Comparison of a Pair of Synthetic Tea-Catechin-Derived Epimers: Synthesis, Antifolate Activity, and Tyrosinase-Mediated Activation in Melanoma

Article abstract of DOI:10.1002/cmdc.201000482

Despite bioavailability issues, tea catechins have emerged as promising chemopreventive agents because of their efficacy in various animal models. We synthesized two catechin-derived compounds, 3-O-(3,4,5-trimethoxybenzoyl)-(-)-catechin (TMCG) and 3-O-(3,4,5-trimethoxybenzoyl)-(-)-epicatechin (TMECG), in an attempt to improve the stability and cellular absorption of tea polyphenols. The antiproliferative and pro-apoptotic activities of both compounds were analyzed with various cancer cell systems, and TMCG, which was easily synthesized in excellent yield, was more active than TMECG in both melanoma and non-melanoma cell lines. TMCG was also a better inhibitor of dihydrofolate reductase and was more efficiently oxidized by tyrosinase, potentially explaining the difference in activity between these epimers.

Full text of DOI:10.1002/cmdc.201000482

MED
DOI: 10.1002/cmdc.201000482
Comparison of a Pair of Synthetic Tea-Catechin-Derived
Epimers: Synthesis, Antifolate Activity, and Tyrosinase-
Mediated Activation in Melanoma
Magalꢀ Sꢁez-Ayala,[a] Luis Sꢁnchez-del-Campo,[a] Marꢀa F. Montenegro,[a] Soledad Chazarra,[a]
Alberto Tꢁrraga,[b] Juan Cabezas-Herrera,[c] and Josꢂ Neptuno Rodrꢀguez-Lꢃpez*[a]
Despite bioavailability issues, tea catechins have emerged as
promising chemopreventive agents because of their efficacy in
various animal models. We synthesized two catechin-derived
compounds, 3-O-(3,4,5-trimethoxybenzoyl)-(ꢀ)-catechin (TMCG)
and 3-O-(3,4,5-trimethoxybenzoyl)-(ꢀ)-epicatechin (TMECG), in
an attempt to improve the stability and cellular absorption of
tea polyphenols. The antiproliferative and pro-apoptotic activi-
ties of both compounds were analyzed with various cancer cell
systems, and TMCG, which was easily synthesized in excellent
yield, was more active than TMECG in both melanoma and
non-melanoma cell lines. TMCG was also a better inhibitor of
dihydrofolate reductase and was more efficiently oxidized by
tyrosinase, potentially explaining the difference in activity be-
tween these epimers.
Introduction
We have recently shown that the ester-bonded gallate cate-
chins isolated from green tea, epigallocatechin-3-gallate
(EGCG) and epicatechin-3-gallate (ECG), are potent inhibitors of
dihydrofolate reductase (DHFR) activity in vitro at concentra-
tions found in the serum and tissues of green tea drinkers.[1]
Since this first report describing the antifolate activity of tea
polyphenols, several studies by us and other research groups
have confirmed this activity[2–4] and reported that EGCG inhibits
DHFR from a variety of biological sources.[5–8] Recently, screen-
ing of DHFR binding drugs by MALDI-TOF MS demonstrated
that EGCG is an inhibitor of DHFR with a relative affinity be-
tween that of pyrimethamine and methotrexate (MTX).[8] How-
ever, the excellent anticancer properties of tea catechins are
significantly limited by their poor bioavailability, which is relat-
ed to their low stability in neutral or slightly alkaline solutions
and their inability to easily cross cellular membranes.[9] In an at-
tempt to solve these bioavailability problems, we synthesized
cavity tissues.[13] As part of our ongoing efforts to develop new
tea-derived compounds, we synthesized a trimethoxybenzoyl
analogue of CG (TMCG, 8; Scheme 1). This compound shares
bioavailability advantages with its epimer, TMECG, due to their
similar hydrophobicities, and is simpler and more economical
to synthesize. This enabled us to compare the epimeric differ-
ences between TMECG and TMCG with respect to DHFR inhibi-
tion, tyrosinase activation, and antiproliferative action against
various cancer cell systems.
Results and Discussion
Comparative synthesis of TMECG and TMCG
Synthesis of TMECG, starting from the commercially available
catechin, was previously described by our research group.[10]
The reaction sequence was designed to avoid problems associ-
ated with unspecific blockage of the 3-hydroxy group of epica-
a
3,4,5-trimethoxybenzoyl analogue of ECG (TMECG, 6;
Scheme 1) that exhibits high antiproliferative activity against
malignant melanoma but considerably lower activity against
other epithelial cancer cell lines.[10] This compound acts as a
prodrug against melanoma and is selectively activated by the
specific melanocyte enzyme tyrosinase.[11] Upon activation,
TMECG generates a stable quinone methide that strongly and
irreversibly inhibits DHFR. TMECG treatment induces apoptosis
in melanoma cells and results in the downregulation of anti-
apoptotic Bcl-2, the upregulation of pro-apoptotic Bax, and
the activation of caspase-3.[12]
[a] M. Sꢀez-Ayala, Dr. L. Sꢀnchez-del-Campo, Dr. M. F. Montenegro,
Dr. S. Chazarra, Dr. J. N. Rodrꢁguez-Lꢂpez
Department of Biochemistry and Molecular Biology A
School of Biology, University of Murcia
Avda. Teniente Flomesta, no. 5, 30003 Murcia (Spain)
Fax: (+34)868-884147
E-mail: neptuno@um.es
[b] Prof. A. Tꢀrraga
Department of Organic Chemistry
Faculty of Chemistry, University of Murcia
Avda. Teniente Flomesta, no. 5, 30003 Murcia (Spain)
Because the major polyphenols present in tea have epicate-
chin configurations, many of the studies designed to elucidate
the biological activity of these tea catechins have been per-
formed with epicatechin derivatives. However, (ꢀ)-catechin
gallate (CG), which is a minor polyphenol in green tea, also in-
hibits the proliferation of cancer cells derived from human oral
[c] Dr. J. Cabezas-Herrera
Research Unit of Clinical Analysis Service
University Hospital “Virgen de la Arrixaca”
Ctra. Madrid-Cartagena, 30120 Murcia (Spain)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201000482.
440
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 440 – 449

Synthetic Tea-Catechin-Derived Epimers
Scheme 1. Synthesis of TMECG (6) and TMCG (8): a) BnBr, K2CO3, DMF, ꢀ108C!RT, overnight; b) Dess–Martin periodinane, CH2Cl2, RT, 3 h; c) l-Selectride,
nBu4NCl, THF, ꢀ788C, 3 h; d) 3,4,5-trimethoxybenzoyl chloride, 4, CH2Cl2, DMAP, RT, 18 h; e) H2, 20% Pd/C, THF/MeOH (3:1), RT, 14 h.
techin following the benzylation reaction with benzyl bromide
and potassium carbonate.[14,15] Therefore, all compounds (both
catechin and epicatechin configurations) were synthesized fol-
lowing the multi-step reaction sequence shown in Scheme 1.
