Regioselective lipase-catalyzed synthesis of 3-o-acyl derivatives of resveratrol and study of their antioxidant properties

Article abstract of DOI:10.1021/jf903210q

One of the approaches to increasing the bioavailability of resveratrol is to protect its 3-OH phenolic group. In this work, regioselective acylation of resveratrol at 3-OH was achieved by transesterification with vinyl acetate catalyzed by immobilized lipase from Alcaligenes sp. (lipase QLG). The maximum yield of 3-O-acetylresveratrol was approximately 75%, as the lipase also catalyzes its further acetylation affording the diester 3,4′-di-O- acetylresveratrol and finally the peracetylated derivative. Long saturated and unsaturated fatty acid vinyl esters were also effective as acyl donors with similar regioselectivity. In contrast, lipase B from Candida antarctica catalyzes the acylation of the phenolic group 4′-OH with 80% yield and negligible formation of higher esters. The analysis of the antioxidant properties showed that the Trolox equivalent antioxidant capability (TEAC) values for the acetyl and stearoyl derivatives at 3-OH were, respectively, 40% and 25% referred to resveratrol. The addition of an acyl chain in the 3-OH position caused a higher loss of activity compared with that at the 4′-OH.

Full text of DOI:10.1021/jf903210q

J. Agric. Food Chem. 2010, 58, 807–813 807
DOI:10.1021/jf903210q
Regioselective Lipase-Catalyzed Synthesis of 3-O-Acyl
Derivatives of Resveratrol and Study of Their Antioxidant
Properties
†
,
ORRES,† ANA
P
OVEDA,‡ JESUS
J
IMENEZ-BARBERO,‡ ANTONIO
,†
B
ALLESTEROS
´
P
AMELA
T
AND
FRANCISCO J. PLOU*
†
ꢀ
Instituto de Catalisis y Petroleoquı
´
mica, CSIC, Cantoblanco, Marie Curie 2, 28049 Madrid, Spain and
‡
ꢀ
Centro de Investigaciones Biologicas, CSIC, Ramiro de Maeztu, 28040 Madrid, Spain
One of the approaches to increasing the bioavailability of resveratrol is to protect its 3-OH phenolic
group. In this work, regioselective acylation of resveratrol at 3-OH was achieved by transesterifica-
tion with vinyl acetate catalyzed by immobilized lipase from Alcaligenes sp. (lipase QLG). The
maximum yield of 3-O-acetylresveratrol was approximately 75%, as the lipase also catalyzes its
further acetylation affording the diester 3,40-di-O-acetylresveratrol and finally the peracetylated
derivative. Long saturated and unsaturated fatty acid vinyl esters were also effective as acyl donors
with similar regioselectivity. In contrast, lipase B from Candida antarctica catalyzes the acylation of
the phenolic group 40-OH with 80% yield and negligible formation of higher esters. The analysis of
the antioxidant properties showed that the Trolox equivalent antioxidant capability (TEAC) values for
the acetyl and stearoyl derivatives at 3-OH were, respectively, 40% and 25% referred to resveratrol.
The addition of an acyl chain in the 3-OH position caused a higher loss of activity compared with
that at the 40-OH.
KEYWORDS: Antioxidants; acylation; lipases; regioselectivity; polyphenols; vinyl esters; Trolox
equivalent antioxidant activity; TEAC
INTRODUCTION
disease observed among wine drinkers (French paradox). At
present, resveratrol is under phase-II clinical trials for the
prevention of colon cancer (4).
Antioxidants protect cells against the effects of harmful free
radicals and play an important role in preventing many human
diseases (e.g., cancer, atherosclerosis, stroke, rheumatoid arthri-
tis, neurodegeneration, inflammatory disorders, and diabetes)
and aging itself (1-4). The study of antioxidants is of great
interest for the role they play in protecting living systems against
lipid peroxidation and other anomalous molecular modifica-
tions (5-7). The modification of natural antioxidants to improve
their chemical, oxidative or thermal stability, bioavailability, and/
or pharmaceutical efficacy yields a series of semisynthetic anti-
It has been well reported that minimal chemical modification
on the stilbenenucleus may cause large variationsinthe biological
activity and, more specifically, in the antitumor properties of
resveratrol (16, 20-22). Lipophilic derivatives of resveratrol and
in particular esters bearing acyl chains may be important in view
of their enhanced affinity with lipophilic membrane constituents,
thus increasing their bioavailability (23). Some acylated deriva-
tives have shown higher cell-growth inhibition toward DU-145
human prostate cancer cells than resveratrol itself (20).
oxidants (e.g., tocopheryl acetate and L-ascorbyl palmitate) with
A reported problem with resveratrol is its limited bioavailabil-
ity owing to its fast metabolism in the liver yielding the less-active
3-sulfate and 3-glucuronide derivatives (24). As a consequence,
the circulating resveratrol has a serum half-life of 8-14 min (25).
One of the approaches to increasing bioavailability is to obtain
resveratrol analogues or derivatives with comparable activity that
cannot be sulfated or glucuronated. For that reason, the regiose-
lective modification of the 3-OH is of great interest (24).
As the phenolic groups of resveratrol at positions 3 and 40
exhibit very similar reactivity (the phenolic groups at 3 and 5 are
chemically equivalent, as resveratrol is a symmetric molecule), the
extraordinary selectivity of enzymes (26) is being exploited for its
regiospecific acylation (20), oxidation (27), or glycosylation (28).
Thus, acylation of resveratrol by Candida antarctica lipase B with
vinyl acetate afforded regioselectively 40-O-acetyl-resveratrol
great impact in the industry (8, 9).
