Cysteinyl-flavan-3-ol conjugates from grape procyanidins. Antioxidant and antiproliferative properties.

Article abstract of DOI:10.1016/S0968-0896(02)00127-X

New bio-based antioxidant compounds have been obtained by depolymerisation of grape polymeric flavanols in the presence of cysteine. Their preparation and purification, as well as their antiradical/antioxidant and antiproliferative properties are reported

Full text of DOI:10.1016/S0968-0896(02)00127-X

Bioorganic & Medicinal Chemistry 10 (2002) 2497–2509
Cysteinyl-flavan-3-ol Conjugates from Grape Procyanidins.
Antioxidant and Antiproliferative Properties
J. L. Torres,a,* C. Lozano,a,c L. Julia,b F. J. Sanchez-Baeza,b J. M. Anglada,b
J. J. Centellesc and M. Cascantec
aDepartment of Peptide and Protein Chemistry, Institute for Chemical and Environmental Research (IIQAB-CSIC),
Jordi Girona 18-26, 08034 Barcelona, Spain
bDepartment of Biological Organic Chemistry, Institute for Chemical and Environmental Research (IIQAB-CSIC),
Jordi Girona 18-26, 08034 Barcelona, Spain
c
´
Department of Biochemistry and Molecular Biology, University of Barcelona, Martı i Franques 1-11, 08028 Barcelona, Spain
`
Received 7 December 2001; accepted 2 April 2002
Abstract—New bio-based antioxidant compounds have been obtained by depolymerisation of grape polymeric flavanols in the
presence of cysteine. Their preparation and purification, as well as their antiradical/antioxidant and antiproliferative properties are
reported. 4b-(S-cysteinyl)epicatechin 5, 4b-(S-cysteinyl)catechin 6 and 4b-(S-cysteinyl)epicatechin 3-O-gallate 7 were efficiently
purified from the crude depolymerised mixture by cation-exchange chromatography and preparative reversed-phase chroma-
tography. The new compounds were more efficient than the underivatised (ꢀ)-epicatechin 1 as scavengers of the 1,1-diphenyl-
2-picrylhydrazyl free radical (DPPH) and weak growth inhibitors of human colon carcinoma HT29 cells. The order of antiradical
and antiproliferative efficiency was 7 >5 ꢁ6 >1, the same for both assays. # 2002 Elsevier Science Ltd. All rights reserved.
Introduction
In humans, the excess of ROS, which can be either
radicals or non-radical compounds, has been related to
numerous diseases and dietary plant antioxidants such
as carotenoids and flavonoids are considered to help the
internal defence mechanisms against unwanted oxid-
ations.1,2 Particularly flavanols from tea, grape or pine
bark are potent free radical scavengers of interest as
preventative agents against cancer and cardiovascular
diseases.3ꢀ7 The capacity to arrest the cell cycle and/or
promote apoptosis may contribute to the chemo-
preventative effect of some members of the flavanoid
family.8,9
Much attention is currently being paid to the putative
benefits of plant polyphenols as complements to the
organism’s antioxidant defence systems. While evolving
from anaerobic microorganisms, aerobic life forms had
to establish a delicate balance between the benefits and
risks of using oxygen to obtain energy. The respiratory
chain inevitably produces reactive intermediates such as
the superoxide anion radical (Oꢂ2ꢀ) which lead to other
reactive species potentially harmful to biological mole-
cules such as DNA, lipids and proteins.1 Life organisms
have developed enzymatic systems to regulate the levels
of all these reactive oxygen species (ROS). Superoxide
dismutase, catalase and glutathione peroxidase which
scavenge the superoxide anion, H2O2 and organic
hydroperoxides, respectively, are major detoxifying
enzymes.1 Occasionally, the defence systems are over-
whelmed by an excess of ROS produced by illnesses,
aging or by external factors such as air pollution,
smoking or UV radiation.
The major flavanols in green tea are monomeric cate-
chins, that is (ꢀ)-epigallocatechin 3-O-gallate, (ꢀ)-epi-
gallocatechin, (ꢀ)-epicatechin (1) and (ꢀ)-epicatechin
3-O-gallate (3).10 Grape and pine bark extracts also
contain monomers and are rich in oligomeric (ꢁ2–7
residues) and polymeric (>7 residues) flavanols.7,11ꢀ14
Polymerised flavanols are called proanthocyanidins
because they yield anthocyanidins upon depolymeri-
sation under acidic conditions. In the presence of a
strong nucleophile, proanthocyanidins yield two kinds of
compounds: flavan-3-ols and their derivatives at position
4, coming from the polymer’s terminal and extension
*Corresponding author. Tel.: +34-93-400-61-12; fax: +34-93-204-59-04;
e-mail: jltqbp@iiqab.csic.es
0968-0896/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00127-X

2498
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
units, respectively15 (Scheme 1). This procedure, with
toluene-a-thiol as nucleophile, is currently used to esti-
mate the degree of polymerisation and galloylation of
proanthocyanidins.11,12,16,17
of the amino function on the cysteinyl moiety. The goal
was to have the underivatised monomers and the
cysteinyl conjugates elute in two separate groups by
means of two consecutive isocratic steps, namely elution
with initial buffer and with NaCl containing buffer,
respectively.
The acid cleavage of proanthocyanidins in the presence
of other thiols (2-hydroxyethylthiol, 2-aminoethylthiol,
Scheme 1) has been recently used to obtain new com-
pounds with antiradical/antioxidant activity and mod-
ified physico-chemical properties.18,19 The combination
of antioxidant polyphenols with other natural products
is a promising approach in the search for new safe
compounds with diversified applications.
First, the elution conditions were optimised at semi-
preparative scale (6 mL bed volume). The buffers pH
was adjusted to 2.25 to make sure that the carboxylic
acid function was completely protonated, and the con-
jugates were retained on the resin through their net
positive charges. Hydrophobic interactions with the
stationary phase were eliminated with 30% (v/v) EtOH
added as co-solvent. More ethanol resulted in early
release of the charged conjugates. Moreover, the
amount of salt necessary to elute the charged conjugates
was minimised. The final optimal conditions were: [A],
4.75 bed volumes of 20 mM sodium phosphate pH 2.25
buffer/EtOH (7:3) for the elution of monomers; [B], 4.75
bed volumes of 20 mM sodium phosphate pH 2.25 buffer/
EtOH (13:7) 100 mM NaCl for the elution of the con-
jugates together with the excess of cysteine from the
cleavage reaction.
Here, we describe the preparation and purification of a
new family of bio-based thio-derivatives of flavan-3-ols
(5–7) obtained by depolymerisation of procyanidins20 in
the presence of the natural amino acid cysteine. The
reaction also yielded the underivatised terminal units
1–3. The cleavage conditions used were essentially those
described before for aminoethylthio-derivatives such as
4.19 The antiradical/antioxidant activity in the DPPH
assay and the antiproliferative activitity on a human colon
carcinoma cell line of the new derivatives are evaluated.
The cation-exchange separation was scaled-up on a
glass column (21 ꢃ 2.5 i.d., 105 mL bed volume) packed
with the same stationary phase. The maximum load was
found to be 500 mL (ca. 500 mg GAE, gallic acid
equivalents by the Folin–Ciocalteu’s method, initial
polyphenols) of crude cleavage mixture which was pro-
cessed in seven repetitive runs. Subsequently, com-
pounds 5–7 were purified by preparative reversed-phase
Results
Purification
The cysteinyl flavan-3-ols were separated from the crude
depolymerisation mixture on a strong cation-exchange
resin (MacroPrepTM High S 50 mm) by taking advantage

