Synthesis and structure-activity relationship of 3-O-acylated (-)-epigallocatechins as 5α-reductase inhibitors

Article abstract of DOI:10.1016/j.ejmech.2010.10.011

A series of 3-O-acylated (-)-epigallocatechins were synthesized and their inhibition of steroid 5α-reductase was studied. They were prepared from the reaction of EGCG with tert-butyldimethylsilyl chloride followed by reductive cleavage of the ester bond. The resultant (-)-epigallocatechins penta-O-tert-butyldimethylsilyl ether was esterified with different fatty acids then desilylated to provide the corresponding products. The activity of 3-O-acylated (-)-epigallocatechins increased with the increasing carbon numbers of the fatty acid moiety, reaching maximum for 16 carbon atoms (compound 4h) with an IC₅₀ of 0.53 μM, which was ～12-fold more potent than EGCG (IC₅₀ = 6.29 μM). Introduction of monounsaturated fatty acid provided the most potent compound 6 (IC₅₀ = 0.48 μM), which showed moderate anti-tumor activity in vivo.

Full text of DOI:10.1016/j.ejmech.2010.10.011

European Journal of Medicinal Chemistry 45 (2010) 6068e6076
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis and structureeactivity relationship of 3-O-acylated
(e)-epigallocatechins as 5 -reductase inhibitors
a
Shu Fu Lin a, Yu-Hsiang Lin a, Mengju Lin a, Yi-Feng Kao a, Ru-Wen Wang a, Li-Wei Teng a,
Shih-Hsien Chuang a, Jia-Ming Chang a, Ta-Tung Yuan a, Kuo Chu Fu a, Kuan Pin Huang a, Ying-Shuen Lee a,
Chao-Cheng Chiang a, Sheng-Chuan Yang a, Chun-Liang Lai a, Chu-Bin Liao a, Paonien Chen a,
Young-Sun Lin a, Kuei-Tai Lai a, Hung-Jyun Huang a, Ju-Ying Yang a, Chia-Wei Liu a, Win-Yin Wei a,
,
Chi-Kuan Chen a, Richard A. Hiipakka b, Shutsung Liao b, Jiann-Jyh Huang a
*
a Development Center for Biotechnology, No. 101, Lane 169, Kangning Street, Xizhi City, Taipei County, Taiwan 221, ROC
b The Ben May Department for Cancer Research, The University of Chicago, 929 East 57th Street, CIS W325F, Chicago, IL 60637, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 3-O-acylated (e)-epigallocatechins were synthesized and their inhibition of steroid 5a-reduc-
Received 26 May 2010
Received in revised form
11 October 2010
Accepted 11 October 2010
Available online 15 October 2010
tase was studied. They were prepared from the reaction of EGCG with tert-butyldimethylsilyl chloride
followed by reductive cleavage of the ester bond. The resultant (e)-epigallocatechins penta-O-tert-butyl-
dimethylsilyl ether was esterified with different fatty acids then desilylated to provide the corresponding
products. The activity of 3-O-acylated (e)-epigallocatechins increased with the increasing carbon numbers
of the fatty acid moiety, reaching maximum for 16 carbon atoms (compound 4h) with an IC50 of 0.53
M). Introduction of monounsaturated fatty
M), which showed moderate anti-tumor activity
mM,
which was w12-fold more potent than EGCG (IC50 ¼ 6.29
m
Keywords:
EGCG
Catechin
acid provided the most potent compound 6 (IC50 ¼ 0.48
in vivo.
m
Ó 2010 Elsevier Masson SAS. All rights reserved.
