Antioxidant activity of (all-E)-lycopene and synthetic apo-lycopenoids in a chemical model of oxidative stress in the gastro-intestinal tract

Article abstract of DOI:10.1039/c1nj20437h

Lycopene, the main pigment of tomato, is known to have protective effects on health. Its metabolites could also be involved in these effects. Potentially bioactive lycopene metabolites namely apo-10′-lycopenol, apo-10′-lycopenal, apo-14′-lycopenol, apo-14′-lycopenal, and apo-11-lycopenoids with alcohol, carboxylic acid, aldehyde and ethyl ester terminal groups were obtained by organic synthesis as (all-E) stereoisomers using HWE condensation reactions. The ability of (all-E)-lycopene and the synthesized apo-lycopenoids to inhibit lipid peroxidation was tested in a chemical model of postprandial oxidative stress in the gastric compartment. Oxidative stress was generated by metmyoglobin, the main form of dietary iron (from red meat), which is able to catalyse the peroxidation of linoleic acid under mildly acidic conditions. In this model, apo-6′-lycopenal and apo-8′-lycopenal were better inhibitors of lipid peroxidation than (all-E)-lycopene itself. For the apo-lycopenoids, a long unsaturated chain and a terminal carboxylic acid group both favour the antioxidant activity. The short-chain apo-lycopenoic acid, apo-14′-lycopenoic acid, was shown to behave like a hydrophilic antioxidant, i.e. by reducing hypervalent iron forms of metmyoglobin. Thus, besides the polyene chain length, the terminal group of the apo-lycopenoids is expected to deeply influence the lipophilic/hydrophilic balance of the molecule (and consequently its distribution between the aqueous and lipid phases) and its affinity for the heme cavity. It can thus be concluded that the polyene chain length and terminal group are two important parameters modulating the mechanism by which apo-lycopenoids express their antioxidant activity. The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012.

Full text of DOI:10.1039/c1nj20437h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
NJC
Cite this: New J. Chem., 2012, 36, 575–587
www.rsc.org/njc
PAPER
Antioxidant activity of (all-E)-lycopene and synthetic apo-lycopenoids in
a chemical model of oxidative stress in the gastro-intestinal tractw
Pascale Goupy,ab Eric Reynaud,ab Olivier Danglesab and Catherine Caris-Veyrat*ab
Received (in Montpellier, France) 23rd May 2011, Accepted 30th October 2011
DOI: 10.1039/c1nj20437h
Lycopene, the main pigment of tomato, is known to have protective effects on health. Its
metabolites could also be involved in these effects. Potentially bioactive lycopene metabolites
namely apo-100-lycopenol, apo-100-lycopenal, apo-140-lycopenol, apo-140-lycopenal, and
apo-11-lycopenoids with alcohol, carboxylic acid, aldehyde and ethyl ester terminal groups were
obtained by organic synthesis as (all-E) stereoisomers using HWE condensation reactions. The
ability of (all-E)-lycopene and the synthesized apo-lycopenoids to inhibit lipid peroxidation was
tested in a chemical model of postprandial oxidative stress in the gastric compartment. Oxidative
stress was generated by metmyoglobin, the main form of dietary iron (from red meat), which is
able to catalyse the peroxidation of linoleic acid under mildly acidic conditions. In this model,
apo-60-lycopenal and apo-80-lycopenal were better inhibitors of lipid peroxidation than
(all-E)-lycopene itself. For the apo-lycopenoids, a long unsaturated chain and a terminal
carboxylic acid group both favour the antioxidant activity. The short-chain apo-lycopenoic acid,
apo-140-lycopenoic acid, was shown to behave like a hydrophilic antioxidant, i.e. by reducing
hypervalent iron forms of metmyoglobin. Thus, besides the polyene chain length, the terminal
group of the apo-lycopenoids is expected to deeply influence the lipophilic/hydrophilic balance of
the molecule (and consequently its distribution between the aqueous and lipid phases) and its
affinity for the heme cavity. It can thus be concluded that the polyene chain length and terminal
group are two important parameters modulating the mechanism by which apo-lycopenoids
express their antioxidant activity.
not yet been clearly shown and is still a matter of debate.10
A major problem is the low concentration of carotenoids in
Introduction
Dietary intake of tomatoes and tomato-derived products has
been shown to be associated with a lower risk of chronic
diseases such as cancer or cardiovascular disease.1–4 Lycopene
plasma, at most of the order of 1 mM, or in organs (compared
with recognized antioxidants like a-tocopherol in prostate)11 as
a result of their limited absorption through the intestinal
(Table 1), the main carotenoid pigment of tomato, could be
barrier. However, in humans, it has been demonstrated that
involved in these protective effects, however mechanisms
underlying its biological activity in vivo are not fully eluci-
lycopene concentrations in the stomach could reach 50 mM only
20 min after ingestion of a meal containing 33.6 g of tomato
puree (10 mg of lycopene).12 The gastric concentration of lycopene
dated. Lycopene metabolites are supposed to participate in the
biological effects attributed to lycopene.6
The antioxidant activity of carotenoids is one possible
was still 40 mM and 25 mM after 1 h and 2 h, respectively. Thus,
it has been hypothesized that the antioxidant action of poorly bio-
mechanism underlying the beneficial effects of carotenoids
available micronutrients like carotenoids and polyphenols could
in vivo.7,8 b-Carotene is indeed an essential antioxidant in green
plants, protecting them from singlet dioxygen produced in
excess of light. In vitro, lycopene has been shown to be the
most efficient carotenoid for the quenching of singlet dioxygen.9
However, the antioxidant activity of carotenoids in humans has
be relevant mainly in the gastro-intestinal tract.13,14 Indeed, lipid
peroxidation induced by dietary iron can be quite fast in the gastric
compartment due to the presence of nonlimiting dioxygen concen-
trations and to the acidic pH, which makes iron ions more redox-
active. Besides the loss of essential fatty acids, the peroxidation of
dietary polyunsaturated lipids has the additional drawback of
producing bioavailable and potentially toxic lipid oxidation pro-
ducts, which are considered possible risk factors in atherogenesis.15
Hence, an experimental model of oxidative stress in the
gastro-intestinal tract may be a biologically relevant model for
mimicking the antioxidant activity of carotenoids in vivo.13,16
a INRA, UMR408 Safety and Quality of Plant Products,
F-84000 Avignon, France. E-mail: catherine.caris@avignon.inra.fr;
Fax: +33 (0)4-32-72-24-92; Tel: +33 (0)4-32-72-24-89
b University of Avignon, UMR408, F-84000 Avignon, France
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c1nj20437h
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 575

View Article Online
Table 1 Chemical structures of (all-E)-lycopene and the apo-lycopenoids investigated
a
Compounds
Short name
Molecular Structure
lmax nm in CHCl3
(All-E)-lycopene (31)c
(all-E)-lyc.
458, ꢀ4ꢀ8ꢀ5, 518, %III/II = 60
345
Apo-11-lycopenal
11-CHO
Apo-11-lycopenoic acid
Apo-11-lycopenol
11-COOH
11-CH2OH
11-COOEt
140-CHO
276, shb 323
ꢀꢀꢀ
sh 276, ꢀ2ꢀ8ꢀ6, sh 297
Ethyl apo-11-lycopenoate
Apo-140-lycopenal
318
422
Apo-140-lycopenoic acid
Apo-140-lycopenol
140-COOH
140-CH2OH
140-COOMe
410
375, ꢀ3ꢀ8ꢀ7, 398, %III/II = 64
411
Methyl apo-140-lycopenoate
Apo-120-lycopenal
120-CHO
454
Apo-100-lycopenal
100-CHO
474
Apo-100-lycopenoic acid (504.4)c
Apo-100-lycopenol (504.3)c
Ethyl apo-100-lycopenoate
100-COOH
100-CH2OH
100-COOEt
80-CHO
454
405, ꢀ4ꢀ2ꢀ7, 457, %III/II = 79
455
Apo-80-lycopenal (491)c
Apo-60-lycopenal (471)c
486
504
60-CHO
a
b
c
Main maximum underlined. Shoulder. Name and/or number from Carotenoids Handbook.5
Lycopene-derived products are believed to play a key role
cannot conclude that apo-lycopenoids are authentic human
metabolites of lycopene.
in biological effects once attributed only to lycopene.6,17
Lycopene metabolites formed by oxidative cleavage of the
hydrocarbon chain (apo-lycopenoids) have been found in
animals fed lycopene-enriched diets, like apo-80-lycopenal
and apo-120-lycopenal in rat liver18 and apo-100-lycopenol in
lung of ferrets possibly issued from a cleavage of (5-Z) and
(13-Z) lycopene isomers by intestinal carotenoid monoxygenase
(CMO2).19 Apo-100-lycopenal could be converted to both
alcohol (apo-100-lycopenol) and acid (apo-100-lycopenoic acid)
in the presence of NADPH and a post-nuclear fraction of ferret
liver. More recently, apo-60-, apo-80-, apo-100-, apo-120-, apo-
140- and apo-150-lycopenals were detected in the plasma
of humans who had consumed tomato juice.20 Interestingly,
these apo-lycopenals have also been found in food products,
especially in processed food (e.g., tomato paste) where they
are more concentrated than in fresh food as a consequence of
plant tissue disruption during processing.20 Hence, so far, one
Several lycopene-derived molecules have been shown to be bio-
logically active in vitro. Acyclo-retinoic acid (apo-150-lycopenoic
acid), the linear analogue of retinoic acid, which is formally
issued from a central cleavage of lycopene, was shown to
increase gap junction communication, although only at high
concentrations.21 Apo-100-lycopenoic acid, a metabolite issued
from an excentric oxidative cleavage of lycopene, has been shown
to inhibit lung carcinogenesis both in vivo and in vitro, via the
activation of the Nrf2 transcription factor and the induction of
phase II detoxifying/antioxidant enzymes.22 Dialdehydes formed
by double oxidative cleavage of the carotenoid chain have been
shown to be more potent than lycopene at activating the
antioxidant response element (ARE) transcription system.23
It was thus shown that lycopene metabolites might induce
phase-II enzymes by the same mechanism as the one
demonstrated for carotenoids24 and even more efficiently.
c
576 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Recently, we demonstrated that apo-140-lycopenoic acid and
apo-100-lycopenoic acid are potential endogenous lycopene
metabolites involved in the activation of hormone receptors
and in the regulation of CMO1 and CMO2 expression.25
While the antioxidant effect of lycopene has been extensively
studied, similar data on lycopene metabolites are scarce. In one
study a decrease in endogenous reactive oxygen species (ROS)
production has been observed after treatment of BEAS-2B cells
with apo-100-lycopenoic acid, suggesting a possible indirect
antioxidant effect.22,26 To our knowledge, direct antioxidant
effects of lycopene metabolites have not been described yet.