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2499
Scheme 1. Depolymerisation of proanthocyanidins (polymeric flavan-3-ols). The arrows indicate putative polymerisation positions.
high-performance liquid chromatography (RP-HPLC)
as stated in the experimental section. Briefly, the mix-
ture obtained by cation exchange was desalted and
fractionated by a CH3CN gradient in 0.1% (v/v) tri-
fluoroacetic acid (TFA). Then, each fraction, rich in one
of the cysteinyl derivatives, was re-chromatographed
with CH3CN gradients in triethylamine phosphate
(TEAP) buffers and 0.1% (v/v) TFA to yield 99.5%
pure (by RP-HPLC) compounds. The stereochemistry
at C-2, C-3 and C-4 of compounds 5–7 was assigned
agreement with the literature, the 4b derivatives were
the major isomers obtained irrespective of 2,3-stereo-
chemistry.15,21
Free radical scavenging activity
The new cysteinyl flavan-3-ol derivatives are potent free
radical scavenging agents (Fig. 1) in the 1,1-diphenyl-
2-picrylhydrazyl free radical (DPPH) assay.22,23 The
results are expressed as ED50 as defined in the experi-
mental section. The inverse of ED50 is a measure of the
antiradical power (ARP). By multiplying the ED50 by
two, the stoichiometric value (theoretical concentration
of antioxidant to reduce 100% of the DPPH) is
obtained. The inverse of this value represents the
moles of DPPH reduced by one mole of antioxidant
and gives an estimate of the number of hydrogen
atoms involved in the process. Table 1 summarises the
parameters obtained from the DPPH assay. The new
conjugates were more efficient than (ꢀ)-epicatechin 1
and Trolox (6-hydroxy-2,5,7,8-tetramethyl-chroman-
2-carboxylic acid), being the order of antiradical power
1
from the hydrogen coupling constants measured by H
15
NMRand following Haslam and co-workers.
In
Table 1. Antiradical power and stoichiometry from the DPPH assay
Compound
ARP
(1/ED50
Stoichiometric
value
H atoms
per moleculea
)
Trolox
4.2
6.2
8.3
7.7
20
0.48
0.32
0.24
0.26
0.10
2.1
3.1
4.2
3.8
10
1
5
6
7
Figure 1. Free radical scavenging activity of the cysteinyl derivatives
of flavan-3-ols in the DPPH assay. Absorbance (A) at 517 nm is a
measure of the amount of free radical remaining in solution. (1ꢀA/
A0)ꢃ100 represents the percentage of DPPH already reacted with the
antioxidant. Each point is the mean of three determinations.
aMoles reduced DPPH per mole antioxidant.

2500
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
7 >5 ꢁ6 >1 >Trolox.24 We obtained a stoichiometric
value of 0.32 for (ꢀ)-epicatechin, corresponding to the
reduction of ca. three DPPH molecules per molecule of
1. The cysteinyl derivatives of epicatechin (5) and cate-
chin (6) were able to reduce roughly one more molecule
of DPPH than the underivatised (ꢀ)-epicatechin 1.
Gallate containing compound 7 is clearly the most effi-
cient of the new molecules.
Nuclear magnetic resonance (NMR) deuterium exchange
1
When recording the standard H NMRspectra of deri-
vatives 4–7 using acetone-d6/D2O as solvent, we
observed a systematic discrepancy in the area corres-
ponding to the signals of hydrogen atoms at positions 6
and 8 which we attributed to deuterium exchange with
the solvent. In order to confirm this, we recorded the 1H
and 13C NMRspectra of two freshly prepared samples,
one using regular compound 4 and a second sample
using a preparation of the same compound generated by
three successive operations of dilution in D2O and lyophi-
lisation. The spectrum corresponding to the first sample
showed the normal signals with only a little decrease in the
integral for H-6 and H-8. In the second sample no signal
for H-6, H-8, C-6 and C-8 was detected while the rest of
the spectrum did not show any other change. A closer look
at the 13C NMRspectrum revealed two low intensity sig-
nals with the characteristic pattern for monodeuterated
carbons near the positions of C-6 and C-8.
The observed hydrogen exchange was unexpectedly fast
and we decided to follow the kinetics of the deuteration
1
process by H NMRunder controlled conditions (sol-
vent, concentration, temperature, pD) using (ꢀ)-epica-
techin 1 as a control. The spectra were obtained using a
mixture of D2O/acetone-d6 buffered at pD=6 (see
details in Experimental). For compounds 4 and 5 the
deuterium exchange at positions H-6 and particularly
H-8 was very fast compared to that of 1 (Figs 2 and 3).
The deuteration first order pseudokinetic data were fit-
ted using a simple exponential decay model and a good
correlation (R2 higher than 0.96) for the six sets of data
was obtained. The observed half life for H-8 and H-6
signals was respectively of 3851 and 5369 min for com-
pound 1; 511 and 911 min for compound 4; 57 and 150
min for compound 5. In all cases the exchange process
proceeded until total hydrogen exchange.
Bond dissociation enthalpies (BDE)
We have performed theoretical calculations at B3LYP
level of theory27,28 to estimate the BDEs of several X–H
Figure 2. 1H NMRspectra expansion of the aromatic hydrogen
region. (a) Compound 1; (b) compound 4 and (c) compound 5. All
three spectra were recorded under identical conditions after 6 h in
D2O/acetone-d6 (3:1 v/v) at pD=6 (ammonium chloride buffer). Sig-
nal at ca. 6.05 ppm corresponds to H-6 and signal at ca. 5.95 ppm
corresponds to H-8.25,26 The signals at lower field and insensitive to
the deuterium exchange correspond to H-20, H-50 and H-60.
Figure 3. Time evolution for the H-6 and H-8 1H NMRsignal areas
for compounds 1, 4, 5. Relative signal is the signal area referred to the
average one-hydrogen area for the further hydrogen signals. Data
were adjusted using a single exponential decay. R2 was higher than
0.96 in all cases.

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2501
Figure 4. Calculated bond dissociation enthalpies for compound 5
obtained at B3LYP level of theory.
Figure 6. EPRspectra of the stable radical generated from 4 b-(S-
cysteinyl)epicatechin 5 (10 mM) by a 1-day treatment with 1 M aqu-
eous NaOH (a) and after ultraviolet irradiation with a high-pressure
mercury lamp for 2 min (b) and 10 min (c).
Figure 5. EPRspectra of NaOH (1 M) aqueous solutions of ( ꢀ)-epi-
catechin 1 (10 mM) (a) and 4b-(S-cysteinyl)epicatechin 5 (10 mM)
(b) at different times.
Figure 7. HT29 colorectal adenocarcinoma cell proliferation in
response to treatment with increasing doses (X axis) of 2-ami-
noethylthio and cysteinyl derivatives of flavan-3-ols. & (4), * (5), !
(6), ~ (7) and ^ (1). The viability and proliferation are expressed as
per cent of untreated control cells (meanꢄSD).
bonds of compound 5. These BDEs are expected to be
directly related to the ability of flavanols to scavenge
reactive oxygen species and radicals such as DPPH by
donating hydrogen atoms. Our computed values for
compound 5 are displayed in Figure 4, which shows that
the BDEs at the C-2 and C-4 positions (80.6 and 80.8
kcal/mol, respectively) are of the same order than the
BDE of the OH at C-40 (80.7 kcal/mol).
radicals detected within minutes which rapidly dis-
appeared and were coincident with those observed by
other authors29 (Fig. 5a), 5 generated within hours a
simple and much more stable radical which presents a
six-line pattern with relative intensities that correspond
to hfs constants from three protons with values of 2.67,
1.55 and 1.62 G (Fig. 5b). These signals slowly
decreased over a period of several days and were
reduced significantly upon irradiation with UV light
(Fig. 6). The same radical was detected for the 2-ami-
noethylthio derivative 4. In no case, this paramagnetic
species was detected from (ꢀ)-epicatechin 1.
Formation of free radicals
Flavanols are air oxidised under aqueous basic condi-
tions and the intermediate radicals formed may be
detected by electron paramagnetic resonance
(EPR).29ꢀ31 The EPRspectra of diluted (10 mM) solu-
tions of (ꢀ)-epicatechin 1 and its cysteinyl derivative 5
in 1 M aqueous NaOH, obtained at different times,
revealed significant differences (Fig. 5). While 1 generated