5a-Reductase inhibitor
Prostate cancer
(e)-Epigallocatechins
1. Introduction
role in prostate cancer [16], and its inhibition is thus helpful to the
chemoprevention of prostate cancer [17]. EGCG also possesses 5
a-
(e)-Epigallocatechin-3-gallate (EGCG, Fig. 1) is the most abundant
catechin in green tea and has been intensively investigated for its
biological activities [1]. In vitro and in vivo experiments have
demonstrated the anti-oxidation, anti-inflammation, anti-bacterial,
anti-viral, and anti-cancer properties of EGCG, which contribute to the
major health benefits of tea [2]. Recent studies focused on its che-
moprevention and anti-tumor properties. In animal models, EGCG is
found to inhibit all stages of carcinogenesis in breast, lung, skin,
prostate, and colon cancers [3,4]. Aclinical trialalsoshows apromising
result for the chemoprevention of prostate cancer by tea catechins in
volunteers with high-grade prostate intraepithelial neoplasia [5].
reductase inhibitory activity with an IC50 value in the micromolar
range [18] in addition to several cellular mechanisms with regard to
chemoprevention [19]. Furthermore, intraperitoneal injection of
EGCG, is found to inhibit the growth or induces the regression of
human prostate cancer xenografts in nude mice [20].
For the discovery of more potent EGCG-based derivatives for
the inhibition of
5a-reductase, several 3-O-acylated (e)-epi-
gallocatechins are synthesized and reported in our previous study
[21]. Some of them are more active than EGCG. However, the cases
are limited and their structureeactivity relationship is not exten-
sively explored. In this paper, we performed a more detail study for
this class of inhibitors (see the general structure in Fig. 1). A new
efficient synthetic method was developed for their preparation and
was suitable to prepare sufficient compounds for in vivo study.
Among the derivatives, (e)-epigallocatechin-3-O-palmitoleate (6)
Steroid 5
lating testosterone to 5
natural androgen [6,7] which is associated with the normal growth,
development, and function of the prostate [8e10]. 5 -Reductase is
a-reductase catalyzes the intracrine conversion of circu-
a-dihydrotestosterone (DHT), the most potent
a
involved in various human disorders [11e15] and plays an important
was most potent for the inhibition 5
a-reductase with an IC50 value
of 0.48 M. Using a PC-3 human prostate cancer xenograft model in
m
SCID mice, compound 6 exhibited moderate anti-cancer activity
and was more potent than EGCG.
* Corresponding author. Tel.: þ886 2 2695 6933x2501; fax: þ886 2 2695 7474.
E-mail address: lukehuang0226@gmail.com (J.-J. Huang).
0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2010.10.011

S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
6069
Table 1
Inhibition of 5
chain saturated fatty acids at C3-O position.
OH
OH
a
-reductase by 3-O-acylated (e)-epigallocatechins with straight-
OH
OH
OH
OH
2
HO
O
4
OH
2
HO
R =
O
4
3
OH
OH
3
O
O
HO
O
OH
OH
OH
OH
O
O
R
O
or olefin (n = 0-20)
OH
( )n
( )n
OH
O
4
(-)-epigallocatechin-3-gallate
(EGCG)
3-O-acylated (-)-epigallocatechin
Fig. 1. Structure of EGCG and 3-O-Acylated (e)-Epigallocatechins
Compound n ¼ Name of CH3(CH2)nCOOH IC50
(
mM) or %inhibition at 5.0 mM
4a
4b
4c
4d
4e
4f
4g
4h
4i
0
2
4
6
8
Acetic acid
Butyric acid
Caproic acid
Caprylic acid
Capric acid
(3.0%)
(35%)
(46%)
(49.3%)
2.94
0.77
0.60
0.53
0.73
2. Chemistry
The preparation of 3-O-acylated (e)-epigallocatechins 4e16
started from the reaction of EGCG with TBDMSCl in DMF in the
presence of imidazole (Scheme 1). The reaction provided EGCG
octa-O-TBDMS (1) in 91% yield [22]. Reductive cleavage of the C3-O
ester bond by use of LiAlH4 in THF afforded EGC penta-O-TBDMS (2)
in 80% yield. The free hydroxyl group in compound 2 was esterified
with different C2eC22 saturated fatty acids using DCC and DMAP to
give the silylated 3-O-acylated (e)-epigallocatechins 3. Use of
aqueous HF in THF for de-silyating compound 3 gave the corre-
sponding 3-O-acylated (e)-epigallocatechins 4 presented in Table 1.