For the synthesis of TMCG (8), isomer 1 was esterified with pre-
viously prepared[10] 3,4,5-trimethoxybenzoyl chloride 4 in a di-
chloromethane solution in the presence of DMAP, producing 7
in high yield. In the final step, the benzyl groups were re-
moved by hydrogenolysis to produce final compound 8 in
high yield and purity. TMECG and TMCG have a common first
synthetic step, but yields of the other synthetic steps are sig-
nificantly different. The overall yield of 8 for the alkylation and
deprotection steps was 88%; however, an overall yield of 16%
was obtained for 6 in the epimerization of C3 (oxidation and
reduction), alkylation, and deprotection steps. The difference
between these yields was due to the stereoselective reduction
of 2, which results in only moderate yield and purity and re-
quires further purifications, thereby lowering the yield. The ab-
sence of this limiting step makes the synthesis of 8 both sim-
pler (only three steps) and more economical (only common re-
agents).
Activity in non-melanoma cells
In studying the antiproliferative activity of TMECG, we noted
Figure 1. Antiproliferative effects of natural and synthetic catechins in
cancer cells. A) Half-maximal inhibitory concentration (IC50) of TMCG and
TMECG against several melanoma and non-melanoma cells after six days of
that this compound was more active against melanoma than
against other epithelial cancer cell lines.[10,11] TMECG inhibited
the growth of human breast (MDA-MB-231), lung (N417), and
colon (Caco-2) cancer lines with half-maximal inhibitory con-
centrations (IC50) at six days of 21ꢁ1.5, 18ꢁ2.1, and 33ꢁ
3.1 mm, respectively (Figure 1A). The high concentrations of
TMECG required to inhibit the growth of these cells suggest
treatment. Differences between the effects of TMCG and TMECG were statis-
tically significant in all treated cells (p<0.05). EGCG data are included for
comparison. B) Time-dependent effect of TMCG (10 mm) on the growth of
non-melanoma cancer cells. At each time point, the percentage of cell
growth was calculated with respect to the growth of an untreated control
(100%).
ChemMedChem 2011, 6, 440 – 449
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
441

J. N. Rodrꢁguez-Lꢂpez et al.
MED
that this compound would not be therapeutically useful. How-
ever, TMCG was much more active against these cancer cell
lines and significantly inhibited cell growth with IC50 values at
six days of 5.9ꢁ0.5, 6.6ꢁ0.6, and 6.2ꢁ0.4 mm against MDA-
MB-231, N417, and Caco-2, respectively (Figure 1A). The time-
dependent effect of TMCG on non-melanoma cancer cell
growth can be visualized in Figure 1B. In addition to inhibiting
cell proliferation, chemotherapeutic agents should ideally
induce apoptosis; therefore, TMCG was analyzed to determine
its ability to induce apoptosis. TMCG induced apoptosis in
these epithelial cancer cell lines at a relatively low concentra-
tion, as demonstrated by the significant morphological
changes induced by treatment including cell shrinkage, loss of
cell–cell contact, and fragmentation of plasmatic and nuclear
membranes (Figure 2A). To confirm the apoptotic activity of
TMCG, Annexin-V and propidium iodide were used to examine
early and late stages of apoptosis, respectively. Conjugated An-
nexin-V–fluorescein was used to determine the translocation of
phosphatidylserine from the inner part of the plasma mem-
brane to the outer layer, which is an early feature of apoptosis;
propidium iodide was used to stain the DNA of cells in very
late stages of apoptosis. Figure 2B shows histograms of MDA-
MB-231 cells stained with Annexin-V–fluorescein and propidi-
um iodide as obtained by flow cytometry. We detected a con-
centration-dependent increase in the total number of apoptot-
ic cells reaching ~40% upon treatment with 40 mm TMCG for
96 h. Together, these data indicate that TMCG could be an ef-
fective anticancer agent with growth inhibitory and apoptotic
effects.
frequently been suggested that the pro-oxidant action of poly-
phenols may be important to their anticancer and apoptosis-
inducing properties.[16] TMECG and TMCG possess similar anti-
oxidant and pro-oxidant properties (Table 1), indicating that
Table 1. Mechanistic studies to explain the differential action of TMCG
and TMECG in non-melanoma cells.
Possible mechanism
EGCG[a]
TMECG
TMCG
Antioxidant
TEAC[b] [mm]
Pro-oxidant
4.8ꢁ0.3
149ꢁ8
1.9ꢁ0.2
11.4ꢁ2
2.0ꢁ0.2
12.1ꢁ3
NADPH consumption
[nmminꢀ1
]
Inhibition of proteasome
IC50 [mm]
Inhibition of DHFR
Ki [mm]
0.2ꢁ0.1[c]
>40
>40
1.2ꢁ0.1
33ꢁ3
2.1ꢁ0.2
110ꢁ9
0.9ꢁ0.1
18ꢁ2
Ki* [nm]
[a] EGCG data are included for comparison. [b] TEAC=trolox equivalent
antioxidant capacity. [c] Data taken from reference [20].
these properties are not responsible for the differential biologi-
cal effects of these two compounds in non-melanoma epithe-
lial cancer cells. A number of additional mechanisms, including
the impact of EGCG on a wide range of molecular targets that
influence cell growth and apoptosis, have been proposed as
causes for the anticancer effects of this polyphenolic com-
pound.[1,17–19] The 3-gallyl moiety of catechins is essential to
the modulation of several molecular targets.[1,17–19] Because
EGCG inhibits the chymotrypsin-like activity of the proteasome
(Table 1), inhibition of this multicatalytic protease was pro-
posed as a general mechanism for the biological effects of tea
catechins.[18–20] However, methylation of the 3-gallyl moiety
suppresses the proteasome-inhibitory function of green tea
polyphenols.[20] Correspondingly, TMECG and TMCG do not sig-
nificantly inhibit the chymotrypsin-like activity of purified
rabbit 20S proteasome (Table 1). Although inhibition of the
Mechanistic studies to explain the differential action of
TMCG and TMECG against non-melanoma cells
Although many epidemiological and laboratory studies sup-
port the beneficial health effects of green tea consumption,
the exact mechanism of action of its polyphenolic compounds
is subject to continuous debate. Most plant polyphenols pos-
sess both antioxidant and pro-oxidant properties, and it has
Figure 2. Induction of apoptosis by TMCG in cancer cells. A) Morphological aspect of untreated MDA-MB-231 and Caco-2 cells compared with those subjected
to five days of treatment with 20 mm TMCG. B) Histograms of MDA-MB-231 cells stained with Annexin-V–fluorescein and propidium iodide (PI) with and with-
out TMCG treatment. Dot plots show percentages of early apoptotic cells (Annexin-V+/PIꢀ) and late apoptotic cells (Annexin-V+/PI+). Histograms show gates
indicating percentages of total apoptotic cells.
