Feruloyl dioleoylglycerol antioxidant capacity in phospholipid vesicles

Article abstract of DOI:10.1021/jf100356a

Ferulic acid and its esters are known to be effective antioxidants. Feruloyl dioleoylglycerol was assessed for its ability to serve as an antioxidant in model membrane phospholipid vesicles. The molecule was incorporated into single-lamellar vesicles of 1,2-dioleoyl-sn-glycero-3- phosphocholine at 1 and 5 mol fractions. Employing a lipid peroxidation inhibition assay, feruloyl dioleoylglycerol was demonstrated to express an oxidation protection ratio relative to Trolox of 0.94 and 0.74 at the 1% and 5% incorporation levels, respectively. The impact of feruloyl dioleoylglycerol incorporation on vesicle integrity was examined by determining calcein-cobalt complex leakage rates. Vesicle leakage was not influenced at 22 or 37 °C with 5% feruloyl dioleoylglycerol incorporation in comparison to that of vesicles lacking feruloyl dioleoylglycerol. Resonance energy transfer analysis showed that the closest approach distance between feruloyl dioleoylglycerol and 1,2-dioleoyl-sn-glycero-3-phospho-l-serine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) was approximately 31 A, which indicated that feruloyl dioleoylglycerol was thoroughly distributed throughout the bilayer plane. Conformational analysis determined that feruloyl dioleoylglycerol has a splayed conformation in which its feruloyl moiety is not closely contacted by its oleoyl groups. Feruloyl dioleoylglycerol integrates into the bilayer with its feruloyl moiety oriented close to the hydrophilic/lipophilic interface and its oleoyl groups extended deeply in the membrane. These findings indicate that feruloyl dioleoylglycerol expresses antioxidant activity by intercepting aqueous-phase free radicals as they penetrate the bilayer. This article not subject to U.S. Copyright. Published 2010 by the American Chemical Society.

Full text of DOI:10.1021/jf100356a

5842 J. Agric. Food Chem. 2010, 58, 5842–5850
DOI:10.1021/jf100356a
Feruloyl Dioleoylglycerol Antioxidant Capacity in
Phospholipid Vesicles
‡
,
J
OSEPH A. LASZLO,*,† KERVIN O. EVANS,† KARL E. VERMILLION
§
AND
M
ICHAEL
A
PPELL
†Renewable Product Technology, USDA-Agricultural Research Service, National Center for Agricultural
Utilization Research, 1815 N. University Street, Peoria, Illinois 61604, ‡Functional Foods Research,
USDA-Agricultural Research Service, National Center for Agricultural Utilization Research,
1815 N. University Street, Peoria, Illinois 61604, and §Bacterial Foodborne Pathogens and Mycology,
USDA-Agricultural Research Service, National Center for Agricultural Utilization Research,
1815 N. University Street, Peoria, Illinois 61604
Ferulic acid and its esters are known to be effective antioxidants. Feruloyl dioleoylglycerol was
assessed for its ability to serve as an antioxidant in model membrane phospholipid vesicles. The
molecule was incorporated into single-lamellar vesicles of 1,2-dioleoyl-sn-glycero-3-phosphocholine
at 1 and 5 mol fractions. Employing a lipid peroxidation inhibition assay, feruloyl dioleoylglycerol was
demonstrated to express an oxidation protection ratio relative to Trolox of 0.94 and 0.74 at the 1%
and 5% incorporation levels, respectively. The impact of feruloyl dioleoylglycerol incorporation on
vesicle integrity was examined by determining calcein;cobalt complex leakage rates. Vesicle
leakage was not influenced at 22 or 37 °C with 5% feruloyl dioleoylglycerol incorporation in
comparison to that of vesicles lacking feruloyl dioleoylglycerol. Resonance energy transfer analysis
showed that the closest approach distance between feruloyl dioleoylglycerol and 1,2-dioleoyl-sn-
˚
glycero-3-phospho-L-serine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) was approximately 31 A, which
indicated that feruloyl dioleoylglycerol was thoroughly distributed throughout the bilayer plane.
Conformational analysis determined that feruloyl dioleoylglycerol has a splayed conformation in
which its feruloyl moiety is not closely contacted by its oleoyl groups. Feruloyl dioleoylglycerol
integrates into the bilayer with its feruloyl moiety oriented close to the hydrophilic/lipophilic interface
and its oleoyl groups extended deeply in the membrane. These findings indicate that feruloyl
dioleoylglycerol expresses antioxidant activity by intercepting aqueous-phase free radicals as they
penetrate the bilayer.
KEYWORDS: Ferulic acid; triglyceride; antioxidant; membrane bilayer; resonance energy transfer
(RET); density functional theory; molecular modeling; nuclear Overhouser effect spectroscopy
(NOESY)
INTRODUCTION
hydroxy acids (1). Just as this multitude of cinnamate conjugates
has broad utility in plants, animals too may have many avenues to
exploit these molecules.
The cinnamates ferulic acid (3-(30-methoxy-40-hydroxyphenyl)-
propenoic acid) and caffeic acid (3-(30,40-dihydroxyphenyl)-pro-
penoic acid) are common and plentiful dietary components (1,2).
Derived from plant sources, ferulic and caffeic acids come as both
free acids and various esters thereof. Their esters are found
predominately as architectural components of the hemicellulose
and lignin fractions of the plant cell wall, as well as in suberin and
cutin waxy surfaces of leaves and other plant parts. Feruloyl esters
of ω-hydroxyalkanoic acids and long-chain alkan-1-ols have been
isolated from potato periderm (3, 4). A caffeoyl-ω-hydroxyfatty
acyl glycerol was isolated from cotton (5). Feruloyl glycerols have
been identified (3, 6). Ferulic acid is found esterified to phytoster-
ols in grain products such as corn and rice bran oils (7). Caffeic
acid is widely found as conjugates of quinic acid and related
Because of their anticipated low toxicity, cinnamate derivatives
have been closely examined for their medicinal value (8). Alkyl
esters of ferulic and caffeic acid are able to perform as anti-
inflammatories and cancerogenesis suppressors. Murakami and
co-workers (9, 10) found that 2-methyl-1-butyl ferulate attenu-
ated the inflammatory events of inducible nitric oxide synthase/
cyclooxygenase-2 expression and superoxide anion generation.
This molecule also suppressed phorbol ester-induced Epstein-
Barr virus activation and papilloma development in mouse skin.
Long-chain alkyl esters of caffeic acid inhibit nitric oxide produc-
tion inmurinemacrophagecells, suggestingananti-inflammatory
activity (11). Ferulate and caffeic alkyl esters demonstrate anti-
tumor proliferation activity with colon breast, lung, and gastric
human cell lines, as well as inhibition of COX-1 and COX-2
enzymes (12). All these activities are attributable to the cellular
*Corresponding author. Tel: 309-681-6322. Fax: 309-681-6686.