Resveratrol (trans-3,5,40-trihydroxystilbene) (1) is a poly-
phenolic phytochemical found in grapes and many plants (10)
that is biosynthesized in response to pathogenic attack or stress
conditions. It possesses a variety of antioxidant (6), anti-inflam-
matory (11, 12), antitumor (13,14), cardioprotective (15), neuro-
protective (16), and immunomodulatory (17) bioactivities, as well
as a delaying effect on aging (18). As a phenolic compound,
resveratrol contributes to the antioxidant potential of red
wine (19) and may be related with the decrease in coronary heart
*To whom correspondence should be addressed. Departamento de
ꢀ
ꢀ
Biocatalisis, Instituto de Catalisis y Petroleoquımica, CSIC, Canto-
´
blanco, Marie Curie 2, 28049 Madrid, Spain. Phone: þ34-91-5854869.
Fax: þ34-91-5854760. E-mail: fplou@icp.csic.es.
© 2009 American Chemical Society
Published on Web 12/17/2009
pubs.acs.org/JAFC

808 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Torres et al.
(20, 29, 30); however, reactions with longer acyl donors such
as vinyl decanoate and vinyl cinnamoate were notably slower.
Regioselective derivatization of resveratrol at positions 3,5 was
achieved by a chemo-enzymatic procedure based on standard
chemical peracetylation followed by lipase-catalyzed alcoholysis
of the 40-acyl group in organic solvents (30).
As part of our interest in regioselective biotransformations of
polyhydroxylated compounds (31-34) using immobilized en-
zymes (35, 36), we have investigated the one-step enzymatic
synthesis of acylated resveratrol by lipase-catalyzed transesteri-
fication. Our results suggest that the regioselectivity of the process
can be controlled by an adequate selection of the biocatalyst (37).
Alltech), and integration was carried out using the Varian Star LC
workstation 6.41. For the analysis of resveratrol acetylation, the column
was a Lichrospher 100 RP8 (4.6 ꢀ 125 mm, 5 μm, Analisis Vinicos), and
the mobile phase was 70:30 (v/v) H2O/methanol (H2O contained 0.1% of
acetic acid) at 1 mL/min for 5 min. Then, a gradient from this mobile phase
to 50:50 (v/v) H2O/methanol was performed in 5 min, and this eluent was
maintained during 15 min. The resveratrol esters were quantified by
measuring the absorbance at 310 nm, and ELSD was used to follow the
formation of free fatty acid. For resveratrol stearate synthesis, the column
was Mediterranea-C18 (4.6 ꢀ 150 mm, 5 μm, Teknokroma, Spain). The
mobile phase was 90:10 (v/v) methanol/H2O (H2O contained 0.1% of
formic acid) at 1.5 mL/min.
HPLC/MS. Samples were analyzed by HPLC (model 1100, Agilent
Technologies) coupled to a mass spectrometer (model QSTAR pulsar,
Applied Biosystems). Chromatographic conditions were as described
above, except for the mobile phase that contained 1% (v/v) formic acid,
and the flow rate was lowered to 0.6 mL/min. Ionization was performed by
electrospray, and the mass spectrometer had a hybrid QTOF analyzer.
Nuclear Magnetic Resonance (NMR). NMR spectra of the differ-
ent compounds were recorded on a Bruker DRX 500 spectrometer using
DMSO as solvent. A temperature of 298 K was employed with concentra-
tions around 10 mM. Chemical shifts were reported in ppm and referenced
versus the solvent signal. Vicinal proton-proton coupling constants were
estimated from first order analysis of the spectra. The 2D-TOCSY
experiment (60 ms mixing time) was performed using a data matrix of
256 ꢀ 1K to digitize a spectral width of 5000 Hz. Four scans were used per
increment, with a relaxation delay of 2 s. 2D-NOESY (600 ms) and 2D-T-
ROESY experiments (500 ms) used the standard sequences provided by
the manufacturer and the data matrixes described above, with 32 and 48
scans per increment, respectively. In all cases, squared cosine-bell apodiza-
tion functions were applied in both dimensions. The spectral widths for the
HSQC spectra were 5000 and 18000 Hz for the 1H and 13C-dimensions,
respectively. The number of collected complex points was 1028 for the 1H-
dimension with a recycle delay of 2 s. The number of transients was 16, and
256 time increments were always recorded in the 13C-dimension. The
J-coupling evolution delay was set to 3.2 ms. A squared cosine-bell
apodization function was applied in both dimensions. Prior to Fourier
transform, the data matrixes were zero filled up to 1024 points in the 13C-
dimension. The spectral widths for the HMBC spectra were 5000 and
25000 Hz for the 1H and 13C-dimensions, respectively. The number of
collected complex points was 1028 for the 1H-dimension with a recycle
delay of 2 s. The number of transients was 64, and 256 time increments
were always recorded in the 13C-dimension. The J-coupling evolution
delay was set to 66 ms. Prior to Fourier transform, the data matrixes were
zero filled up to 1024 points in the 13C-dimension.
MATERIALS AND METHODS
Materials. Resveratrol from Polygonum cuspidatum was purchased in
Shanghai Seebio Biotechnology. Immobilized lipases from C. antarctica
B (Novozym 435), Thermomyces lanuginosus (Lipozyme TL IM), and
Rhizomucor miehei (Lipozyme RM IM) were kindly donated by Novo-
zymes A/S. Lipase from Pseudomonas cepacia (lipase PS IM) was
purchased from Amano. Immobilized lipases from Alcaligenes sp.
(lipases PLG and QLG) were kindly donated by Meito Sangyo Co.
