Dietary Fiber of Cereals after 1 M NaOH Treatmenta
Oxidative modification of low-density lipoproteins (LDL) has
been recognized to play a central role in early stage of
atherosclerosis (1). Consequently, a growing interest is devoted
to the intake of antioxidant-rich foods, such as fresh fruits,
vegetables, and beverage plants. Catechols and phenolic com-
pounds, which are widely distributed in nature (2, 3), offer a
great number of pharmacologically interesting molecules such
as flavonoids (quercetin, myricetin, rutin, kaempferol) (4),
caffeic acid derivatives, and 4-hydroxycinnamic compounds
(ferulic and sinapic acids) (5, 6). Among these compounds,
4-hydroxycinnamic derivatives are known to be absorbed across
the gastrointestinal barrier, then conjugated and eliminated in
human urine (7-10). In plants, these compounds undergo
oxidative cross-coupling leading to the corresponding 8-8
noncyclic, 8-8 cyclic, 8-5, 5-5, and 8-O-4 dehydrodimers, natural
precursors of lignin (11-14). The content of dehydrodimers
isolated from natural material varied with both hydrolysis
conditions and species considered (Table 1) (15, 16). Because
8-8 cyclicb
8-8 noncyclicb
8-5b
8-O-4b
5-5b
barley
maize
oats
rye
millet
33
27
37
30
39
17
7
19
8
12
17
17
43
19
17
25
6
3
17
31
24
21
16
20
5
a Data previously reported in ref 20. b Expressed as the percentage of each
dimer over the total ferulic acid dehydrodimers.
they can also act as cross-linking agents between polysaccha-
rides and lignin (17-20), 4-hydroxycinnamic acid derivatives
modify the mechanical properties of cell walls, inducing
insolubility of cereal dietary fibers (21).
Previous studies have shown that, in all cereal insoluble
dietary fibers, the 8-5 dehydrodimers predominated, whereas
in cereal soluble dietary fibers, the 8-8 coupled dimers became
the major ones (21). Similarly, 8-8 noncyclic and cyclic sinapic
acid dehydrodimers were found in wild rice insoluble dietary
fibers (22), while 8-8 noncyclic dehydroferulate was shown to
play a key role in conferring thermal stability of cell-cell
* Corresponding author. Phone: 33 01 53 73 96 46. Fax: 33 01 44 07
35 88. E-mail: martine.largeron@univ-paris5.fr.
† Unite´ Mixte de Recherche 8638 CNRS-Universite´ Paris 5.
‡ Laboratoire de Biochimie Me´tabolique et Clinique (EA 3617).
§ Unite´ Mixte de Recherche 5802 CNRS-Universite´ Bordeaux 1.
10.1021/jf052923p CCC: $33.50
© 2006 American Chemical Society
Published on Web 02/07/2006

Protective Effects of 4-Hydroxycinnamic Ethyl Ester Derivatives
J. Agric. Food Chem., Vol. 54, No. 5, 2006 1899
3 and
+
Table 2. Partition Coefficient and Antioxidant Activity toward Cu2 or AAPH-Induced LDL Oxidation of 4-Hydroxycinnamic Ethyl Esters 1
−
Relative Dehydrodimers 4−8
-
1
a Concentration of compounds 1
was incubated with 4 mM AAPH, at 37
c Probucol, a hypocholesterolemic compound, was evaluated in the same test systems for comparison.
−
8 leading to 50% decrease of the amount of TBARS concentration produced after 250 min of incubation. LDL (0.06 g of protein
‚
L
)
-
7
-3
°
C, in the presence of 10 to 10 M solutions of the compounds in DMSO for 250 min. b Data previously reported in ref 40.
adhesion (23). More recently, an 8-8 cyclic/8-O-4 dehydroferulic
acid was isolated from maize bran (24).
were found less active, in agreement with the final sequence
order of activity 5 ≈ 6 > 4 > 7 ≈ 3 > 2 > 8 ≡ probucol > 1.
At the same time, many studies attempted to determine the
antioxidant properties of 4-hydroxycinnamic acid derivatives,
especially their capacity to protect LDL from oxidation (25-
29). Many in vitro assays used the copper-catalyzed LDL
oxidation model as a mimic of the in vivo oxidation process
(30-32). These assays measure a combination of radical
scavenging and transition metal chelation. Radical scavenging
by phenols occurs by reducing the reactive oxygen species or
lipid peroxyl radicals by means of hydrogen atom donation from
free hydroxyl radicals, giving rise to phenoxyl radicals. Metal
chelation would be considered as prevention means of the
peroxidation, either by sequestering metal ions or by restricting
the access of metal ions toward lipid hydroperoxides (6, 33-
35). As part of our continuing research of efficient antioxidants
(36-38), we showed that electrochemical oxidative coupling
of 4-hydroxycinnamic ester derivatives constituted a straight-
forward one-pot method for the biomimetic synthesis of natural
lignin precursors (39). Consecutively, we reported the protective
effects against copper-induced LDL oxidation, exerted by a
series of 4-hydroxycinnamic ethyl ester derivatives 1-3 and
related dehydrodimers 4-8 (Table 2) chosen as model of their
corresponding saccharidic esters (40). Our results showed that,
on the whole, dehydrodimers derivatives 4-8 acted as better
antioxidants than their monomeric counterparts 1-3. In par-
ticular, the noncyclized 8-8 dehydrodimers 5 and 6 were
characterized as very potent antioxidants, while the correspond-
ing cyclized 8-8 dehydrodimer (dihydronaphthol) derivatives
As the protection of LDL against oxidation may take place
through metal chelation as well as radical scavenging (31-35,
41-44), the antioxidant activity of the series studied was also
tested in the trolox equivalent antioxidant capacity (TEAC)
assay, which measured the cation radical 2,2′-azino-bis-(3-ethyl-
benzothiazoline-6-sulfonate) ABTS+• scavenging properties.