The aim of the present study was to study the ability of
(all-E)-lycopene and apo-lycopenoids with different chain lengths
and terminal groups (alcohol, aldehyde, carboxylic acid and
ester) to inhibit lipid peroxidation in a simple chemical model
of postprandial oxidative stress in the gastric compartment.16
Oxidative stress was generated by metmyoglobin, one of the
main forms of dietary iron (red meat), which is able to catalyse
the peroxidation of linoleic acid in mildly acidic conditions. This
model allowed us to demonstrate that hydrophilic and lipophilic
antioxidants display distinct inhibition mechanisms27 and more
recently was used to discriminate the antioxidant action of red
wine pigments with close structures.28
of this condensation involves a cyclic phosphate intermediate
and leads to a new carbon–carbon double bond with a diastereo-
isomeric (E/Z) ratio of 3.5 and a yield of 82%. Silica gel
chromatography did not allow us to efficiently separate the
11-COOEt diastereoisomers, so the crude mixture was reduced
using DIBAH. After purification by silica gel chromatography,
apo-11-lycopenol (11-CH2OH) was obtained as an (E) isomer
(purity 97%, 3% (Z) isomer) with a yield of 16%. A diastereo-
isomeric mixture of 11-CH2OH was used for the preparation of
apo-11-lycopenal (11-CHO) by oxidation of the allylic alcohol.
After purification by silica gel chromatography, 11-CHO was
obtained in 30% yield as an (E) isomer (purity 96%, 4% (Z)
isomer). Finally, apo-11-lycopenoic acid (11-COOH) was
obtained as a pure (E) isomer by mild oxidation of 11-CHO
using Ag2O with a yield of 90%. Saponification of crude
11-COOEt gave a diastereoisomeric mixture of 11-COOH that
could not be purified to obtain the pure (E) isomer.
Apo-140-lycopenoids. Reduction of methyl apo-140-lycopenoate25
(140-COOMe) by DIBAH and subsequent crystallization pro-
vided pure (E) apo-140-lycopenol (140-CH2OH) in 35% yield
(Scheme 2A). Apo-140-lycopenal (140-CHO) was obtained as a
pure (E) isomer by oxidation of 140-CH2OH using MnO2 with
a yield of 44%.
Apo-100-lycopenoids. Apo-100-lycopenol (100-CH2OH) was
obtained by reduction of ethyl apo-100-lycopenoate25 (100-COOEt)
in the presence of an excess of the mild reducing agent DIBAH
(Scheme 2B). After purification, 100-CH2OH was obtained with a
yield of 31% as a pure (E) isomer. Oxidation of 100-CH2OH using
an excess of MnO2 in ethyl acetate provided apo-100-lycopenal
(100-CHO), which was crystallized in a pure (E) isomer with a yield
of 44%.
Results and discussion
Synthesis of apo-11-, apo-140- and apo-100-lycopenoids
We have previously obtained apo-lycopenoids by direct oxidation
of lycopene with KMnO4.29 More recently, we described the
organic synthesis of apo-100-lycopenoic acid (100-COOH) and
apo-140-lycopenoic acid (140-COOH) and their corresponding
methyl or ethyl esters.25 In the present work, a similar strategy
based on the Horner–Wadsworth–Emmons (HWE) reaction
was developed to build the hydrocarbon skeleton of apo-11-
lycopenoids (Table 1). Reduction and subsequent oxidation
reactions were applied to the apo-lycopenoic esters of the 140
and 100 series previously obtained in order to prepare the
corresponding apo-lycopenols, apo-lycopenals and apo-lycopenoic
acids (Table 1). Special attention was paid to obtaining
apo-lycopenoids of (all-E) configuration.
Spectral characteristics of (all-E)-lycopene and apo-lycopenoids
(All-E)-lycopene displays a UV-visible spectrum with a typical
three-band (I–III) fine structure (458, ꢀ4ꢀ8ꢀ5, 518 nm, in CHCl3,
%III/II = 60) (Fig. 1 and Table 1). This fine structure is also
present in the UV-visible spectrum of 100-CH2OH (%III/II = 79)
and the shorter apo-lycopenol 140-CH2OH (%III/II = 64),
whereas it has almost completely disappeared in the shortest
apo-lycopenol 11-CH2OH (Fig. 1A). Thus, the vibrational fine
structure of lycopene is maintained in the longest apo-lycopenols
only. By contrast, the fine structure is strongly weakened in the
UV-visible spectra of apo-lycopenoic acids and esters and has
completely disappeared in the spectra of apo-lycopenals as shown
Apo-11-lycopenoids. Ethyl apo-11-lycopenoate (11-COOEt)
was prepared by elongation of the carbon chain of pseudo-
ionone (PI) upon treatment with one equivalent of triethyl-
phosphonoacetate (HWE reaction) (Scheme 1). The mechanism
Scheme 1 Synthesis of ethyl apo-11-lycopenoate (11-COOEt), apo-11-lycopenol (11-CH2OH), apo-11-lycopenal (11-CHO) and apo-11-lycopenoic
acid (11-COOH). Reagents and conditions: (a) triethylphosphonoacetate, NaH, THF; (b) DIBAH, THF; (c) MnO2, AcOEt; and (d) Ag2O,
NaOH, EtOH.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 577

View Article Online
Scheme 2 (A) Synthesis of apo-140-lycopenol (140-CH2OH) and apo-140-lycopenal (140-CHO). (B) Synthesis of apo-100-lycopenol (100-CH2OH)
and apo-100-lycopenal (100-CHO). Reagents and conditions: (a) DIBAH, THF; (b) MnO2, AcOEt.
for apo-100-lycopenoids (Fig. 1B, results not shown for the other
apo-lycopenoid series). It is known that conjugation of the
polyene chromophore of carotenoids to a carbonyl group induces
a loss of the vibrational fine structure (though in nonpolar
solvents the fine structure may be retained).30 The maximum
absorption wavelength increases in the series 100-CH2OH (405,
ca. 130 nM sꢀ1 (Fig. 2). In the presence of lycopene and the
apo-lycopenoids, the accumulation of CDs proceeds according
to two phases: a lag phase followed by a propagation phase
(Fig. 2). The antioxidant markedly slows down the accumulation
of CDs during the lag phase whose duration increases with the
antioxidant concentration and reflects the period during which
most of the carotenoid is consumed (Fig. 2). Similar kinetic
profiles were observed for all apo-lycopenoids as shown in Fig. 3
for the apo-100- and apo-140-series, except for apo-11-lycopenoids,
which are poor antioxidants in this model (results not shown).
However, significant differences between short-chain and long-
chain apo-lycopenoic acids can be outlined. Indeed, whereas
100-COOH exerts its inhibitory activity mainly by prolonging
the lag phase, 140-COOH partly inhibits the peroxidation by
lowering the rate of the propagation phase (Fig. 3A), which might
indicate a different inhibition mechanism (see below). It is also
remarkable that, at the end of the lag phase, short-chain anti-
oxidants (140 series) are more spared than long-chain analogues,
which typically are totally consumed (Table 2).
427, 457 nm), 100-COOH (454 nm), 100-COOEt (455 nm) and
ꢀꢀꢀ
100-CHO (474 nm). These results reflect the conjugation of the
carbonyl group with the polyene chain and the stronger attractive
mesomeric effect of an aldehyde vs. a carboxylic acid or ester.
Moreover, the difference in lmax values between 100-CHO and
100-CH2OH is higher than that observed between apo-100-
b-carotenal (434 nm) and apo-100-b-carotenol (386, 408,
ꢀꢀꢀ
432 nm)30 in ethanol (47 nm vs. 26 nm). This difference might
be explained by restrictions in conjugation brought about by the
cyclic terminal group of apo-b-carotenoids and by the different
solvents used (chloroform vs. ethanol).
As expected, a clear hypsochromic shift is observed for apo-
lycopenals (Fig. 1C) when the number of conjugated double
bonds decreases with lmax values going from 504 nm for apo-
60-lycopenal to 345 nm for apo-11-lycopenal. Simultaneously,
a hypochromic shift is also observed.