(Japan). Tripropionin was from Acros. Vinyl acetate, 2-methyl-2-butanol
(t-amyl alcohol), R-tocopherol, Trolox (6-hydroxy-2,5,7,8-tetramethy-
chroman-2-carboxylic acid), ABTS (2,20-azinobis(3-ethylbenzothiazo-
line-6-sulfonate), and potassium persulfate were purchased from Sigma-
Aldrich. Vinyl stearate was from TCI-Europe. Vinyl oleate was from
ABCR GmbH & Co (Germany). Solvents were dried over 3 A molecular
sieves (water content was less than 0.2% v/v by Karl Fischer). All other
reagents were of the highest available purity and were used as purchased.
Determination of Enzyme Activity. The hydrolytic activity was
measured titrimetrically at pH 8.0 and 30 °C using a pH-stat (Mettler,
Model DL 50). The reaction mixture contained tripropionin (0.3 mL, final
concentration 80 mM), acetonitrile (0.7 mL), and Tris-HCl buffer (19 mL,
1 mM, pH 8.0) containing NaCl (0.1 M). The immobilized biocatalyst was
then added and the pH automatically maintained at 8.0 using 0.1 N NaOH
as titrant. Experiments were done in triplicate. One enzyme unit (U) was
defined as that catalyzing the formation of 1 μmol of fatty acid per min.
General Procedure for Enzymatic Reactions on an Analytical
Scale. Resveratrol (59.2 mg, 50 mM) and vinyl acetate (0.345 mL,
750 mM) or vinyl stearate (1.16 g, 750 mM) were incubated in 2-methyl-
2-butanol (5 mL) in sealed 30 mL dark vials under nitrogen at 40 °C with
150 rpm orbital shaking (SI50, Stuart Scientific). The biocatalyst was
added to a final concentration of 150 mg/mL. Aliquots (200 μL) were
withdrawn at intervals, filtered using an Eppendorf tube containing a
Durapore 0.45 μm filter, and the progress of the reaction analyzed by
HPLC.
TEAC Assay. To measure the antioxidant activity of the new com-
pounds, we used the Trolox equivalent antioxidant capability (TEAC)
assay that was previously described by Re et al. (38), with some modifica-
tions to adapt to 96 well plates. This assay is based on the ability of
antioxidants to reduce the ABTS radical. Briefly, ABTS (7 mM final
concentration) was added to an aqueous solution of 2.45 mM potassium
persulfate and kept in the dark at room temperature for 15 h to obtain the
ABTS radical, which was stable for 2 days. The ABTS•þ solution was
diluted with ethanol to get an absorbance of 0.70 ((0.02) at 734 nm and
equilibrated at room temperature. In each well, 20 μL of a solution of
Trolox (standard) or of the antioxidants (0.5-10 μM) in ethanol was
added to 230 μL of adjusted ABTS•þ solution. The decrease of absorbance
of the ABTS•þ solution was monitored at 734 nm during 6 min using a
microplate reader (model Versamax, Molecular Devices), and the decrease
of absorbance (ΔA734nm) for each concentration was determined using the
area under the curve. The concentration vs ΔA734nm curve was plotted for
the different compounds and used to calculate the equivalent Trolox
concentration. The TEAC value was determined as the ratio between the
slopes of concentration-ΔA734nm curves for the corresponding antioxi-
dant and Trolox.
Preparative-Scale Enzymatic Reactions. The reaction mixture
contained resveratrol (570 mg, 250 mM), vinyl acetate (3.45 mL,
3.75 M), lipase QLG (1.5 g), and 2-methyl-2-butanol (6.55 mL). The
mixture was incubated under conditions similar to those described in the
analytical scale procedure and monitored by HPLC. After 30 h, the
mixture was cooled, filtered, the solvent evaporated, and the crude
preparation loaded onto a silica-gel column using heptane/ethyl acetate
(3:1) as eluent, yielding pure 3-O-acetyl-resveratrol (3) and 3,40-di-O-
acetyl-resveratrol (5). For the acylation with vinyl stearate, the reaction
mixture contained resveratrol (171 mg, 150 mM), vinyl stearate (3.5 g,
2.25 M), lipase QLG (750 mg) or Novozym 435 (1 g), and 2-methyl-2-
butanol (5 mL). The mixture was incubated under conditions similar to
those described in the analytical scale procedure and monitored by HPLC.
After 72 h, the mixture was cooled, filtered, the solvent evaporated, and the
crude product purified by chromatography on a silica-gel column using
heptane/ethyl acetate (2:1) as eluent, yielding 3-O-stearoyl-resveratrol (7)
with lipase QLG and 40-O-stearoyl-resveratrol (8) with Novozym 435.
HPLC Analysis. HPLC analysis was performed using a ternary pump
(model 9012, Varian) coupled to a thermostatized (25 °C) autosampler
(model L-2200, VWR International). The temperature of the column was
ꢀ
was performed using a photodiode array detector (ProStar, Varian) in
series with an evaporative light scattering detector (ELSD, model 2000ES,
RESULTS AND DISCUSSION
Enzyme Screening for Acetylation of Resveratrol. Several lipases
and esterases were used for a preliminary HPLC screening of the
acetylation of resveratrol (1) in 2-methyl-2-butanol (2M2B)
kept constant at 45 °C (MEF-01 oven, Analisis Vınicos, Spain). Detection
´

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 809
Table 1. Screened Immobilized Lipases That Catalyzed the Acetylation of Resveratrol
microorganism
support
commercial name
supplier
hydrolytic activity (U/g)a
resveratrol selectivity
Candida antarctica (lipase B)
Thermomyces lanuginosus
Rhizomucor miehei
macroporous acrylic resin
granulated silica
Novozym 435
Lipozyme TL IM
Lipozyme RM IM
lipase PS IM
lipase PLG
Novozymes A/S
Novozymes A/S
Novozymes A/S
Amano
1725
2490
119
40-OH
3-OH > 40-OH
40-OH > 3-OH
3-OH > 40-OH
3-OH = 40-OH
3-OH
macroporous anion exchange resin
diatomaceous earth
Pseudomonas cepacia
Alcaligenes sp.