Then, we observed an inversion in the sequence order of activity
which became the following: probucol > 8 > 7 > 4 > 1 > 6
≈ 5 > 3 > 2 (40). So, in copper-induced LDL oxidation, 8-8
noncyclic dehydrodimers were found to be the most efficient
protectors against copper-induced oxidation of human LDL,
while in the TEAC aqueous system, 8-8 cyclic dehydrodimers
exhibited the strongest scavenging properties against the ABTS+•
cation radical.
As these two assays differed from both the nature of the phase
and that of the oxidation initiator, we thought that a more
comprehensive investigation of the antioxidant capacity of the
studied compounds would be desirable to undoubtedly determine
which of the dehydrodimers, cyclized or not, could be consid-
ered as the most attractive derivative. Consequently, we sought
to evaluate the radical scavenging properties of the title
compounds through their respective ability to inhibit 2,2′-azo-
bis-(2-amidinopropane)dihydrochloride (AAPH)-induced oxida-
tion of LDL, thus allowing a more rigorous comparison with
the copper-induced oxidation assay. Secondarily, we envisaged
a second antioxidant mechanism of the title compounds,
consecutive to copper chelate formation.

1900 J. Agric. Food Chem., Vol. 54, No. 5, 2006
Neudo¨rffer et al.
MATERIALS AND METHODS
of compounds 1-8, in deaerated acetonitrile containing 1 equiv
(compounds 1-3) or 2 equiv (compounds 4-8) of tetramethylammo-
nium hydroxide. The absorption spectra of aliquots were recorded in
the range of 200-600 nm, after each addition of copper solution.
Isolation of Resulting Compounds. Among the compounds studied,
similar behavior was observed (1 vs 2 and 3; 4 vs 5 and 6; 7 vs 8).
Consequently, the reactions of Cu2+ with compounds 1, 4, 7 were not
presented.
Chemicals. AAPH, a water-soluble source of peroxyl radicals, was
a generous gift from A. Kontush and was obtained from Polysciences
(Warrington, PA). BHT was obtained from Carlo-Erba (Rodano, Italy).
Tetramethylammonium hydroxide (25 wt % solution in methanol), Cu-
(NO3)2‚2.5 H2O and Probucol were supplied from Sigma-Aldrich (Saint
Quentin Fallavier, France). Acetonitrile (anhydrous analytical grade)
was purchased from SDS (Peypin, France). Compounds 1-8 were
prepared according to our published procedures (39, 40).
Reaction of Cu2+ with Compound 2: Method A. A solution of Cu-
(NO3)2 (0.1 M) in acetonitrile was added stepwise, under stirring at
room temperature, to a 2.0 mM solution (100 mL) of compound 2 (50.4
mg, 0.2 mmol) and tetramethylammonium hydroxide (84 µL, 0.2
Isolation of Human Low-Density Lipoproteins (LDLs). LDL was
isolated from the pooled plasma of healthy normolipidemic human
subjects in the presence of EDTA (1.08 mM) by sequential ultracen-
trifugation (1.019 < d <1.050), according to the method of Havel et
al. (45). For oxidation experiments, LDL was dialyzed in the dark, for
18 h at 4 °C, against 10 mM sodium phosphate buffer pH 7.4, containing
150 mM sodium chloride (PBS). After determination of protein
concentration by a pyrogallol technique (Elitech Diagnostics, Sees,
France) (46), LDL was diluted to a final protein concentration of 0.06
mmol), in deaerated acetonitrile. After addition of 1.4 equiv of Cu2+
,
the final solution was poured into a 0.5 M acetic acid buffered aqueous
solution of pH ≈ 4.5 (50 mL). The resulting mixture was concentrated
to 50 mL under reduced pressure, at 40 °C, and extracted with ethyl
acetate (100 mL). The organic phase was dried over anhydrous sodium
sulfate, and the solvent was removed under reduced pressure, at 40
°C. Flash chromatography of the residue on silica gel, with a toluene-
acetone (85:15) mixture as the eluent, afforded 8-8 noncyclic dehy-
drodimer 5 (28.0 mg, 55%) and 8-8 cyclic dehydrodimer 8 (6.5 mg,
13%) as pale yellow oils, together with 8% of the starting material 2
(4.0 mg).
g L-1
.
Kinetics of AAPH-Induced LDL Oxidation. The effects of
antioxidants on the kinetics of lipid oxidation of human LDL were
assessed by spectrophotometric monitoring, at 234 nm, of conjugated
diene lipid hydroperoxide formation (47), during AAPH-induced
oxidation (4 mM AAPH, 10 mM Chelex-treated PBS pH 7.4, 37 °C,
in the dark) (48). The tested compounds, dissolved in DMSO (10 µL),
were added to dialyzed LDL (final volume 1 mL, 0.06 g of protein‚L-1),
with final concentrations of 5 µM just before the introduction of the
AAPH solution. LDL oxidized in the presence of DMSO constituted a
reference. Absorbance at 234 nm was recorded every 10 min, during
600 min, on a Uvikon Kontron spectrophotometer (reference, 4 mM
AAPH in 10 mM PBS, pH 7.4) and the differential absorbance (∆A )
A - At0) was calculated. Three phases could be distinguished from the
absorbance change pattern, i.e., lag phase of conjugated diene formation,
propagation phase, and termination phase. During the lag phase, the
lipophilic endogenous antioxidant derivatives protected the polyun-
saturated fatty acids of the LDL against oxidation. After their
consumption, the lipid peroxidation process underwent the propagating
chain reaction phase.
Antioxidant Activity upon LDL Oxidation Evaluated by TBARS.
The antioxidant activity of each compound was also estimated through
their ability to inhibit the formation of products of lipid peroxidation
during AAPH-induced LDL oxidation. Oxidation was performed with
dialyzed LDL (0.06 g of protein‚L-1 final concentration), by the above-
described method (4 mM AAPH, 10 mM Chelex-treated PBS pH 7.4,
37 °C, in the dark) (48). A nonoxidized LDL sample, incubated in the
absence of AAPH and in the presence of 200 µM EDTA, constituted
the blank control, while addition of 10 µL of DMSO was used for
reference. Oxidation was stopped by adding an EDTA-BHT solution
(200 µM EDTA and 20 µM BHT final concentration) and cooling in
an ice bath. After a 250 min period of oxidation, the final concentration
of TBARS was determined by the spectrofluorometric method of Yagi
(λexc ) 515 nm, λem ) 548 nm), with 1,1′,3,3′-tetraethoxypropane
as the standard and the nonoxidized sample as the blank (49). For each
compound, 10 concentrations, ranging from 10-7 to 10-3 M, were tested
in duplicate, the variability between the results being less than 5%.