2502
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
Table 2. Antiproliferative potency against human colorectal adeno-
carcinoma HT29 cells
reported the use of an agarose-based gel (SP SepharoseTM
high performance) and acetonitrile (CH3CN) as co-sol-
vent for the separation of flavanol conjugates with
cysteamine.19 The monomers were washed away with
10% (v/v) CH3CN in the starting eluent and the amine
containing conjugates were eluted with a gradient of salt
to eliminate the electrostatic interaction with the resin
and a simultaneous gradient of CH3CN. The co-solvent
gradient allowed the separation of the epicatechin and
catechin derivatives from the more hydrophobic gallate
containg conjugate. This facilitated the subsequent pur-
ification of the products. The main disadvantage of this
procedure is the high eluting volumes needed to obtain a
baseline separation. In the present case we decided to
minimise the eluting volume by raising the amount of
co-solvent at the cost of having all conjugates eluting
together. Further modifications introduced in this step
were the use of a bulk stationary phase (MacroPrepTM
High S 50 mm) based on a methacrylate co-polymer and
EtOH instead of CH3CN. The appropriate amount of
co-solvent in the starting eluent was found to be 30%
(v/v) EtOH. Less co-solvent resulted in unacceptable
retention of (ꢀ)-epicatechin 3-O-gallate 3 and more
EtOH resulted in early elution of the charged conjugates
5 and 6. This effect has also been observed in the cation-
exchange purification of hydrophobic cationic surfac-
tants32 and might be due to the influence of the co-solvent
on the stabilisation of ions pairs with the buffer. After
elution of the monomers we took advantage of this
effect and increased the amount of EtOH to help
recover the conjugates in the presence of salt (100 mM
NaCl). At the end we came out with a set of conditions
that yielded two groups of compounds, namely under-
ivatised monomers and cysteinyl conjugates, with mini-
mum buffer, co-solvent and salt consumption.
Compound
na
Mean IC50
(mM)
SD
1
4
5
6
7
4
6
6
6
6
825.7
408.0
406.6
421.2
218.9
12.2
11.6
21.0
26.5
6.7
an, number of experiments performed; SD, standard deviation.
Antiproliferative activity
The effect of compounds 4–7 and (ꢀ)-epicatechin 1 on
the proliferation of a human carcinoma cell line (HT29)
using an MTT assay was examined. The results showed
a reduction in cell proliferation in a dose-dependent
fashion after treatment with the flavan-3-ols at the
reported concentrations for 72 h (Fig. 7). Table 2 shows
the mean IC50 values and corresponding SD obtained.
It should be noted that IC50 was calculated with respect
to the total number of cells in the control after 72 h of
proliferation (20,000 cells/well). At the IC50, the final
number of treated cells was approximately 10,000 cells/
well, which was 4 times higher than the number of cells
at the beginning of the experiment. Thus, the IC50 cal-
culated in this paper is indicative of the concentration of
product that inhibits the proliferation by 50%. The
lower the IC50, the more potent the compound is. The
order of antiproliferative potency followed the same
pattern obtained from the antiradical activity assay: 7
>4 ꢁ5 ꢁ6 >1.
Discussion
Antiradical/antioxidant activity
Purification
In the present paper, we have focused on the free radical
scavenging activity of the cysteinyl derivatives in solu-
tion using a well known stable free radical (DPPH).
While the overall efficiency of polyphenolic antioxidants
against lipid and protein oxidation may depend on
many factors including the physico-chemical properties
of the environment, the DPPH test provides a simple
measure of intrinsic antioxidant efficiencies.23,33,34
The efficient cation-exchange separation of the cysteinyl
conjugates from the complex crude cleavage mixture
depends on several factors, particularly on how suc-
cessfully the hydrophobic interactions with the sta-
tionary phase are handled. If no interaction other than
electrostatic were to take place, the underivatised
monomers together with any other uncharged material
would be washed away with the starting aqueous buffer
while positively charged amine containing derivatives
would be retained and subsequently eluted with salt. In
many instances though, some amount of an organic co-
solvent is needed to either lessen or eliminate hydro-
phobic interactions. In our hands, when no solvent was
added to the eluents all molecules were hydrophobically
retained on the resin. Particularly, more than 10 bed
volumes of washing buffer were needed to release the
uncharged (ꢀ)-epicatechin 3-O-gallate 3. Moreover,
other materials, including coloured species from the
grape extract, were also retained and gradually released
over the entire chromatographic process. Addressing
this point is crucial for preparative purposes when
volumes must be kept to a minimum and cleaner mix-
tures mean easier purifications. We have previously
It has been described that the oxidation of (+)-catechin
2 by DPPH takes place in two kinetic steps, one fast
step when the most labile hydrogen atoms are abstrac-
ted and simple dimers formed, and a second slow step
when dimers are further oxidised and polymerised
(Scheme 2, catechin).33 The electron-rich positions 6
and 8 are putatively involved in the polymerisation
process by nucleophilic additions on flavanol derived
o-quinones.33 Altogether the process results in the
reduction of more than two molecules of DPPH per
molecule of (+)-catechin. Regeneration of phenolic
hydroxyls (Scheme 2, third step) by polymerisation
appears to be the reason for the high number of H
atoms or electrons involved in the scavenging activity of
polyphenols under a variety of other experimental set-
ups such as enzyme catalysed oxidation followed by

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2503
30
either product purification35 or EPRmeasurements,
and electrochemical oxidation.36
process, we introduced changes in the appropriate
acquisition parameters with the same outcome: H-6 and
H-8 signals had completely vanished from the old sam-
ples. The exchange was corroborated by 13C NMRwith
a sample of 4 previously lyophilised in the presence of
D2O as described in Experimental and by a kinetic
study at controlled pD for compounds 1, 4 and 5 (Figs 2
and 3). It is known that phloroglucinol derivatives are
sensitive to the presence of Lewis and protic (low pH)
acids reactive towards electrophiles.37 When spectra of
those compounds were recorded in solvents with inter-
changeable deuterium atoms (e.g., D2O) the aromatic
ring hydrogen atoms could be exchanged by deuterium.
In flavan-3-ols hydrogen atoms at positions 6 and 8
have similar properties but deuterium exchange is more
difficult.37 This exchange may be related to the well-
known aromatic electrophilic substitution or keto-enol
The introduction of sulphur atom containing moieties
into position 4 of the flavan system have resulted in
derivatives (4–7) able to reduce at least one more mole-
cule of DPPH than (ꢀ)-epicatechin 1. The differences in
antiradical efficiencies may be related to an enhanced
nucleophilic character of positions 6 and 8 in ring A.
Our evidence comes from NMRstudies. While the 1H
NMRspectrum of ( ꢀ)-epicatechin 1 remained almost
unaltered in acetone-d6/D2O, the signals corresponding
to protons 6 and 8 from compounds 4, 5 disappeared
progressively over a period of 1–72 h, suggesting that
these protons were exchanged by deuterium. To rule out
detection problems arising from long relaxation times of
those atoms or due to a dynamic chemical exchange
Scheme 2.