Alcaligenes sp.
3965
120
granulated diatomaceous earth
granulated diatomaceous earth
Meito Sangyo
Meito Sangyo
lipase QLG
695
a Measured in the hydrolysis of tripropionin.
depicted in Supporting Information, showing the correlation of
the quaternary aromatic carbon resonances to the key protons in
the corresponding ring.
The use of lipase QLG initially leads to one compound that has
a nonsymmetric spectrum in ring A (the three-substituted one;
Figure 2). It has to be mentioned that resveratrol shows a
symmetric spectrum for both rings. One acetate group was clearly
visible in the spectra with two free hydroxyl groups, one for ring A
and one for ring B. The existence of mono acetyl substitution was
also confirmed by MS analysis. NOESY spectra showed contacts
between the acetate group and hydrogens H2 and H4 belonging
to ring A. The HMBC showed all the expected peaks for 3-O-
acetyl-resveratrol (3). A new compound was formed when the
concentration of compound 3 was high enough; its MS and NMR
spectra showed the presence of two acetates. The spectrum also
showed the loss of symmetry in ring A, while the nuclei belonging
to ring B showed important shifting of their resonances. The
HMBC spectra confirmed the presence of one acetate moiety in
each ring, thus permitting us to assign this molecule as 3,40-di-O-
acetyl-resveratrol (5). The NOE data indicated close contacts
between each acetyl group and one of the resveratrol skeleton
rings.
We confirmed that the immobilized lipase B from C. antarctica
(Novozym 435) was highly specific for the phenolic group at
40-OH (20), which is less sterically hindered than the 3-OH (29),
yielding 40-O-acetyl-resveratrol (2). In fact, the only previous
report for the synthesis of monoacetate at 3-OH was carried out
by a multistep procedure based on the chemical peracetylation
of resveratrol followed by regiospecific alcoholysis catalyzed by
P. cepacia and C. antarctica lipases (30).
The lipases from Thermomyces lanuginosus (Lipozyme TL IM)
and Pseudomonas cepacia (Lipase PS IM) afforded mixtures of
3- and 40-regioisomers, with a major content of the monoacetate
at 3. A lipase from a different strain from Alcaligenes sp. (PLG)
yielded a nearly equimolar mixture of both derivatives. Lipase
from Rhizomucor miehei (Lipozyme RM IM) synthesized the
40-regioisomer as the main product. The reaction also took place
in other solvents such as tert-butanol, isopropyl ether, and
2-pentanone, or even when using vinyl acetate as solvent and
acyl donor at the same time.
Figure 1 shows the HPLC chromatograms with the three
lipases that synthesized 3-O-acetyl-resveratrol as the main pro-
duct. The molar ratio between the two monoesters was 18:1 for
lipase QLG, 2.5:1 for Lipase PS IM, and 1.5:1 for TL IM. As
shown in Figure 3, the three biocatalysts also formed the diester
3,40-di-O-acetyl-resveratrol (5) as well as a small amount of
3,5-di-O-acetyl-resveratrol (4).
Figure 1. HPLC chromatograms showing the formation of the two mono-
acetates (2 and 3) and the two diacetates (4 and 5) of resveratrol using as
biocatalysts (I) lipase QLG; (II) lipase PS IM; and (III) lipase TL IM.
Reaction conditions: 50 mM resveratrol, 750 mM vinyl acetate, 150 mg/mL
biocatalyst, 2M2B, and 40 °C. The chromatograms correspond to a
reaction time of 216 h.
under the following conditions: 50 mM resveratrol, 750 mM vinyl
acetate, 100 mg/mL biocatalyst, and 40 °C. The solvent 2M2B
was selected for the screening because of the notable solubility of
resveratrol and vinyl acetate. Six commercial immobilized lipases
were found to catalyze this process (Table 1). Two peaks
(corresponding to compounds 2 and 3) were observed in the
monoester fraction of the HPLC chromatograms (Figure 1), as
well as two peaks corresponding to the diesters 4 and 5. The ratio
between the two monoesters in the HPLC chromatograms was
dependent on the nature of the biocatalyst. In particular, the
lipase QLG from Alcaligenes sp. almost exclusively synthesized
compound 3, whereas the immobilized lipase B from C. antarctica
(Novozym 435), which is the preferred biocatalyst for many
esterification and transesterification processes including the mod-
ification of antioxidants (39-41), formed the other isomer (2).
We scaled-up the reactions with lipase QLG and Novozym 435
in order to isolate and characterize the different products by mass
spectrometry and 2D-NMR. The position of the substitution was
unequivocally deduced by using 2D-NMR experiments. A com-
bination of homo (NOESY and ROESY)- and heteronuclear
sequences (HSQC and HMBC; see the Materials and Methods
section for details) was employed for this task, assisted by NOE-
type experiments. Indeed, it was not possible to deduce the exact
nature of the substitution using chemical shift analysis alone. The
reported data for resveratrol were also employed to further
support the obtained conclusions (42). The 1H and 13C chemical
shifts for the different compounds are gathered in Supporting
Information. The key sections of the HMBC spectra are also
Considering that the positions of the phenolic substituents on
the aromatic rings of resveratrol may play an important role in its
biological activity (43) and that protection of 3-OH is one of the
approaches to improve resveratrol bioavailability, we focused our
attention on lipase QLG as a simple and selective alternative to
obtain monoesters at 3-OH. This biocatalyst is obtained by
immobilizing a lipase from Alcaligenes sp. (MW 31,000) on

810 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Torres et al.