The presence of DMSO, or of the studied compounds at 10-3 M, did
not significantly interfere on the calibration curve used for the assay.
The concentration (IC50) leading to 50% decrease of the amount of
TBARS formed after oxidation for 250 min (compared to the oxidation
in the presence of DMSO) was estimated by linear regression analyses.
Reaction of Cu2+ with Compound 3: Method B. A solution of Cu-
(NO3)2 (0.1 M) in acetonitrile was added stepwise, under stirring, at
room temperature, to a 2.0 mM solution (100 mL) of compound 3 (60.8
mg, 0.2 mmol) and tetramethylammonium hydroxide (84 µL, 0.2
mmol), in deaerated acetonitrile. After addition of 1.2 equiv of Cu2+
,
evaporation of the solvent under reduced pressure, at 40°C, gave a pale
yellow oil. Analysis by 1H NMR showed that the bisquinonemethide 9
was the sole product of the reaction. Spectroscopic data for compound
9 have been previously reported (39).
Reaction of Cu2+ with Compounds 5, 6, and 8: Method A (vide
supra) was applied to compounds 5, 6, and 8 (0.2 mmol), introducing
2 equiv of tetramethylammonium hydroxide (168 µL, 0.4 mmol). After
addition of respectively 1.7, 0.6, and 2.2 equiv of copper, the starting
materials 5, 6, and 8 were recovered, after chromatography on silica
gel, in 77%, 79%, and 78% yields, respectively.
Binding Constants Determinations. The binding constants were
determined using the LETAGROP-SPEFO program. This program
calculates the global binding constant â ) K1K2K3 ... (and the successive
association constants Ki) for a chemical scheme by iterative comparison
of calculated data with experimental data, searching for the global
minimum of the error function. Equations that do not fit the data are
rejected. This program is a multicompartmental approach to spectral
fitting, which allows simultaneous fitting at many wavelengths (51,
52).
RESULTS AND DISCUSSION
Radical Scavenging Properties and Partitioning Proper-
ties. The radical scavenging properties of compounds 1-8 were
tested in a nonaqueous system using the 2,2′-azo-bis-(2-
amidinopropane) dihydrochloride (AAPH)-induced LDL oxida-
tion. In the AAPH-mediated test, lipid peroxyl radicals are
generated indirectly within the lipidic phase, consecutively to
spontaneous decomposition of the azo compound producing
AAPH-derived peroxyl radicals. This metal-independent oxida-
tion measures the ability of antioxidants to intercept lipid peroxyl
radicals and thereby to break the free radical chain reactions.
The effects of the different monomeric and dimeric 4-hy-
droxycinnamic ethyl esters derivatives 1-8 (5 µM) on the
kinetics of AAPH-mediated oxidation reaction of LDL were
determined by measuring conjugated diene formation at 234
nm. For comparison, Probucol, a well-known hypocholester-
olemic and lipophilic antioxidant compound, was used as a
reference antioxidant. As shown in Figure 1a, monomeric
compounds 1-3 exhibited moderate antioxidant properties,
while noncyclic 8-8 dehydrodimers 4-6 inhibited more strongly
Partition Coefficient Determination (Log P). Log P values, which
express the partitioning of the compounds in an n-octanol-water
system, were calculated using Crippen’s fragmentation through the CS
Chem Draw Pro (7.0.1 Ultra, Cambridge Soft Company) program (50).
In this program, the contribution of each atom to molar hydrophobicity
was evaluated within a molecular database of experimental partitioning
values, using a constrained least-squares technique.
Interactions of Cu2+ with Compounds 1-8. Electronic Absorption
Spectroscopy. A solution of Cu(NO3)2 (0.1 M) in acetonitrile was added
in portions, under stirring, at room temperature, to 0.1 mM solutions

Protective Effects of 4-Hydroxycinnamic Ethyl Ester Derivatives
J. Agric. Food Chem., Vol. 54, No. 5, 2006 1901
Then, it appeared that the discrepancy in the order of activities
noted between the data obtained through LDL copper-mediated
oxidation (giving the noncyclized dehydrodimers 5 and 6 as
the most active compounds) and the previous TEAC aqueous
assay (showing the cyclic dehydrodimers 7 and 8 as the strongest
radical scavengers) was suppressed. Indeed, the radical scaveng-
ing properties of the cyclic dehydrodimers 7 and 8 notably
decreased in the AAPH lipidic assay, while, in contrast, related
noncyclic dehydrodimers 5 and 6 were found to be the best
radical scavengers. These would inhibit LDL oxidation through
radical scavenging, likely due to the high stability of the
corresponding phenoxyl radical, consecutive to delocalization
of charge and spin over the whole molecular framework.
Accordingly, the lowering of radical scavenging observed with
cyclic dehydrodimers 7 and 8 would be consecutive to a
decrease of the delocalization of charge and spin over a sole
phenyl ring.
Interestingly, all compounds substituted by tert-butyl groups
on the aromatic rings, probucol together with compounds 3 and
6, exhibited low IC50 values (Table 2). At this point, the
contribution of partitioning should also be taken into account.
The partition coefficient P has been the most frequently used
physicochemical parameter to rapidly evaluate the permeability
of molecules through biological membranes. The log P values
of the 4-hydroxycinnamic ester derivatives 1-8 were calculated
in octanol-water mixture, according to the Crippen’s fragmen-
tation method. We found that 3 (log P ) 5.55) and especially
6, with the highest log P value (10.47), also exhibited the highest
antioxidant activity in the lipidic AAPH system, while 7 and 8
(log P ) 3.12 and 2.87, respectively) were notably less active.