442
www.chemmedchem.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 440 – 449

Synthetic Tea-Catechin-Derived Epimers
proteasome by EGCG may be biologically significant in several
isolated cell systems, as a general mechanism of action for tea
polyphenols it does not account for much of the epidemiologi-
cal data, as liver methylation of catechins is one of the major
biotransformation reactions under physiological conditions.[21]
Therefore, metabolic methylation of catechin, leading to me-
thylated EGCG, may alter the biological activities of this com-
pound.[22]
In contrast, epimeric differences, which could explain the
biological data, were observed with respect to the inhibition of
DHFR, another potential target of green tea polyphenols.[1,6] Ki-
netic analyses indicated that TMCG was a more efficient inhibi-
tor of this enzyme than TMECG. Analysis of the binding of
these polyphenols to human DHFR using fluorescence quench-
ing indicated that TMCG bound to the enzyme with a lower
dissociation constant (Ki) than TMECG (Table 1). Pre-incubation
of human DHFR with TMECG or TMCG confirmed that both
compounds had the characteristics of slow-binding inhibitors
of human DHFR (Figure 3A),[1,5,6] but pre-incubation of the
enzyme with TMCG had a more profound effect on enzymatic
activity. Complete kinetic analysis, assuming isomerization to a
slowly dissociating inhibition complex (E+I,EI,E*I),[5] dem-
onstrated that the overall inhibition constant (Ki*), which is af-
fected by further EI-complex reactions, decreased dramatically
when TMCG acted as an inhibitor of human DHFR (Table 1). To-
gether, the results indicate that the EI,E*I transition was irre-
versible to a greater extent with TMCG, which possesses a cat-
echin configuration.
Figure 3. Kinetics and molecular modeling of human DHFR inhibition by
TMECG and TMCG. A) Effect of pre-incubation times and catechin concentra-
*
*
tions on the inhibition of DHFR by TMECG ( ) and TMCG ( ). To assess the
effect of pre-incubation time, experiments were performed in the presence
of 40 mm TMECG or TMCG. To assess the effect of catechin concentration,
DHFR was pre-incubated with TMECG or TMCG for 10 min. B) View of
TMECG and TMCG modeled into the folate binding site of human DHFR. The
folate active site of human DHFR is a hydrophobic pocket ~15 ꢅ wide, in
which the only polar side chain is the Glu30 carbonyl group.[39] Other resi-
dues composing this active site are Ile7, Leu22, Gln35, Trp24, Asn64,
Val115, and Thr136. Only residues in the active site or in its proximity that
interact with TMCG and TMECG are highlighted. Phe34, which is located
near TMCG ring D, is not labeled for the purpose of clarity. Specific hydrogen
bonds between TMECG and Glu30 and between TMCG and Glu30 and
Asn64 are indicated (a).
Molecular modeling
In silico molecular modeling experiments performed in our lab-
oratory indicated that TMECG bound to human DHFR in a fash-
ion similar to the binding of EGCG, another natural tea phenol
with an epicatechin configuration.[1,11] The most notable inter-
action between TMECG and DHFR was a hydrogen bond be-
tween the ring A hydroxy group of TMECG and Glu30 in the
enzyme active site (O–O distance 2.94 ꢅ) (Figure 3B). Compari-
son of a range of DHFR structures containing folate or other
inhibitors shows that a majority of the TMECG lies within the
consensual substrate/inhibitor envelope, with the exception of
the non-ester dihydroxybenzoyl moiety (ring B), which is locat-
ed in the proximity of Phe31. To accommodate this ring, the
Leu22 side chain adopts a different orientation than in the
original DHFR structure used in TMECG modeling (PDB ID:
1S3V).[23] The ester-bound gallate moiety (ring D) of TMECG is
accommodated in an amphiphilic region of DHFR involving
residues Gln35, Asn64, and Leu67.
moving away from Phe31, allowing Leu22 to adopt its usual
position within the human DHFR structure. Ring B of TMCG is
also located within hydrogen bonding distance of Gln35 (O–O
distance 2.96 ꢅ) (Figure 3B). This residue can form a hydrogen
bond with the a-carboxylate group of the glutamyl moiety of
MTX in mouse and human DHFRs,[24,25] and its mutation can
yield catalytically active MTX-resistant mutants.[26] Therefore,
the presence of an additional hydrogen bond between Gln35
and the ring B hydroxy group of TMCG could stabilize the
enzyme–inhibitor complex. Finally, the trimethoxylated moiety
of TMCG (ring D) can interact with the protein through Trp57
and Phe34 (Figure 3B). A similar orientation was discovered in
several quinazolinone analogues,[27] where binding to DHFR
was stabilized through hydrogen bonds with Thr56, Trp57,
and Phe58. Although our molecular modeling studies did not
predict additional hydrogen bonds at these positions, the
proximity of the methoxy groups of ring D to Trp57 and
As expected, TMCG adopts a different orientation in the
active site of human DHFR (Figure 3B). Although the ring A
phenolic group maintained a hydrogen bond interaction with
the side chain of Glu30 (O–O distance 2.91 ꢅ), rings B and D
occupied different positions within the enzyme active site. Sev-
eral observations suggest that the position of TMCG in the
human DHFR active site may be more favorable than that of
TMECG, which could explain the higher DHFR inhibitory poten-
cy of TMCG. Ring B adopts a more favorable position by
ChemMedChem 2011, 6, 440 – 449
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
443

J. N. Rodrꢁguez-Lꢂpez et al.
MED
Phe34 indicated that these hydrophobic interactions would
favor binding of TMCG.
Tyrosinase activation of TMECG and TMCG in melanoma
cells
Despite the observed differences between the actions of these
two epimers against non-melanoma cells, both compounds ex-
hibited higher activities against melanoma cells (Figure 1). As
has been described for TMECG,[10–12,28] TMCG downregulated
DHFR in SK-MEL-28 melanoma cells (as demonstrated by
mRNA and protein expression) and modulated Bax/Bcl-2 ex-
pression to a ratio favoring apoptosis (Figure 4). We have re-
cently reported that the elevated activity of TMECG against
melanoma was due to its cellular activation by tyrosinase.[11] Ki-
netic and spectroscopic data indicated that tyrosinase oxidized
TMECG to its corresponding o-quinone, which quickly evolved
through a series of chemical reactions to a quinone methide
(QM) with high stability over a wide pH range.[11] The QM was
found to be a potent irreversible inhibitor of human DHFR,
and this highly stable product may be responsible for the high
activity of TMECG against melanoma cells.[11] This hypothesis
was also confirmed by experiments designed to show the ac-
tivity of natural catechin EGCG in melanoma cells. EGCG was
moderately active toward SK-MEL-28 melanoma cells (Figure 1)
and was unable to induce apoptosis (Figure 4A). Cells treated
with EGCG showed an evident increase in their melanin con-
tent, which may be related to the instability of the quinonic
products derived from oxidation of EGCG by tyrosinase at the
unprotected ring D hydroxy groups. EGCG cannot produce a
stable QM after oxidation by tyrosinase, and it appears that
these oxidation products can incorporate directly into mela-
nins.