Control experiments without linoleic acid (LH) or metmyo-
globin (MbFeIII) or both were carried out. For each anti-
oxidant (initial concentration = 10 mM, except lycopene, 2 mM),
the percentage of consumption was evaluated at a fixed time
period corresponding to the end of the lag phase of metmyoglobin-
induced lipid peroxidation. A non-negligible consumption of
(all-E)-lycopene and apo-lycopenoids (between 7 to 17%)
was observed in the phosphate buffer containing 2 mM Tween
20 at pH 5.8 and 37 1C under gentle stirring (Table 2). In such
conditions, antioxidant degradation could be due to photo-
oxidation and/or autoxidation initiated by metal ion traces
present in buffer and peroxides contaminating Tween 20.31
Linoleic acid or MbFeIII alone slightly accelerates apo-lycopenoid
consumption. This could reflect the slow reduction of MbFeIII
into MbFeII (myoglobin) by the antioxidants and the ‘spontaneous’
co-oxidation of linoleic acid and antioxidants. By contrast,
the simultaneous presence of LH and MbFeIII dramatically
accelerates apo-lycopenoid consumption, thus indicating that
this process essentially reflects the inhibition by apo-lycopenoid
of the metmyoglobin-induced peroxidation of linoleic acid. For
apo-lycopenals, the longer the chain length, the longer the lag
Inhibition of linoleic acid peroxidation
(All-E)-lycopene and apo-lycopenoids were assayed for their
ability to inhibit the peroxidation of linoleic acid induced by
dietary heme iron (metmyoglobin) in a mildly acidic micellar
solution. The selected pH of 5.8 is representative of the gastric
compartment during the early step of digestion12 when oxida-
tive processes are most likely to take place. The accumulation
of conjugated dienes (CDs, mainly lipid hydroperoxides16)
as primary products of lipid oxidation was monitored by
UV-visible spectroscopy at 234 nm in the absence (uninhibited
peroxidation) or presence of (all-E)-lycopene or apo-lycopenoids
at different concentrations. Simultaneously, the consumption of
the antioxidants displaying a convenient UV-visible absorption
(l Z 400 nm) was also monitored (see for example results for
80-CHO in Fig. 2). The A(234 nm) vs. time curve of uninhibited
peroxidation is quasi-linear with a peroxidation rate (R0p) of
c
578 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 2 Accumulation of conjugated dienes (full lines, detection at
234 nm) and simultaneous consumption of the antioxidant (dotted lines,
detection at 485 nm) during the metmyoglobin-induced peroxidation of
linoleic acid in the presence of 80-CHO. 80-CHO concentrations are:
0 (0), 2 (1), 4 (2), 6 (3), 8 (4) and 10 mM (5). Linoleic acid (0.7 mM),
metmyoglobin (100 nM), pH 5.8 phosphate buffer containing 2 mM
Tween 20, 37 1C.
higher antioxidant activity (Fig. 4). For each concentration
tested, differences between antioxidants were significant,
except for 140-CHO and 120-CHO. Moreover, due a slight
curvature in the T vs. antioxidant concentration (C), T can be
satisfactorily fitted against a quadratic function of C (not
shown): T/T0 = 1 + aC + bC2 to extract a IC50 value for
each antioxidant. IC50 parameter is defined as the antioxidant
concentration leading to T = 2T0: IC50 = (O(a2 + 4b) ꢀ a)/2b.28
The lower the IC50 value, the more effective the antioxidant. It is
noteworthy that two apo-lycopenoids are more antioxidant than
lycopene itself: 60-CHO followed by 80-CHO (Table 3). High
IC50 values of apo-11-lycopenoids confirm their very low anti-
oxidant activity. More importantly, comparisons based on DT
(Fig. 4) or on IC50 both lead to the same conclusions: apo-
lycopenals with longer chains are more antioxidant than shorter
ones. Although limited to a smaller number of antioxidants, the
same trend is observed for acids, alcohols and esters. For a fixed
chain length (100 and 140 series), the influence of the terminal
group on the antioxidant activity is: alcohol o ester r aldehyde o
carboxylic acid. It must be outlined that apo-lycopenoic acids
are amphiphilic antioxidants with an anionic carboxylate end.
Their likely location at the interface of micelles seems to improve
their efficiency at inhibiting the peroxidation mechanism.
Fig. 1 UV-visible spectra of (all-E)-lycopene and apo-lycopenoids in
CHCl3. Concentration was 10 mM except for 11-CH2OH, 25 mM.
(A) (All-E)-lycopene (dotted line) and apo-lycopenols: 100-CH2OH
(full line); 140-CH2OH (circles); 11-CH2OH (full circles). (B) Apo-100-
lycopenoids series: 100-COOEt (full circles); 100-CHO (triangles);
100-COOH (full line). (C) Apo-lycopenals: 60-CHO (circles); 80-CHO
(full circles); 100-CHO (triangles); 120-CHO (full triangles); 140-CHO
(squares); 11-CHO (full squares).
phase, the longer the half-life of the antioxidant and the higher
the extent of antioxidant consumption at the end of the lag
phase. Within both 100 and 140 series, the lag phase of
peroxidation and half-life of the antioxidant both increase as
follows: apo-lycopenol o apo-lycopenoate o apo-lycopenal o
apo-lycopenoic acid. Those trends reflect the relative efficiency
of the selected antioxidants.
Mechanisms of inhibition
Theoretical considerations. HOMO energies (E(HOMO)) of
lycopene and apo-lycopenoids were estimated by semi-empirical
quantum mechanics calculations as an approximation of the
ionisation potential (Ip): Ip = ꢀE(HOMO). The Ip value is a
measure of the intrinsic electron-donating capacity. For apo-
lycopenals, the longer the chain length, the lower the ionisation
potential and the lower the IC50 (Table 3). This correlation
suggests that the antioxidant activity of apo-lycopenals may be
directly linked to their reducing capacity. This is also true for
esters. However, for the same chain length (100, 140 and 11 series),
acids, esters and aldehydes have very similar ionisation potentials,
whereas Ip values are lower for apo-lycopenols. This trend is
For a global estimation of the antioxidant efficiency based
on both the lag and propagation phases and involving several
antioxidant concentrations, a IC50 parameter was calculated as
follows. For each antioxidant concentration C, T is defined as
the period of time needed to accumulate a given CD concen-
tration after addition of MbFeIII (e.g., that corresponding to a
0.7 increase in the absorbance at 234 nm from its value at time
zero). In the absence of antioxidant, T becomes T0. For a fixed
antioxidant concentration, a larger DT = T ꢀ T0 means a
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 579

View Article Online
Fig. 3 Accumulation of conjugated dienes (detection at 234 nm) during inhibition of the metmyoglobin-induced peroxidation of linoleic acid by
apo-140-lycopenoids (full lines, numbers in normal) and apo-100-lycopenoids (dotted lines, numbers in italic). Apo-lycopenoids concentrations are: 0 (0),
2 (1), 4 (2), 6 (3), 8 (4) and 10 mM (5). Linoleic acid (0.7 mM), metmyoglobin (100 nM), pH 5.8 phosphate buffer containing 2 mM Tween 20, 37 1C.
(A) Inhibition by apo-lycopenoic acids. (B) Inhibition by apo-lycopenals. (C) Inhibition by apo-lycopenoates. (D) Inhibition by apo-lycopenols.
Table 2 Kinetic parameters of antioxidant (AO) consumption at pH 5.8 and 37 1C (detection at the maximum absorption wavelength of each
compound). Initial concentrations: antioxidant (CAOi), 10 mM, except (all-E)-lyc., 2 mM, Tween 20, 2 mM, LH, 0.7 mM, MbFeIII, 100 nM
% AOb consumed at the end of the lag phasec
Without LH
and MbFeIII
With LH
alone
With MbFeIII
alone
With LH and
MbFeIII
Half-life of
Compound
l
max/nma
antioxidantb,d/s
Lag phaseb,c/s
(all-E)-lyc.
60-CHO
480
504
485
454
454
426
440
405
450
405
425
371
8 ꢁ 0
18 ꢁ 4
42 ꢁ 2
39 ꢁ 7
20 ꢁ 1
21 ꢁ 2
14 ꢁ 1
25 ꢁ 3
19 ꢁ 1
19 ꢁ 1
13 ꢁ 1
20 ꢁ 1
17 ꢁ 1
34 ꢁ 1
30 ꢁ 1
27 ꢁ 1
22 ꢁ 3
23 ꢁ 2
15 ꢁ 1
35 ꢁ 2
26 ꢁ 1
22 ꢁ 1
12 ꢁ 1
32 ꢁ 2
29 ꢁ 3
94 ꢁ 0
98 ꢁ 1
92 ꢁ 1
96 ꢁ 2
76 ꢁ 2
60 ꢁ 1
88 ꢁ 1
68 ꢁ 1
95 ꢁ 3
62 ꢁ 3
95 ꢁ 1
84 ꢁ 0
158 ꢁ 27
1389 ꢁ 37
1274 ꢁ 37
1001 ꢁ 129
832 ꢁ 37
849 ꢁ 19
1089 ꢁ 74
1039 ꢁ 49
542 ꢁ 18
650 ꢁ 74
392 ꢁ 19
377 ꢁ 18
375 ꢁ 28
17 ꢁ 1
16 ꢁ 1
14 ꢁ 1
13 ꢁ 4
10 ꢁ 1
16 ꢁ 1
13 ꢁ 1
12 ꢁ 1
7 ꢁ 0
3626 ꢁ 112
2398 ꢁ 204
1612 ꢁ 110
1429 ꢁ 202
1066 ꢁ 44
2525 ꢁ 71
1604 ꢁ 69
1287 ꢁ 65
941 ꢁ 24
80-CHO
100-CHO
120-CHO
140-CHO
100-COOH
140-COOH
100-COOEt
140-COOMe
100-CH2OH
140-CH2OH
9 ꢁ 1
1075 ꢁ 27
763 ꢁ 26
10 ꢁ 1
c
a
b
In micellar solution. n = 2. Estimated by graphical determination from the A(234 nm) vs. time curves recorded in the presence of both
d
MbFeIII and linoleic acid. For a given antioxidant, the same time interval is considered in the control experiments. Period of time needed to
consume half the initial antioxidant concentration in the presence of both MbFeIII and linoleic acid.
not in accordance with the weaker antioxidant activity of
apo-lycopenols compared to the other apo-lycopenoids (Fig. 4
and Table 3). It also does not reflect the higher potency of
apo-lycopenoic acids. For data interpretation, ionisation potentials
are thus of limited value and other criteria must also be con-
sidered, among which the partition of antioxidants between
the lipid phase, interface and aqueous phase (including possible
binding to metmyoglobin and its activated forms involved
in the initiation of lipid peroxidation). Hence, log P values
(P: n-octanol/water partition coefficient) were calculated to
assess the relative affinity of lycopene and its derivatives for
the lipid vs. aqueous phase (Table 3). As expected, the log P
values of aldehydes on the one hand and esters on the other
hand increase with the hydrocarbon chain length. As their anti-
oxidant potency concomitantly increases, it can be suggested
that these compounds act within the lipid phase by scavenging
lipid-derived radicals. A similar correlation does not hold for
alcohols and acids. In particular, 100-COOH and 140-COOH
are equally potent at inhibiting lipid peroxidation while the
former is much more lipophilic (higher log P) than the latter.