In contrast, when we compared two derivatives exhibiting very
close values of log P, 4 and 5 for example (log P ) 3.40 and
3.15, respectively), we found that the former possessed a
markedly lower antioxidant activity than the latter. Therefore,
although high lipophilicity improved the antioxidant potential
activity of 4-hydroxycinnamic ethyl ester derivatives by increas-
ing their propensity to penetrate into lipidic bilayers, it was not
sufficient to explain the variation of the IC50, in the series 4-8.
Consequently, we turned our attention toward the chelating
properties of the compounds.
Interactions of Cu2+ Ions with Compounds 1-8. Transition
metals are strongly implicated in the production of highly
reactive hydroxyl radicals through the superoxide-driven Fenton
reaction, as well as in the direct reductive decomposition of
lipid hydroperoxides, to give alkoxyl and lipid peroxyl radicals
during the propagation step. Consequently, antioxidative com-
pounds could provide protection against copper-induced LDL
oxidation, through reduction of Cu2+ ions into Cu+ ions or by
chelation of Cu2+ ion, alternatively. To clarify the ability of
4-hydroxycinnamic ethyl ester derivatives 1-3 and related
dehydrodimers 4-8 to react with cupric ion, their interaction
was assessed by electronic absorption spectroscopy, and the
resulting products were isolated and characterized by 1H NMR.
As the phenolate anion species proved to be a better substrate
than the neutral form for oxidation or metal chelation, the study
was performed in acetonitrile containing a stoichiometric amount
of tetramethylammonium hydroxide.
Figure 1. Kinetics of formation of conjugated dienes in the absence or
presence of 5
absorbance at 234 nm (reference, 4 mM AAPH in 10 mM PBS, pH 7.4;
1 cm).
µM phenolic compounds 1−8, measured by the differential
l
)
the conjugated diene formation (Figure 1b) than did the
corresponding cyclic 8-8 dehydrodimers 7 and 8 (Figure 1c).
Note that (a) compounds 3 and 6, possessing tert-butyl substit-
uents, delayed LDL oxidation to a greater extent than did
probucol; (b) the dimeric derivatives, except for compound 8,
are more potent antioxidants than the corresponding monomeric
4-hydroxycinnamic derivatives; (c) compound 8 had a biphasic
effect. First, it acted as a moderate antioxidant increasing the
lag time, and then it behaved as a prooxidant inducing the
formation of conjugated diene (Figure 1c). Similarly, previous
studies have shown that the syringic acid (4-hydroxybenzoic
compound bearing the 2,6-dimethoxyphenol framework) also
induces LDL oxidation, in the copper-mediated test (26).
The protective effects of the different monomeric and dimeric
4-hydroxycinnamic ethyl ester derivatives 1-8 were also
determined by the TBARS assay. According to the above data,
the concentration of oxidative products reached a steady state
after 250 min of incubation. The IC50 values represent the
concentration of phenol derivative which leads to 50% decrease
of the amount of TBARS produced at this time-point. As shown
in Table 2, the order of effectiveness within the series studied
was as follows: probucol > 3 ≈ 5 > 6 > 8 > 2 > 1 ≡ 7 > 4.
Although the TBARS method was aspecific (it measures a
multistep process and, if malonaldehyde is the purported product
(53), it would only be formed from fatty acids having three or
more double bonds), in our case its use appeared necessary to
compare the present results to those previously published (40).
Monomeric Compounds 1-3. Stepwise addition of Cu(NO3)2
to monomeric compound 3 induced a decrease in the UV-vis
absorption band shown by the anionic form at 437 nm, while
new bands developed at 293 and 314 nm, respectively (Figure
2). After addition of 1.1 equiv of Cu2+ ion, the initially yellow
solution became colorless and the spectral evolution reached a
steady state. Spectral changes showed two isosbestic points at

1902 J. Agric. Food Chem., Vol. 54, No. 5, 2006
Neudo¨rffer et al.
-4
Figure 2. UV−vis absorption spectra of a 10 M solution of compound
3, in acetonitrile, in the presence of 1 equiv of tetramethylammonium
hydroxide; spectrophotometric titration with Cu(NO3)2, up to 1.1 equiv.
+
Each spectrum was recorded after addition of 0.1 equiv of Cu2 (cell
thickness l
) 0.1 cm).
Scheme 1. Dimerization Reaction of 3,5-di-tert-Butyl-4-hydroxy
-
Cinnamate 3
-
Figure 3. Titration of compound 2 (10 4 M), in acetonitrile, in the presence
+
of 1 equiv of tetramethylammonium hydroxide, with Cu2 (cell thickness
+
l
)
0.1 cm). (a) [Cu2
]
)
0, 0.10, 0.20, 0.30, 0.40, 0.44, 0.48, 0.52, 0.56,
-
4
+
0.6, and 0.70
×
×
10 M; (b) [Cu2
]
)
0.8, 0.9, 1.0, 1.1, 1.2, 1.3, and 1.4
-
4
10 M.
252 and 354 nm indicating that a simple equilibrium between
two species was operating. The two new bands observed at 293
and 314 nm can be assigned to the bisquinonemethide 9, by
comparison with the UV-vis absorption spectrum of an
authentic sample (39). Furthermore, the formation of 9 was
confirmed after evaporation of the solvent (acetonitrile) and
1
analysis of the isolated solid by H NMR, which identified it
as the bisquinonemethide 9. This result indicates that compound
3 did not form a chelate but acted as a reducing agent toward
cupric ions, generating cuprous ions together with the bisquino-
nemethide 9 through the dimerization of the phenoxyl radical
3• (Scheme 1).