2504
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
tautomerism applied to the case of aromatic phenol
rings with highly increased p-electron density. Our data
show clearly that the substitution of the 4b hydrogen of
flavanols by cysteamine or cysteine through a thioether
bond exerts a dramatic influence over the properties of
the contiguous aromatic ring A. In the pseudo first
order kinetic conditions used (great excess of the
electrophile reagent), the difference in exchange half-
lives may be directly related to variations in the acti-
vation energy barriers. This effect may be mediated by
the formation of intramolecular hydrogen bonds
between the phenol groups and the functional groups on
the sulphur containing moiety. Other mechanisms
directly related with the sulphur atom may also be
involved. In any case our results show that positions 6
and particularly 8 of compounds 4–7 posses an
increased ability to react with electrophilic species (e.g.,
water). This may result in a higher capacity to regener-
ate polyphenolic hydroxyls through polymerisation
(Scheme 2) that would explain why derivatives 4–6 are
able to provide more H atoms than (ꢀ)-epicatechin 1.
Compound 7 is the most effective of the conjugates as
expected by the presence of the pyrogallol group con-
taining gallate ester moiety, which provides additional
H atoms and polymerisation positions.29,38
pattern multiplet coming from 4 and 5. This pattern
differs from the spectra of radicals generated from
1-substituted 2,3,4-trihydroxybenzenes, 1-substituted
3,4,6-trihydroxybenzenes and 1-substituted 3,4,5-tri-
hydroxybenzenes.40 The stable radical detected for the
sulphur containing derivatives might alternatively be
formed on ring A or C. In view of the bond dissociation
energies of the different carbon-hydrogen bonds
obtained by theoretical calculations (Fig. 4), a radical
on C-4 is likely to be formed under our experimental
conditions. This radical would be coupled to three pro-
tons at positions 3, 6 and 8 which might explain the
EPRpattern recorded (Figs 5b and 6a). This point is
currently being looked into in our Laboratories. Apart
from those described so far, other reactions may occur
in an aqueous basic medium, namely oxidation to sulf-
oxides and nucleophilic attack of the amino group onto
quinones, which may influence the formation and/or
stabilisation of radicals. In any case the thio derivatives
tend to generate relatively stable radicals that last for days
and remain sensitive to further oxidation triggered by UV
light (Fig. 6) and, presumably, by reactive oxygen species.
An alternative explanation for the enhanced capacity of
the new derivatives to transfer hydrogen atoms would
involve the substituted position 4 on ring C. Mechan-
istic studies on radical oxidation of catechins by
2,20-azobis(2-aminopropane)hydrochloride (AAPH) and
semiempirical calculations show that hydrogen atoms
other than phenolic (i.e., H-2 on ring C) may be
abstracted from flavanols.38,39 Using more accurate
theoretical calculations, which have been shown to
compare quite well with the experimental values,28 we
obtained a BDE for the C–H bond at position 4 of
compound 5 similar to the values for the C–H bond at
position 2 and the O–H bond at position 40 (Fig. 4).
This result is suggesting that the hydrogen atom on C-4
may be abstracted by oxidising agents such as DPPH
and might contribute to the enhanced antiradical power
of the thiol containig derivatives.
Antiproliferative activity
It has been reported that flavan-3-ols are active against
proliferation of cancer cells including colon carcinoma
cell lines, being (ꢀ)-epicatechin 1 and (+)-catechin 2 the
least efficient of the monomers.4,41,42 The novel 2-amino-
ethylthio and cysteinyl derivatives 4–7 proved to be
more efficient than 1 in HT29 human carcinoma cells. It
has been proposed that the antiproliferative activity of
catechins (flavanols) and other flavonoids comes from
the inhibition of the kinase activity of proteins involved
in cell cycle regulation43ꢀ46 rather than any interaction
with any antioxidant defence mechanism. The conjugates
might have higher affinity than intact flavan-3-ols for the
appropriate protein binding site. Alternatively, conjuga-
tion might influence the transport of the flavanols into the
cell to reach putative intracellular binding sites. This point
is also being looked into in our Laboratories. Curiously,
the order of efficiency of the molecules on both the anti-
radical and antiproliferative assays was identical. Whether
this is just a coincidence or not is something that will have
to be clarified. Thorough studies on cell cycle disruption
and apoptosis are also under way.
EPRexperiments of spontaneous air oxidation of 4, 5
and (ꢀ)-epicatechin 1 in 1 M aqueous NaOH also evi-
denced differences between derivatised and under-
ivatised flavanols. The multiplet (Fig. 5b), with a Lande
factor g=2.0041, detected when 4 and 5 (10 mM) were
analysed corresponds to the coupling of the unpaired
electron with three protons with constant values of 2.67,
1.55 and 1.62 G. In a very extensive work by Jensen and
Pedersen31 about oxidation of (ꢀ)-epicatechin 1, (+)-
catechin 2 and catechinic acid in basic solutions they
concluded that oxidation of (ꢀ)-epicatechin led to the
stable radicals 8 and 9 in alkaline aqueous DMSO
solution, and the generation of only 8 in alkaline aqueous
EtOH solution. These radicals appeared to be species
resulting from hydroxylation on ring B and rearrange-
ment reactions on rings A and C of (ꢀ)-epicatechin. We
have confirmed that oxidation of (ꢀ)-epicatechin 1 in
concentrated (50 mM) 1 M aqueous NaOH solution
generates radical 8, and in no case gives the six-line
Conclusions
Potent cysteine conjugates with natural antioxidants of
the flavanol type have been obtained by acid cleavage of
polymeric procyanidins followed by cation-exchange