Oxidation of TMCG and TMECG by tyrosinase is predicted to
generate the same final product, as proton-catalyzed hydroly-
sis of ring C would generate a freely rotating carbon (C3)
which should prevent epimeric differences in the QM product
(Scheme 2). To confirm that TMECG and TMCG generate the
same quinonic product after tyrosinase oxidation, both sub-
strates were oxidized in vitro using mushroom tyrosinase as a
catalyst. The final products of corresponding oxidations were
analyzed and compared using several spectroscopic tech-
niques. Tyrosinase oxidized TMECG and TMCG to stable final
products, which varied in color from yellow to orange in a pH-
dependent manner. The products had similar spectroscopic
properties, with lmax at 275/412 nm at acidic pH and 275/
470 nm at higher pH values (pKa =6.9) (Figure 5A). UV/Vis
spectroscopy data indicated that, as represented in Scheme 2,
both TMEGC and TMCG generated the same QM product fol-
lowing tyrosinase oxidation. Mass spectroscopy confirmed
these results, and the spectra of both final oxidation products
exhibited the same molecular ion peak. High-performance
liquid chromatography–mass spectrometry (HPLC–MS) re-
vealed that the molecular weights of the compounds were
498.7 and 498.8 for TMECG and TMCG, respectively, corre-
sponding to the calculated mass of the QM product depicted
in Scheme 2 (Figure 5B). Both molecules were analyzed by
Figure 4. Effect of TMCG on SK-MEL-28 melanoma cells. A) Morphological as-
pects of untreated SK-MEL-28 cells (control) relative to those subjected to
five days of treatment with TMCG and EGCG (both at 20 mm). B) Effect of
TMCG on DHFR mRNA and protein expression in SK-MEL-28 cells. Data were
obtained by real-time PCR and western blot analysis of samples of SK-MEL-
28 cells treated with 10 mm TMCG for three days (mRNA) and five days (pro-
tein). C) TMCG treatment resulted in a decrease in Bcl-2 and an increase in
Bax, resulting in a significant increase in the Bax/Bcl-2 ratio that favors apop-
tosis. Data were obtained by western blot analysis from samples of SK-MEL-
28 cells treated with 50 mm TMCG for five days. In all cases, protein and/or
mRNA levels were normalized with respect to b-actin and to their respective
controls (100%). In panels B) and C), *p<0.05 relative to their respective
controls (ꢀTMCG).
MS–MS and produced the same daughter ion peaks at m/z
363 and m/z 287, corresponding to loss of the dihydroxyben-
zoyl or trimethoxybenzoyl moiety, respectively.
TMCG was slightly more active than TMECG at inhibiting cell
growth (IC50 at six days: 1.5ꢁ0.2 and 2.9ꢁ0.3 mm for TMCG
and TMECG, respectively) (Figure 1) and inducing apoptosis
(Figure 6A) in melanoma cells. Because bioavailability is not af-
fected by epimerization,[29] we analyzed the oxidation of these
compounds by tyrosinase to elucidate the cause of their differ-
ent degrees of activity against melanoma. The catalytic effi-
444
www.chemmedchem.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 440 – 449

Synthetic Tea-Catechin-Derived Epimers
of QM inside melanoma cells was analyzed using HPLC–MS–
MS of cell extracts after treatment with TMCG or TMECG. The
concentration of accumulated QM in SK-MEL-28 cells was sig-
nificantly higher (3.6-fold) in cells treated with 10 mm TMCG for
24 h than the concentration of accumulated QM in cells treat-
ed with TMECG under the same conditions (Figure 6B). A con-
trol experiment in which MDA-MD-231 breast cancer cells were
treated with TMCG or TMECG under the same conditions dem-
onstrated that QM was formed only in cells containing the
melanocytic enzyme tyrosinase.
Conclusions
Some natural catechins, such as ECG or EGCG, inhibit cancer
cell proliferation.[30–33] To avoid therapeutic problems associated
with poor stability and low cellular uptake of these com-
pounds, derivative compounds TMECG and TMCG were syn-
thesized. TMCG may be more appropriate and effective than
TMECG for the treatment of non-melanoma epithelial cancer
cells. The synthesis of TMCG was simpler and more economical
than the synthesis of TMECG, and its effectiveness in inhibiting
cell growth and inducing apoptosis in human breast, colon,
and lung cell lines was significantly higher than that observed
for TMECG. To understand the differences in the actions of
these epimeric compounds, we tested various possibilities and
observed large differences in their abilities to inhibit human
DHFR. Although both compounds exhibited characteristics of
slow-binding inhibitors, TMCG was sixfold more potent than
TMECG, as deduced from their overall inhibition constants
(Table 1). A crucial factor in the inhibition process is the disso-
Scheme 2. Reaction sequences indicating the oxidation of TMECG and
TMCG by tyrosinase (TYR) and the formation of quinone methide (QM) spe-
cies.
ciency of tyrosinase toward TMCG (6.9 minꢀ1 mmꢀ1) was four-
fold higher than for TMECG (1.7 minꢀ1 mmꢀ1), indicating that
tyrosinase activation of TMCG may be favored over TMECG in
melanoma cells. To confirm these observations, the formation
Figure 5. Analysis and comparison of the final products generated from the oxidation of TMECG and TMCG by mushroom tyrosinase (TMECG-QM and TMCG-
QM, respectively). A) UV/Vis absorption spectra of TMECG-QM and TMCG-QM at various pH values. B) Mass spectra of TMCG-QM and MS–MS daughter ion
mass spectra of the molecular ion peak at m/z 499. TMECG-QM generated the same MS and MS–MS spectroscopic results.
ChemMedChem 2011, 6, 440 – 449
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
445

J. N. Rodrꢁguez-Lꢂpez et al.
MED
cells. Antifolate prodrugs de-
signed to be specifically activat-
ed in tumor cells represent an
attractive alternative that could
prevent these undesirable side
effects.[34] From this point of
view, TMECG would be consid-
ered a better prodrug against
melanoma. The decreased anti-
folate character of the TMECG
prodrug relative to TMCG
(Table 1) would favor specific ac-
tivity against melanoma cells
and prevent unspecific side ef-
fects in rapidly dividing healthy
cells.