E-mail: Joe.Laszlo@ars.usda.gov.
pubs.acs.org/JAFC
Published on Web 04/15/2010
This article not subject to U.S. Copyright. Published 2010 by the American Chemical Society

Article
J. Agric. Food Chem., Vol. 58, No. 9, 2010 5843
incorporation efficacies of the lipophilic cinnamates. However,
significant cytotoxicity occurs with long-chain alkyl esters of
ferulic or caffeic acids (11, 13). It thus may be valuable to look
for cinnamate derivatives displaying lower cytotoxic potential
while retaining the indicated beneficial attributes.
product was an oil at room temperature. 1H NMR (d6-acetone): δ (ppm)
7.65 (1 H, d, dCH-Ar), 7.37 (1 H, s, Ar-H), 7.16 (1 H, d, Ar-H), 6.89 (1
H, d, Ar-H), 6.43 (1 H, d, CHdCH-Ar), 5.36 (m, 2-CH2CHdCHCH2),
4.1-4.5 (4 H, m, 2-CH2O), 3.95 (3 H, s, OCH3), 2.35 (m, 2-CH2CH2CO2),
2.04 (m, 2-(CH2CH2)2CHdCH), 1.63 (m, 2-CH2CH2CH3), 1.38 (m, 18-
CH2CH2CH2), 0.90 (m, 2-CH3). 13C{1H} NMR (d6-acetone): δ (ppm)
172.3 (CH2CH2CO2), 166.2 (CHdCHCO2), 149.4 (Ar), 148.0 (Ar), 145.7
(Ar-CHd), 130.0 (2-CH2CHdCHCH2), 126.5 (Ar), 123.3 (Ar), 115.2 (Ar),
114.1(CHdCHCO2), 110.4(Ar), 69.2 (CH(CH2O-)2), 61.9 (CH(CH2O-)2),
55.6 (CH3O), 33.7 (2-CH2CH2CO2), 31.6 (2-CH2CH2CH3), 29.1 (aliphatic),
27.0 (2-CH2CHdCHCH2), 24.8 (2-CH2CH2CHdCH CH2CH2), 22.7
(2-CH3CH2-), 13.5 (2-CH3-). Electrospray ionization mass spectroscopy
(ESI-MS): 795.7 [M - H]- (calcd C49H80O8: 796.6). FDOG molar absor-
ptivity: 1.93 ꢀ 104 M-1 cm-1 at 325 nm in ethanol.
Lipid peroxidation products of polyunsaturated fatty acids
play a major role in genotoxicity of the cell (14). Free radical-
initiated lipid peroxidation causes the loss of membrane structure
and so too gives rise to toxic products such as 4-hydroxyalkenals
and other aldehydes that in principle can move from the site of
origin and produce effects in distant proteins and nucleic acids.
Such distal molecular structures are protected by antioxidant
quenching of free-radical propagation events in membranes. The
physiological influences evinced by lipophilic cinnamates may in
part be a result of their substantial antioxidant capacity. Lipo-
philic cinnamate esters of vegetable oils are readily produced by
enzymatic transesterification (16). In vitro tests of feruloyl
dioleoylglycerol (FDOG) demonstrated that these compounds
have significant antioxidant capacity (17). However, there is
much to learn about the physiological performance of such
cinnamate esters.
A detailed understanding of the biological mechanisms by
which lipophilic cinnamate esters elicit their various effects is
lacking. Single-lamellar phospholipid vesicles can serve as con-
venient models for cellular membranes. Demonstration of anti-
oxidant incorporation into such vesicles should be central to
having the expected lipid protection effect. Furthermore, minimal
loss of vesicle integrity with antioxidant insertion may serve as a
measure of low cytotoxicity potential. FDOG was examined in
the present work to determine whether it incorporates into
phospholipid vesicles with the retention of antioxidant capacity
and without membrane integrity disruption. The localization of
the feruloyl group within the phospholipid bilayer was investi-
gated to determine its proximity to potential oxidation-sensitive
lipid double-bond sites. Conformational analysis was employed
to assess FDOG structure and orientation within the bilayer.
On the basis of 1H NMR and heteronuclear multiple bond correlation
(HMBC) analysis, it was determined that of the two potential FDOG
regioisomers, 88% was 1(3)-feruloyl-2,3(1)-dioleoyl glycerol and 12% was
2-feruloyl-1,3-dioleoyl glycerol (see Supporting Information).
DPPH Assay. The free radical scavenging capability of FDOG, was
determined using the DPPH assay with modifications as described by
Choo and Birch (17). DPPH spontaneously forms a free radical in
solution. FDOG (2.5 mM) was mixed with 0.2 mM DPPH in ethanol.
Ferulic acid and ethyl ferulate were also examined for comparison. DPPH
radical reduction was followed spectrophotometrically at 517 nm for 30 min
at 2-min intervals using a Shimadzu 1240 UV-vis spectrophotometer. The
reaction was conducted at room temperature (approximately 23 °C).
Triplicate runs of antioxidant at each concentration were performed. The
percentage of DPPH radical remaining over time was calculated as the
sample absorbance divided by that of a control DPPH sample at the same
time, multiplied by 100.
Lipid Preparation and Formation of Unilamellar Phospholipid
Vesicles. Lipids and vesicles were prepared in a manner similar to that
used by MacDonald and co-workers (19) with some modifications. Lipids
were prepared from appropriate amounts of DOPC in chloroform and
FDOG (at 1- or 5-mol % of total lipids) in ethanol added to clean amber
vials and then gently mixed. Chloroform and ethanol were removed by
warming and gently blowing argon into the vials for several minutes until a
dried lipid film was present. Residual solvent was removed by keeping the
dried films under vacuum overnight. The dried lipid films were stored
under argon at -20 °C until needed.
To form phospholipid vesicles, the dried lipid films were allowed to
equilibrate to ambient temperature, rehydrated in the appropriate buffer,
and mixed periodically over an hour. The multilamellar vesicles that
formed from the rehydration were then put through five freeze-thaw
cycles (using 2-propanol on dry ice and water heated to 60 °C). These
vesicles were then extruded 11 times through two, stacked, polycarbonate
100-nm filters in a LiposoFast extruder (Avestin, Inc.; Ottawa, Canada) to
form unilamellar vesicles (approximately 125 nm in diameter). Vesicles
were covered with argon, protected from light, and stored at 4 °C until
used.
Vesicle Leakage (Calcein-Cobalt). Vesicles with and without
FDOG were evaluated for membrane integrity using a leakage assay
described by Evans (20). The column buffer was modified to include
17 mM NaCl for osmotic balance during the separation of the vesicles
from the unincorporated calcein-cobalt complex. The leakage assay was
designed such that the low fluorescent signal of the entrapped calcein-
cobalt complex increased as vesicles leaked their contents, and cobalt was
stripped from calcein by external EDTA. Vesicle leakage was monitored
spectroscopically on a Fluorolog 3-21 fluorometer. The excitation and
emission wavelengths were 490 and 520 nm, respectively, with 2-nm slit
widths apiece. Samples were continuously stirred in 1-cm path length
quartz cuvettes. Measurements were recorded for 30 min at 5-s intervals.
Experiments were conducted in triplicate at 22 and 37 °C using a total lipid
concentration of 0.15 mM. Leakage percentage was determined from the
intensity of the fluorescent signal, 100 ꢀ (Ft - Fo)/(Fmax - Fo), where Ft
was the fluorescent signal over time, Fo was the extrapolated initial signal
at time zero, and Fmax was the fluorescent signal at the end of the
experiment after complete rupture by addition of 1% v/v aqueous Triton
X-100 solution (20 μL added to 2 mL to make a final concentration of
0.01% v/v Triton X-100). The apparent rate constant k was obtained by
fitting to the equation: Ft = Fmax (1 - exp(-kt)).
MATERIALS AND METHODS
Reagents and Materials. Ethyl ferulate was obtained from Sinova
Corp. (Ningbo, China). Triolein, ferulic acid 2,2-diphenyl-1-picrylhydra-
zyl (DPPH), cobalt(II) chloride hexahydrate, ethylenediaminetetraacetic
acid disodium dehydrate (EDTA), 2,20-azobis(2-amidinopropane) dihy-
drochloride (AAPH), and Triton X-100 were purchased from Sigma-
Aldrich (St. Louis, MO). Novozym 435 was purchased from Novozymes
through Brenntag Great Lakes (Chicago, IL). 1,2-Dioleoyl-sn-glycero-3-
phosphocholine (DOPC), 1, 2-dioleoyl-sn-glycero-3-phospho(TEMPO)-
choline (TEMPO-PC), 1-palmitoyl-2-stearoyl-(5-DOXYL)-sn-glycero-3-
phosphocholine (5DOXYL-PC), 1-palmitoyl-2-stearoyl-(12-DOXYL)-
sn-glycero-3-phosphocholine (12DOXYL-PC), and 1,2-dioleoyl-sn-gly-
cero-3-phospho-L-serine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (ammonium
salt) (NBD-PS) were purchased from Avanti Polar Lipids (Alabaster, AL).