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2505
and RP-HPLC purification. The derivatisation appears
to improve the free radical scavenging capacity of fla-
vanols by fostering the regeneration of active hydroxyl
groups upon polymerisation and/or by facilitating the
abstraction of the hydrogen atom geminal to the sul-
phur atom. The new conjugates are also weak inhibitors
of colon carcinoma cells proliferation. The cysteinyl
flavanols, together with the previously described cystea-
mine conjugates incorporate chemical functions (amine,
carboxylic acid) and include activated positions (8 and 6
on ring A) which may facilitate the preparation of other
products with improved properties.
volume) packed in-house with MacroPrepTM High S 50
mm (BioRad Laboratories, Hercules, CA, USA). For
preparative purposes, the same stationary phase was
packed into a flash chromatography type glass column
(21 ꢃ 2.5 cm i.d., ca. 105 mL bed volume). Preparative
RP-HPLC chromatography was performed on
a
Waters (Milford, USA) Prep LC 4000 pumping system
with a Waters PrepPack1 1000 module fitted with a
PrepPack1 Waters cartridge (30 ꢃ 4.7 cm i.d.) filled
with VYDACTM (The Separations Group, Hesperia,
USA) C18, 300 A pore size, 15–20 mm particle size
stationary phase. A flow splitter was placed after the
column and detection was done by an analytical Merck-
Hitachi (Darmstadt, Germany) L-4000 UV detector.
Semipreparative runs were performed on the same
pumping system fitted with a Perkin–Elmer C18 column
(25 ꢃ 2 cm i.d.). Analytical RP-HPLC was performed
on either a Kontron Analytical system (Kontron
Experimental
Materials
The primary source of procyanidins was the byproduct
from pressing destemmed Parellada grapes (Vitis vini-
fera) (Bodegas Miguel Torres, S.A., Vilafranca del
Penedes, Spain) which were extracted and fractionated
as described.19,20 Water and solvents used were: analy-
tical grade MeOH (Panreac, Montcada i Reixac, Spain)
for the acid cleavage reaction and DPPH assay; deio-
nised water and bulk EtOH (Momplet y Esteban, Bar-
celona, Spain) for semipreparative and preparative
cation-exchange chromatography; milli-Q1 water and
HPLC grade CH3CN (E. Merck, Darmstadt, Germany)
for analytical RP-HPLC; deionised water and pre-
parative grade CH3CN (Scharlau, Barcelona, Spain) for
preparative and semipreparative RP-HPLC; Milli-Q1
water for electron paramagnetic resonance (EPR) stud-
ies. Deuterated solvents for nuclear magnetic resonance
(NMR) were from SDS (Peypin, France). Cysteine
hydrochloride (Aldrich, Steinheim, Germany) was of
synthesis grade. Acetic acid, 37% HCl (E. Merck) and
NaCl (Carlo Erba, Milano, Italy) were of analytical
grade. Triethylamine (E. Merck) was of buffer grade.
Trifluoroacetic acid (TFA, Fluorochem, Derbyshire,
UK) biotech grade was distilled in-house. 1,1-Diphenyl-
2-picrylhydrazyl free radical (DPPH) (95%) and gallic
acid (97%) were from Aldrich (Gillingham-Dorset, UK),
6-hydroxy-2,5,7,8-tetramethyl-chroman-2-carboxylic acid
(Trolox) (97%) from Aldrich (Milwaukee, USA) and
(ꢀ)-epicatechin from Sigma Chemical CO, (Saint Louis,
MO, USA).
Instruments, Basel, Switzerland) fitted with
a
TM
VYDAC C18, 300 A pore size, 5 mm particle size, 250
ꢃ 4.6 mm i.d. column, or a Smart1 System (Amersham-
Pharmacia Biotech) equipped with a m Peak Monitor
(Amersham-Pharmacia Biotech) and fitted with a mRPC
C2/C18 SC 2.1/10 (100 ꢃ 2.1 mm i.d.) column (Amer-
sham-Pharmacia Biotech).
Mass spectrometry, NMR and EPR measurements
Electrospray–mass spectrometry (ES–MS) analyses
were recorded on a VG-Quattro1 system from Fisons
Instruments (Altricham, UK). The carrier solution was
water/CH3CN (1:1) containing 1% (v/v) formic acid. 1H
and 13C NMRspectra were observed with a Unity-300
spectrometer from Varian (Palo Alto, CA, USA) for
acetone-d6/D2O solutions using standard acquisition
conditions at a regulated temperature of 25 ꢅC. Con-
trolled deuterium exchange experiments were performed
in D2O/acetone-d6 (100% deuteration degree, 3:1 v/v)
buffered at pD=6 (measured with standard pH elec-
trode) with deuterated ammonium chloride (150 mM
final concentration in the solvent mixture). All the glass
material used in the experiment was thoroughly washed
with the deuterated solvent mixture to remove all traces
of labile hydrogen. Compounds (ca. 5 mg) were dis-
solved in 0.75 mL of the buffered solvent mixture (con-
centration between 15 and 20 mM) just before starting
the NMRacquisition. Spectra were acquired at 30 ꢅC
using short pulse widths (15ꢅ) and 32 scans (for each
kinetic data point) with 4 s interpulse time in order to
assure a similar integral to concentration factor for all
spectral signals. In the case of deuterium substituted
samples, the carbon spectra were acquired using an
interpulse delay of 5 s and a pulse of 25ꢅ until a mini-
mum signal:noise ratio of 40:1 for the quaternary aro-
matic ring carbons signals was achieved.
Fetal bovine serum (FBS) was purchased from Gibco
(Invitrogen, Carlsbad, CA, USA). Dulbecco’s phos-
phate buffer saline (PBS) and trypsin-EDTA solution C
(0.05% trypsin & EDTA 1:5000 in PBS) were from
Biological Industries (Kibbutz Beit Haemek, Israel).
MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetra-
zolium bromide) and DMSO (dimethyl sulfoxide) were
from Sigma Chemical CO (Saint Louis, MO, USA).
EPRmeasurements were performed on a Varian (Palo
Alto, CA, USA) E-109 spectrometer. The freshly pre-
pared basic solutions (1 M NaOH) of the samples were
degassed and transferred to a quartz capillary tube, fit-
ted into the cavity of the spectrometer working in the X
band at room temperature, and the changes in the spec-
tra were recorded with time. g Values of the radicals were
Chromatographic equipment and columns
Semi-preparative cation-exchange chromatography was
performed on a FPLC1 system (Amersham-Pharmacia
Biotech, Uppsala, Sweeden) fitted with a Omnifit
(Cambridge, UK) column (8 ꢃ 1 cm i.d., ca. 6 mL bed