Experimental Section
Materials
Highly purified tea EGCG (>95%)
was purchased from Sigma Chemi-
cal Co. (Madrid, Spain). Human
DHFR was expressed in Bombyx
mori chrysalides and purified by
MTX-affinity
chromatography.[35]
Enzyme concentration was deter-
mined by MTX titration of enzyme
fluorescence.[36] NADPH and dihy-
drofolic acid (DHF) were obtained
from Sigma. Mushroom tyrosinase,
Bcl-2, b-actin, and DHFR antibodies
were purchased from Sigma, and
the antibody against Bax was from
Santa Cruz Biotechnology (Santa
Cruz, CA, USA).
Figure 6. Differences between the actions of TMECG and TMCG in melanoma cells. A) Induction of apoptosis by
TMECG and TMCG in SK-MEL-28 melanoma cells after six days of treatment. The differences in apoptosis induced
by the two compounds were statistically significant for treatments with 5 and 10 mm drug (*p<0.05); ns: not sig-
nificant (left panel). Time-dependent apoptosis induction of TMCG (10 mm) on SK-MEL-28 (middle panel). Effect of
~
*
10 mm ( ) and 20 mm ( ) TMCG on SK-MEL-28 growth determined by the MTT assay and compared with an un-
&
treated control ( ) (right panel). B) Accumulation of QM species in SK-MEL-28 and MDA-MB-231 cells after 24 h
treatments with TMCG and TMECG (both at 10 mm). The left panel represents the HPLC–MS chromatograms.
Synthesis
ciation constant of the E*I complex, which was found to be ir-
All reactions were carried out using solvents that were dried by
routine procedures. 1H and 13C NMR spectra were recorded on
Bruker Avance 300 MHz and Bruker Avance 400 MHz instruments.
The following abbreviations are used to represent the multiplicity
of the signals: s (singlet), d (doublet), dd (double doublet), m (mul-
tiplet), and q (quaternary carbon atom). Chemical shifts are given
reversible to a greater extent when using TMCG. A possible ex-
planation for the irreversibility of this slowly dissociating
enzyme complex is provided by molecular modeling.
The differential action of TMCG and TMECG against melano-
ma cells cannot be explained by the differences in their abili-
ties to inhibit DHFR. Both compounds are prodrugs that are se-
lectively activated in melanoma by the melanogenic enzyme
tyrosinase, which transforms TMCG and TMECG to the same
QM product. Therefore, the slight but statistically significant
differences in the action of these drugs against melanoma are
due to the different specific activities of tyrosinase toward
TMCG and TMECG. In terms of activity, TMCG was a more effec-
tive drug for the treatment of melanoma; however, TMECG
would be a more appropriate prodrug in regard to tumor se-
lectivity. Antifolate compounds are designed to inhibit DHFR
and act specifically during DNA and RNA synthesis, making
them more toxic to rapidly dividing cells. This characteristic
also results in antifolates that are unspecific for tumor cells
and produces adverse side effects in rapidly dividing healthy
1
with reference to the signals of (CH3)4Si in H and 13C NMR spectra.
Electrospray (ES) mass spectra were recorded on Agilent 6220 TOF
and Agilent VL spectrometers. Elemental analyses were performed
on a Carlo Erba EA-1108 elemental analyzer. Melting points were
determined on a Kofler hot-plate melting point apparatus and are
uncorrected. Compounds 1, 2, and 3 were obtained using experi-
mental procedures described elsewhere,[14] and their spectral data
correlate with previously reported data.[14,18] Compounds used in
biological tests possess purity higher than 98% as determined by
elemental analysis.
5,7,3’,4’-Tetra-O-benzyl-3-(3’’,4’’,5’’-trimethoxybenzoyl)-(ꢀ)-cate-
chin (7): A solution of 4 (1.41 g, 6.14 mmol) in dry CH2Cl2 (5 mL)
was added dropwise under nitrogen atmosphere to a solution con-
taining 1 (2 g, 3.07 mmol) and 4-dimethylaminopyridine (DMAP;
0.94 g, 7.68 mmol) in the same solvent (30 mL). The reaction mix-
446
www.chemmedchem.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 440 – 449

Synthetic Tea-Catechin-Derived Epimers
ture was stirred at room temperature for 18 h. A solution of satu-
rated sodium bicarbonate (40 mL) was added, and the mixture was
extracted twice with EtOAc (2ꢆ30 mL). The organic layers were
then extracted twice with H2O (2ꢆ30 mL) and dried with anhy-
drous magnesium sulfate, and the solvent was removed under
vacuum. The resulting yellow oil was subjected to silica gel column
chromatography using n-hexane/EtOAc/CH2Cl2 (6:6:2) as a solvent.
The solvent was removed under reduced pressure, and the solid
was recrystallized from Et2O/n-hexane to obtain a white solid
(yield=98%). Rf =0.76 (n-hex/EtOAc/CH2Cl2 =6:6:2); mp: 123–
(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT;
Sigma) cell proliferation assay.[37] For this assay, cells were plated in
a 96-well plate at a density of 1000–2000 cells per well. Com-
pounds were added once at the beginning of each experiments.
Apoptosis assays: Apoptosis induction was assessed by analyzing
cytoplasmic histone-associated DNA fragmentation with a kit from
Roche Diagnostics (Barcelona, Spain). Apoptosis was determined as
the specific enrichment of mono- and oligonucleosomes released
into the cytoplasm and was calculated by dividing the absorbance
of treated samples by the absorbance of untreated controls. The
Annexin-V–FLUOS Staining Kit from Roche Diagnostics was used to
detect cell apoptosis. Annexin-V staining was performed according
to the manufacturer’s protocols. After washing with PBS, cells were
resuspended in 100 mL of Annexin-V–FLUOS labeling solution (con-
taining PI and Annexin-V–fluorescein) and incubated for 15 min at
room temperature in the dark. Cells were analyzed by flow cytom-
etry in a Beckman Coulter Epics XL flow cytometer.