Employment of unsaturated carboxylic acids, cycloalkyl carboxylic
acid, and aromatic carboxylic acid to esterify compound 3 also
provided the corresponding 3-O-acylated (e)-epigallocatechins
5e16 as shown in Table 2.
10 Lauric acid
12 Myristic acid
14 Palmitic acid
16 Stearic acid
18 Arachidic acid
20 Behenic acid
4j
4k
EGCG
(47.9%)
(28.3%)
6.29
e
Gallatic acid
(compound 4h, with palmitic acid, C16:0) with an IC50 of 0.53
m
M,
which was around 12-fold more potent than EGCG (IC50 ¼ 6.29
mM).
The inhibitory activity started to decrease when n ꢀ 16. For n ꢀ 18,
the activity decreased dramatically. Both compounds 4j (n ¼ 18)
and 4k (n ¼ 20) did not exhibit significant 5
a-reductase inhibitory
activities (IC50 > 5.0
mM).
3. In vitro studies
The 5 -redutase inhibitory activities for (e)-epigallocatechins
a
acylated with unsaturated fatty acids (compounds 5e13), with
cyclic carboxylic acids (compound 14), and with aromatic carbox-
ylic acids (compounds 15 and 16) were shown in Table 2. Among
compounds 5e7 which were derived from monounsaturated fatty
acids, EGC-palmitoleate (6, C16:1 for the fatty acid) was the most
The inhibitory activity of 3-O-acylated (e)-epigallocatechins
4e16 toward 5
lished method using rat liver microsomal 5
The percent inhibition value at a concentration of 5.0
a
-reductase was determined according to a pub-
-reductase [18,21,23].
M, or for
a
m
more potent compounds, the IC50 values are presented in Table 1
and Table 2. For compound 4 where its C3-O position was acyl-
ated with straight-chain saturated fatty acids, the inhibitory
activities increased with the increasing numbers of carbon atoms in
the fatty acids. The inhibitory activities reached maximum at n ¼ 14
potent compound with an IC50 of 0.48
similar to its saturated analogue EGC-palmitate (4h,
IC50 ¼ 0.53 M). The potency of compound 5 (IC50 ¼ 2.47 M, with
C10:1 fatty acid) was slightly improved in comparison with its
saturated analogue 4e (IC50 ¼ 2.94 M, with C10:0 fatty acid). EGC-
(E)-geranate (8, IC50 ¼ 2.24 M) containing two double bonds was
also more active than its saturated analogue 4e (IC50 ¼ 2.94 M).
EGCe -linoleic acid (9, with C18:3 fatty acid) had almost identical
mM. The activity of 6 was
m
m
m
m
m
OR
OR
g
OR
OR
OR
OR
inhibitory activity as its saturated analogs EGC-stearate (4i, with
C18:0 fatty acid) and EGC-(Z)-oleate (7, with C18:1 fatty acid). For
EGCe13-cis-retinoic acid (10) with five double bonds at the C3-O
position, the potency was similar to compound 9 and its analogs.
The potency for compounds 11e13 acylated by fatty acids with five
to six carbon atoms was not significant (IC50 > 5.0 mM), which were
consistent with results of compounds 4b and 4c derived from
saturated fatty acids with less than six carbons.
RO
O
RO
O
a
b
EGCG
O
OH
OR
OR
OR
OR
O
2
R = TBDMS
1
OR
OR
OH
OR
OR
OH
OH
Replacement of the gallate group in EGCG with lipophilic
cycloalkyl or aromatic carboxylates, including 4-tert-butylcyclo-
RO
O
HO
O
d
c
hexylcarboxylate,
enhanced the 5
14e16 in Table 2). Compared with EGCG (IC50 ¼ 6.29
values for compounds 14e16 decreased to 1.19, 1.32 and 1.60
respectively. In these cases, introduction of the lipophilic carbox-
ylates generated more potent inhibitors.
p-methybenzoate
-reductase inhibitory activities (see compound
M), the IC50
M,
or
p-chlorobenzoate,
O
O
a
OR
OH
O
R
O
R'
m
4
(R = -(CH2)nCH3, n = 0-20)
3
m
5−16
(R = alkenyl, aryl, cycloalkyl)
Scheme 1. Synthesis of 3-O-acylated (e)-epigallocatechins as 5a-reductase inhibitors.