2506
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
determined with DPPH (g=2.0037) as standard.
Measurement conditions: microwave power, 5 mW;
modulation amplitude, 0.1 G; modulation 100 kHz;
scan width, 20 G. High values of gain were necessary
to detect very low concentrations of magnetic species.
set-up the separation conditions at milligram scale. The
preparative separations were performed on a 105 mL
bed volume column loaded with the same stationary
phase. The eluents were [A]: 20 mM sodium phosphate,
pH 2.25 buffer/EtOH (7:3) and [B]: 20 mM sodium
phosphate, pH 2.25 buffer/EtOH (13:7), 100 mM NaCl.
The column was equilibrated with eluent [A], loaded
with the quenched depolymerised mixture (500 mL) and
washed with [A] (500 mL, 4.75 bed volumes). The
retained flavan-3-ol-derivatives were released with 500
mL (4.75 bed volumes) of eluent [B]. The column was
then washed with 7.14 bed volumes (750 mL) of 20 mM
sodium phosphate buffer, pH 2.25/EtOH (3:2), 1 M
NaCl. The operation was repeated (7 times total) until
the whole mixture was consumed. The separation pro-
cess was monitored by analytical RP-HPLC on a
Theoretical calculations
The program Gaussian 9847 was used to perform all
calculations described in this paper. The bond dissocia-
tion enthalpies were computed following the theoretical
approach pointed out by Wright et al.28 In a first step,
we carried out geometry optimisations and vibration
frequencies calculations for X–H and X radicals by
using the semiempirical PM3 method. In a second step,
the electronic energies were obtained by performing
single point calculations at the optimised geometries by
means of density functional theory (DFT) using the
B3LYP functional.27 The bond dissociation enthalpies
at 298 K were then computed as the difference between
the B3LYP energy of the X+H radicals minus the
B3LYP energy of the XH molecule, plus the corres-
ponding thermochemical corrections obtained at the
PM3 level of theory. The B3LYP calculations were per-
formed using the 6–31G(d,p) basis set,48 with the
p-exponent on hydrogen changed to 1.0 as optimised by
Wright et al.28 Moreover, in computing the thermo-
chemical corrections, we have also scaled the zero point
energy (ZPE) by 0.947 as suggested by Wright et al.28 It
is worth noting that our results show that the PM3
derived BDEs are systematically underestimated in the
8–30 kcal/mol range, with respect to the more accurate
B3LYP values. We obtained computed BDEs of 69.2,
65.9, 60.3, 65.0, 67.5, 71.4 and 71.7 kcal/mol for posi-
tions 2, 3, 4, 30, 40, 5 and 7, respectively, at PM3 level of
theory, which are lower than those displayed in Fig. 4,
calculated at B3LYP level of theory. The reason for this
discrepancy is that the high level of delocalisation of the
radicals is not accurately described at the semiempirical
level of theory.
TM
VYDAC C18 column eluted with a binary system, [C]
0.10% (v/v) aqueous TFA, [D] 0.09% (v/v) TFA in
water/CH3CN (2:3) under isocratic conditions 14% [D]
(5 and 6), 22% [D] (compound 7), at a flow rate of 1.5
mL/min and detection at 215 nm, 0.016 absorbance
units full scale (aufs). The eluates containing com-
pounds 5–7 were pooled (3.5 L), the solvent volume was
reduced under vacuum down to 1.8 L, and water was
added up to a final volume of 2.6 L. The RP-HPLC
profile of the pooled eluates as well as the initial and
final washing steps were recorded on the same analytical
system under gradient conditions 8–23% [D] over 45
min at a flow rate of 1.5 mL/min with detection at
215nm.
The mixture containing compounds 5–7 was fractio-
nated on a preparative RP-HPLC cartridge filled with
TM
VYDAC C18 stationary phase by a CH3CN gradient
in 0.1% (v/v) aqueous TFA (2–14% CH3CN over 45
min). The solution (2.6 L) was processed in three por-
tions of ca. 900 mL each. Fractions enriched in each of
the three compounds were obtained: fraction I, 2–5%
CH3CN, compound 6; fraction II, 5–8% CH3CN, com-
pound 5; fraction III, 10–13% CH3CN, compound 7.
Treatment of procyanidins with acid in the presence of
cysteine
Purification of the S-cysteinyl derivatives
The 4b-(S-cysteinyl)flavan-3-ols were purified from
fractions I–III by preparative RP-HPLC and identified
by mass spectrometry and nuclear magnetic resonance.
The solvent (water saturated with EtOAc) was elimi-
nated from an aliquot (400 mL, 4 g GAE, 6 g estimated
polyphenols by weight, coming from 3.2 kg of grape
byproduct) of the source of procyanidins.20 The pellet
was then dissolved in MeOH (400 mL) and dried. This
operation was repeated three times in order to eliminate
moisture. The resulting syrupy residue was dissolved in
MeOH (400 mL) and a solution of cysteine hydro-
chloride (20 g) and 37% HCl (10 mL) in MeOH (400
mL) was added. The mixture was kept at 65 ꢅC for 20
min under agitation. The reaction was then quenched
with cold water (3.2 L).
4ꢀ-(S-Cysteinyl)epicatechin 5. Fraction II (6.7 L) from
reversed-phase fractionation was concentrated (5 L final
volume) under vacuum to eliminate most of the
CH3CN, loaded onto the preparative cartridge and
eluted using a CH3CN gradient (0–12% over 60 min) in
triethylamine phosphate pH 2.25 buffer, at a flow rate
of 100 mL/min, with detection at 230 nm. Compound 5
eluted at 4–7% CH3CN. Analysis of the fractions was
accomplished under isocratic conditions in 0.10% (v/v)
TM
aqueous TFA/CH3CN using the VYDAC C18 col-
umn, solvent system, flow rate and detection described
above with isocratic elution at 14% [D]. The pure fractions
were pooled, diluted with water (1:1) and re-chroma-
tographed on the same cartridge by a CH3CN gradient
(2 to 11% over 30 min) in 0.10% (v/v) aqueous TFA.
After combining the eluates and lyophilization, the
Cation exchange separation of the cysteinyl derivatives
from the depolymerised mixture and fractionation by
preparative RP-HPLC
Semipreparative runs on a 6 mL bed volume column
loaded with MacroPrepTM High S resin were used to