Figure 2. Scheme of the acetylation of resveratrol catalyzed by lipase from Alcaligenes sp. (lipase QLG).
Figure 4. HPLC chromatograms showing the formation of the two mono-
stearates of resveratrol using (I) lipase QLG; (II) lipase PS IM; and (III)
lipase TL IM. Screening conditions: 50 mM resveratrol, 750 mM vinyl
stearate, 150 mg/mL biocatalyst, and 40 °C. The chromatograms corre-
spond to a reaction time of 138 h.
synthesized, whereas the formation of diester 5 was almost
negligible. The maximum yield of the monoester with lipase
QLG reached nearly 75% (on a molar basis) in 12 h, whereas
for diester 5, the maximum yield (approximately 60%) was
achieved at 160 h. With Novozym 435, a maximum yield of
80% for monoester 2 was obtained in 50 h.
Resveratrol Acylation with Vinyl Stearate. We tested the acyla-
tion of resveratrol with saturated and unsaturated long-chain
acyl-donors. The monoester selectivity was similar to that ob-
served in the acetylation process. Figure 4 shows the HPLC
chromatograms with lipases from Alcaligenes sp. (lipase QLG),
P. cepacia (Lipase PS IM) and T. lanuginosus (Lipozyme TL IM),
in the transesterification with vinyl stearate. As shown in Figure 4,
the acylation with vinyl stearate catalyzed by lipase QLG
afforded one major compound (7). The NMR and MS spectra
showed that this was a monosubstituted molecule, whose HMBC
spectrum was indeed similar to that of compound 3. The careful
HMBC analysis suggested that this molecule was 3-O-stearoyl-
resveratrol (7), which was further confirmed by the existence of
NOE contacts and the spatial proximity between the CH2 moiety
of the fatty acid chain and hydrogens H-2 and H-4 at the aromatic
ring A. Thus, the nature of 7 was unequivocally confirmed.
Novozym 435 also formed a monosubstituted major compound,
whose NMR spectra were different from those of monostearate 7.
The spectra showed symmetry for both rings, and the HMBC and
NOESY analysis (NOEs between the CH2 moiety of the fatty
acid chain and hydrogens H-30 and H-50 at the aromatic ring B)
indicated that this molecule was 40-O-stearoyl-resveratrol (8).
The highest selectivity toward the monoester 3-O-stearoyl-
resveratrol (7) was obtained again with lipase QLG, whereas PS
Figure 3. Kinetics of resveratrol acetylation in 2M2B catalyzed by (A)
lipase QLG and (B) Novozym 435. Experimental conditions: 50 mM
resveratrol, 750 mM vinyl acetate, 150 mg/mL biocatalyst, and 40 °C.
granulated diatomaceous earth. We observed by scanning elec-
tron microscopy (SEM) that the particle size was quite uniform
for this biocatalyst (approximately 500 μm) (see Supporting
Information). Up to now, the main application of this immobi-
lized lipase has been the resolution of racemic mixtures (44).
Kinetics of Resveratrol Acetylation. Figure 3 shows the progress
of the acetylation reaction with both lipase QLG and Novozyme
435. Lipase QLG initially formed 3-O-acetyl-resveratrol (3).
When the concentration of 3 was high enough, the synthesis of
diester 3,40-di-O-acetyl-resveratrol (5) was the main process.
Finally, the triester 3,40,5-tri-O-acetyl-resveratrol (6) was formed.
With the lipase B from C. antarctica, monoester 40 (2) was

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 811
Figure 7. Effect of antioxidant concentration on the decrease of ABTS•þ
absorbance for the acetyl derivatives of resveratrol.
Figure 5. Kinetics of resveratrol acylation with vinyl stearate in 2M2B
catalyzed by lipase QLG and Novozym 435. Experimental conditions: 50
mM resveratrol, 750 mM vinyl stearate, 150 mg/mL biocatalyst, 150 rpm,
and 40 °C.
Table 2. TEAC values of different acyl derivatives of resveratrol
compound
TEAC
slope (ΔA734/μM)
R2
resveratrol
1.79
0.69
1.64
0.33
0.16
0.44
1.02
5.41
2.08
4.96
1.01
0.50
1.32
3.07
0.975
0.989
0.962
0.985
0.992
0.991
0.990
3-O-acetyl-resveratrol
40-O-acetyl-resveratrol
3,40-di-O-acetyl-resveratrol
3,5,40-tri-O-acetyl-resveratrol
3-O-stearoyl-resveratrol
40-O-stearoyl-resveratrol
Figure 6. Progress of the ABTS•þ absorbance in the presence of different
antioxidants (5 μM).
IM and TL IM yield mixtures of the two isomers with the
3-derivative as the major product. We observed that lipase
QLG also regioselectively afforded derivatives at 3-OH bearing
unsaturated fatty acids (e.g., oleic acid; data not shown). Lipases
PLG and RM IM afforded mixtures of 3- and 40-regioisomers,
with a major content in the 40-isomer. Figure 5 shows the progress
of the acylation reaction with both lipase QLG and Novozyme
435. The formation of diesters with both enzymes was almost
negligible, probably due to steric hindrance. The maximum yield
of the monoester with lipase QLG reached 55% in 160 h, whereas
with Novozym 435, a maximum yield of 35% was obtained in
160 h. It was reported that lipase B from C. antarctica presented
a low reactivity with acyl chains longer than acetate; Teng et al.
reported 10% yield in the acylation of resveratrol with vinyl
decanoate and negligible reaction with vinyl cinnamate (29).