High purity calcein and 4,4-difluoro-5-(4-phenyl-1,3-butadienyl)-4-bora-
3a,4a-diaza-s-indacene-3-undecanoic acid (C11-Bodipy 581/591) were pur-
chased from Invitrogen (Carlsbad, CA). Potassium phosphate monobasic,
potassium phosphate dibasic, Sephadex G-75 column beads, columns, and
sodium chloride were all obtained from Fisher Scientific.
Feruloyl Dioleoylglycerol (FDOG) Preparation. Ethyl ferulate
(2.5 g), triolein (10 g), and immobilized Candida antarctica lipase B
(2.5 g of Novozym 435) were combined and shaken at 60 °C for one
week (15, 18). Novozym 435 was removed by filtration (nylon filter). The
reaction product was subjected to flash chromatography using a C18
reverse phase column and a 60:40 v/v acetonitrile/acetone solvent, with UV
detection at 340 nm. The FDOG containing peak was subjected to further
purification by HPLC using a Luna semipreparative phenylhexyl column
(250 mm ꢀ10 mm, 10 μm; Phenomenex, Torrance, CA) developed with
methanol at a flow rate of 5 mL/min. Details of analytic HPLC and mass
spectroscopy procedures have been given previously (18). The FDOG

5844 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Laszlo et al.
Lipid Oxidation Inhibition. Lipid oxidation inhibition by FDOG
was measured by the lipid peroxidation inhibition capacity (LPIC) assay,
which employs AAPH as a free-radical initiator and the lipidic reporter
molecule C11-Bodipy (21). DOPC and C11-Bodipy were combined in a
vial, dried to a film, and then rehydrated in 20 mM Tris-HCl, pH 7.4, to
give a concentration of 5.0 mM DOPC and 4.8 μM C11-Bodipy. FDOG (at
1- or 5-mol % of DOPC) from ethanol was included in the preparation of
the DOPC/C11-Bodipy lipid film prior to the formation of vesicles, in those
instances where FDOG LPIC was determined. Unilamellar vesicles were
prepared from rehydrated lipids as described above. Samples were diluted
to a final concentration of 2.5 mM DOPC and 2.4 μM C11-Bodipy and
allowed to equilibrate to 37 °C for 15 min prior to measurement. From the
time of AAPH addition (40 mM final concentration), fluorescence
intensity (excitation at 540 nm, emission at 595 nm) was measured every
minute for 180 min. Experiments were conducted in triplicate. The relative
fluorescence intensity was calculated by dividing the reading at each
minute by the initial fluorescence at the beginning of the experiment
(Ft/Fo). Background subtraction was accomplished using vesicles contain-
ing only DOPC.
should experience the most quenching from the nearest nitroxide group,
therefore allowing calculation of the distance between feruloyl group and
the bilayer center. Nitroxide-labeled lipids were separately incorporated at
15 mol % into the membrane in conjunction with FDOG at various
concentrations. Measurements were conducted in potassium phosphate
buffer at pH 7.0. Fluorescence emission spectra were recorded using an
excitation wavelength of 330 nm. The excitation and emission slit widths
were 5 nm. A background emission was subtracted using DOPC vesicles,
with or without nitroxide-labeled lipids as appropriate.
The difference in quenching of FDOG by the nitroxide-labeled lipids
indicates a difference in depth penetration by FDOG in the membrane.
Accordingly, the quenching differences allow the calculation of the
probable location of FDOG in a membrane using the pallalax equa-
tion (29) as follows:
zcf ¼ Lc1 þ ½ - lnðF1=F2Þ=πC - L2 ꢁ=2L21
ð3Þ
21
In the parallax eq 3, zcf is the distance the fluorophores are from
the center of the membrane, F1 is the normalized emission fluorescence
intensity (normalized in respect to vesicles with FDOG only) in the pre-
sence of the shallow quencher, F2 is the normalized fluorescence intensity
in the presence of the deep quencher, Lcl is the distance between the shallow
quencher and the center of the bilayer, C is the concentration of quenchers
per area (mole fraction of nitroxide-labeled lipids/area of lipid), and L21 is
the difference in depth between quenchers. The area of the phosphatidyl-
The water-soluble antioxidant Trolox was used for comparison. Trolox
was combined with DOPC and C11-Bodipy, dried, and then rehydrated as
described above. A Trolox concentration range of 10 to 125 μM (before
dilution with AAPH) was employed. The peroxyl radical scavenging
capacity of FDOG relative to Trolox (relative antioxidation activity)
was calculated using the following equation:
2
˚
choline lipids was taken to be 70.1 A per molecule (30), and the distances
˚
from the bilayer center for the nitroxide group were assumed as 19.5 A for
ðAUCSample - AUC0Þ molTrolox
¼
ꢀ
ð1Þ
˚
˚
TEMPO-PC, 12.2 A for 5DOXYL-PC, and 5.9 A for 12DOXYL-PC (28).
Preliminary tests found that the best results using the strongest quenching
pairs (shallow versus deep quenchers) were TEMPO-PC/DOPC versus
5DOXYL-PC/DOPC for vesicles containing 1.0 mol % FDOG and
5DOXYL-PC/DOPC versus 12DOXYL-PC/DOPC for vesicles with
5.0-mol % FDOG.
ðAUCTrolox - AUC0Þ molSample
where AUCSample is the area under the fluorescence curve (AUC) for
vesicles containing antioxidant (Trolox or FDOG), AUCTrolox is the area
under the curve for vesicles containing 50 μM Trolox (thereby arbitrarily
assigned a relative antioxidation activity value of 1.0), AUC0 is the area
under the curve for vesicles containing only Bodipy, and molSample is the
molar concentration of the antioxidant (22). The AUC was calculated by
three different methods: those described by Cao and co-workers (23),
Naguib (24), and Zhang and co-workers (21).
FDOG Bilayer Distribution Using Fluorescent Probe Energy
Transfer. The mobility and distribution of FDOG in the bilayer was
determined from the resonance energy transfer technique (25,26). FDOG
was treated as the donor and NBD-PS as the acceptor. Vesicles contained
1-mol % FDOG with 0.01-, 0.1-, or 1.0-mol % NBD-PS, or 5-mol %
FDOG with 0.05-, 0.5-, or 5.0-mol % NBD-PS. Measurements were
conducted in 10 mM potassium phosphate buffer, pH 7.0, at 37 °C.
Fluorescence emission spectra were recorded using an excitation wave-
length of 330 nm. The excitation and emission slit widths were 5 nm.
Spectral scattering and NBD-PS contributions to the emission spectra
were taken into account by subtracting the spectra of vesicles containing
the appropriate concentration of NBD-PS absent FDOG.
Conformational Analysis. NMR Analysis. 1H nuclear Overhauser
effect spectroscopy (NOESY) was conducted on a Bruker Avance 500
spectrometer using a 5-mm BBO probe with Z-gradient and Topspin 1.3
software. The sample was dissolved in CDCl3 and degassed with bubbling
argon. Spectra were acquired at 300.0 K. Chemical shifts are reported as
parts per million from tetramethylsilane based on the lock solvent. The
NOESY spectrum was acquired with a mixing time of 0.5 s using the
Bruker pulse program noesyetgp. There were 96 acquisitions per slice, and
the total experiment time was approximately 19 h.