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2507
resulting solid (0.9 g) was dissolved in tryethylamine
phoshate buffer pH 5.54 and loaded again onto the
preparative cartridge previously equilibrated in the
same buffer. Compound 5 was eluted by a CH3CN gra-
dient (0–12% over 90 min). The pure fractions were
pooled and desalted by a fast CH3CN gradient in 0.1%
(v/v) aqueous TFA on the same cartridge. 4b-(S-cystei-
nyl)epicatechin 5 (550 mg) was obtained as the tri-
fluoroacetate by lyophilisation. ES–MS positive ions, m/z
410.0 (M+1)+, calculated for C18H20N1O8S1 (M+H)+
aqueous TFA/CH3CN using the column, solvent sys-
tem, flow rate and detection described above with elu-
tion at 19% [D]. The best fractions were pooled, diluted,
re-loaded onto the cartridge and eluted with a CH3CN
gradient (2–14% over 60 min) in triethylamine phos-
phate pH 5.46 buffer. The purest fractions were pooled
and re-chromatographed with a CH3CN gradient
(9–21% over 30 min) in 0.1% (v/v) aqueous TFA. After
lyophilisation, 4b-(S-cysteinyl)epicatechin 3-O-gallate 7
(75 mg) was obtained as the trifluoroacetate. ES–MS,
positive ions, m/z 561.9 (M+1)+ calculated for
1
410.1. H NMR(acetone- d6+3 drops D2O, 300 MHz):
1
d 3.98 (1H, d J=2.0 Hz, H-4 3, 4-trans configuration);
4.08 (1H, dd J=2.0, 0.9 Hz, H-3); 4.45 (1H, m, S–CH2–
CH<); 5.09 (1H, s, H-2 2, 3-cis configuration); 5.89
(1H, d J=2.1 Hz, H-8); 6.10 (1H, d, J=2.1 Hz, H-6);
6.77–6.86 (2H, m, H-50, H-60); 7.07 (1H, d J=2.1 Hz, H-
20). Purity (>99.5%) was ascertained by RP-HPLC on
a mRPC C2/C18, 3 mm column; elution, [C] 0.10% (v/v)
aqueous TFA, [D] 0.09% (v/v) TFA in water/CH3CN
(2:3), gradient 8 to 23% [D] over 45 min at a flow rate of
200 mL/min with simultaneous detection at 214, 280 and
320 nm.
C25H24N1O12S1 (M+H)+ 562.1. H NMR(acetone- d6
+3 drops D2O, 300 MHz): d 4.15 (1H, d J=2.1 Hz, H-4
3, 4-trans configuration); 4.37 (1H, m, S–CH2–CH<);
5.28 (1H, 2m, H-3); 5.39 (1H, bs, H-2 2, 3-cis configur-
ation); 6.01 (1H, d J=2.1 Hz, H-8); 6.13 (1H, d, J=2.1
Hz, H-6); 6.78 (1H, d J=8.1 Hz, H-50); 6.90 (1H, dd
J=2.1, 8.1 Hz, H-60); 6.97 (2H, s, galloyl H); 7.11 (1H,
d J=2.1 Hz, H-20). Purity (>99.5%) was ascertained by
RP-HPLC on the system described for compound 5.
Antiradical activity
The free radical scavenging activity was evaluated by
the DPPH method.22,23 The samples (0.1 mL) were
added to aliquots (3.9 mL) of a solution made up with
4.8 mg DPPH in 200 mL of MeOH and the mixture
incubated for 1 h at room temperature. The initial con-
centration of DPPH, approximately 60 mM, was calcu-
lated for every experiment from a calibration curve
made by measuring the absorbance at 517 nm of stan-
dard samples of DPPH at different concentrations. The
equation of the curve was Abs517nm=11,345 ꢃ CDPPH
as determined by linear regression. The results were
plotted as the degree of absorbance disappearance at
517 nm [(1ꢀA/A0)ꢃ100] against mmoles of sample divi-
ded by the initial mmoles of DPPH. Each point was
acquired in triplicate. A dose-response curve was
obtained for every product. The results were expressed
as the efficient dose ED50 given as mmoles of product
able to consume half the amount of free radical divided
by mmoles of initial DPPH.
4ꢀ-(S-Cysteinyl)catechin 6
Fraction I from reversed-phase fractionation was con-
centrated as stated for 5, loaded onto the preparative
cartridge and eluted using a CH3CN gradient (0–12%
over 90 min) in triethylamine phosphate pH 2.25 buffer,
at a flow rate of 100 mL/min, with detection at 230 nm.
Analysis of the fractions was accomplished under iso-
cratic conditions in 0.10% (v/v) aqueous TFA/CH3CN
using the VYDACTM C18 column, solvent system, flow
rate and detection described above with isocratic elution
at 10% [D]. The best fractions were pooled, diluted, re-
loaded onto the cartridge and eluted with a CH3CN
gradient (0–12% over 90 min) in triethylamine phos-
phate pH 5.30 buffer. The purest fractions were pooled,
desalted with a steep CH3CN gradient in 0.1% (v/v)
aqueous TFA and lyophilised. Then the preparation
was re-chromatographed on a semipreparative Perkin–
Elmer C18 cartridge eluted with 6 and 9% CH3CN in
0.10% (v/v) aqueous TFA under isocratic conditions.
After pooling the best fractions and lyophilization, 4b-
(S-cysteinyl)catechin 6 (74 mg) was obtained as the tri-
fuoroacetate. ES–MS, positive ions, m/z 409.9
(M+1)+, calculated for C18H20N1O8S1 (M+H)+ 410.1.
1H NMR(acetone- d6+ 3 drops D2O, 300 MHz): d 4.15
(1H, 2d, J=9.7, 4.5, H-3); 4.28 (1H, d J=4.5, H-4 3,
4-cis configuration); 4.4 (1H, m, S–CH2–CH<); 4.80
(1H, d J=9.7 Hz, H-2 2, 3-trans configuration); 5.76
(1H, d J=2.1 Hz, H-8); 6.12 (1H, d, J=2.1 Hz, H-6);
6.78 (2H, m, H-50, H-60); 6.93 (1H, d J=1.2 Hz, H-20).
Purity (99.9%) was ascertained by RP-HPLC on the
system described for compound 5.
Cell culture
HT29 cells (colorectal adenocarcinoma) were purchased
from American Type Culture Collection (ATCC) and
cultured in Dulbecco modified Eagle’s medium
(DMEM) supplemented with 10% (v/v) heat-inactivated
FBS, 2 mM l-glutamine and antibiotics: 100 U/mL
penicillin and 100 mg/mL streptomycin (Biological
Industries). Cells were grown in an isolated 37 ꢅC–5%
CO2 tissue incubator compartment and the medium was
changed every 3 days. HT29 cell cultures used in this
study were free of mycoplasm infection as shown by the
Gen-probe rapid mycoplasm detection system prior to
the treatments with the samples.
4ꢀ-(S-Cysteinyl)epicatechin 3-O-gallate 7. Fraction III
from reversed-phase fractionation was concentrated as
stated for 5, loaded onto the preparative cartridge and
eluted using a CH3CN gradient (2–14% over 60 min) in
triethylamine phosphate pH 2.25 buffer, at a flow rate
of 100 mL/min, with detection at 230 nm. Fractions
were analysed under isocratic conditions in 0.10% (v/v)
Proliferation assay
Cell growth was determined using the MTT assay.49
Cells were seeded into 96-well plates at 2.5 ꢃ 103 cells/
well density and incubated for 24 h prior to addition of