Antioxidant Activity. The antioxidant activity of the new
derivatives was analyzed. We initially compared the antioxidant
capability of resveratrol with other phenolic antioxidants such as
R-tocopherol and Trolox (used as standard). Figure 6 shows a
different behavior between resveratrol and R-tocopherol. The
latter was able to complete the reaction in 1 min; in contrast,
resveratrol showed a further small inhibitory effect even after
Figure 8. Effect of antioxidant concentration on the decrease of ABTS•þ
absorbance for the stearoyl derivatives of resveratrol.
7 min of reaction. Resveratrol also showed a higher activity than
Trolox and R-tocopherol, which may be related to the fact that
resveratrol has three hydroxyl groups (hydrogen donors),
whereas R-tocopherol and Trolox present only one (6).
The new acetyl derivatives showed antioxidant activity, but it
decreased when increasing the degree of substitution of the phe-
nolic groups. The results of the assay are presented in Figure 7,
and the TEAC values are summarized in Table 2. We observed
that the addition of an acyl chain in the position 3-OH caused a
higher loss of activity compared with that at 40-OH; the TEAC
value for the acetyl derivative at 3-OH was 40% that of resver-
atrol. Thus, 3-OH seems to play an important role in antioxidant
activity. Stivala et al. reported that the hydroxyl group in the

812 J. Agric. Food Chem., Vol. 58, No. 2, 2010
Torres et al.
40-position is required for antioxidant activity (45). It is note-
worthy that the ethanolic solution of resveratrol and the 40-OH
derivatives developed a yellow color (typical of resveratrol
oxidation) during the assay, which was not observed with the
monoesters at 3-OH; this observation may be related to the intrin-
sic stability of the different compounds. Regarding the stearoyl
derivatives (Figure 8), the effect of chemical modification of
aromatic rings on the antioxidant activity was more pronounced
than in the acetylation. Again, the substitution at 3-OH gave rise
to a higher decrease of activity than the corresponding activity
at 40-OH.
Conclusions. Immobilized lipase from Alcaligenes sp. (lipase
QLG) is able to catalyze in one step the regioselective synthesis
of resveratrol fatty acid esters in the 3-OH phenyl group. The
resulting derivatives bearing different saturated or unsaturated
acyl-groups are suitable for in vitro and in vivo structure-activity
relationship studies. As the stability of resveratrol in serum is
extremely low because of its fast metabolism in the liver resulting
in the chemical modification of the 3-OH, the novel 3-O-acyl
derivatives may exhibit improved bioavailability and pharmaco-
logical properties.
(10) Gonzalez-Barrio, R.; Beltran, D.; Cantos, E.; Gil, M. I.; Espin, J. C.;
Tomas-Barberan, F. A. Comparison of ozone and UV-C treatments
on the postharvest stilbenoid monomer, dimer, and trimer induction
in var. ‘Superior’ white table grapes. J. Agric. Food Chem. 2006, 54,
4222–4228.
(11) Geronikaki, A. A.; Gavalas, A. M. Antioxidants and inflammatory
disease: Synthetic and natural antioxidants with anti-inflammatory
activity. Comb. Chem. High Throughput Screening 2006, 9, 425–442.
(12) de la Lastra, C. A.; Villegas, I. Resveratrol as an anti-inflammatory
and anti-aging agent: Mechanisms and clinical implications. Mol.
Nutr. Food Res. 2005, 49, 405–430.
(13) Busquets, S.; Ametller, E.; Fuster, G.; Olivan, M.; Raab, V.; Argiles,
J. M.; Lopez-Soriano, F. J. Resveratrol, a natural diphenol, growth
in an experimental reduces metastatic cancer model. Cancer Lett.
2007, 245, 144–148.
(14) Larrosa, M.; Tomas-Barberan, F. A.; Espin, J. C. The grape and
wine polyphenol piceatannol is a potent inducer of apoptosis
in human SK-Mel-28 melanoma cells. Eur. J. Nutr. 2004, 43,
275–284.
(15) Wu, J. M.; Wang, Z. R.; Hsieh, T. C.; Bruder, J. L.; Zou, J. G.;
Huang, Y. Z. Mechanism of cardioprotection by resveratrol, a
phenolic antioxidant present in red wine (review). Int. J. Mol.
Med. 2001, 8, 3–17.
(16) Jung, J. C.; Lim, E.; Lee, Y.; Kang, J. M.; Kim, H.; Jang, S.; Oh, S.;
Jung, M. Synthesis of novel trans-stilbene derivatives and evaluation
of their potent antioxidant and neuroprotective effects. Eur. J. Med.
Chem. 2009, 44, 3166–3174.
(17) King, R. E.; Bomser, J. A.; Min, D. B. Bioactivity of resveratrol.
Compr. Rev. Food Sci. Food Saf. 2006, 5, 65–70.
(18) Shakibaei, M.; Harikumar, K. B.; Aggarwal, B. B. Resveratrol
addiction: To die or not to die. Mol. Nutr. Food Res. 2009, 53,
115–128.
ACKNOWLEDGMENT
We thank Drs. Juan Carlos Espın (CEBAS, CSIC, Murcia,
´
Spain), Juan Carlos Morales (IIQ, CSIC, Sevilla, Spain), and
Isabel Medina (IIM, CSIC, Vigo, Spain) for technical informa-
tion and suggestions. We are grateful to Meito Sangyo Co., Ltd.