Molecular Modeling. Initial structures were built using the AMBER
force field as implemented in the HyperChem v7.52 program (Hypercube
Inc., Gainsville, Florida). This force field approach was demonstrated to
be suitable for the conformational analysis of related lipids (31). Quantum
chemical calculations were carried out using Parallel Quantum Solutions
hardware and software (v3.2) (Parallel Quantum Solutions, Fayetteville,
Arkansas). Select conformations were investigated using Becke’s three
parameter functional with Lang-Yang-Paar functionals (B3LYP) at the
6-311þþg(d,p) level. Unless otherwise noted, geometry optimizations
were performed with the convergence criteria set at 1 ꢀ 10-6 Hartree
and a gradient of less than 3 ꢀ 10-4 a.u. using the eigenvector following
algorithm. The Conductor-like screening model (COSMO) continuum
solvation model was employed for implicit solvation (chloroform) (32).
The additional atomic radius for solvent accessible surfaces was set at
An overlap of the excitation of NBD-PS (acceptor) with the emission
spectrum of FDOG (donor) indicated that resonance energy transfer was
possible (see Supporting Information, Figure 8). This spectral overlap
€
allowed the determination of the Forster radius (Ro) according to the
€
Forster equation:
1=6
Ro ¼ 0:211½K2n- 4QDJðλÞꢁ
ð2Þ
˚
2.48 A with the electrical permittivity of the dielectric continuum set at
4.90. Results were imaged using Hyperchem v7.52.
where QD is the quantum yield of FDOG in the absence of NBD-PS, n is
the refractive index of the bilayer, J(λ) is the overlap integral between
FDOG and NBD-PS, and κ is the orientation factor of the transition
dipoles of FDOG and NBD-PS (assumed to be 2/3 for dynamic random
averaging for donor-acceptor pair). The distance from the bilayer center
RESULTS
DPPH Antioxidant Capacity. The antioxidant activity of a
phenolic compound is attributable to its scavenger activity
toward several free radicals, such as alkyl, alkoxyl, and peroxyl
radicals (R•, RO•, and ROO•, respectively). Such radicals are
formed during the oxidation of fats and oils. The rapidity and
completeness with which an antioxidant reacts with the stable free
radical DPPH• in organic solvent is a common measure of
antioxidant capacity. Figure 1 demonstrates that FDOG had an
antioxidant capacity greater than that of ferulic acid or ethyl
ferulate on an equivalent molar basis. Triolein does not quench
˚
for the NBD group was assumed to be 18.8 A at pH 7.0 (27) .
Feruloyl Group Bilayer Location Using Nitroxide-Labeled Lipids.
The depth of the feruloyl group of FDOG within the vesicle phospholipid
bilayer was determined by parallax analysis at 37 °C (28, 29). Briefly,
FDOG fluorescence intensity was reduced by collisional quenching caused
by nitroxide-labeled phosphatidylcholine (PC). The nitroxide-labeled
lipids used had the nitroxide quencher positioned in the headgroup region
(TEMPO-PC), just below the headgroup region (5DOXYL-PC), or
farther down the acyl chain (12DOXYL-PC). Conceptually, FDOG

Article
J. Agric. Food Chem., Vol. 58, No. 9, 2010 5845
Figure 1. Time course for the reaction of ferulate species with the DPPH
free radical. The concentrations were 2.5 mM for FDOG, ferulic acid, or
ethyl ferulate. Error bars represent one standard deviation from the mean
(n = 3).
DPPH• (17), indicating that the feruloyl moiety was responsible
for the observedscavenging activity. In our hands, FDOG was far
more reactive with DPPH• than previously reported (17). Using
the same solvent (ethanol) and 25 mM FDOG, a 10-fold higher
concentration of FDOG than employed herein, Choo and
Birch (17) observed poor scavenging activity (approximately
35% DPPH• remaining after 30 min). They found 25 mM ferulic
acid to be much more capable than FDOG as a DPPH• scavenger
(less than 10% DPPH• remaining after 2 min with ferulic acid).
The reason for the DPPH activity disparity in FDOG results is
unclear.
Lipid Peroxidation Inhibition. The ability of FDOG to express
its antioxidant properties under conditions more relevant to an in
vivo molecular environment was examined. FDOG was incorpo-
rated into unilamellar phosphocholine vesicles and subjected to
oxidative conditions using the free-radical initiator AAPH. C11-
Bodipy, which fluoresces until oxidized, served as a reporter lipid
within the vesicles to follow the progression of the free-radical
attack on the membrane. The conventional antioxidant Trolox
also was examined to provide a basis for comparison.
Figure 2. Comparison of lipid peroxidation inhibition capacities of Trolox
and FDOG in DOPC vesicles at 37 °C. Upper panel: relative fluorescence
changes of C11-Bodipy in the presence of 40 mM AAPH and Trolox at
various concentrations. The insert is a plot of the net area under the curve
(ΔAUC) versus Trolox concentration. Lower panel: relative fluorescence
changes of C11-Bodipy in the presence of 40 mM AAPH and FDOG at
various concentrations (0, 1, and 5 mol % of total lipids). The insert is a plot
of the ΔAUC versus FDOG concentration.
decay rate also decreased with increased FDOG concentration,
and the ΔAUC for FDOG demonstrated a linear correlation with
concentration (Figure 2, lower panel, insert).
The ratio of antioxidant activity of 1- and 5-mol % FDOG
relative to Trolox (eq 1) was determined to be 0.94 and 0.74,
respectively (Table 1). This lower relative antioxidant activity
value and the lack of a plateau region indicated that FDOG,
although able to retard oxidation, was not as capable of an
antioxidant as Trolox under the conditions of the assay. How-
ever, these results confirmed that FDOG retains its antioxidation
capacity when incorporated into a model phospholipid bilayer.
Vesicle Leakage. The ability of FDOG to destabilize phospho-
lipid vesicles was determined by monitoring the release of
entrapped contents (calcein-cobalt complex). Figure 3 shows
that the leakage ofDOPC vesicles at 22°C proceeded equally with
or without FDOG present. Further analysis of the kinetics
confirmed that vesicle leakage was not altered by FDOG inclu-
sion in the membrane at this temperature (Table 2). At a higher,
physiologically significant temperature (37 °C), vesicles exhibited
a slightly increased leakage rate, as expected (33). Again, FDOG
incorporation at 1- and 5-mol % levels did not significantly (P >
0.05) alter the leakage rate relative to that of vesicles lacking
FDOG (Table 2). Small changes in leakage rates are ascribed to
Figure 2 (upper panel) shows the time-dependent relative
fluorescence of C11-Bodipy in the presence of 40 mM AAPH
and varying Trolox concentrations. With no AAPH present, the
relative fluorescence of C11-Bodipy remained steady, demonstrat-
ing the stability of the probe under these conditions. Upon the
addition of AAPH, the relative fluorescence rapidly decreased, as
would be expected when no antioxidant protection of C11-Bodipy
was provided. With Trolox present, a plateau in the relative
fluorescence developed and extended with increased concentra-
tion of Trolox, as demonstrated by Zhang and co-workers (21). A
linear correlation between the net area under the curve (ΔAUC)
and Trolox concentration was also evident (Figure 2, upper panel,
insert).
Figure 2 (lower panel) shows the relative fluorescence of C11-
Bodipy in the presence of FDOG at two different concentrations
(1 and 5 mol %). FDOG exhibited no plateau phase in the relative
fluorescence like that provided by Trolox. However, the decay
rate of the C11-Bodipy relative fluorescence with FDOG present
was slower than the decay rate without FDOG. The fluorescence

5846 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Laszlo et al.