For further exploration of the structureeactivity relationship of
the class of compounds toward the inhibition of 5a-reductase, we
Reagents and conditions: (a) TBDMSCl, imidazole, DMF, 91% yield; (b) LiAlH4, THF, 80%
yield; (c) R0COOH, DCC, DMAP, CH2Cl2, (d) HF, THF.

6070
S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
Table 2
synthesized peracylated EGCG 17 and permethylated EGCG 18 as
shown in Fig. 2. Compounds 17 and 18 were almost inactive with
8.4% and 0.5% inhibition at 5.0 M, respectively. Replacement of the
Inhibition of 5a-reductase by 3-O-acylated (e)-epigallocatechins with unsaturated,
cyclic, or aromatic carboxylic acids at C3-O position.
m
OH
ester moiety in 3-O-acylated (e)-epigallocatechins with a carba-
mate moiety decreased inhibitory activity (i.e. 15/19 and 16/20).
Introduction of the lipophilic naphthyl carbamate to the C3-O
OH
OH
HO
O
position (compound 21) did not improve 5
activity significantly.
a-reductase inhibitory
O
OH
O
R
4. Ligand-based pharmacophore identification
5 16
-
Catalyst/HipHop approach was used to evaluate the common
features required for the inhibition of 5a-reductase activity by 3-O-
acylated (e)-epigallocatechins [24]. EGCG, 4ae4k, and 5e21 were
subjected for pharmacophore model generation based on common
chemical features. The resulting most reasonable pharmacophore
model contained ten common features as shown in Fig. 3A: four
hydrogen-bond donors, one hydrophobic group, and five excluded
volumes. The most potent compound 6 was mapped onto the
model as shown in Fig. 3B.
Compound Structure of RCOOH
IC50 (mM) or %
inhibition at 5
mM
5
2.47
HO2C
(CH2)6CH3
6
0.48
0.74
2.24
HO2C(H2C)7 (CH2)5CH3
HO2C(H2C)7 (CH2)7CH3
5. In vivo study
7
8
For the study of anti-cancer activity of 3-O-acylated (e)-epi-
gallocatechins in vivo, compound 6 (20 mg/kg), EGCG (20 mg/kg
and 80 mg/kg), and vehicle control (15% PEG 400 and 5% EtOH in
H2O) were injected intraperitoneally daily into PC-3 bearing SCID
mice for 14 days. The tumor size was measured every two days for
28 days (see Fig. 4). Tumor size at day 28 was reduced by 38% and
32% with the administration of compound 6 at 20 mg/kg and EGCG
at 80 mg/kg, respectively. No significant anti-cancer effects of
dosing EGCG at 20 mg/kg were observed. There was no body weight
loss or drug-related death in all treatment groups. Though PC-3
cancer cell line is androgen-insensitive [25], compound 6 showed
a moderate anti-cancer activity in vivo.
O
HO
9
0.78
0.76
HOOC(H2C)7
CH2
CH2
CH2CH3
10
O
HO
11
(29.1%)
O
6. Discussion
HO
6.1. Chemistry
12
13
14
(36.0%)
(48.8%)
1.19
O
Several methods have been reported for the synthesis of 3-O-
acylated (e)-epigallocatechins by use of EGC as the starting material
[26] or through total synthesis [27]. In this study, we chose EGCG as
the starting material since it is the most abundant catechin from tea
leaves and commercially available. Furthermore, it is suitable for
larger scale preparation of the compounds for evaluation of the anti-
cancer activity in vivo.