-
Figure 4. Titration of compound 6 (10 4 M), in acetonitrile, in the presence
2+
of 2 equiv of tetramethylammonium hydroxide, with Cu
(cell thickness
+
-4
l
)
0.1 cm). [Cu2
]
)
0, 0.1, 0.2, 0.3, 0.4, 0.5, and 0.6
×
10 M.
Comparatively, the spectral changes observed during the
addition of Cu(NO3)2 to a solution of compound 2 were more
complex: up to 0.7 equiv of copper, interaction of Cu2+ ions
with the phenolate anion 2- induced hypochromic and hypso-
chromic shifts of the initial band, from 436 to 393 nm (Figure
3a). Further addition of Cu2+ up to 1.3 equiv resulted in the
observation of four new bands (or shoulders) at 316, 333, 472,
and 506 nm, respectively. The lack of well-defined isosbestic
points suggested that concomitant reactions take place (Figure
3b). Accordingly, a preparative scale experiment allowed the
isolation of the 8-8 noncyclic dehydrodimer 5 as the major
product (55%), the 8-8 cyclic dehydrodimer 8 (13%) as the
minor product, together with 8% of the starting material.
Compound 1 behaved similarly, but the isolation of the resulting
compounds was not attempted as the oxidation of compound 1
simultaneously generated five coupling products (8-8 noncyclic,
8-8 cyclic, 8-5, 5-5, and 8-O-4 dehydrodimers) (39).
cupric ions, generating cuprous ions together with the dehy-
drodimer oxidized forms.
Dehydrodimers 4-8. A distinct behavior was observed with
8-8 linear dehydrodimer 6, although spectral changes recorded
in the course of titration were close to those found for the parent
monomer 3 (vide supra) (Figure 4). A steady state was reached
after addition of only 0.6 equiv of cupric ions, while a new
band increased at 322 nm. This spectral evolution, characterized
by the presence of three isosbestic points at 245, 263, and 358
nm, indicated that the stoichiometry of the reaction remained
unchanged and that no secondary reaction occurred during this
time. However, it was not possible to isolate an oxidized product,
as extraction only regenerated the starting material in 79% yield.
These results indicated that the Cu2+ ion was not reduced but
likely bridged by the 8-8 dehydrodimer 6 ligand. The resulting
chelate would be expected to decompose in water during the
workup. At this point, to substantiate the formation of the copper
complex we attempted to model the spectral changes with the
Finally, as previously reported in the case of ferulic and
sinapic acid (6), monomeric compounds 1-3 did not act as
chelators but, in contrast, revealed a reducing activity toward

Protective Effects of 4-Hydroxycinnamic Ethyl Ester Derivatives
J. Agric. Food Chem., Vol. 54, No. 5, 2006 1903
Figure 5. LETAGROP plot for the determination, in acetonitrile, of the
+
stoichiometry and association constants of compound 6 with Cu2 from
UV titration between 304 and 488 nm (24 wavelengths and 7 different
salt concentrations: geometrical symbols, experimental data; solid lines,
calculated spectra; Ri, weighted residuals distribution).
+
Table 3. Binding Constant Values for Dehydrodimers 4
−
8 and Cu2
-
Figure 6. Titration of compound 5 (10 4 M), in acetonitrile, in the presence
Ions
of 2 equiv of tetramethylammonium hydroxide, with Cu2 (cell thickness
+
compounds
log K1:1
log K1:2
log K2:1
log â2
+
-4
l
)
0.1 cm). (a) [Cu2
]
)
0.0, 0.1, 0.2, 0.3, 0.4, 0.5, and 0.6
) 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, and 1.7 × 10
×
10 M;
(b) [Cu2
]
4
5
6
7
8
5.65
5.81
4.65
5.04
8.63
10.30
10.85
8.63
11.76
10.55
+
-4
M.
6.72
5.81
5.04
4.74
whatever the compound considered within the 8-8 noncyclic
dehydrodimer series, the chelate formed was a 2:1 ligand-to-
copper complex stoichiometry.
help of the known LETAGROP-SPEFO method (51, 52). This
program calculates, for a given chemical scheme (see eqs
hereafter), the global binding constant â2 ) K1K2 by iterative
comparison of calculated data with experimental data, searching
for the global minimum of the error function. Equations that
do not fit the data are rejected. As shown in Figure 5, a very
good agreement between experimental and theoretical spectra
was obtained for the formation of a 2:1 dehydrodimer-copper
chelate, with a global binding constant log â2 ) 8.63 (Table
3).
In the case of 8-8 cyclic dehydrodimers 7 and 8, spectral
shifts were very close to those described above. The spectral
evolution involved two steps, and more than 75% of the starting
material was recovered at the end of the reaction. However,
the stoichiometry of the chelates formed was different: the
curves were correctly fitted by considering the generation of
1:1 and 1:2 ligand-to-copper complexes simultaneously.
L + Cu2+ / LCu2+
LCu2+ + Cu2+ / L(Cu2+)2
(L ≡ compound 7 or 8)
K1:1
K1:2
K1:1K1:2 ) â2
L + Cu2+ / LCu2+
(L ≡ compound 6)
K1:1
Concluding Remarks. In this work, we have provided new
insight into the antioxidant capacity of a series of 4-hydroxy-
cinnamic ethyl ester and related dehydrodimers. Taking into
account both the ability of the compounds to inhibit the metal-
dependent and metal-independent oxidation of LDL, the parti-
tioning properties together with the chelating properties, several
conclusions could be drawn:
In the AAPH-induced LDL oxidation, the monomeric com-
pounds, except for highly lipophilic derivative 3, exhibited the
lowest radical scavenging activities of the series studied. In
agreement with previous studies concerning dietary compounds
or eugenol compounds (6, 44), they were not endowed with
copper chelating properties, but, in contrast, they reduced cupric
ions to cuprous ions, affording the corresponding dehydrodimers
compounds 4-8, for which they could be considered as
precursors.