1
1248C; H NMR (CDCl3, 300 MHz): d=7.31–7.12 (m, 20H, Ph), 7.02
(s, 2H, H2’’ and H6’’), 6.97 (d, 1H, 4J=1.8 Hz, H2’), 6.88 (dd, 1H,
3J=8.2 Hz, 4J=1.8 Hz, H6’), 6.80 (d, 1H, 3J=8.2 Hz, H5’), 6.21 (d,
4
4
1H, J=2.4 Hz, H6), 6.19 (d, 1H, J=2.4 Hz, H8), 5.38 (m, 1H, H3),
5.04 (s, 2H, CH2O), 5.00 (m, 1H, H2), 4.98 (s, 2H, CH2O), 4.93 (s, 4H,
2ꢆCH2O), 3.79 (s, 3H, OCH3), 3.72 (s, 6H, OCH3), 3.05 (m, 1H,
Hgem, H4), 2.76 ppm (m, 1H, Hgem, H4); 13C NMR (CDCl3, 75 MHz):
d=165.1 (q, -COO), 158.8 (q, Ar-O), 157.6 (q, Ar-O), 154.9 (q, Ar-O),
152.7 (2ꢆq, Ar-O), 149.0 (q, Ar-O), 148.9 (q, Ar-O), 142.2 (q, Ar-O),
136.9 (q, PhCH2), 136.8 (q, PhCH2), 136.7 (2ꢆq, PhCH2), 130.9 (q,
C1’), 128.5 (CH, PhCH2), 128.4 (CH, PhCH2), 128.3 (2ꢆCH, PhCH2),
127.9 (CH, PhCH2), 127.8 (CH, PhCH2), 127.7 (2ꢆCH, PhCH2), 127.4
(CH, PhCH2), 127.3 (CH, PhCH2), 127.1 (2ꢆCH, PhCH2), 124.8 (q,
C1’’), 120.0 (CH, C6’), 114.7 (CH, C5’), 113.4 (CH, C2’), 106.7 (CH, C2’’
and C6’’), 101.4 (q, C4a), 94.3 (CH, C6), 93.7 (CH, C8), 78.4 (CH, C2),
71.3 (CH2, CH2Ph), 71.1 (CH2, CH2Ph), 70.1 (CH, C3), 70.0 (CH2,
CH2Ph), 69.8 (CH2, CH2Ph), 60.8 (CH3, OCH3), 56.1 (CH3, OCH3),
24.6 ppm (CH2, C4); ESMS m/z (%) 845.3 ([M++1], 100); Anal. calcd
for C53H48O10 (844.3): C 75.34, H 5.73, found: C 75.21, H 5.84.
Antioxidant activity: The Trolox equivalent antioxidant capacity
(TEAC) for each catechin was determined as described else-
where.[10]
NADPH oxidation by catechins: The pro-oxidant activity of catechins
was determined by their NADPH oxidation capacity.[10] The rate of
NADPH (0.1 mm) oxidation at 378C was calculated in the presence
of 50 mm catechins in sodium-phosphate buffer (pH 7.4) by follow-
ing the decrease in absorbance of NADPH at 340 nm in a Perki-
nElmer Lambda-35 spectrophotometer.
Inhibition of purified 20S proteasome activity: The chymotrypsin-like
activity of the 20S proteasome was measured by incubating 30 ng
of purified rabbit 20S proteasome (Sigma) with 40 mm of the fluo-
rescent peptide substrate Suc-Leu-Leu-Val-Tyr-AMC with and with-
out synthetic catechins.
3-O-(3,4,5-Trimethoxybenzoyl)-(ꢀ)-catechin (8): Under normal
pressure, a solution of 7 (1.5 g, 1.77 mmol) and 10% Pd/C (0.05 g
Pd, 0.47 mmol) in THF/MeOH (3:1) (40 mL) was treated with H2.
The solution was stirred for 14 h at room temperature and then fil-
tered through a Celite pad, which was washed afterward with
CH2Cl2/MeOH (9:1) (200 mL). The solvent was removed under
vacuum, and the resulting solid was recrystallized from Et2O (90%
yield). Rf =0.18 (n-hexane/EtOAc/CH2Cl2 =6:6:2); mp: 109–1108C;
1H NMR ([D6]acetone, 400 MHz): d=8.42 (bs, 1H, OH), 8.19 (bs, 1H,
OH), 7.99 (bs, 1H, OH), 7.98 (bs, 1H, OH), 7.13 (s, 2H, H2’’ and H6’’),
7.00 (d, 1H, 4J=1.8 Hz, H2’), 6.87 (dd, 1H, 3J=8.1 Hz, 4J=1.8 Hz,
DHFR activity assays: The activity of DHFR in the absence or pres-
ence of catechins was determined at 258C by following the de-
crease in the absorbance of NADPH and DHF at 340 nm as de-
scribed elsewhere.[5] Experiments to determine the recovery of
enzyme activity after inhibition by pre-incubation with catechins
were performed as follows. DHFR (165 nm) was pre-incubated for
10 min at 258C in the buffer mixture containing catechins at vari-
ous concentrations. Aliquots (20 mL) of the incubation mixture
were then diluted 50-fold into a reaction mixture containing the
buffer mixture, NADPH (100 mm), and DHF (20 mm) to give a final
enzyme concentration of 3.3 nm. Recovery of enzyme activity was
followed by continuous monitoring at 340 nm.
3
4
H6’), 6.81 (d, 1H, J=8.1 Hz, H5’), 6.08 (d, 1H, J=2.4 Hz, H6), 5.97
4
(d, 1H, J=2.4 Hz, H8), 5.28 (m, 1H, H3), 5.05 (m, 1H, H2), 3.81 (s,
6H, OCH3), 3.75 (s, 3H, OCH3), 3.14 (m, 1H, Hgem, H4), 2.75 ppm
(m, 1H, Hgem, H4); 13C NMR ([D6]acetone, 100 MHz): d=165.6 (q,
-COO), 158.0 (q, Ar-O), 157.2 (q, Ar-O), 156.5 (q, Ar-O), 153.9 (q, Ar-
O), 145.8 (q, Ar-O), 145.7 (q, Ar-O), 143.2 (q, Ar-O), 131.0 (q, C1’),
126.0 (q, C1’’), 119.5 (CH, C6’), 115.8 (CH, C5’), 114.7 (CH, C2’), 107.6
(CH, C2’’ and C6’’), 99.3 (q, C4a), 96.4 (CH, C6), 95.4 (CH, C8), 79.3
(CH, C2), 71.7 (CH, C3), 60.5 (CH3, CH3O), 56.4 (CH3, CH3O),
25.8 ppm (CH2, C4); ESMS m/z (%) 483.6 ([M+ꢀ1], 100); Anal. calcd
for C25H24O10 (484.1): C 61.98, H 4.99, found: C 61.96, H 5.11.
Fluorescence studies: Dissociation constants for the binding of
TMECG and TMCG to free human DHFR were determined by fluo-
rescence titration in an automatic scanning FluoroMax-3 spectro-
fluorimeter (Jobin Ybon, Horiba, Edison, NJ, USA) with 1.0 cm light
path cells and a 150 W Mercury–Xenon light source. Formation of
a binary complex between the enzyme and the ligand was fol-
lowed by measuring the quenching of tryptophan fluorescence of
the enzyme upon addition of microliter volumes of a concentrated
stock solution of ligand. Fluorescence emission spectra were re-
corded when human DHFR fluorescence was excited at 290 nm,
and titrations were performed as described elsewhere.[5]
1H and 13C NMR, HMQC, and mass spectra for compounds 7 and 8
are provided in the Supporting Information.