HO
O
HO
O
O
O
OR
OH
HO
OR
OR
OH
OH
tBu
Me
Cl
RO
O
HO
O
15
16
1.32
1.60
O
O
HO
HO
OR
OR
OR
OH
R
O
O
N
H
OR
: R = Ac (0.5%)
18: R = Me (8.4%)
19
: R = p-Me-Ph- (0.8%)
17
20: R = p-Cl-Ph- (19.1%)
21
: R = 2-naphthyl- (28.7%)
Fig. 2. Inhibitory activity of peracylated and permethylated EGCG, and EGC-carba-
mates toward 5 -reductase. The numbers in the parentheses are the percent inhibition
at 5.0 M.
a
m

S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
6071
Fig. 3. (A) Pharmacophore generated using Catalyst methodology that describes the inhibition of 5a-reductase by 3-O-acylated (e)-epigallocatechins. Phamacophore features are
color-coded (hydrogen-bond donor, light red; hydrophobic, light blue; excluded volume, black). (B) Compound 6 mapped onto the pharmacophore. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article).
The traditional method to derive (e)-epigallocatechins esteri-
fied at the C3-O position involves the use of benzyl chloride [28] to
mask the hydroxyl groups in EGCG followed by saponifying the
gallate group by use of methanolic NaOH. However, the yield was
lower than that from the use of TBDMSCl and epimerization takes
place at C-2 position due to the use of the strong base [29]. In
addition, the methodology is not suitable for unsaturated fatty
acids since hydrogenation is required to remove the benzyl groups
and the condition also reduces the double or triple bonds. For the
deprotection of TBDMS groups in compound 3, we first tried the
use of TBAF as the desilylating agent. However, only trace amount of
the corresponding products were obtained. Since it is reported that
EGCG and (e)-epigallocatechin is extremely unstable in alkaline
solutions (pH > 8) while stable in acidic solutions (pH < 4) [30], we
used HF instead of TBAF to maintain the acidic condition to obtain
the products in satisfactory yields.
6.2. Biological studies
Since blocking all the hydroxyl groups in EGCG dramatically
reduced the activity (see the %inhibition of compounds 17 and 18 in
Fig. 2), the hydroxyl groups in EGCG was important to interact with
5a-reductase and might act as hydrogen bonding donors. Neverthe-
less, the contribution of each hydroxyl group could not be assessed. A
selective protection method should be used to rule out their different
roles. It could be postulated that the gallate group was not important
since its replacement with fatty acids retained or improved the 5a-
reductase inhibitory activity (see Tables 1 and 2). More evidences
come from the compounds derived from the replacement of gallate
groupinEGCGwithclose-in-size tert-butylcyclohexyl, methylphenyl,
or chlorophenyl group (see compounds 14e15 in Table 2). Their 5
reductase inhibitory was more potent. Though the structure of 5
reductase is unknown, it would be reasonable to assume a hydro-
phobic pocket to interact with the long-chain fatty acid in 3-O-acyl-
ated (e)-epigallocatechins. Previous kinetic study has shown that
EGCG does not compete with testosterone or NADPH for binding to
a
a
-
-
microsomal 5a-reductase; the inhibition could not be overcome by
increasing the concentration of NADPH or testosterone [18]. The 3-O-
acylated (e)-epigallocatechins would not be substrate analogs and
bound to the different pocket from testosterone and NADPH.
The range of inhibitory activity (w1e2 log units) for 3-O-acylated
(e)-epigallocatechins toward 5a-reductase was not sufficient to
generate a meaningful activity-based (predictive) pharmacophore
model using Catalyst/Hypogen technology. Catalyst/HipHop
approach was thus used as an alternative and the model generated
was presented in Fig. 3. For providing more information on the pocket
bound with the compounds (e.g., through receptor surface analysis)
[31,32], the synthesis of more diverse compounds is required. The
roles of hydroxyl groups at the different position of 3-O-acylated
(e)-epigallocatechin should be addressed and the receptor surface
around them needs to be probed by different functionalities.