This is also true for 100-CH2OH and 140-CH2OH. It can thus be
suggested that the relative antioxidant efficiency of 140-COOH
c
580 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Table 3 IC50 (inhibition of metmyoglobin-induced peroxidation of
linoleic acid), Ip (ionisation potential) and log P (P: n-octanol/water
partition coefficient) values of (all-E)-lycopene and the apo-lycopenoids
investigated
Compounds
IC50/mMa
Ip = ꢀE(HOMO)/eV
log P
(all-E)-lyc.
60-CHO
1.94 ꢁ 0.14
0.94 ꢁ 0.08
1.27 ꢁ 0.10
2.35 ꢁ 0.21
3.21 ꢁ 0.46
3.47 ꢁ 0.11
14.81 ꢁ 1.57
1.86 ꢁ 0.24
1.65 ꢁ 0.09
20.11 ꢁ 7.58
2.62 ꢁ 0.15
3.41 ꢁ 0.14
27.10 ꢁ 8.44
3.66 ꢁ 0.35
3.61 ꢁ 0.10
11.41 ꢁ 0.89
7.74
7.91
7.95
8.02
8.09
8.22
8.81
8.04
8.25
8.87
8.02
8.21
8.80
7.84
7.98
8.40
11.11
7.52
7.18
6.32
5.97
5.11
3.39
6.58
5.37
3.65
7.18
5.64
4.25
6.62
5.42
3.70
80-CHO
100-CHO
120-CHO
140-CHO
11-CHO
100-COOH
140-COOH
11-COOH
100-COOEt
140-COOMe
11-COOEt
100-CH2OH
140-CH2OH
11-CH2OH
a
Data are expressed as mean ꢁ SD (n = 3–4).
Fig. 5 Estimation of energy changes for the addition of MeOOꢂ
(a simple model for the peroxyl radicals derived from linoleic acid) to
100-CH2OH (full circles) and 100-CHO (empty circles) as a function of
the site of addition on the unsaturated carbon chain (Catom number =
C5 to C110).
Fig. 4 Antioxidant activity of apo-lycopenoids in metmyoglobin-
induced peroxidation of linoleic acid. For the same concentration,
different letters point to significant differences between apo-lycopenoids
(ANOVA, post-hoc S–N–K, p o 0.05). (A) Inhibition by apo-lycopenals.
(B) Inhibition by apo-100-lycopenoids. (C) Inhibition by apo-140-
lycopenoids. From black to white bars, antioxidant concentrations are:
10, 8, 6, 4, and 2 mM.
group in the conjugation). This difference is most pronounced
for addition at C140.
– Peroxyl addition is most favourable at C110 (C-atom linked
to the terminal group) because of maximal electron delocalization
along the polyene chain. A similar situation occurs with addition
at the other end (C5) but seems less favourable.
(and to a lesser degree 140 CH2OH) reflects a specific protecting
mechanism (see below).
– Peroxyl addition is less favourable in the middle of the
chain, especially when the C-atoms involved are methylated
(C9, C13). It is also unfavourable at C6 and C120 as conjugation
in the corresponding radicals reaches a minimum.
Moreover, as an alternative to electron transfer, scavenging
of the propagating lipid peroxyl radicals by addition of the
latter to the polyene moiety of lycopene and apo-lycopenoids
may occur as already demonstrated for b-carotene.32 Hence,
the energy changes for the addition of MeOOꢂ (a simple model
for the peroxyl radicals derived from linoleic acid) to 100-CH2OH
and 100-CHO were estimated and plotted as a function of the site
of addition (Fig. 5). The following comments can be made:
– The energy change is generally more exothermic with
100-CHO as a likely consequence of the extended delocalization
in the radical adducts formed (participation of the aldehyde
The relatively unfavourable energy changes accompanying
radical addition to 100-CH2OH is consistent with its being a
poorer antioxidant (despite being a better electron donor) than
100-CHO in our model.
In conclusion, there is a general trend toward more lipophilic
apo-lycopenoids being better inhibitors of heme-induced lipid
peroxidation in agreement with their higher affinity for micelles
(higher log P value). As more lipophilic apo-lycopenoids also
display a longer conjugated hydrocarbon chain, they are generally
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 581

View Article Online
better electron donors. However, the favourable influence of
terminal groups with an electron-withdrawing mesomeric effect
(CHO, COOH, COOEt) vs. CH2OH suggests that the antioxidant
mechanism in micelles mainly proceeds via addition of the lipid
peroxyl radicals on the polyene moiety of apo-lycopenoids (rather
than electron transfer to the same radicals). The relative efficiency
of short-chain apo-lycopenoids with polar terminal groups
(mainly, 140-COOH) escapes those interpretations and has now
to be further analyzed.
Reduction of hypervalent iron species derived from metmyoglobin.
Our recent works16,27,33 have shown that MbFeIII efficiently
initiates the peroxidation of LH in micelles by forming an
ill-defined one-electron-oxidised form (simply noted MbFeIV for
convenience) in its reaction with initial traces of lipid hydroper-
oxides (LOOH). MbFeIV can then react with a second LOOH
molecule or a LH molecule to form the propagating lipid peroxyl
radicals (LOOꢂ). Depending on their hydrophilic–lipophilic
balance, antioxidants were also shown to act as inhibitors of pro-
pagation (scavenging of LOOꢂ in the lipid phase) or inhibitors of
initiation via reduction of MbFeIV. The latter mechanism was in
agreement with the ability of several water-soluble antioxidants
to rapidly reduce ferrylmyoglobin (MbFeIVQO).16,33–35 Direct
interaction between the heme crevice and anionic antioxidants
like ascorbate and the polyphenol chlorogenic acid was also
demonstrated.36 Although probably different from MbFeIV,
ferrylmyoglobin will be taken as a model of the peroxidation-
initiating species as this iron-oxo species has many advantages:
it is relatively stable, conveniently detected in the visible range,
and easily prepared by reacting MbFeIII with H2O2. It was thus
interesting to check whether some relatively polar apo-lycopenoids
could reach the heme and reduce the iron-oxo centre, as already
demonstrated in the case of the water-soluble carotenoid
crocetin.37 The reactions were run in the same conditions as
those with linoleic acid i.e. in the presence of Tween 20.
When H2O2 (1 equiv.) is added to a solution of MbFeIII, this
species is readily converted into MbFeIVQO, which has a
characteristic absorption band at 590 nm (Fig. 6A). After
A(590 nm) reached a plateau, (all-E)-lycopene or apo-lycopenoids
were added and changes in A(590 nm) were monitored. A signi-
ficant decay was observed only after addition of 140-COOH
(Fig. 6B), indicating that this sole derivative is able to reduce
MbFeIVQO back to MbFeIII. In particular, the other apo-140-
lycopenoids were inactive, thus showing that the carboxyl group
is essential in this reaction. However, compared with quercetin, a
potent flavonol antioxidant, 140-COOH appeared much less
efficient (Fig. 7). From a fitting of the A(590 nm) vs. time curves
relative to MbFeIVQO formation from the equimolar mixture of
MbFeIII and H2O2, a value of 104 Mꢀ1 cmꢀ1 was obtained for the
molar absorption coefficient of MbFeIVQO (in good agreement
with the literature37). This value was used for estimating the
MbFeIVQO and MbFeIII concentrations at the time of antioxidant
addition. An exponential fitting of the A(590 nm) vs. time curves
obtained after addition of an antioxidant in excess gave
apparent first-order rate constants (kobs), from which second-
order rate constants of ferrylmyoglobin reduction (kr) were
deduced: kobs = krC with C = initial antioxidant concen-
tration. One obtained: kr = 1150 (ꢁ210) and 7.7 (ꢁ1.1) Mꢀ1 sꢀ1
for quercetin and 140-COOH, respectively (37 1C, n = 2).
Fig. 6 (A) Spectral changes during the formation of ferrylmyoglobin.
Metmyoglobin (100 mM) + H2O2 (100 mM), pH 7.4 phosphate buffer
containing 2 mM Tween 20, 37 1C. (B) Spectral changes during
reduction of ferrylmyoglobin by 140-COOH (300 mM).
Fig. 7 Time dependence of the absorbance at 590 nm following the
addition of 140-COOH (full circles, 100 mM: 1, 200 mM: 2, 300 mM: 3)
or quercetin (empty triangles, 100 mM: 4, 200 mM: 5) to a solution of
ferrylmyoglobin (31–35 mM) in a pH 5.8 phosphate buffer containing
2 mM Tween 20, 37 1C. Control without antioxidant: 0.
The value for quercetin can be compared with the one estimated
at 10 1C in the absence of Tween 20: kr = 280 Mꢀ1 sꢀ1 38
.