Table 1. Relative Antioxidant Capacity of Trolox and FDOG in the LPIC Assay
relative antioxidant capacitya
Table 2. Rates of Calcein-Cobalt Complex Leakage from DOPC Vesicles
with or without Incorporated FDOG
temperature (°C)
FDOG (mol %)
leakage rate constant (ms-1
)
concentration
(μM)
AUC
method A
AUC
method B
AUC
method C
antioxidant
22
22
22
0
1
5
2.24 ( 0.32
2.29 ( 0.16
2.32 ( 0.35
Trolox
Trolox
Trolox
FDOG
FDOG
10
1.03
1.00
1.04
0.94
0.74
1.03
1.00
1.04
0.94
0.75
1.03
1.00
1.04
0.94
0.74
50
125
25
37
37
37
0
1
5
2.57 ( 0.07
3.65 ( 0.78
2.40 ( 0.64
125
a Relative antioxidant activity calculated according to eq 1. Method A: AUC
calculated according to Zhang and co-workers (21). Method B: AUC calculated
according to Cao and co-workers (23). Method C: AUC calculated according to
Naguib (24).
Figure 3. Time dependence of calcein-cobalt leakage from the DOPC
vesicle without and with 1- or 5-mol % FDOG at 22 °C. Data displayed
represent the mean values of triplicate analyses.
changes in membrane fluidity (33). An order of magnitude
increase in leakage rates would be expected for small pore
formation, and instantaneous leakage would be observed with
vesicle disruption (34). Therefore, over the temperature range
explored FDOG did not induce any substantial disturbance of the
phospholipid bilayer fluidity or intactness that could be correlated
with changes in the release rate of vesicle-entrapped contents.
Mobility and Distribution in the Membrane. That membrane-
incorporated FDOG displays antioxidant capacity and does not
adversely modify membrane permeability suggests the molecule is
mobile and distributed throughout the membrane plane. To test
this hypothesis, the average distance between FDOG and other
vesicle phospholipids was determined. The NBD-PS excitation
spectrum and FDOG fluorescence emission spectrum were found
to overlap (Supporting Information, Figure 8). Therefore, these
two molecules made a good resonance energy transfer pair
(nonradiative dipole-dipole coupling) that could be used to
Figure 4. Fluorescence spectra of FDOG at (upper panel) 1 mol % and
(lower panel) 5 mol % in phospholipid vesicles with and without quenchers.
The composition of the vesicles was DOPC/FDOG (no quencher, 99:1 or
99:5), DOPC/TEMPO-PC/FDOG, DOPC/5DOXY-PC, or 12DOXYL-PC
(all at 84:15:1 or 80:15:5).
€
calculate the Forster distance, i.e., the distance at which the
Figure 8). This means that the equation for energy transfer
efficiency (E = 1 - FDA/FD, where FDA and FD are the fluo-
rescence of the donor in the presence and absence of the acceptor,
respectively) is valid. Therefore, the energy transfer efficiency was
used to calculate the closest approach distance between FDOG
energy transfer efficiency is 50%, between FDOG and NBD-PS.
€
To calculate the Forster distance, a corrected emission spectrum
and the quantum yield of FDOG were needed. On the basis of the
method described by Peng and co-workers (35) and using 9,10-
diphenylanthracene as the reference standard (quantum yield of
0.95 (36)), the quantum yield of FDOG was determined to be 0.55
˚
and NBD-PS to be approximately 31 A. This is well within the
range of closest approach distances found for other fluorescent
lipids (26).
€
in ethanol and 0.08 in DOPC vesicles. The Forster distance for
FDOG in DOPC vesicles was therefore calculated to be approxi-
mately 29.8 A.
When the sample was excited at 330 nm, nearly all of the fluo-
rescence at 530 nm was from NBD-PS (Supporting Information,
Energy transfer can occur between donor-acceptor pairs that
are both in the same bilayer plane (leaflet) and the opposing
bilayer plane. Therefore, the closest approach distance deter-
mined here is an average of energy transfer occurring within the
˚

Article
J. Agric. Food Chem., Vol. 58, No. 9, 2010 5847
Figure 5. NOESY analysis of FDOG intramolecular interactions. The data were collected with FDOG dissolved in CDCl3 using a 500 MHz instrument.
same bilayer plane and between opposite planes. Assuming that
energy transfer between FDOG and NBD-PS does occur between
pairs within the same bilayer plane and pairs in opposite planes,
then the closest approach distance determined here suggests that
FDOG and NBD-PS in opposite bilayer planes are able to pass
over the top of one another. The interpretation that FDOG and
NBD-PS can pass over the top of one another further suggests
that FDOG doesnot have restricted movementwithinthe bilayer.
The closest approach distance further excludes the possibility that
FDOG formed domains, such as those observed with cholesterol
˚
in bilayers. Typical domain sizes are 100 to 2000 A. It would be
expected that if FDOG formed domains the closest approach
distance between NBD-PS and FDOG would fall more within the
range of typical domain sizes. Therefore, FDOG at 1- and 5-mol
% appeared to be mobile and uniformly dispersed throughout the
bilayer, not sequestered in isolated domains such as those
observed with cholesterol in phospholipid bilayers.
Feruloyl Group Membrane Depth. FDOG incorporated at
1- and 5-mol % in DOPC vesicles had a strong fluorescent signal,
as can be observed in Figure 4. Fluorescence quenching of FDOG
was exhibited by phosphatidylcholine molecules bearing nitroxide
groups at various depths within the vesicle membrane. This
allowed the distance from the center of the bilayer, zcf, of FDOG’s
fluorophore, the feruloyl moiety, to be estimated (eq 3). Distances
˚
˚
of 16.8 ( 1.4 A and 15.7 ( 0.9 A were calculated for 1- and 5-mol %
FDOG, respectively. These values are not significantly different
˚
(P = 0.2); therefore, a mean value of 16.3 ( 1.2 A was assumed.
This distance places the feruloyl moiety within the beginning of
the hydrophilic region of the DOPC bilayer just above the
hydrophobic layer, but not fully extended into the aqueous
media (30, 37).
Conformational Analysis. NOESY analysis of FDOG was
conducted to identify the salient intramolecular interactions
(Figure 5). The NOESY spectrum (Supporting Information,
Figure 7) showed the expected interactions of the feruloyl
moiety’s olefin protons with each other and to the aromatic
protons. The methoxy protons interact with an aromatic proton.
The glyceryl protons also have the expected cross-talk peaks, as
Figure 6. Conformers of FDOG and ferulic acid (f1, f2, and f3). Relative
energies are in kcal/mol. Carbon atoms are displayed in green, oxygen
atoms in red, and hydrogen atoms in white.

5848 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Laszlo et al.
Figure 7. Molecular representation of FDOG embedded in a DOPC bilayer (5-mol % FDOG). FDOG is represented in brown. For DOPC, carbon atoms are
displayed in green, oxygen atoms in red, nitrogen atoms in blue, and hydrogen atoms in white. Calculations were carried out using the AMBER force field to
position the molecular components. A simple, two-molecule deep representation of the bilayer was extracted to facilitate visualization, which accounts for the
apparent void spaces in the image.
˚
do the oleoyl moieties, but there was no indication that the
feruloyl group interacts, within the 4.5-A limit of detection, with
2.32 A, and this conformation is within 0.28 kcal/mol of the
preferred conformation f1 (ΔEChloroform = 0.13 kcal/mol). These
calculations matched the NOESY analysis and suggest that
FDOG can adopt two stable orientations of the ferulic acid
moiety (Figure 6, f1 and f3). The entire FDOG molecule opti-
mized at the PM3 level is shown in Figure 6. Again, these
calculations matched the NOESY analysis, which found no
interaction of the oleic acid groups with the feruloyl moiety.