(Tokyo, Japan) for samples of lipases QLG and PLG. We thank
Ramiro Martınez (Novozymes A/S, Madrid, Spain) for supply-
´
(19) Fremont, L. Minireview: Biological effects of resveratrol. Life Sci.
2000, 66, 663–673.
ing lipase samples and suggestions.
(20) Cardile, V.; Lombardo, L.; Spatafora, C.; Tringali, C. Chemo-
enzymatic synthesis and cell-growth inhibition activity of resveratrol
analogues. Bioorg. Chem. 2005, 33, 22–33.
Supporting Information Available: Characterization data of
the new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
(21) Chillemi, R.; Sciuto, S.; Spatafora, C.; Tringali, C. Anti-tumor
properties of stilbene-based resveratrol analogues: Recent results.
Nat. Prod. Commun. 2007, 2, 499–513.
LITERATURE CITED
(1) Noguchi, N.; Watanabe, A.; Shi, H. L. Diverse functions of anti-
oxidants. Free Radical Res. 2000, 33, 809–817.
(2) Fang, Y. Z.; Yang, S.; Wu, G. Y. Free radicals, antioxidants, and
nutrition. Nutrition 2002, 18, 872–879.
(22) de Lima, D. P.; Rotta, R.; Beatriz, A.; Marques, M. R.; Montenegro,
R. C.; Vasconcellos, M. C.; Pessoa, C.; de Moraes, M. O.;
Costa-Lotufo, L. V.; Sawaya, A. C. H. F.; Eberlin, M. N. Synthesis
and biological evaluation of cytotoxic properties of stilbene-based
resveratrol analogs. Eur. J. Med. Chem. 2009, 44, 701–707.
(23) Fragopoulou, E.; Nomikos, T.; Karantonis, H. C.; Apostolakis, C.;
Pliakis, E.; Samiotaki, M.; Panayotou, G.; Antonopoulou, S.
Biological activity of acetylated phenolic compounds. J. Agric. Food
Chem. 2007, 55, 80–89.
(24) Kang, S. S.; Cuendet, M.; Endringer, D. C.; Croy, V. L.; Pezzuto,
J. M.; Lipton, M. A. Synthesis and biological evaluation of a library
of resveratrol analogues as inhibitors of COX-1, COX-2 and NF-
kappa B. Bioorg. Med. Chem. 2009, 17, 1044–1054.
(25) Yu, C. W.; Shin, Y. G.; Chow, A.; Li, Y. M.; Kosmeder, J. W.; Lee,
Y. S.; Hirschelman, W. H.; Pezzuto, J. M.; Mehta, R. G.; van
Breemen, R. B. Human, rat, and mouse metabolism of resveratrol.
Pharm. Res. 2002, 19, 1907–1914.
(26) Alcalde, M.; Ferrer, M.; Plou, F. J.; Ballesteros, A. Environmental
biocatalysis: from remediation with enzymes to novel green pro-
cesses. Trends Biotechnol. 2006, 24, 281–287.
(3) Mantovani, G.; Maccio, A.; Madeddu, C.; Mura, L.; Gramignano,
G.; Lusso, M. R.; Murgia, V.; Camboni, P.; Ferreli, L.; Mocci, M.;
Massa, E. The impact of different antioxidant agents alone or in
combination on reactive oxygen species, antioxidant enzymes and
cytokines in a series of advanced cancer patients at different sites:
Correlation with disease progression. Free Radical Res. 2003, 37,
213–223.
(4) Espin, J. C.; Garcia-Conesa, M. T.; Tomas-Barberan, F. A. Nutra-
ceuticals: Facts and fiction. Phytochemistry 2007, 68, 2986–3008.
(5) Lo Nostro, P.; Capuzzi, G.; Romani, A.; Mulinacci, N. Self-
assembly and antioxidant properties of octanoyl-6-O-ascorbic acid.
Langmuir 2000, 16, 1744–1750.
(6) Kerem, Z.; Chetrit, D.; Shoseyov, O.; Regev-Shoshani, G. Protec-
tion of lipids from oxidation by epicatechin, trans-resveratrol, and
gallic and caffeic acids in intestinal model systems. J. Agric. Food
Chem. 2006, 54, 10288–10293.
(7) Kapich, A. N.; Galkin, S.; Hatakka, A. Effect of phenolic acids on
manganese peroxidase-dependent peroxidation of linoleic acid and
degradation of a non-phenolic lignin model compound. Biocatal.
Biotransform. 2007, 25, 350–358.
(8) Torres, P.; Kunamneni, A.; Ballesteros, A.; Plou, F. J. Enzymatic
modification for ascorbic acid and alpha-tocopherol to enhance their
stability in food and nutritional applications. Open Food Sci. J. 2008,
2, 1–9.
(27) Nicotra, S.; Cramarossa, M. R.; Mucci, A.; Pagnoni, U. M.; Riva, S.;
Forti, L. Biotransformation of resveratrol: Synthesis of trans-dehy-
drodimers catalyzed by laccases from Myceliophtora thermophyla
and from Trametes pubescens. Tetrahedron 2004, 60, 595–600.
(28) Shim, H.; Hong, W.; Ahn, Y. Enzymatic preparation of phenolic
glucosides by Streptococcus mutans. Bull. Korean Chem. Soc. 2003,
24, 1680–1682.