In a previous work,16 we have shown that the inhibition of
metmyoglobin-induced lipid peroxidation by quercetin essen-
tially proceeds in the aqueous phase via a fast reduction of an
activated iron species (distinct from ferrylmyoglobin) produced
by reaction of metmyoglobin with the lipid hydroperoxides. We
hypothesize that 140-COOH could act by a similar mechanism.
c
582 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
was purchased from CaroteNature (CaroteNature GmbH,
Lupsingen, Switzerland). Apo-100-lycopenoic acid, ethyl apo-100-
lycopenoate, apo-140-lycopenoic acid and methyl apo-140-
lycopenoate were obtained in our laboratory by multi-step
organic synthesis.25 Diisobutylaluminium hydride (DIBAH),
manganese dioxide, ammonium formate, sylver oxide (Ag2O),
sodium hydride (NaH), 3 A molecular sieves, horse heart metmyo-
globin (MbFeIII, type II, MW ca. 17600 g molꢀ1), linoleic acid,
quercetin and Tween 20 (polyoxyethylenesorbitan monolaurate)
were purchased from Sigma-Aldrich (Sigma-Aldrich, Saint-
Quentin Fallavier, France). Triethyl phosphonoacetate and
benzophenone were purchased from Alfa Aesar (Alfa Aesar
France, Schiltigheim, France). Buffer salts and all other
chemicals were of analytical grade. The phosphate buffers
(10 mmol, pH 5.8 or 6.8), prepared with grade water (18 MO)
prepared using a Millipore Q-Plus water, were passed through a
column of Chelex-100 chelating resin (Bio-Rad, Marnes-la-
Coquette, France) to remove contaminating free metal ion
traces. Silica gel (Si 60, 40–63 mm) was obtained from Merck
(Merck France, Limonest, France). Celite was purchased from
VWR (VWR International SAS, Fontenay sous Bois, France).
Chloroform (CHCl3) and methanol (MeOH) were of HPLC
grade. All other solvents used were of analytical grade. The
tetrahydrofurane (THF) used for chemical reactions was freshly
distilled over Na–benzophenone. Ethyl acetate (AcOEt), ethanol,
hexane, dichloromethane (CH2Cl2) and toluene were dried over
3 A molecular sieves (previously dried at 150 1C for 4 h).
Scheme 3 Proposed mechanism for the inhibition of the metmyoglobin-
induced peroxidation of linoleic acid by 100-COOH and 140-COOH.
In particular, 140-COOH could bind to the iron centre through
its carboxylate group (Scheme 3). This hypothesis is comforted
by the shape of the A(234 nm) vs. time curves recorded during
the inhibition of lipid peroxidation by 140-COOH, which does
not show a well-defined lag phase (Fig. 3A) as usually obtained
for chain-breaking antioxidants acting in the lipid phase.26 On
the opposite, apo-lycopenoids with longer chains like 100-COOH
and its analogues with different terminal groups must be too
lipophilic to directly interact with activated metmyoglobin, thus
staying in the lipid phase where they scavenge lipid peroxyl
radicals (Scheme 3), as previously demonstrated for the typical
chain-breaking antioxidant a-tocopherol.16,27
Analyses
Conclusion
HPLC-MS. HPLC-MS analyses were performed using a
Hewlett Packard 1100 apparatus coupled to a UV/Vis diode
array detector and a Micromass platform LCZ 4000 mass
spectrometer. Mass analyses were performed in the positive
electrospray ionization mode with a capillary voltage of 15 V
and a desolvation temperature of 250 1C. The column used
was a 150 mm ꢃ 2.1 mm i.d., 3 mm, YMC C30 column, with a
10 mm ꢃ 2 mm i.d. guard column of the same material
(Interchim, Montlucon, France). The temperature of the
column was set at 40 1C. The solvent system was a gradient
of A (25 mmol ammonium formate + 2.6 mmol formic acid in
water), B (25 mmol ammonium formate in MeOH) and C
(methyl tert-butyl ether), with 60% B, 40% A at t = 0; 40% A,
60% B at t = 1.25 min; 30% A, 70% B at t = 2.50 min; 25% A,
70% B, 5% C at t = 3.75 min; 20% A, 70% B, 10% C at
t = 5 min; 5% A, 10% B, 85% C at t = 15 min; and 5% A,
10% B, 85% C at t = 25 min (flow rate = 0.5 ml minꢀ1).
In this work, we demonstrated the ability of lycopene and apo-
lycopenoids to inhibit the peroxidation of linoleic acid induced
by heme iron in acidic micellar solutions. 60-CHO, which has the
longest unsaturated chain among the apo-lycopenoids tested, was
the best inhibitor, followed by 80-CHO. It is important to note
that both long-chain apo-lycopenals, which have been evidenced
in fresh tomato39 and in tomato products,20 are more antioxidant
than lycopene itself in our model. For the same chain length, apo-
lycopenoic acids come up as the most active molecules, which
might be due to a more favourable hydrophilic/lipophilic balance
allowing their location at the lipid–water interface. Both the chain
length and terminal group of apo-lycopenoids thus govern their
antioxidant activity and the mechanisms involved. In particular,
140-COOH behaves more as a hydrophilic antioxidant possibly
reacting directly with hypervalent iron species in the aqueous
phase whereas 100-COOH, its analogue with a longer chain,
might exert its antioxidant action in the lipid phase by quenching
lipid peroxyl radicals. Experiments are ongoing in our laboratory
to better understand the different mechanisms involved in the
antioxidant activity of apo-lycopenoids.
TLC. Analytical thin layer chromatography (TLC) was
performed on silica gel plates (Merck Kieselgel 60 F254) with
UV detection at 254 or 365 nm.
GC-MS. Volatile products were monitored by GC-MS
(CP2010; Shimadzu, Kyoto). Samples (1 ml) were injected onto
a UBWAX 52 CB capillary column (30 m, 0.25 mm i.d., 0.5 mm
thickness). The injection port was operated in split mode (R =
1/10). The carrier gas (He) velocity was constant (60 cm sꢀ1).
The initial oven temperature of 60 1C was increased at a rate of
5 1C per min to 230 1C. This final temperature was maintained
for 15 min. The mass spectrometer was operated in the electron
Experimental
Materials
(All-E)-lycopene, apo-80-lycopenal and apo-120-lycopenal were
a gift from DSM (DSM Nutritional Products Ltd., Kaiseraugst,
Switzerland). Pseudoionone was given by BASF (BASF
Aktiengesellschaft, Ludwigshafen, Germany). Apo-60-lycopenal
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 583

View Article Online
impact mode at 70 eV with continuous scans (every 0.5 s) in the
range m/z 29–350. Data were analyzed using the GC-MS
solution software.
HPLC-UV/Vis-MS: tR 13.02 min; lmax/nm 396, 418, 444
ꢀꢀꢀ
(main absorption band is underlined), m/z 379.5 ([M + H]+),
361.5 ([M + H]+-H2O).
NMR. H and 13C NMR spectra were recorded on a Bruker
1
(2E,4E,6E,8E,10E,12E,14E,16E,20E)-4,9,13,17,21-Pentamethyl-
Advance DRX 500 MHz spectrometer. Chemical shifts are given
in ppm relative to peak solvent (CDCl3).40 J values are given
in Hz. Signal assignments were deduced from a combination of
heteronuclear single quantum coherence (HSQC), distortionless
enhancement by polarization transfer quantum (DEPTQ 135),
nuclear overhauser effect spectroscopy (NOESY), correlation
spectroscopy (COSY) and heteronuclear multiple bond coherence
(HMBC) spectra.
docosa-2,4,6,8,10,12,14,16,20-nonaenal
(apo-100-lycopenal,
100-CHO). Compound 100-CH2OH (90 mg, 0.24 mmol) was
dissolved in AcOEt (5 ml). After addition of MnO2 (510 mg,
5 mmol), the heterogeneous mixture was vigorously stirred at
room temperature for 2 h. Then, the mixture was filtered off
through Celite, dried over Na2SO4 and concentrated. The
residue was purified by filtration through a silica gel chromato-
graphic column (eluent hexane/AcOEt (2/8, v/v)). Compound
100-CHO was obtained as a red powder (40 mg, 44%).
dH (500 MHz; CDCl3) 1.61 (3H, s, C(21)Me), 1.69 (3H, s,
C(22)H3), 1.82 (3H, s, C(17)Me), 1.96 (3H, s, C(4)Me), 1.98
(3H, s, C(13)Me), 2.01 (3H, s, C(9)Me), 2.12 (4H, m, C(18,
19)H2), 5.11 (1H, m, C(20)H), 5.95 (1H, d, J 11.0, C(16)H),
6.18 (1H, dd, J 7.8 and 15.3, C(2)H), 6.18 (1H, d, J 11.6,
C(12)H), 6.25 (1H, d, J 15.1, C(14)H), 6.28 (1H, d, J 12.1,
C(8)H), 6.36 (1H, d, J 14.9, C(10)H), 6.53 (1H, dd, J 11.0 and
15.1, C(15)H), 6.61 (1H, m, C(6)H), 6.61 (1H, m, C(5)H), 6.73
(1H, dd, J 11.6 and 14.9, C(11)H), 6.86 (1H, m, C(7)H), 7.15
(1H, d, J 15.3, C(3)H), 9.58 (1H, d, J 7.8, C(1)H).
HRMS. High-resolution mass analysis was carried out using
a QStar Elite (Applied Biosystems SCIEX) with positive
electrospray-TOF ionisation. High-resolution mass analysis
was carried out on an Ultraflextreme TOF/TOF (Bruker) with
NALDIt ionisation for compound 140-CH2OH.
Synthesis of apo-100-, apo-140- and apo-11-lycopenoids
All reactions were carried out under dim light and argon
atmosphere.