LCu2+ + L / L2Cu2+
K2:1 K1:1K2:1 ) â2
The interaction of the Cu2+ ion with compounds 4 and 5 was
studied under the same experimental conditions. Spectral
changes obtained in the case of compound 5 markedly differed
from those recorded with compound 6. Two steps could be
distinguished in the course of the titration. From 0 to 0.6 equiv,
the initial band at 435 nm, characteristic of the anionic form
5-, decreased. Simultaneously, a new band around 365 nm
increased (Figure 6a). Further addition of Cu2+ ion up to 1.7
equiv induced the disappearance of the latter in favor of two
new bands at 473 and 506 nm (Figure 6b), and a steady state
was obtained up to 1.7 equiv of Cu 2+ ion. After extraction, we
recovered the starting material in 77% yield as the sole
compound. LETAGROP-SPEFO calculations confirmed the
simultaneous formation of two types of chelates, a 2:1 and a
1:1 ligand-to-copper chelate. The values of their binding constant
are reported in Table 3. Compound 4 behaved similarly. Thus,
In the subseries of dehydrodimers, two cases had to be
considered: the cyclic dihydronaphthol dehydrodimers 7 and
8 were found to be low to moderate inhibitors of both metal-

1904 J. Agric. Food Chem., Vol. 54, No. 5, 2006
Neudo¨rffer et al.
(6) Briante, R.; Febbraio, F.; Nucci, R. Antioxidant properties of
low molecular weight phenols present in the mediterranean diet.
J. Agric. Food Chem. 2003, 51, 6975-6981.
independent and metal-dependent oxidation of LDL. These
results solved the problem of the deviating effects observed
earlier (40) between the TEAC assay and the copper-mediated
oxidation of LDL assay, thus demonstrating that the observed
difference was essentially due to the change in the phase (strictly
aqueous or not), rather than in the oxidation mediator (cupric
ion or stable radical). Conversely, although the cyclic dehy-
drodimers 7 and 8 behaved as low and moderate radical
scavengers respectively, they were found to be effective metal
chelators toward cupric ions, affording up to 1:2 ligand-to-
copper complexes. The comparison between the powerful
chelating properties with the moderate activity found in the
copper-mediated LDL oxidation test, which evaluated the
exhaustive antioxidant activity including both radical scavenging
and metal chelation, indicated that the contribution of metal
chelation would not be the major process in the antioxidant
activity measured in this assay.
This assumption was substantiated when considering the
related noncyclic diphenol dehydrodimers 5 and 6. As suggested
by preliminary results concerning the 8-8 noncyclic diferulic
acid (27), these compounds proved to be the most potent
antioxidants, both in the metal-independent and metal-dependent
oxidation of LDL assays. Furthermore, they chelated cupric ion
with a lower efficiency than that of the cyclic dehydrodimer
counterparts 7 and 8, since they only gave complexes with a
2:1 ligand-to-copper stoichiometry. It is important to consider
that these noncyclized diphenol dehydrodimers were mainly
present in the diet as ester-linked compounds in the insoluble
fraction of the cell walls. Consequently, at the opposite of their
parent monomers, they cannot be released into the intestine after
consumption of high-bran cereal, and then are not detected in
substantial amount in human plasma (7, 9, 28). So, they are
unable to directly contribute to some of the beneficial effects
of a phenolic-rich diet, though they could be spontaneously
produced in lipidic phase from their corresponding parent
monomers.
(7) Andreasen, M. F.; Kroon, P. A.; Williamson, G.; Garcia-Conesa,
M.-T. Intestinal release and uptake of phenolic antioxidant
diferulic acids. Free Radical Biol. Med. 2001, 31, 304-314.
(8) Rechner, A. R.; Kuhnle, G.; Bremner, P.; Hubbard, G. P.; Moore,
K. P.; Rice-Evans, C. A. The metabolic fate of dietary polyphe-
nols in humans. Free Radical Biol. Med. 2002, 33, 220-235.
(9) Kern, S. M.; Bennett, R. N.; Mellon, F. A.; Kroon, P. A.; Garcia-
Conesa, M.-T. Absorption of hydroxycinnamates in humans after
high-bran cereal consumption. J. Agric. Food Chem. 2003, 51,
6050-6055.
(10) Rechner, A. R.; Smith, M. A.; Kuhnle, G.; Gibson, G. R.;
Debnam, E. S.; Srai, S. K. S.; Moore, K. P.; Rice-Evans, C. A.
Colonic metabolism of dietary polyphenols: influence of struc-
ture on microbial fermentation products. Free Radical Biol. Med.
2004, 36, 212-225.
(11) Ralph, J.; Quideau, S.; Grabber, J. H.; Hatfield, R. D. Identifica-
tion and synthesis of new ferulic acid dehydrodimers present in
grass cell walls. J. Chem. Soc., Perkin Trans. 1 1994, 3485-
3498.
(12) Saulnier, L.; Thibault, J. F. Ferulic acid and diferulic acids as
components of sugar-beet pectins and maize bran heteroxylans.
J. Sci. Food Agric. 1999, 79, 396-402.
(13) Ward, G.; Hadar, Y.; Bilkis, I.; Konstantinovsky, L.; Dosoretz,
C. G. Initial steps of ferulic acid polymerization by lignin
peroxidase. J. Biol. Chem. 2001, 276, 18734-18741.
(14) Bunzel, M.; Funk, C.; Steinhart, H. Semipreparative isolation
of dehydrodiferulic and dehydrotriferulic acids as standard
substances from maize bran. J. Sep. Sci. 2004, 27, 1080-1086.
(15) Grabber, J. H.; Hatfield, R. D.; Ralph, J. Diferulate cross-links
impede the enzymatic degradation of nonlignified maize walls.
J. Sci. Food Agric. 1998, 77, 193-200.
(16) Marita, J. M.; Vermerris, W.; Ralph, J.; Hatfield, R. D. Variations
in the cell wall composition of maize brown mibrid mutants. J.
Agric. Food Chem. 2003, 51, 1313-1321.
(17) Grabber, J. H.; Ralph, J.; Hatfield, R. D. Cross-linking of maize
walls by ferulate dimerization and incorporation into lignin. J.
Agric. Food Chem. 2000, 48, 6106-6113.
Finally, in the series studied, it could be concluded that the
noncyclized diphenol dehydrodimers behaved as the most
attractive antioxidants through the main antioxidant mechanism
involving radical scavenging.