(29) Teng, R. W.; Bui, T. K. A.; McManus, D.; Armstrong, D.; Mau,
S. L.; Bacic, A. Regioselective acylation of several polyhydroxylated
natural compounds by Candida antarctica lipase B. Biocatal. Bio-
transform. 2005, 23, 109–116.
(9) Sabally, K.; Karboune, S.; St Louis, R.; Kermasha, S. Lipase-
catalyzed synthesis of phenolic lipids from fish liver oil and dihy-
drocaffeic acid. Biocatal. Biotransform. 2007, 25, 211–218.

Article
J. Agric. Food Chem., Vol. 58, No. 2, 2010 813
(30) Nicolosi, G.; Spatafora, C.; Tringali, C. Chemo-enzymatic prepara-
tion of resveratrol derivatives. J. Mol. Catal. B: Enzym. 2002, 16,
223–229.
(39) Antoniotti, S.; Fernandez, X.; Dunach, E. Reaction design for
evaluation of the substrate range of hydrolases. Biocatal. Biotrans-
form. 2008, 26, 228–234.
(31) Reyes-Duarte, D.; Lopez-Cortes, N.; Ferrer, M.; Plou, F. J.;
Ballesteros, A. Parameters affecting productivity in the lipase-
catalysed synthesis of sucrose palmitate. Biocatal. Biotransform.
2005, 23, 19–27.
(32) Plou, F. J.; Cruces, M. A.; Ferrer, M.; Fuentes, G.; Pastor, E.;
Bernabe, M.; Christensen, M.; Comelles, F.; Parra, J. L.; Ballesteros,
A. Enzymatic acylation of di- and trisaccharides with fatty acids:
choosing the appropriate enzyme, support and solvent. J. Biotechnol.
2002, 96, 55–66.
(33) Torres, P.; Reyes-Duarte, D.; Lopez-Cortes, N.; Ferrer, M.;
Ballesteros, A.; Plou, F. J. Acetylation of vitamin E by Candida
antarctica lipase B immobilized on different carriers. Proc. Biochem.
2008, 42, 145–153.
(34) Alcalde, M.; Ferrer, M.; Plou, F. J. Environmental biocatalysis:
From remediation with enzymes to novel green processes. Biocatal.
Biotransform. 2007, 25, 113.
(35) Torres, P.; Datla, A.; Rajasekar, V. W.; Zambre, S.; Ashar, T.;
Yates, M.; Rojas-Cervantes, M. L.; Calero-Rueda, O.; Barba, V.;
Martinez, M. J.; Ballesteros, A.; Plou, F. J. Characterization and
application of a sterol esterase immobilized on polyacrylate epoxy-
activated carriers (Dilbeads). Catal. Commun. 2008, 9, 539–545.
(36) Berrio, J.; Plou, F. J.; Ballesteros, A.; Martin, A. T.; Martinez, M. J.
Immobilization of Pycnoporus coccineus laccase on Eupergit C:
Stabilization and treatment of oil mill wastewaters. Biocatal. Bio-
transform. 2007, 25, 130–134.
(37) Torres, P.; Plou, F. J.; Ballesteros, A. Enzymatic procedure for the
one-step preparation of 3-O-acyl resveratrol and 3,40-di-O-acyl
resveratrol. Spanish Patent 200803184, 2008.
(38) Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.;
Rice-Evans, C. Antioxidant activity applying an improved ABTS
radical cation decolorization assay. Free Radical Biol. Med. 1999, 26,
1231–1237.
(40) Los Rios, A. P.; Hernandez-Fernandez, F. J.; Martinez, F. A.;
Rubio, M.; Villora, G. The effect of ionic liquid media on acti-
vity, selectivity and stability of Candida antarctica lipase B in
transesterification reactions. Biocatal. Biotransform. 2007, 25,
151–156.
(41) Ferrer, M.; Soliveri, J.; Plou, F. J.; Lopez-Cortes, N.; Reyes-Duarte,
D.; Christensen, M.; Copa-Patino, J. L.; Ballesteros, A. Synthesis of
sugar esters in solvent mixtures by lipases from Thermomyces
lanuginosus and Candida antarctica B, and their antimicrobial
properties. Enzyme Microb. Technol. 2005, 36, 391–398.
(42) Commodari, F.; Khiat, A.; Ibrahimi, S.; Brizius, A. R.; Kalkstein, N.
Comparison of the phytoestrogen trans-resveratrol (3,40,5-
trihydroxystilbene) structures from X-ray diffraction and solution
NMR. Magn. Reson. Chem. 2005, 43, 567–572.
(43) Seeram, N. P.; Kulkarni, V. V.; Padhye, S. Sources and Chemi-
stry of Resveratrol. In Resveratrol in Health and Disease;
Aggarwal, B. B., Shishodia, S., Eds.; CRC Press: Boca Raton, FL,
2007; pp 17-32.
(44) Yuasa, Y.; Fukaya, H.; Yuasa, Y. Stereoselective and enantioselec-
tive syntheses of the four stereoisomers of muscol from (3RS)-
muscone. Helv. Chim. Acta 2007, 90, 977–982.
(45) Stivala, L. A.; Savio, M.; Carafoli, F.; Perucca, P.; Bianchi, L.;
Maga, G.; Forti, L.; Pagnoni, U. M.; Albini, A.; Prosperi, E.;
Vannini, V. Specific structural determinants are responsible for the
antioxidant activity and the cell cycle effects of resveratrol. J. Biol.
Chem. 2001, 276, 22586–22594.
Received for review September 11, 2009. Revised manuscript received
November 17, 2009. Accepted November 23, 2009. We are grateful to
Comunidad de Madrid for a research contract to P.T. This research was
supported by the Spanish CSIC (Project 200680F0132).