Synthesis of apo-100-lycopenoids
(2E,4E,6E,8E,10E,12E,14E,16E,20E)-4,9,13,17,21-Pentamethyl-
docosa-2,4,6,8,10,12,14,16,20-nonaen-1-ol (apo-100-lycopenol,
100-CH2OH). Ethyl apo-100-lycopenoate (100-COOEt)25 (287 mg,
0.68 mmol) was dissolved in dry THF (5 ml). A solution of
DIBAH (1.7 mmol) in toluene (1.4 ml) was carefully added
over 30 min at 0 1C. The mixture was stirred for 4 h, then a
MeOH/water (3 ml, 1/1, v/v) mixture was slowly added at
0 1C. The gel formed was destroyed by addition of a 2 M
potassium sodium tartrate aqueous solution (10 ml) and
vigorous stirring for 10 h. Then, the product was extracted
with diethylether (3 ꢃ 10 ml). The combined organic phases
were dried over Na2SO4 and concentrated. The crude product
was purified by filtration through a silica gel chromatographic
column (eluent hexane/AcOEt (1/1, v/v)), then crystallized
from CH2Cl2/hexane. Compound 100-CH2OH was obtained
as a red powder (80 mg, 31%).
dC (125 MHz; CDCl3) 12.9 (C(4)Me), 13.1 (C(9)Me), 13.1
(C(13)Me), 17.1 (C(17)Me), 17.8 (C(21)Me), 25.8 (C(22)), 26.8
(C(19)), 40.4 (C(18)), 124.0 (C(20)), 125.7 (C(15)), 125.8 (C(16)),
127.2 (C(2)), 127.2 (C(11)), 128.6 (C(6)), 131.3 (C(12)), 131.7
(C(8)), 131.8 (C(21)), 133.8 (C(4)), 135.3 (C(14)), 135.5 (C(7)),
136.8 (C(10)), 137.7 (C(13)), 140.2 (C(9)), 140.3 (C(17)), 141.3
(C(5)), 156.7 (C(3)), 193.8 (C(1)).
HRMS: m/z 377.2823 ([M + H]+) (377.2839 calculated for
C27H37O).
HPLC-UV/Vis-MS: tR 13.61 min, lmax/nm 458, m/z 377.5
([M + H]+).
Synthesis of apo-140-lycopenoids
(2E,4E,6E,8E,10E,12E,16E)-5,9,13,17-Tetramethyloctadeca-
2,4,6,8,10,12,16-heptaen-1-ol (apo-140-lycopenol, 140-CH2OH).
A solution of DIBAH (3.5 mmol) in toluene (3.5 ml) was
carefully added over 1 h to a solution of methyl apo-140-
lycopenoate (140-COOMe)25 (500 mg, 1.47 mmol) in dry
THF (20 ml) at 0 1C and the resulting mixture stirred for 3 h
at 0 1C. The reaction was quenched by careful addition of
MeOH/water (5 ml, 1/1, v/v) at 0 1C. The gel thus formed was
dissociated by addition of aqueous 2 M potassium sodium
tartrate (20 ml) and vigorous stirring for 10 h. Then, the
product was extracted with diethylether (3 ꢃ 10 ml) and the
combined organic phases dried over Na2SO4 and concentrated.
The residue was purified by filtration through a silica gel
chromatographic column (eluent hexane/AcOEt (1/1, v/v)),
then crystallized from CH2Cl2/hexane. Compound 140-CH2OH
was obtained as an orange powder (163 mg, 35%).
dH (500 MHz; CDCl3) 1.39 (1H, t, J 5.85, OH), 1.61 (3H, s,
C(21)Me), 1.69 (3H, s, C(22) H3), 1.82 (3H, s, C(17)Me), 1.91
(3H, s, C(4)Me), 1.97 (6H, s, C(9, 13)Me), 2.12 (4H, m,
C(18, 19) H2), 4.25 (2H, t, J 6.1, C(1)H), 5.11 (1H, m, C(20)H),
5.88 (1H, dt, J 6.1 and 15.6, C(2)H), 5.95 (1H, d, J 11.0, C(16)H),
6.18 (1H, d, J 11.5, C(12)H), 6.21 (1H, d, J 11.5, C(8)H), 6.23 (1H,
d, J 11.0, C(5)H), 6.25 (1H, d, J 15.2, C(14)H), 6.34 (1H, d, J 15.6,
C(3)H), 6.35 (1H, d, J 15.0, C(10)H), 6.49 (1H, dd, J 11.0 and
15.2, C(15)H), 6.59 (1H, m, C(6)H), 6.61 (1H, m, C(7)H),
6.64 (1H, m, C(11)H).
dC (125 MHz; CDCl3) 13.0 (C(4)Me), 13.0 (C(9)Me), 13.0
(C(13)Me), 17.1 (C(17)Me), 17.9 (C(21)Me), 25.9 (C(22)), 26.8
(C(19)), 40.4 (C(18)), 64.1 (C(1)), 124.1 (C(20)), 125.0 (C(15)),
125.5 (C(11)), 125.9 (C(16)), 127.5 (C(2)), 129.6 (C(6)), 130.6
(C(7)), 131.6 (C(12)), 131.9 (C(21)), 132.4 (C(8)), 132.5 (C(5)),
134.9 (C(4)), 135.5 (C(14)), 136.4 (C(13)), 136.5 (C(3)), 136.9
(C(9)), 137.4 (C(10)), 139.7 (C(17)).
dH (500 MHz; CDCl3) 1.4 (1H, t, J 5.8, HO), 1.61 (3H, s,
C(17)Me), 1.69 (3H, s, C(18)H3), 1.82 (3H, s, C(13)Me), 1.94
(3H, s, C(5)Me), 1.95 (3H, s, C(9)Me), 2.12 (4H, m, C(14,
15)H2), 4.25 (2H, t, J 5.9, C(1)H2), 5.11 (1H, m, C(16)H), 5.89
(1H, dt, J 5.9 and 14.9, C(2)H), 5.95 (1H, d, J 11.0, C(12)H),
6.12 (1H, d, J 11.4, C(4)H), 6.16 (1H, d, J 11.4, C(8)H),
HRMS: m/z 379.2978 ([M + H]+) (379.2995 calculated for
C27H39O).
c
584 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
6.24 (1H, d, J 15.0, C(10)H), 6.31 (1H, d, J 14.9, C(6)H), 6.49
(1H, dd, J 11.0 and 15, C(11)H), 6.63 (2H, m, C(3, 7)H).
dC (125 MHz; CDCl3) 12.8 (C(5)Me), 13.0 (C(9)Me), 17.1
(C(13)Me), 17.8 (C(17)Me), 25.8 (C(18)), 26.8 (C(15)), 40.4
(C(14)), 63.9 (C(1)), 124.1 (C(16)), 125.1 (C(11)), 125.7 (C(7)),
125.8 (C(12)), 128.4 (C(3)), 130.5 (C(8)), 131.1 (C(17)), 131.3
(C(4)), 132.4 (C(2)), 135.4 (C10)), 136.4 (C(9)), 136.7 (C(5)),
137.1 (C(6)), 139.7 (C(13)).
HRMS (For 140-CH2OH high-resolution mass analysis was
carried out on an Ultraflextreme TOF/TOF (Bruker) with
NALDIt ionisation): m/z 312.2441 [M]+ (312.2448 calculated
for C22H32O).
compound 11-COOEt as a light-yellow oil (2.1 g, 82%,
(E/Z) = 3.5 (determined by HPLC)).
dH (500 MHz; CDCl3) 1.29 (3H, t, J 7.1, CH2CH3), 1.61
(3H, s, C(11)Me), 1.69 (3H, s, C(12)H3), 1.85 (3H, s, C(7)Me),
2.13 (4H, m, C(8, 9)H2), 2.32 (3H, s, C(3)Me), 4.16 (2H, q,
J 7.1, CH2CH3), 5.09 (1H, m, C(10)H), 5.62 (1H, s, C(2Z)H),
5.74 (1H, s, C(2E)H), 5.96 (1H, d, J 11.1, C(6)H), 6.18 (1H, d,
J 15.3, C(4)H), 6.84 (1H, dd, J 11.1 and 15.3, C(3)H).
dC (125 MHz; CDCl3) 14.1 (C(3)Me), 14.9 (CH2CH3), 17.4
(Me-C(7)), 18.2 (Me-C(11)), 26.4 (C(12)), 27.0 (C(9)), 40.8
(C(8)), 59.9 (CH2CH3), 116.5 (C(2Z)), 118.4 (C(2E)), 123.9
(C(10)), 125.2 (C(6)), 131.5 (C(5)), 132.2 (C(11)), 133.6 (C(4)),
144.1 (C7)), 153.2 (C(3)), 167.3 (C(1)).
HPLC-UV/Vis-MS: tR 11.28 min, lmax/nm 348, 366, 387,
ꢀꢀꢀ
m/z 313.3 ([M + H]+), 295.2 ([M + H]+-H2O).
HRMS: m/z 263.2003 ([M + H]+) (263.2005 calculated for
C17H27O2).
HPLC-UV/Vis-MS: tR(2Z) 9.76 min and tR(2E) 10.40 min,
(2E,4E,6E,8E,10E,12E,16E)-5,9,13,17-Tetramethyloctadeca-
2,4,6,8,10,12,16-heptaenal (apo-140-lycopenal, 140-CHO). MnO2
(175 mg, 2 mmol) was added to a solution of 140-CH2OH (26 mg,
0.08 mmol) in AcOEt (1 ml). The heterogeneous mixture was
vigorously stirred at room temperature for 1 h, then filtered off
through Celite, dried over Na2SO4 and concentrated. The residue
was purified by filtration through a silica gel chromatographic
column (eluent hexane/AcOEt (2/8, v/v)). Compound 140-CHO
was obtained as an orange powder (11 mg, 44%).
l
max/nm 318, m/z 263.2 ([M + H]+). GC-MS: tR(2Z) 34.3 min
and tR(2E) 34.9 min, m/z 262 (M+).
(2E,4E,6E)-3,7,11-Trimethyldodeca-2,4,6,10-tetraen-1-ol (apo-
11-lycopenol, 11-CH2OH). A solution of DIBAH (1.9 mmol) in
hexane (1.9 ml) was carefully added over 1 h to a solution of
11-COOEt (210 mg, 0.75 mmol) in dry THF (5 ml) at 0 1C and
the resulting mixture stirred for 3 h at room temperature.