(18) Grabber, J. H.; Ralph, J.; Hatfield, R. D. Model studies of
ferulate-coniferyl alcohol cross-product formation in primary
maize walls: implications for lignification in grasses. J. Agric.
Food Chem. 2002, 50, 6008-6016.
(19) Bunzel, M.; Ralph, J.; Lu, F.; Hatfield, R. D.; Steinhart, H.
Lignins and ferulate-coniferyl alcohol cross-coupling products
in cereal grains. Grasses. J. Agric. Food Chem. 2004, 52, 6496-
6502.
(20) Renger, A.; Steinhart, H. Ferulic acid dehydrodimers as structural
elements in cereal dietary fibre. Eur. Food Res. Technol. 2000,
211, 422-428.
(21) Bunzel, M.; Ralph, J.; Marita, J. M.; Hatfield, R. D.; Steinhart,
H. Diferulates as structural components in soluble and insoluble
cereal dietary fibre. J. Sci. Food Agric. 2001, 81, 653-660.
(22) Bunzel, M.; Ralph, J.; Kim, H.; Lu, F.; Ralph, S. A.; Marita, J.
M.; Hatfield, R. D.; Steinhart, H. Sinapate dehydrodimers and
sinapate-ferulate heterodimers in cereal dietary fiber. J. Agric.
Food Chem. 2003, 51, 1427-1434.
(23) Parker, C. C.; Parker, M. L.; Smith, A. C.; Waldron, K. W.
Thermal stability in chinese water chesnut may be dependent
on 8,8′-diferulic acid (aryltetralyn form). J. Agric. Food Chem.
2003, 51, 2034-2039.
ABBREVIATIONS USED
AAPH, 2,2′-azo-bis-(2-amidinopropane) dihydrochloride;
ABTS+, 2,2′-azino-bis-(3-ethylbenzothiazoline-6-sulfonic acid)
diammonium salt; BHT, butylhydroxytoluene; LDL, low-density
lipoprotein; TEAC, trolox equivalent antioxidant capacity;
TBARS, thiobarbituric acid-reactive substances.
LITERATURE CITED
(1) Berliner, J. A.; Heinecke, J. W. The role of oxidized lipoproteins
in atherogenesis. Free Radical Biol. Med. 1996, 20, 707-727.
(2) Robbins, R. J. Phenolic acids in foods: an overview of analytical
methodology. J. Agric. Food Chem. 2003, 51, 2866-2887.
(3) Finley, J. W. Proposed criteria for assessing the efficacy of cancer
reduction by plants foods enriched in carotenoids, glucosinolates,
polyphenols and selenocompounds. Ann. Bot. (London) 2005,
95, 1075-1096.
(24) Funk, C.; Ralph, J.; Steinhart, H.; Bunzel, M. Isolation and
structural characterisation of 8-O-4/8-O-4- and 8-8/8-O-4-coupled
dehydrotriferulic acids from maize. Phytochemistry 2005, 363-
371.
(4) Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Structure-
antioxidant activity relationships of flavonoids and phenolic
acids. Free Radical Biol. Med. 1996, 20, 933-956.
(25) Meyer, A. S.; Donovan, J. L.; Pearson D. A.; Waterhouse, A.
L.; Frankel, E. N. Fruit hydroxycinnamic acids inhibit human
low-density lipoprotein oxidation in vitro. J. Agric. Food Chem.
1998, 46, 1783-1787.
(5) Andreasen, M. F.; Christensen, L. P.; Meyer, A. S.; Hansen, A° .
Content of phenolic acids and ferulic acid dehydrodimers in 17
rye (Secale cereale L.) varieties. J. Agric. Food Chem. 2000,
48, 2837-2842.

Protective Effects of 4-Hydroxycinnamic Ethyl Ester Derivatives
J. Agric. Food Chem., Vol. 54, No. 5, 2006 1905
(26) Natella, F.; Nardini, M.; Di Felice, M.; Scaccini, C. Benzoic
and cinnamic acid derivatives as antioxidants: structure-activity
relation. J. Agric. Food Chem. 1999, 47, 1453-1459.
(27) Garcia-Conesa, M. T.; Wilson, P. D.; Plumb, G. W.; Ralph, J.;
Williamson, G. Antioxidant properties of 4,4′-dihydroxy-3, 3′-
dimethoxy-â,â′-bicinnamic acid (8-8-diferulic acid, noncyclic
form). J. Sci. Food Agric. 1999, 79, 379-384.
(28) Andreasen, M. F.; Landbo, A.-K.; Christensen, L. P.; Hansen,
Å.; Meyer, A. S. Antioxidant effects of phenolic rye (Secale
cereale L.) extracts, monomeric hydroxycinnamates, and ferulic
acid dehydrodimers on human low-density lipoproteins. J. Agric.
Food Chem. 2001, 49, 4090-4096.
(29) Lee, W. S.; Baek, Y.-I.; Kim, J.-R.; Cho, K.-H.; Sok, D.-E.;
Jeong, T.-S. Antioxidant activities of a new lignan and a
neolignan from Saururus chinensis. Bioorg. Med. Chem. Lett.
2004, 14, 5623-5628.
(30) Patel, R. P.; Darley-Usmar, V. M. Molecular mechanisms of the
copper-dependent oxidation of low-density lipoprotein. Free
Radical Res. 1999, 30, 1-9.
(31) Burkitt, M. J. A critical overview of the chemistry of copper-
dependent low-density lipoprotein oxidation: roles of lipid
hydroperoxides, R-tocopherol, thiols and ceruloplasmin. Arch.
Biochem. Biophys. 2001, 394, 117-135.
(32) Van Acker, S. A. B. E.; Van den Berg, D.-J.; Tromp, M. N. J.
L.; Griffioen, D. H.; Van Bennekom, W. P.; Van der Vidj, W.
J. F.; Bast, A. Structural aspects of antioxidant activity of
flavonoids. Free Radical Biol. Med. 1996, 20, 331-342.