Biology
Tyrosinase assays: Catechin oxidation, catalyzed by mushroom tyro-
sinase, was followed at 440 nm (isosbestic point) using a Perkin-
Elmer Lambda-35 spectrophotometer (Waltham, MA, USA). Experi-
ments were performed in acetate buffer (pH 5.0).
Cell cultures: Human cancer cell lines (SK-MEL-28, MDA-MB-231,
N417, and Caco-2) were obtained from the American Type Culture
Collection (ATCC) and were maintained in appropriate culture
media supplemented with 10% fetal calf serum and antibiotics
under standard tissue culture conditions. Cell viability was evaluat-
ed by a colorimetric assay for mitochondrial function using the 3-
ChemMedChem 2011, 6, 440 – 449
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
447

J. N. Rodrꢁguez-Lꢂpez et al.
MED
Real-time PCR: Real-time PCR analysis was carried out as described
elsewhere,[10] using the following primers for the amplification of
human genes: dhfr (forward: 5’-ATG CCT TAA AAC TTA CTG AAC
AAC CA-3’; reverse: 5’-TGG GTG ATT CAT GGC TTC CT-3’); b-actin
(forward: 5’-AGA AAA TCT GGC ACC ACA CC-3’; reverse: 5’-GGG
GTG TTG AAG GTC TCA AA-3’).
Statistical analysis
In all experiments, the mean ꢁ standard deviation (SD) values for
three to five determinations in triplicate were calculated. Statistical-
ly significant differences were evaluated using the Student’s t-test.
Differences were considered statistically significant at p<0.05.
Western blotting: Cells were lysed by sonication in PBS pH 7.4 con-
taining 1% NP-40, 1% Triton X-100, 0.5% sodium deoxycholate,
0.1% SDS, and protease inhibitor cocktails. After centrifugation
(15000 g, 20 min), soluble proteins were separated by SDS-PAGE,
transferred to nitrocellulose membranes, and analyzed by immuno-
blotting (ECL Plus, GE Healthcare).
Acknowledgements
This work was supported in part by grants from the Ministerio de
Ciencia e Innovaciꢂn (MICINN), Spain (project SAF2009-12043-
C02-01) and the Fundaciꢂn Sꢃneca, Regiꢂn de Murcia (FS-RM),
Spain (project 08595/PI/08) to J.N.R.-L. and J.C.-H., and from the
MICINN (project CTQ2008-01402) and the FS-RM (project 04509/
GERM/06) to A.T. J.C.-H. is contracted by the Translational Cancer
Research Group (Fundaciꢂn para la Formaciꢂn e Investigaciꢂn
Sanitarias (FFIS); Murcia, Spain). M.S.-A. has a fellowship from
the MICINN. L.S.-d.-C. has a fellowship from the FS-RM. S.C. is
contracted by the program “Torres Quevedo” from the MICINN,
and M.F.M. is contracted by an agreement with the Fundaciꢂn de
la Asociaciꢂn EspaÇola contra el Cꢀncer (FAECC).
In silico molecular modeling
Molecular modeling was carried out using the CAChe software
package v. 7.5 (Fujitsu, Krakow, Poland). In searching for available
ligand-bound human DHFR structures in the PDB,[38] we identified
a 1.8 ꢅ structure (PDB ID: 1S3V)[27] that was the best available
structural match for TMECG and TMCG. Hydrogen atoms were
added to the DHFR molecules prior to docking procedures. TMECG
and TMCG were built and energy minimized on CAChe. The molec-
ular geometries of both compounds were optimized using MM3
molecular mechanics methods until the root mean square (RMS)
gradient value was smaller than 0.1 kcalmolꢀ1. The fastest and easi-
est method for docking a ligand into an active site is to superim-
pose the ligand on to a bound ligand already in the active site and
then delete the bound ligand. Then, using the position of (R)-6-
{[methyl-(3,4,5-trimethoxyphenyl)amino]methyl}-5,6,7,8-tetrahydro-
quinazoline-2,4-diamine as a guide, compounds were docked into
this protein structure, and the energy of the inhibitor–protein com-
posite was then minimized using MM3.
Keywords: antifolate activity
prodrugs · tea catechins
·
epimers
·
melanoma
·
[1] E. Navarro-Perꢁn, J. Cabezas-Herrera, F. Garcꢀa-Cꢁnovas, M. C. Durrant,
R. N. F. Thorneley, J. N. Rodrꢀguez-Lꢃpez, Cancer Res. 2005, 65, 2059–
2064.
[2] N. C. Alemdaroglu, S. Wolffram, J. P. Boissel, E. Closs, H. Spahn-Lang-
guth, P. Langguth, Planta Med. 2007, 73, 27–32.
[3] E. Navarro-Perꢁn, J. Cabezas-Herrera, L. Sꢁnchez del Campo, J. N. Rodrꢀ-
guez-Lꢃpez, Int. J. Biochem. Cell Biol. 2007, 39, 2215–2225.
[4] M. D. Navarro-Martꢀnez, F. Garcꢀa-Cꢁnovas, J. N. Rodrꢀguez-Lꢃpez, J. Anti-
microb. Chemother. 2006, 57, 1083–1092.
UV/Vis spectroscopy
[5] E. Navarro-Perꢁn, J. Cabezas-Herrera, A. N. P. Hiner, T. Sadunishvili, F.
Garcꢀa- Cꢁnovas, F. J. N. Rodrꢀguez-Lꢃpez, Biochemistry 2005, 44, 7512–
7525.
[6] M. Spina, M. Cuccioloni, M. Mozzicafreddo, F. Montecchia, S. Pucciarelli,
A. M. Eleuteri, E. Fioretti, M. Angeletti, Proteins 2008, 72, 240–251.
[7] T. T. Kao, K. C. Wang, W. N. Chang, C. Y. Lin, B. H. Chen, H. L. Wu, G. Y.
Shi, J. N. Tsai, T. F. Fu, Drug Metab. Dispos. 2008, 36, 508–516.
[8] P. Hannewald, B. Maunit, J. F. Muller, Anal. Bioanal. Chem. 2008, 392,
1335–1344.
UV/Vis absorption spectra of TMECG-QM and TMCG-QM at differ-
ent pHs were recorded on a UV/Vis PerkinElmer Lambda-35 spec-
trophotometer with a spectral bandwidth of 1 nm at a scan speed
of 960 nmminꢀ1. Experiments were performed in various buffers
over pH range 5.0–9.0, with the pH of the reaction measured
before and after the experiment.
[9] J. Hong, H. Lu, X. Meng, J. H. Ryu, Y. Hara, C. S. Yang, Cancer Res. 2002,
62, 7241–7246.