The reaction was quenched by careful addition of MeOH/water
(5 ml, 1/1, v/v) at 0 1C. The gel thus formed was dissociated by
addition of aqueous 2 M potassium sodium tartrate (20 ml) and
vigorous stirring for 10 h. Then, the product was extracted with
diethylether (3 ꢃ 10 ml) and the combined organic phases dried
over Na2SO4 and concentrated. The residue was purified and
diastereoisomers were separated by an open silica gel chromato-
graphic column (eluent hexane/AcOEt (8/2, v/v)). Compound
11-CH2OH was obtained as a white solid (27 mg, 16%).
dH (500 MHz; CDCl3) 1.61 (3H, s, C(11)Me), 1.69 (3H, s,
C(12)H3), 1.79 (3H, s, C(3)Me), 1.83 (3H, s, C(7)Me), 2.11 (4H,
m, C(8, 9)H2), 4.27 (2H, d, J 7.0, C(1)H2), 5.10 (1H, m, H-C(10)),
5.63 (1H, t, J 7, C(2)H), 5.9 (1H, d, J 10.8, C(6)H), 6.17 (1H, d,
J 15.3, C(4)H), 6.46 (1H, dd, J 10.8 and 15.3, C(5)H).
dC (125 MHz; CDCl3) 12.6 (C(3)Me), 16.8 (C(7)Me), 17.6
(C(11)Me), 25.6 (C(12)), 26.6 (C(9)), 40.1 (C(8)), 59.4 (C(1)),
123.9 (C(10)), 125.0 (C(5)), 125.2 (C(6)), 129.0 (C(2)), 131.7
(C(11)), 134.3 (C4)), 136.9 (C(3)), 139.6 (C(7)).
dH (500 MHz; CDCl3) 1.62 (3H, s, C(17)Me), 1.69 (3H, s,
C(18)H3), 1.83 (3H, s, C(13)Me), 2 (3H, s, C(9)Me), 2.11 (3H, s,
C(5)Me), 2.12 (4H, m, C(14, 15)H2), 5.11 (1H, m, C(16)H), 5.96
(1H, d, J 11.1, C(12)H), 6.17 (1H, dd, J 7.9 and 14.8, C(2)H),
6.20 (1H, d, J 11.5, C(8)H), 6.26 (1H, d, J 15.0, C(10)H),
6.35 (1H, d, J 11.9, C(4)H), 6.38 (1H, d, J 15.0, C(6)H), 6.59
(1H, dd, J 11.1 and 15, C(11)H), 6.91 (1H, dd, J 11.5 and 15,
C(7)H), 7.52 (1H, dd, J 11.8 and 14.8, C(3)H), 9.61 (1H, d,
J 8.0, C(1)H).
dC (125 MHz; CDCl3) 13.3 (C(9)Me), 13.6 (C(5)Me), 17.2
(C(13)Me), 17.8 (C(17)Me), 25.8 (C(18)), 26.8 (C(15)), 40.4
(C(14)), 124.0 (C(16)), 125.7 (C(12)), 126.8 (C(11)), 128.7 (C(4)),
130.1 (C(7)), 130.7 (C(8)), 130.9 (C(2)), 132 (C(17)), 135 (C10)),
135.7 (C(6)), 139.8 (C(9)), 141.2 (C(13)), 146.8 (C(5)), 147.7
(C(3)), 193.7 (C(1)).
HRMS: m/z 311.2369 ([M + H]+) (311.2369 calculated for
C22H31O).
HPLC-UV/Vis-MS: tR 13.61 min, lmax/nm 422, m/z 311.5
([M + H]+).
HRMS: m/z 243.1717 ([M + Na]+) (243.1719 calculated for
C15H24ONa).
HPLC-UV/Vis-MS: tR 7.79 min, lmax/nm 279, m/z 203.1
([M + H]+-H2O).
Synthesis of apo-11-lycopenoids
Ethyl (2E and 2Z,4E,6E)-3,7,11-trimethyldodeca-2,4,6,10-
tetraenoate (ethyl apo-11-lycopenoate, 11-COOEt). Triethyl
phosphonoacetate (10 ml, 49 mmol) was slowly added at 0 1C
in a heterogenous mixture containing 30 ml of dry THF and
NaH (1.96 mg, 48 mmol) previously washed with hexane (3 ꢃ
20 ml). After the end of hydrogen release, pseudoionone
(PI, 2.2 ml, 9.8 mmol) was added over 1 h then the mixture
stirred for 30 h at room temperature. The reaction was then
quenched with 25 ml of saturated aqueous NH4Cl solution and
the product extracted with diethylether (3 ꢃ 10 ml). The
combined organic layers were washed with brine (3 ꢃ 10 ml),
dried over Na2SO4 and concentrated. The residue was purified
on silica gel (eluent EtOAc/hexane (2/98, v/v)) to afford
(2E,4E,6E)-3,7,11-Trimethyldodeca-2,4,6,10-tetraenal (apo-
11-lycopenal, 11-CHO). MnO2 (14 g, 140 mmol) was added to
a solution of diastereoisomeric mixture of 11-CH2OH (1.37 g,
6.2 mmol) in AcOEt (25 ml). The heterogeneous mixture was
vigorously stirred at room temperature for 3 h, then filtered off
through Celite, dried over Na2SO4 and concentrated. The
residue was purified and diastereoisomers were separated by an
open silica gel chromatographic column (eluent hexane/AcOEt
(9/1, v/v)). Compound 11-CHO was obtained as a light-yellow
oil (400 mg, 30%).
dH (500 MHz; CDCl3) 1.61 (3H, s, C(11)Me), 1.69 (3H, s,
C(12)H3), 1.87 (3H, s, C(7)Me), 2.14 (4H, m, C(8, 9)H2),
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 585

View Article Online
2.30 (3H, s, C(3)Me), 5.09 (1H, m, C(10)H), 5.95 (1H, d, J 8,
C(2)H), 6.01 (1H, d, J 11.1, C(6)H), 6.26 (1H, d, J 15.3,
C(4)H), 6.99 (1H, dd, J 11.1 and 15.3, C(5)H), 10.10 (1H, d,
J 8.0, C(1)H).
and apo-lycopenoids, 0.5–10 mM. Two ml of the freshly
prepared solution were transferred to the spectrometer cells.
At time zero, lipid oxidation was initiated by adding 20 ml of
the metmyoglobin solution (final concentration in cell: 100 nM).
The UV-visible spectra were recorded at regular time intervals
(26 s). Each experiment was repeated three or four times at
different antioxidant concentrations.
dC (125 MHz; CDCl3) 13.7 (C(3)Me), 17.7 (C(7)Me), 18.3
(C(11)Me), 26.2 (C(12)), 26.9 (C(9)), 40.7 (C(8)), 123.7 (C(10)),
125.2 (C(6)), 129.1 (C(2)), 132.4 (C(11)), 132.9 (C(5)), 133.3
(C4)), 146.5 (C(7)), 155.3 (C(3)), 191.6 (C(1)).
HRMS: m/z 241.1563 ([M + Na]+) (241.1562 calculated for
C15H22ONa).
Stability of (all-E)-lycopene and apo-lycopenoids
The same experimental conditions were used to estimate the
stability of (all-E)-lycopene (2 mM) and apo-lycopenoids
(10 mM) without linoleic acid or metmyoglobin or both. The
UV-visible spectra were recorded at regular time intervals
(5 s). Each experiment was duplicated at different antioxidant
concentrations.
HPLC-UV/Vis-MS: tR 7.90 min, lmax/nm 344, m/z 219.0
([M + H]+).
(2E,4E,6E)-3,7,11-Trimethyldodeca-2,4,6,10-tetraenoic acid
(apo-11-lycopenoic acid, 11-COOH). Ag2O (3.7 g, 26 mmol)
and NaOH (230 mg, 5.3 mmol) dissolved in EtOH (20 ml) were
added to a solution of 11-CHO (230 mg, 1.1 mmol) in toluene
(20 ml). The heterogeneous mixture was vigorously stirred at
room temperature for 20 h, then filtered off through Celite. The
filtrate was washed with diluted HCl solution (3 ꢃ 10 ml) then
dried over Na2SO4 and the solvent removed under vacuum.
Compound 11-COOH was obtained as an amorphous yellow
solid (221 mg, 90%).
Reduction of ferrylmyoglobin
Ferrylmyoglobin (MbFeIVQO) was previously formed in the
spectrometer cell by adding 20 ml of a 5 mM aqueous solution
of H2O2 (concentration determined using e = 39.4 Mꢀ1 cmꢀ1
at 240 nm, final concentration in the cell: 100 mM) to 1 ml of a
100 mM MbFeIII solution in 10 mM phosphate buffer containing
2 mM Tween 20 (pH 7.4). The spectral changes featuring the
conversion of MbFeIII into MbFeIVQO were recorded in the
visible range until stability (2–3 minutes). Then, 1 ml of a con-
centrated solution of an antioxidant (apo-lycopenoids or quercetin)
in the same phosphate buffer–Tween 20 solution was added
in the cell (final concentrations: 100, 200 and 300 mM) and
the reduction of MbFeIVQO back to MbFeIII was monitored
at 590 nm. Each experiment was duplicated at different
antioxidant concentrations.
dH (500 MHz; CDCl3) 1.61 (3H, s, C(11)Me), 1.69 (3H, s,
C(12)H3), 1.85 (3H, s, C(7)Me), 2.13 (4H, m, C(8, 9)H2), 2.34
(3H, s, C(3)Me), 5.09 (1H, m, C(10)H), 5.77 (1H, s, C(2)H),
5.98 (1H, d, J 10.9, C(6)H), 6.21 (1H, d, J 15.1, C(4)H), 6.90
(1H, dd, J 10.9 and 15.1, C(5)H).
dC (125 MHz; CDCl3) 14.2 (C(3)Me), 17.4 (C(7)Me), 17.9
(C(11)Me), 25.9 (C(12)), 26.6 (C(9)), 40.4 (C(8)), 117.1 (C(2)),
123.7 (C(10)), 125.0 (C(6)), 132.2 (C(5)), 132.2 (C(11)), 133.4
(C4)), 145.0 (C(7)), 155.9 (C(3)), 172.0 (C(1)).