A model showing the FDOG in a dioleoyl phosphatidylcholine
membrane bilayer is given in Figure 7. The length of extended
˚
either the glyceryl or acyl fatty acid features of the molecule. The
lack of intramolecular interactions between the feruloyl moiety
and the glycerol backbone or the other acyl chains is in marked
contrast to that reported for ferulates esterified directly to
straight, long-chain alkyl groups (31, 38). NOESY experiments
are relatively insensitive; therefore, there were no discernible
cross-peaks associated with the 12% 2-feruloyl-1,3-dioleoylglycerol
regioisomer present in the sample.
FDOG is a large lipid with a feruloyl moiety, glycerol linker,
and two oleic acid chains (Figure 5). Smaller but structurally
similar molecules have been investigated using the B3LYP density
functional at the 6-31G(d,p) level, including caffeate and gallate
esters (13,39), ferulic acid (38,40-42), and hydroxycinnamic acid
derivatives (43, 44). Initial attempts to geometry optimize the
complete structures of FDOG were complicated by incomplete
convergence, possibly due to shallow minimization wells asso-
ciated with the oleic chains. To address this issue, geometry
optimization first was carried out on the feruloyl moiety. A
conformational analysis of ferulic acid was carried out at the
B3LYP/6-311þþG(d,p) level. A preferred conformation, f1
(Figure 6), compared favorably with the experimental NOESY
findings. This structure possesses several signals observable by
˚
FDOG is approximately 33 A, in contrast to the approximate
˚
27 A length of dioleoyl phosphatidylcholine. The feruloyl moiety
oriented so as to be near the hydrophobic/hydrophilic interface,
which is consistent with its placement determined by resonance
energy transfer measurements. The oleoyl groups reach deeply
into the bilayer and can extend across to the opposing leaflet.
Thus the conformational analyses indicated that FDOG exhibits
a splayed structure and that the entire, extended FDOG molecule
can be embedded within the dioleoyl phosphatidylcholine mem-
brane bilayer.
DISCUSSION
Hydroxycinnamates such as ferulic acid have been the subject
of extensive research because of their potential to counteract the
destructive action of reactive oxygen species. Antioxidant efficacy
reflects both its inherent reactivity (antioxidant capacity) and
proximity to those structures requiring protection. Hydroxycin-
namic acids and their hydrophilic esters are expected to be useful
for guarding against oxidation of aqueous phase cellular compo-
nents (e.g., proteins and nucleic acids). Lipophilic hydroxycinna-
mate esters would be expected to perform in membranes. Thus,
drawing marginal distinctions between whether a molecule is a
betterantioxidant inits water-soluble form (e.g., ferulic acid) or in
its lipophilic form (e.g., FDOG), on the basis of measures such as
the DPPH assay, are of limited physiological significance. To
assess antioxidant efficacy for membrane lipids, an antioxidant
˚
NOESY experiments, with a distance of 2.38 A between the
hydrogenofthe methoxy substituent and the C3-H. The distance
between the aromatic C3-H and the C8-H of the olefin is 2.20
˚
˚
A. The conformation features a favorable 2.09 A hydrogen bond
between the hydrogen of the C1 phenolic hydroxyl and the
oxygen of the methoxy substituent. A recent density functional
study investigated conformation f2 of ferulic acid, which lacks
this hydrogen bond interaction (40); however, this structure was
found to be unfavorable, and higher in energy by 6.99 kcal/mol in
vacuo over conformation f1 (5.42 kcal/mol higher in chloroform).
The experimental NMR data indicated NOE signals between
both the C3-H and C5-H with the C7-H of the olefin. The
distance between the C3-H with the C7-H in conformer f3 is

Article
J. Agric. Food Chem., Vol. 58, No. 9, 2010 5849
must demonstrate (1) a free-radical scavenging capacity (such as
by the DPPH method), (2) the ability to partition into membranes
at a significant concentration without disrupting the integrity of
the membrane, (3) an antioxidant capacity while incorporated
into a membrane, and (4) the prevention of ROS initiated
destruction of lipids proximal to the membrane-incorporated
antioxidant.
membranes at significant levels without disrupting the mem-
branes and displayed potent antioxidative capacity. Unlike
simple alkyl esters of ferulic acid, the triacylglycerol structure of
FDOG adopts an extended structure that places the feruloyl
group close to the membrane aqueous/lipid interface. Further-
more, FDOG disperses throughout the membrane, which should
afford greater protection to the membrane lipids and proteins.
FDOG was shown to have a free-radical scavenging capacity
(Figure 1). This is consistent with the findings of others with
lipophilic ferulate esters (17, 31, 45-47) and related phenols
(48-50). A range of ferulate alkyl esters (saturated C7 to C18,
linear, and branched) showed similar DPPH scavenging activ-
ity (31). This indicates that with saturated alkyl groups the side
chain length is not an important influence on antiradical activity
under homogeneous conditions such as the DPPH assay. This is
true for gallate esters as well (48). Esterification of dihydrocaffeic
acid with polyunsaturated alcohols (linolenyl alcohol) and into
polyunsaturated glycerides (linoleyl and linolenyl) results in rather
poor antiradical activity (49, 50). The present work demonstrated
that association of the ferulate moiety with monounsaturated
lipids (FDOG) leads to effective antiradical activity. Monoun-
saturated fatty acids are less prone to oxidation than their poly-
unsaturated counterparts, which may permit the feruloyl moiety an
opportunity to quench radical species not present within itself.
Moreover, FDOG retains its antioxidative capacity when incor-
porated into a bilayer (Figure 2 and Table 1) and disperses evenly
throughout the membrane plane, not forming domains. A posi-
tioning of the feruloyl moiety near the hydrophilic/hydrophobic
interface (Figure 7) may promote greater activity toward a flux of
radicals coming from the aqueous phase. Therefore, FDOG
should be able to protect other entities present in the bilayer from
free radical species.
ACKNOWLEDGMENT
We are indebted to Leslie Smith, Judy Blackburn, and Ray
Holloway for their technical assistance.
Supporting Information Available: S1, HPLC chromatogram
of purified FDOG; S2, electrospray ionization mass spectro-
metry (ESI-MS) analysis of FDOG; explanation of how NMR
wasusedtodistinguish 1(3)-feruloyl-2,3(1)-dioleoyl glycerolfrom
2-feruloyl-1,3-dioleoyl glycerol; S3, proton NMR of FDOG in
spectral range for determining the position of ferulate substitu-
tion; S4, complete FDOG proton NMR spectrum in deuterated
acetone; S5, FDOG HMBC spectrum; S6, 13C assignments for
FDOG;S7, NOESYspectrumofFDOG;S8, fluorescencespectra
of FDOG; Table 1. Basis set dependence of relative energies for
B3LYP geometry optimization calculations of conformations of
ferulic acid. This material is available free of charge via the
Internet at http://pubs.acs.org.
LITERATURE CITED
(1) Clifford, M. N. Chlorogenic acids and other cinnamates - nature,
occurrence and dietary burden. J. Sci. Food Agric. 1999, 79, 362–372.
(2) Zhao, Z.; Moghadasian, M. H. Chemistry, natural sources, dietary
intake and pharmacokinetic properties of ferulic acid: a review. Food
Chem. 2008, 109, 691–702.