HPLC–MS
[10] L. Sꢁnchez-del-Campo, F. Otꢃn, A. Tꢁrraga, J. Cabezas-Herrera, S. Chazar-
ra, J. N. Rodrꢀguez-Lꢃpez, J. Med. Chem. 2008, 51, 2018–2026.
[11] L. Sꢁnchez-del-Campo, A. Tꢁrraga, M. F. Montenegro, J. Cabezas-Herrera,
J. N. Rodrꢀguez-Lꢃpez, Mol. Pharm. 2009, 6, 883–894.
[12] L. Sꢁnchez-del-Campo, J. N. Rodrꢀguez-Lꢃpez, Int. J. Cancer 2008, 123,
2446–2455.
[13] M. Babich, H. L. Zuckerbraun, S. M. Weinerman, Toxicol. Lett. 2007, 171,
171–180.
[14] W. Tꢇckmantel, A. P. Kozikowski, J. J. Romanczyk, Jr., J. Am. Chem. Soc.
1999, 121, 12073–12081.
[15] K. D. Park, S. G. Lee, S. U. Kim, S. H. Kim, W. S. Sun, S. J. Cho, D. H. Jeong,
Bioorg. Med. Chem. Lett. 2004, 14, 5189–5192.
[16] S. Azam, N. Hadi, N. U. Khan, S. M. Hadi, Toxicol. In Vitro 2004, 18, 555–
561.
[17] S. Nam, D. M. Smith, Q. P. Dou, J. Biol. Chem. 2001, 276, 13322–13330.
[18] S. B. Wan, D. Chen, Q. P. Dou, T. H. Chan, Bioorg. Med. Chem. 2004, 12,
3521–3527.
[19] K. Osanai, K. R. Landis-Piwowar, Q. P. Dou, T. H. Chan, Bioorg. Med. Chem.
2007, 15, 5076–5082.
QM was analyzed on a HPLC–MS system consisting of an Agilent
1100 Series HPLC (Agilent Technologies) connected to an Agilent
Ion Trap XCT Plus mass spectrometer (Agilent Technologies) using
an electrospray (ESI) interface. To detect QM in cell extracts, cells
were collected at the end of each incubation period, washed three
times with PBS, and lysed by addition of a buffer containing 2 mm
EDTA, 2 mm EGTA, 20 mm imidazole-HCl, and 50 mm ascorbic acid
(pH 7). Ascorbic acid was included to avoid oxidation of catechins
during the extraction process. After one hour of incubation at 48C,
the lysates were sonicated and centrifuged. The supernatants were
deproteinized by the addition of acetonitrile, and the solution was
centrifuged to recover the supernatants, which were filtered on Mi-
crocon centrifugal filter devices with a mass cutoff of 10000 units
(Millipore). Filtrates were lyophilized and resuspended in 50 mL ace-
tonitrile. The resulting suspensions were centrifuged, and the su-
pernatants were analyzed by HPLC–MS–MS. Analysis was carried
out using the same HPLC–MS system.
448
www.chemmedchem.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 440 – 449

Synthetic Tea-Catechin-Derived Epimers
[20] K. R. Landis-Piwowar, S. B. Wan, R. A. Wiegand, D. J. Kuhn, T. H. Chan,
Q. P. Dou, J. Cell. Physiol. 2007, 213, 252–260.
[32] G. Y. Yang, J. Liao, K. Kim, E. J. Yurkow, C. S. Yang, Carcinogenesis 1998,
19, 611–616.
[21] J. D. Lambert, C. S. Yang, J. Nutr. 2003, 133, 3262S–3267S.
[22] C. Huo, S. B. Wan, W. H. Lam, L. Li, Z. Wang, K. R. Landis-Piwowar, D.
Chen, Q. P. Dou, T. H. Chan, Inflammopharmacology 2008, 16, 248–252.
[23] V. Cody, J. R. Luft, W. Pangborn, A. Gangjee, S. F. Queener, Acta Crystal-
logr. Sect. D 2004, 60, 646–655.
[24] D. K. Stammers, J. N. Champness, C. R. Beddell, J. G. Dann, E. Eliopoulos,
A. J. Geddes, D. Ogg, A. C. North, FEBS Lett. 1987, 218, 178–184.
[25] V. Cody, J. R. Luft, W. Pangborn, Acta Crystallogr. Sect. D 2005, 61, 147–
155.
[33] M. Nihal, N. Ahmad, H. Mukhtar, G. S. Wood, Int. J. Cancer 2005, 114,
513–521.
[34] M. Rooseboom, J. N. M. Commandeur, N. P. E. Vermeulen, Pharmacol.
Rev. 2004, 56, 53–102.
[35] S. Chazarra, S. Aznar-Cervantes, L. Sꢁnchez-del-Campo, J. Cabezas-Her-
rera, W. Xiaofeng, J. L. Cenis, J. N. Rodrꢀguez-Lꢃpez, Appl. Biochem. Bio-
technol. 2010, 162, 1834–1846.
[36] S. L. Smith, P. Patrick, D. Stone, A. W. Phillips, J. J. Burchall, J. Biol. Chem.
1979, 254, 11475–11484.
[26] J. P. Volpato, E. Fossati, J. N. Pelletier, J. Mol. Biol. 2007, 373, 599–611.
[27] S. T. Al-Rashood, I. A. Aboldahab, M. N. Nagi, L. A. Abouzeid, A. A. M.
Andel-Aziz, S. G. Andel-Hamide, K. M. Youssef, A. M. Al-Obaid, H. I. El-
Subbagh, Bioorg. Med. Chem. 2006, 14, 8608–8621.
[37] T. Mosmann, J. Immunol. Methods 1983, 65, 55–63.
[38] H. M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H. Weissig,
I. N. Shindyalov, P. E. Bourne, Nucleic Acids Res. 2000, 28, 235–242.
[39] S. J. Benkovic, C. A. Fierke, A. M. Naylor, Science 1988, 239, 1105–1110.
[28] L. Sꢁnchez-del-Campo, S. Chazarra, M. F. Montenegro, J. Cabezas-Her-
rera, J. N. Rodrꢀguez-Lꢃpez, J. Cell. Biochem. 2010, 110, 1399–1409.
[29] J. Z. Xu, S. Y. V. Yeung, Q. Chang, Y. Huang, Z.-Y. Chen, Br. J. Nutr. 2004,
91, 873–881.
[30] Y. D. Jung, L. M. Ellis, Int. J. Exp. Pathol. 2001, 82, 309–316.
[31] Y. D. Hsuuw, W. H. Chan, Ann. N. Y. Acad. Sci. 2007, 1095, 428–440.
Received: November 10, 2010
Revised: December 23, 2010
Published online on February 7, 2011
ChemMedChem 2011, 6, 440 – 449
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
449