HRMS: m/z 257.1515 ([M + Na]+) (257.1512 calculated for
Data analysis and semi-empirical quantum mechanics
calculations
C15H22O2Na).
HPLC-UV/Vis-MS: tR 5.83 min, lmax/nm 301, m/z 235.3
([M + H]+).
Curve-fittings were carried out on a PC using the Scientist
program (MicroMath, Salt Lake City, USA). Differences of anti-
oxidant activity between (all-E)-lycopene and apo-lycopenoids
were tested for significance by a one-way ANOVA procedure
using Student–Newman–Keuls post-hoc test (XLSTAT for
Microsoft Excel version 2009.2.04). Differences were considered
to be significant at p o 0.05.
Kinetic experiments
All kinetic experiments were recorded by UV-visible spectro-
scopy on a SPECORDs diode-array spectrometer (Analytik
Jena, Germany) or on a Hewlett-Packard 8453 diode-array
spectrometer (Agilent Technologies, France). Solutions in cells
(optical pathlength: 1 cm) were stirred with a magnetic stirrer
and thermostated at 37 (ꢁ1) 1C.
Semi-empirical quantum mechanics calculations were run on
the HyperChem 5.1 software (Hypercube, Waterloo, Canada).
HOMO energies were calculated using the PM3 program.
Energy changes for radical addition (difference between the
energy of adduct and the sum of the energies of starting
polyene and methylperoxyl radical) were calculated using the
AM1 program (UHF mode).
Inhibition of the metmyoglobin-induced peroxidation of
linoleic acid
The experimental conditions used were adapted from already
published procedures.15 Metmyoglobin (MbFeIII, 17.6 mg)
was dissolved in 50 ml of phosphate buffer (10 mM,
pH 6.8). Its concentration was standardized at 10 mM using
e = 7700 Mꢀ1 cmꢀ1 at 525 nm.41 Given volumes of daily
prepared solutions of linoleic acid (7 mM), Tween 20 (40 mM),
(all-E)-lycopene and apo-lycopenoids (1 mM) in CHCl3 were
mixed and the solvent removed under reduced pressure in
darkness and 25 1C. The residue was dissolved immediately in
20 ml of 10 mM phosphate buffer at pH 5.8 under gentle
stirring. The final concentrations in the solution were: linoleic
acid, 0.7 mM; Tween 20, 2 mM; (all-E)-lycopene, 0.5–2.5 mM
Values of log P (P: n-octanol/water partition coefficient)
were calculated using the ChemPro program (ChemDraw Pro
11.0 CambridgeSoft).
Acknowledgements
This work was supported by a grant from the European
Community’s Sixth Framework Program (Lycocard, Integrated
European Project No. 016213). Special thanks are given to
Bettina Wustenberg and Werner Simon (DSM Nutritional
Products Ltd., Kaiseraugst, Switzerland) for their generous gift
c
586 New J. Chem., 2012, 36, 575–587
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
of (all-E)-lycopene, apo-80-lycopenal and apo-120-lycopenal and
to Hansgeorg Ernst (BASF Aktiengesellschaft, Fine Chemicals,
and Biocatalysis Research, Ludwigshafen, Germany) for his
generous gift of pseudoionone. We thank Claude Marfisi
(LEA/Bruker) for the HRMS analysis of 140-CH2OH. We also
thank Christian Ginies for his help with GC-MS analyses and
the Spectropole in Marseille for their assistance in NMR and
HRMS analyses.
19 K. Q. Hu, C. Liu, H. Ernst, N. I. Krinsky, R. M. Russell and
X. D. Wang, J. Biol. Chem., 2006, 281, 19327–19338.
20 R. E. Kopec, K. M. Riedl, E. H. Harrison, R. W. Curley,
D. P. Hruszkewycz, S. K. Clinton and S. J. Schwartz, J. Agric.
Food Chem., 2010, 58, 3290–3296.
21 W. Stahl, J. von Laar, H.-D. Martin, T. Emmerich and H. Sies,
Arch. Biochem. Biophys., 2000, 373, 271–274.
22 F. Lian and X. D. Wang, Int. J. Cancer, 2008, 123, 1262–1268.
23 K. Linnewiel, H. Ernst, C. Caris-Veyrat, A. Ben-Dor, A. Kampf,
H. Salman, M. Danilenko, J. Levy and Y. Sharoni, Free Radical
Biol. Med., 2009, 47, 659–667.
24 A. Ben-Dor, M. Steiner, L. Gheber, M. Danilenko, N. Dubi,
K. Linnewiel, A. Zick, Y. Sharoni and J. Levy, Mol. Cancer Ther.,
2005, 4, 177–186.
Notes and references
1 A. V. R. Rao and S. Agarwal, J. Am. Coll. Nutr., 2000, 19,
563–569.
2 H. D. Sesso, S. M. Liu, J. M. Gaziano and J. E. Buring, J. Nutr.,
2003, 133, 2336–2341.
3 H. D. Sesso, J. E. Buring, E. P. Norkus and J. M. Gaziano, Am. J.
Clin. Nutr., 2004, 79, 47–53.
4 E. Giovannucci, J. Nutr., 2005, 135, 2030S–2031S.
5 G. Britton, S. Liaaen-Jensen and H. Pfander, Carotenoids Handbook,
Birkhauser, 2004, pp. 563.
25 E. Reynaud, G. Aydemir, R. Ruhl, O. Dangles and C. Caris-
Veyrat, J. Agric. Food Chem., 2011, 59, 1457–1463.
26 J. R. Mein, F. Lian and X. D. Wang, Nutr. Rev., 2008, 66, 667–683.
27 P. Goupy, E. Vulcain, C. Caris-Veyrat and O. Dangles,
Free Radical Biol. Med., 2007, 43, 933–946.
28 P. Goupy, A. B. Bautista-Ortin, H. Fulcrand and O. Dangles,
J. Agric. Food Chem., 2009, 57, 5762–5770.
29 C. Caris-Veyrat, A. Schmid, M. Carail and V. Bohm, J. Agric.
Food Chem., 2003, 51, 7318–7325.
6 B. L. Lindshield, K. Canene-Adams and J. W. Erdman, Arch.
Biochem. Biophys., 2007, 458, 136–140.
7 A. J. Young and G. M. Lowe, Arch. Biochem. Biophys., 2001, 385,
20–27.
8 W. Stahl and H. Sies, Mol. Aspects Med., 2003, 24, 345–351.
9 P. Di Mascio, S. Kaiser and H. Sies, Arch. Biochem. Biophys.,
1989, 274, 532–538.
30 G. Britton, in Carotenoids. Volume 1B: Spectroscopy, ed.
G. Britton, S. Liaaen-Jensen and H. Pfander, Birkhauser Verlag,
1995, vol. 1B, pp. 13–62.
31 J. R. Mancuso, D. J. McClements and E. A. Decker, J. Agric. Food
Chem., 1999, 47, 4146–4149.
32 D. C. Liebler and T. D. McClure, Chem. Res. Toxicol., 1996, 9,
8–11.
10 N. I. Krinsky and E. J. Johnson, Mol. Aspects Med., 2005, 26,
459–516.
33 B. Lorrain, C. Dufour and O. Dangles, Free Radical Biol. Med.,
2010, 48, 1162–1172.
11 V. L. Freeman, M. Meydani, S. Yong, J. Pyle, Y. Wan, R. Arvizu-
Durazo and Y. Liao, Am. J. Epidemiol., 2000, 151, 109–118.
12 V. Tyssandier, E. Reboul, J. F. Dumas, C. Bougteloup-Demange,
M. Armand, J. Marcand, M. Sallas and P. Borel, Am. J. Physiol.:
Gastrointest. Liver Physiol., 2003, 284, G913–G923.
13 J. Kanner and T. Lapidot, Free Radical Biol. Med., 2001, 31,
1388–1395.
14 B. Halliwell, K. Zhao and M. Whiteman, Free Radical Res., 2001,
33, 819–830.
15 J. S. Cohn, Curr. Opin. Lipidol., 2002, 13, 19–24.
16 E. Vulcain, P. Goupy, C. Caris-Veyrat and O. Dangles,
Free Radical Res., 2005, 39, 547–563.
34 C. P. Baron and H. J. Andersen, J. Agric. Food Chem., 2002, 50,
3887–3897.
35 C. U. Carlsen, J. S. K. Møller and L. H. Skibsted, Coord. Chem.
Rev., 2005, 249, 485–498.
36 C. U. Carlsen, M. V. Kroger-Ohlsen, R. Bellio and L. H. Skibsted,
J. Agric. Food Chem., 2000, 48, 204–212.
37 L. V. Jorgensen, H. J. Andersen and L. H. Skibsted, Free Radical
Res., 1997, 27, 73–87.
38 L. V. Jorgensen and L. H. Skibsted, Free Radical Res., 1998, 28,
335–351.
39 A. Ben-Aziz, G. Britton and T. W. Goodwin, Phytochemistry,
1969, 12, 2759–2764.
17 J. W. Erdman, N. A. Ford and B. L. Lindshield, Arch. Biochem.
Biophys., 2009, 483, 229–235.
40 H. E. Gottlieb, V. Kotlyar and A. Nudelman, J. Org. Chem., 1997,
62, 7512–7515.
18 M. Gajic, S. Zaripheh, F. R. Sun and J. W. Erdman, J. Nutr., 2006,
136, 1552–1557.
41 A. Mikkelsen and L. H. Skibsted, Z. Lebensm.-Unters. Forsch.,
1995, 200, 171–177.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 575–587 587