FDOG incorporated into model phospholipid bilayers without
an overt disruption of bilayer integrity (calcein;cobalt complex
leakage test). This is clear because the leakage percentage
(Figure 3) and leakage rates (Table 2) are virtually the same for
all measurements. The lack of large variations in the leakage
kinetics indicates that one single step was involved in the leakage
mechanism (51). It is also interpreted that the shape of the leakage
kinetics further indicates that the leakage mechanism was similar
for all cases. Presumably, this means that FDOG does not induce
any packing changes within the bilayer, as tighter packing is
expected to lead to a lower leakage percentage (and possibly a
slower rate) and looser packing is expected to lead to a greater
leakage percentage (and rate). As for the location of FDOG
within the bilayer, the feruoyl group proximity correlated well
with the headgroup of DOPC lipids, suggesting that FDOG
lipophilic behavior is similar to that of DOPC.
The FDOG extended conformation is similar to conforma-
tions studied in recent PM3 semiempirical calculations on cis-
isomers of free unsaturated fatty acids, including oleic acid (52).
However, the extended conformation ofFDOG therefore permits
its seamless insertion into the membrane. This finding is contrary
to the stated expectation of Anselmi and co-workers (38) that a
branched-chain, straight conformation, feruloyl ester would not
fit well into a bilayer. This feature may be an asset that diminishes
the potential of the molecule to express cytotoxic effects.
Any number of disease states and aging phenomena can be
ascribed to a causal initiating influence of reactive oxygen and
nitrogen species on cellular structures. Mitochondrial membranes
are particularly susceptible to oxidative stressors. Lipophilic
antioxidants such as FDOG could diminish membrane free
radicals. Its mode of action would be similar to that of the natural
lipophilic antioxidants vitamin E and lipoic acid. The results of
this study demonstrated that FDOG incorporates into model
(3) Garc-a, J.; Pereira, H. Suberin structure in potato periderm: glycerol,
long-chain monomers, and glyceryl and feruloyl dimers. J. Agric.
Food Chem. 2000, 48, 5476–5483.
(4) Yunoki, K.; Musa, R.; Kinoshita, M.; Tazaki, H.; Oda, Y.; Ohnishi,
M. Presence of higher alcohols as ferulates in potato pulp and its
radical-scavenging activity. Biosci. Biotechnol. Biochem. 2004, 68,
2619–2622.
(5) Schmutz, A.; Jenny, T.; Ryser, U. A caffeoyl-fatty acid-glycerol
ester from wax associated with green cotton fibre suberin. Phyto-
chemistry 1994, 36, 1343–1346.
(6) Cooper, R.; Gottlieb, H. E.; Lavie, D. New phenolic diglycerides
from Aegilops ovata. Phytochemistry 1978, 17, 1673–1675.
(7) Wang, T.; Hicks, K. B.; Moreau, R. Antioxidant activity of
phytosterols, oryzanol, and other phytosterol conjugates. J. Am.
Oil Chem. Soc. 2002, 79, 1201–1206.
(8) Ou, S.; Kwok, K.-C. Ferulic acid: pharmaceutical functions, prepara-
tion and applications in foods. J. Sci. Food Agric. 2004, 84, 1261–1269.
(9) Murakami, A.; Kadota, M.; Takahashi, D.; Taniguchi, H.; Nomura,
E.; Hosoda, A.; Tsuno, T.; Maruta, Y.; Ohigashi, H.; Koshimizu, K.
Suppressive effects of novel ferulic acid derivatives on cellular
responses induced by phorbol ester, and by combined lipopolysac-
charide and interferon-gamma. Cancer Lett. 2000, 157, 77–85.
(10) Murakami, A.; Nakamura, Y.; Koshimizu, K.; Takahashi, D.;
Matsumoto, K.; Hagihara, K.; Taniguchi, H.; Nomura, E.; Hosoda,
A.; Tsuno, T.; Murata, Y.; Won Kim, H.; Kawabata, K.; Ohigashi,
H. FA15, a hydrophobic derivative of ferulic acid, suppresses
inflammatory responses and skin tumor promotion: comparison
with ferulic acid. Cancer Lett. 2002, 180, 121–129.
(11) Uwai, K.; Osanai, Y.; Imaizumi, T.; Kanno, S.; Takeshita, M.;
Ishikawa, M. Inhibitory effect of the alkyl side chain of caffeic acid
analogues on lipopolysaccharide-induced nitric oxide production in
RAW264.7 macrophages. Bioorg. Med. Chem. 2008, 16, 7795–7803.
(12) Jayaprakasam, B.; Vanisree, M.; Zhang, Y.; Dewitt, D. L.; Nair, M.
G. Impact of alkyl esters of caffeic and ferulic acids on tumor cell
proliferation, cyclooxygenase enzyme, and lipid peroxidation.
J. Agric. Food Chem. 2006, 54, 5375–5381.

5850 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Laszlo et al.
~
(13) Fiuza, S. M.; Gomes, C.; Teixeira, L. J.; Girao da Cruz, M. T.; Cordeiro,
(35) Peng, X.; Chen, H.; Draney, D. R.; Volcheck, W.; Schutz-Geschwender,
€
M. N. D. S.; Milhazes, N.; Borges, F.; Marques, M. P. M. Phenolic acid
derivatives with potential anticancer properties: a structure;activity
relationship study. Part 1: Methyl, propyl and octyl esters of caffeic and
gallic acids. Bioorg. Med. Chem. 2004, 12, 3581–3589.
A.; Olive, D. M. A nonfluorescent, broad-range quencher dye for Forster
resonance energy transfer assays. Anal. Biochem. 2009, 388, 220–228.
(36) Morris, J. V.; Mahaney, M. A.; Huber, J. R. Fluorescence quantum
yield determinations. 9,10-Diphenylanthracene as a reference stan-
dard in different solvents. J. Phys. Chem. 1976, 80, 969–974.
(37) Fattal, D. R.; Ben-Shaul, A. Mean-field calculations of chain
packing and conformational statistics in lipid bilayers: comparison
with experiments and molecular dynamics studies. Biophys. J. 1994,
67, 963–995.
(38) Anselmi, C.; Centini, M.; Andreassi, M.; Buonocore, A.; La Rosa,
C.; Maffei Facino, R.; Sega, A.; Tsumo, F. Conformational analysis:
a tool for the elucidation of the antioxidant properties of ferulic acid
derivatives in membrane models. J. Pharm. Biomed. Anal. 2004, 35,
1241–1249.
(39) Bakalbassis, E. G. A density functional theory study of structure-
activity relationships in caffeic and dihydrocaffeic acids and related
monophenols. J. Am. Oil Chem. Soc. 2003, 80, 459–466.
(40) Sebastian, S.; Sundaraganesan, N.; Manoharan, S. Molecular struc-
ture, spectroscopic studies and first-order molecular hyperpolariz-
abilities of ferulic acid by density functional study. Spectrochim.
Acta, Part A 2009, 74, 312–323.
(41) Calheiros, R.; Borges, F.; Marques, M. P. M. Conformational
behaviour of biologically active ferulic acid derivatives. J. Mol.
Struct. THEOCHEM 2009, 913, 146–156.
(14) Comporti, M. Three models of free radical-induced cell injury.
Chem.-Biol. Interact. 1989, 72, 1–56.
(15) Compton, D. L.; Laszlo, J. A.; Berhow, M. A. Lipase-catalyzed
synthesis of ferulate esters. J. Am. Oil Chem. Soc. 2000, 77, 513–519.
(16) Laszlo, J. A.; Compton, D. L.; Eller, F. J.; Taylor, S. L.; Isbell, T. A.
Packed-bed bioreactor synthesis of feruloylated monoacyl- and
diacylglycerols: clean production of a “green” sunscreen. Green
Chem. 2003, 5, 382–386.