2508
J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
the samples, which were prepared by serial dilutions
from a concentrated solution made up in Dulbecco’s
phosphate buffered saline (PBS). Control cultures were
treated with equal volume of PBS as the treated cultures.
After 3 days of culture, the supernatant was aspirated
and 100 mL of filtered MTT (0.2 mg/mL in cell culture
medium) was added to each well. The plates were incu-
bated at 37 ꢅC–5% CO2 during 4 h. The supernatant
was removed, and the blue MTT formazan precipitated
was dissolved in DMSO (100 mL) and optical density
(OD) measured at 550 nm on a multi-well reader
(Merck ELISA System MIOS1).
9. Tan, X. H.; Hu, D. R.; Li, S. R.; Han, Y.; Zhang, Y. L.;
Zhou, D. Y. Cancer Lett. 2000, 158, 1.
10. Harbowy, M. E.; Balentine, D. A. Crit. Rev. Plant Sci.
1997, 16, 415.
11. Prieur, C.; Rigaud, J.; Cheynier, V.; Moutounet, M.
Phytochemistry 1994, 36, 781.
12. Souquet, J.-M.; Cheynier, V.; Brossaud, F.; Moutounet,
M. Phytochemistry 1996, 43, 509.
13. Souquet, J. M.; Labarbe, B.; LeGuerneve, C.; Cheynier,
V.; Moutounet, M. J. Agric. Food Chem. 2000, 48, 1076.
14. Cheynier, V.; Souquet, J.-M.; Le Roux, E.; Guyot, S.;
Rigaud, J. Method. Enzymol. 1999, 299, 178.
15. Thompson, R. S.; Jacques, D.; Haslam, E.; Tanner,
R . J. N.J. Chem. Soc., Perkin Trans. 1 1972, 1387.
16. Rigaud, J.; Perez-Ilzarbe, J.; Ricardo da Silva, J. M.;
Cheynier, V. J. Chromatogr. 1991, 540, 401.
17. Sanoner, P.; Guyot, S.; Marnet, N.; Molle, D.; Drilleau,
J. F. J. Agric. Food Chem. 1999, 47, 4847.
18. Tanaka, T.; Kusano, R.; Kouno, I. Bioorg. Med. Chem.
Lett. 1998, 8, 1801.
19. Torres, J. L.; Bobet, R. J. Agric. Food Chem. 2001, 49, 4627.
20. Polymeric procyanidins came from the residue generated
after extraction of antioxidant polyphenols from grape
pomace. Briefly, the total extract in 70% EtOH was dried,
suspended in water and partitioned with ethyl acetate. The
residual aqueous layer used in this work contained procyani-
dins with a mean degree of polimerisation of 8.5 and a content
of gallate esters (galloylation) of 11%, calculated by thiolysis
with cysteamine and HPLC analysis (Torres, J. L.; Lozano, C.
Chromatographia 2001, 54, 523).
For each compound, a minimum of four experiments
measuring the growth inhibition was conducted and the
meanꢄSD of OD data from the replicated wells was
calculated for each concentration tested. The inhibitory
effect of the products at each concentration was expres-
sed as a percentage [(mean OD treated cells after 72 h of
incubation with the product/mean OD of control cells
after 72 h of incubation with extra-medium instead of
product) ꢃ 100]. The IC50 or sample concentration
causing a 50% reduction in the mean OD value relative
to the control at 72 h of incubation, was estimated using
GraFit 3.00 (Data Analysis and Graphics Program,
Erithacus Software Ltd. Microsoft Corp., Surrey, UK)
curve option: IC50 curve — start at 0.
21. Kolodziej, H. Phytochemistry 1990, 29, 1671.
22. Blois, M. S. Nature 1958, 181, 1199.
Acknowledgements
23. Brand-Williams, W.; Cuvelier, M. E.; Berset, C. Lebensm.-
Wis. u.-Technol. 1995, 28, 25.
24. Free radical scavenging efficiency of 4 was reported in ref
19. ED50 was 0.11 (ARP 9.1). The corresponding stoichio-
metry value is 0.22 and the moles DPPH reduced by mole
antioxidant is 4.5.
25. Shen, C. C.; Chang, Y.-S.; Ho, L.-K. M. Phytochemistry
1993, 34, 843.
26. Hemingway, R. W.; Tobiason, F. L.; McGraw, G. W.;
Steynberg, J. P. Mag. Res. Chem. 1996, 34, 424.
27. Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
28. Wright, J. S.; Carpenter, D. J.; McKay, D. J.; Ingold,
K. U. J. Am. Chem. Soc. 1997, 119, 4245.
29. Guo, Q. N.; Zhao, B. L.; Li, M. F.; Shen, S. R.; Xin, W. J.
Biochim. Biophys. Acta-Lipid Lipid Met. 1996, 1304, 210.
30. Bors, W.; Michel, C.; Stettmaier, K. Archiv. Biochem.
Biophys. 2000, 374, 347.
Financial support (research grants PPQ2000–0688-C05–
03 and-04, and doctoral fellowship to C.L.) from the
Spanish Ministry of Science and Technology is
acknowledged. We are thankful to Dr. Irene Fernandez
from the Servei d’Espectrometria de Masses of the Uni-
versity of Barcelona for the mass spectrometry analyses,
Ms. Avencia Dıez from the Servei d’Espectrometria de
`
Ressonancia Paramagnetica Electronica at IIQAB-CSIC
for the EPRanalyses and Ms. Montserrat Sindreu
`
`
`
from the Servei d’Espectrometria de Ressonancia Mag-
netica Nuclear at IIQAB-CSIC for the NMRanalyses.
´
The calculations described in this work were performed
´
at the Centre de Supercomputacio de Catalunya
(CESCA).
31. Jensen, O. N.; Pedersen, J. A. Tetrahedron 1983, 39, 1609.
32. Torres, J. L.; Piera, E.; Infante, M. R.; Clapes, P. Prep.
Biochem. Biotechnol. 2001, 31, 259.
33. Dangles, O.; Fargeix, G.; Dufour, C. J. Chem. Soc., Per-
kin Trans. 2 1999, 1387.
References and Notes
1. Diplock, A. T.; Charleux, J. L.; Crozier-Willi, G.; Kok,
˜
F. J.; Rice-Evans, C.; Roberfroid, M.; Stahl, W.; Vina-Ribes,
34. Dangles, O.; Fargeix, G.; Dufour, C. J. Chem. Soc., Per-
kin Trans. 2 2000, 1653.
J. Br. J. Nutr. 1998, 80, S77 (Suppl 1).
2. Bravo, L. Nutr. Rev. 1998, 56, 317.
35. Guyot, S.; Vercauteren, J.; Cheynier, V. Phytochemistry
1996, 42, 1279.
3. Yang, C. S.; Lee, M.-J.; Chen, L.; Yang, G. Environ. Health
Perspect. 1997, 105 (Suppl. 4), 971.
4. Yang, G. Y.; Liao, J.; Kim, K.; Yurkow, E. J.; Yang, C. S.
Carcinogenesis 1998, 19, 611.
5. Soleas, G. J.; Diamandis, E. P.; Goldberg, D. M. J. Clin.
Lab. Anal. 1997, 11, 287.
6. Ruf, J. C. Drug Exp. Clin. Res. 1999, 25, 125.
7. Packer, L.; Rimbach, G.; Virgili, F. Free Rad. Biol. Med.
1999, 27, 704.
36. Hotta, H.; Sakamoto, H.; Nagano, S.; Osakai, T.; Tsu-
jino, Y. Biochim. Biophys. Acta- Gen. Subjects 2001, 1526, 159.
37. Kolar, G. F. J. Label. Compd. 1971, 7, 409.
38. Kondo, K.; Kurihara, M.; Miyata, N.; Suzuki, T.;
Toyoda, M. Free Rad. Biol. Med. 1999, 27, 855.
39. Kondo, K.; Kurihara, M.; Miyata, N.; Suzuki, T.;
Toyoda, M. Archiv. Biochem. Biophys. 1999, 362, 79.
40. Stone, T. J.; Waters, W. A. J. Chem. Soc. 1965, 1488.
41. Valcic, S.; Timmermann, B. N.; Alberts, D. S.; Wachter,
¨
G. A.; Krutzsch, M.; Wymer, J.; Guillen, J. M. Anti-Cancer
Drugs. 1996, 7, 461.
8. Ahmad, N.; Feyes, D. K.; Nieminen, A. L.; Agarwal, R.;
Mukhtar, H. J. Nat. Cancer Inst. 1997, 89, 1881.

J. L. Torres et al. / Bioorg. Med. Chem. 10 (2002)2497–2509
2509
42. Chung, L. Y.; Cheung, T. C.; Kong, S. K.; Fung, K. P.;
Choy, Y. M.; Chan, Z. Y.; Kwok, T. T. Life Sci. 2001, 68,
1207.
43. Liang, Y. C.; Lin-shiau, S. Y.; Chen, C. F.; Lin, J. K.
J. Cell. Biochem. 1997, 67, 55.
M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, S.; Clifford, S.; Och-
terski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma,
K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Fores-
man, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu,
G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A.
Gaussian 98, Revision A.6; Gaussian Inc: Pittsburg, PA, 1998.
48. Hariharan, P. C.; Pople, J. A. Theoret. Chim. Acta 1973,
28, 213.
44. Ahmad, N.; Gali, H.; Javed, S.; Agarwal, R. Biochem.
Biophys. Res. Commun. 1998, 247, 294.
45. Liang, Y. C.; Chen, Y. C.; Lin, Y. L.; Lin-Shiau, S. Y.;
Ho, C. T.; Lin, J. K. Carcinogenesis 1999, 20, 733.
46. Liberto, M.; Cobrinik, D. Cancer Lett. 2000, 154, 151.
47. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria,
G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A., Jr; Stratmann, R. E.; Burant, J. C.; Dap-
prich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain,
49. Mosmann, T. J. Immunol. Meth. 1983, 55.