(33) Masuda, T.; Toi, Y.; Bando, H.; Maekawa, T.; Takeda, Y.;
Yamaguchi, H. Structural identification of new curcumin dimers
and their contribution to the antioxidant mechanism of curcumin.
J. Agric. Food Chem. 2002, 50, 2524-2530.
(34) Sundaryono, A.; Nourmamode, A.; Grardat, C.; Fritsch, A.;
Castellan, A. Synthesis and complexation properties of two new
curcuminoid molecules bearing a diphenylmethane linkage. J.
Mol. Struct. 2003, 649, 177-190.
(35) Priyadarsini, K. I.; Maity, D. K.; Naik, G. H.; Kumar, M. S.;
Unnikrishnan, M. K.; Satav, J. G.; Mohan, H. Role of phenolic
O-H and methylene hydrogen on the free radical reactions and
antioxidant activity of curcumin. Free Radical Biol. Med. 2003,
35, 475-484.
(36) Largeron, M.; Lockhart, B.; Pfeiffer, B.; Fleury, M.-B. Synthesis
and in vitro evaluation of new 8-amino-1,4-benzoxazine deriva-
tives as neuroprotective antioxidants. J. Med. Chem. 1999, 42,
5043-5052.
(37) Larget, R.; Lockhart, B.; Pfeiffer, B.; Neudo¨rffer, A.; Fleury,
M.-B.; Largeron, M. Synthesis of novel orthoalkylaminophenol
derivatives as potent neuroprotective agents in vitro. Bioorg. Med.
Chem. Lett. 1999, 9, 2929-2934.
(38) Blattes, E.; Lockhart, B.; Lestage, P.; Schwendimann, L.;
Gressens, P.; Fleury, M.-B.; Largeron, M. Novel 2-alkylamino-
1,4-benzoxazine derivatives as potent neuroprotective agents:
structure-activity relationship studies. J. Med. Chem. 2005, 48,
1282-1286.
(39) Neudo¨rffer, A.; Deguin, B.; Hamel, C.; Fleury, M.-B.; Largeron,
M. Electrochemical oxidative coupling of 4-hydroxycinnamic
esters derivatives: a convenient methodology for the biomimetic
synthesis of lignin precursors. Collect. Czech. Chem. Commun.
2003, 68, 1515-1530.
(40) Neudo¨rffer, A.; Bonnefont-Rousselot, D.; Legrand, A.; Fleury,
M.-B.; Largeron, M. 4-Hydroxycinnamic ethyl ester derivatives
and related dehydrodimers: relationship between oxidation
potential and protective effects against oxidation of low-density
lipoproteins. J. Agric. Food Chem. 2004, 52, 2084-2091.
(41) Brown, J. E.; Khodr, H.; Hider, R. C.; Rice-Evans, C. A.
Structural dependence of flavonoid interactions with Cu2+ ions:
implications for their antioxidant properties. Biochem. J. 1998,
330, 1173-1178.
(42) Miller, N. J.; Castelluccio, C.; Tijburg, L.; Rice-Evans, C. The
antioxidant properties of theaflavins and their gallate esterss
radical scavengers or metal chelators? FEBS Lett. 1996, 392,
40-44.
(43) Cren-Olive´, C.; Teissier, E.; Duriez, P.; Rolando, C. Effect of
catechin o-methylated metabolites and analogues on human LDL
oxidation. Free Radical Biol. Med. 2003, 34, 850-855.
(44) Ito, M.; Murakami, K.; Yoshino, M. Antioxidant action of
eugenol compounds: role of metal ion in the inhibition of lipid
peroxidation. Food Chem. Toxicol. 2005, 43, 461-466.
(45) Havel, R. J.; Eder, H. A.; Bragdon, J. H. The distribution and
chemical composition of ultracentrifugally separated lipoproteins
in human serum. J. Clin. InVest. 1955, 34, 1345-1353.
(46) Watanabe, N.; Kamei, S.; Ohkubo, A.; Yamanaka, M.; Ohsawa,
S.; Makino, K.; Tokuda, K. Urinary protein as measured with a
pyrogallol red-molybdate complex, manually and in a Hitachi
726 automated analyzer. Clin. Chem. 1986, 32, 1551-1554.
(47) Pryor, W. A.; Castle, L. Chemical methods for the detection of
lipid hydroperoxides. Methods Enzymol. 1984, 105, 293-299.
(48) Kontush, A.; Chantepie, S.; Chapman, M. J. Small, dense HDL
particles exert potent protection of atherogenic LDL against
oxidative stress. Atheroscler. Thromb. Vasc. Biol. 2003, 23,
1881-1888.
(49) Yagi, K. Lipid peroxides and human diseases. Chem. Phys. Lipids
1987, 45, 337-351.
(50) Ghose, A. K.; Crippen, G. M. Atomic physicochemical param-
eters for three-dimensional-structure-directed quantitative struc-
ture-activity relationships. 2. Modeling dispersive and hydro-
phobic interactions. J. Chem. Inf. Comput. Sci. 1987, 27, 21-
35.
(51) (a) Sillen, L. G.; Warnquist, B. High-speed computers as a
supplement to graphical methods. 6. A strategy for two-level
LETAGROP adjustment of common and “group” parameters.
Some features that avoid divergence. Ark. Kemi 1968, 31, 315-
339 see also parts 1-4 cited in the references, for a general and
theoretical presentation. (b) Sillen, L. G.; Warnquist, B. High-
speed computers as a supplement to graphical methods. 7. Model
selection and rejection with LETAGROP. Elimination of species
with negative or “insignificant” equilibrium constants. Ark. Kemi
1968, 31, 341-351 and references therein.
(52) Havel, J. Haltafal-Spefo Program; Mazaryle University: Brno,
Moravia, Czech Republic.
(53) Del Rio, D.; Stewart, A. J.; Pellegrini, N. A review of recent
studies on malondialdehyde as toxic molecule and biological
marker of oxidative stress. Nutr. Metab. CardioVasc. Dis. 2005,
15, 316-328.
Received for review November 23, 2005. Accepted January 4, 2006.
JF052923P