(17) Choo, W.-S.; Birch, E. J. Radical scavenging activity of lipophilized
products from lipase-catalyzed transesterification of triolein with
cinnamic and ferulic acid. Lipids 2009, 44, 145–152.
(18) Compton, D. L.; Laszlo, J. A.; Berhow, M. A. Identification and
quantification of feruloylated mono- and diacylglycerols from
vegetable oils. J. Am. Oil Chem. Soc. 2006, 83, 753–758.
(19) MacDonald, R. C.; MacDonald, R. I.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. Small-volume extrusion apparatus for
preparation of large, unilamellar vesicles. Biochim. Biophys. Acta
1991, 1061, 297–303.
(20) Evans, K. O. Room-temperature ionic liquid cations act as short-
chain surfactants and disintegrate a phospholipid bilayer. Colloids
Surf., A 2006, 274, 11–17.
(42) Machado, N. F. L.; Caheiros, R.; Gaspar, A.; Garrido, J.; Broges,
F.; Margques, M. P. M. Antioxidant phenolic esters with potential
anticancer activity: solution equilibria studied by Raman spectro-
scopy. J. Raman Spectrosc. 2009, 40, 80–85.
(21) Zhang, J.; Stanley, R. A.; Melton, L. Lipid Peroxidation inhibition
capacity assay for antioxidants based on liposomal membranes. Mol.
Nutr. Food Res. 2006, 714–724.
(22) Naguib, Y. M. A. A fluorometric method for measurement of
peroxyl radical scavenging activities of lipophilic antioxidants. Anal.
Biochem. 1998, 265, 290–298.
(43) Calheiros, R.; Machado, N. F. L.; Fiuza, S. M.; Gaspar, A.; Garrido,
J.; Milhazes, N.; Borges, F.; Marques, M. P. M. Antioxidant
phenolic esters with potential anticancer activity: A Raman spectro-
scopy study. J. Raman Spectrosc. 2008, 39, 95–107.
(44) Ge, M.; Zhao, H.; Wang, W.; Zhang, Z.; Yu, X.; Li, W. Terahertz
time-domain spectroscopy of four hydroxycinnamic acid derivatives.
J. Biol. Phys. 2006, 32, 403–412.
(23) Cao, G.; Alessio, H. M.; Cutler, R. G. Oxygen-radical absorbance
capacity assay for antioxidants. Free Radical Biol. Med. 1993, 14, 303–311.
(24) Naguib, Y. M. A. Antioxidant activities of astazanthin and related
carotenoids. J. Agric. Food Chem. 2000, 48, 1150–1154.
(25) Davenport, L.; Dale, R. E.; Bisby, R. H.; Cundall, R. B. Transverse
location of the fluorescent probe 1,6-diphenyl-1,3,5-hexatriene in
model lipid bilayer membrane systems by resonance excitation
energy transfer. Biochemistry 1985, 24, 4097–4108.
~
(45) Silva, F. A. M.; Borges, F.; Guimaraes, C.; Lima, J. L. F. C.; Matos,
C.; Reis, S. Phenolic acids and derivatives: studies on the relationship
among structure, radical scavenging activity, and physicochemical
parameters. J. Agric. Food Chem. 2000, 48, 2122–2126.
(26) Wolf, D. E.; Winiski, A. P.; Ting, A. E.; Bocian, K. M.; Pagano, R.
E. Determination of the transbilayer distribution of fluorescent lipid
analogues by nonradiative fluorescence resonance energy transfer.
Biochemistry 1992, 31, 2865–2873.
(27) Mukherjee, S.; Raghuraman, H.; Dasgupta, S.; Chattopadhyay, A.
Organization and dynamics of N-(7-nitrobenz-2-oxa-1,3-diazol-4-
yl)-labeled lipids: a fluorescence approach. Chem. Phys. Lipids. 2004,
127, 91–101.
(28) Kachel, K.; Asuncion-Punzalan, E.; London, E. The location of
fluorescence probes with charged groups in model membranes.
Biochim. Biophys. Acta 1998, 1374, 63–76.
(29) Kondo, M.; Mehiri, M.; Regen, S. L. Viewing membrane-bound
molecular umbrellas by parallax analyses. J. Am. Chem. Soc. 2008,
130, 13771–13777.
(46) Nenadis, N.; Zhang, H.-Y.; Tsimidou, M. Z. Structure-antioxidant
activity relationship of ferulic acid derivatives: effect of carbon side
chain characteristic groups. J. Agric. Food Chem. 2003, 51, 1874–1879.
(47) Fang, X.; Shima, M.; Kadota, M.; Tsuno, T.; Adachi, S. Suppressive
effect of alkyl ferulate on the oxidation of linoleic acid. Biosci.
Biotechnol. Biochem. 2006, 70, 457–461.
(48) Kikuzaki, H.; Hisamoto, M.; Hirose, K.; Akiyama, K.; Taniguchi,
H. J. Antioxidant properties of ferulic acid and its related com-
pounds. J. Agric. Food Chem. 2002, 50, 2161–2168.
(49) Sabally, K.; Karboune, S.; Yeboah, F. K.; Kermasha, S. Lipase-
catalyzed esterification of selected phenolic acids with linolenyl
alcohols in organic solvent media. Appl. Biochem. Biotechnol.
2005, 127, 17–27.
(50) Sabally, K.; Karboune, S.; Kermasha, S. Lipase-catalyzed trans-
esterification of triolein or trilinolenin with selected phenolic acids.
J. Am. Oil Chem. Soc. 2006, 83, 101–107.
(30) Lewis, B. A.; Engelman, D. M. Lipid bilayer thickness varies linearly
with acyl chain length in fluid phosphatidylcholine vesicles. J. Mol.
Biol. 1983, 166, 211–217.
€ €
(51) Thoren, Per E.G.; Soderman, O.; Engstrom, S.; von Corswant, C.
´
(31) Anselmi, C.; Centini, M.; Granata, P.; Sega, A.; Buonocore, A.;
Bernini, A.; Facino, R. M. Antioxidant activity of ferulic acid alkyl
esters in a heterophasic system: a mechanistic insight. J. Agric. Food
Chem. 2004, 52, 6425–6432.
Interactions of novel, nonhemolytic surfactants with phospholipid
vesicles. Langmuir 2007, 23, 6956–6965.
(52) Vysotsky, Y. B.; Belyaeva, E. A.; Fainerman, V. B.; Vollhardt, D.;
Aksenenko, E. V.; Miller, R. Thermodynamics of the clusterization
process of cis isomers of unsaturated fatty acids at the air/water
interface. J. Phys. Chem. B 2009, 113, 4347–4359.
€€
(32) Klamt, A.; Schuurmann, G. COSMO: A new approach to dielectric
screening in solvents with explicit expressions for the screening
energy and its gradient. J. Chem. Soc., Perkin Trans. 1993, 2, 799.
(33) Shimanouchi, T.; Ishii, H.; Yoshimoto, N.; Umakoshi, H.; Kuboi,
R. Calcein permeation across phosphatidylcholine bilayer mem-
brane: effects of membrane fluidity, liposome size, and immobiliza-
tion. Colloids Surf., B 2009, 73, 156–160.
(34) Evans, K. O.; Lentz, B. R. Kinetics of lipid rearrangements during
poly(ethylene glycol)-mediated fusion of highly curved unilamellar
vesicles. Biochemistry 2002, 41, 1241–1249.
Received for review January 29, 2010. Revised manuscript received
April 1, 2010. Accepted April 3, 2010. Names are necessary to report
factually on available data; however, the USDA neither guarantees nor
warrants the standard of the product, and the use of the name by the
USDA implies no approval of the product to the exclusion of others that
may also be suitable.