As shown in Fig. 4, dosing compound 6 at 20 mg/kg [33] for 14
days showed better anti-cancer activity than dosing EGCG at
80 mg/kg and 20 mg/kg in PC-3 xenograft model. The results seem
to correlate to the 5
compounds (IC50 for 6, 0.48
a
-reductase inhibitory activity of the respective
M; IC50 for EGCG, 6.29 M). However,
m
m
other mechanism for the anti-cancer activity of 6 could not be ruled
out since EGCG also alters several signal transduction pathway
associated the proliferation and survival of cells [19]. The better
cell-penetrating ability of compound 6 arisen from the lipophilicity
would also attribute to its better in vivo anti-cancer activity.
Fig. 4. In vivo anti-cancer activity of compound 6 and EGCG for PC-3 human prostate
cancer xenograft.

6072
S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
7. Conclusions
1.0 equiv) was slowly added LiAlH4 (1.44 g, 37.9 mmol, 2.0 equiv) in
an ice bath. The solution was warmed to room temperature and
stirred for 4.0 h. The reaction was quenched with saturated
Na2SO4(aq) in an ice bath and filtered. The cake was washed with
Et2O (100 mL) and the combined filtrates were condensed under
reduced pressure. The residue was purified by gravity column
chromatography using silica gel (3.0% EtOAc in hexanes) to provide
(e)-epigallocatechin penta(tert-butyldimethylsilyl) ether (2, 13.3 g,
15.2 mmol) as viscous liquid in 80% yield: 1H NMR (CDCl3, 500 MHz)
In summary, we have synthesized a series of 3-O-acylated
(e)-epigallocatechins with inhibitory activity toward steroid 5
a-
reductase. -Reductase inhibition increased with increasing
5a
number of carbon atoms of a saturated fatty acid coupled with EGC,
reaching a maximum when the fatty acid was sixteen carbons in
length (palmitic acid, compound 4h, IC50 ¼ 0.53
mM). Replacement
of the saturated fatty acids with unsaturated fatty acid slightly
increased activity, providing the most potent compound 6 with an
d
6.61 (s,1H, ArH), 6.11 (d, J ¼ 2.1 Hz,1H, ArH), 5.96 (d, J ¼ 2.1 Hz,1H,
IC50 of 0.48
m
M, which showed moderate anti-tumor activity on PC-
ArH), 4.84 (s,1H, C2-H), 4.17(brs,1H, C3-H), 2.83e2.86(m, 2H, C4-H),
0.93e1.00 (m, 45H, 5 ꢃ C(CH3)3), 0.12e0.23 (m, 30H, 5 ꢃ Si(CH3)2);
3 human prostate cancer xenograft in vivo.
13C NMR (CDCl3, 125 MHz)
d e4.23, e4.06, e3.77, e3.45, 18.39
8. Experimental
(2 ꢃ C), 18.57, 18.98, 25.89 (2 ꢃ C), 26.36 (2 ꢃ C), 28.39, 66.77, 78.04,
101.50, 103.96, 104.07, 112.26, 129.95, 138.06, 148.94, 155.10 (2 ꢃ C),
155.47; ESI-MS m/z 879 (M þ H)þ.
8.1. Chemistry
(e)-Epigallocatechin-3-gallate (EGCG, >98% purity) were
purchased from Ryss Co. Chemical reagents and starting materials
including imidazole, TBDMSCl, fatty acids, carboxylic acids, and
isocyanates were purchase from Acros Organics or Aldrich. Purifi-
cation by gravity column chromatography was carried out by use of
Merck Silica Gel 60 (particle size 0.063e0.200 mm, 70e230 mesh
ASTM). Infrared (IR) spectra were measured on a PerkinElmer
Spectrum GX FT-IR spectrometer. The wavenumbers reported are
referenced to the polystyrene absorption at 1601 cmꢁ1. Absorption
intensities are recorded by the following abbreviations: s, strong; m,
medium; w, weak. Proton NMR spectra were obtained on a Bruker
AVANCE DRX 500 NMR spectrometer (500 MHz) by use of CDCl3,
CD3OD, or DMSO-d6 as solvent. Carbon-13 NMR spectra were
obtained a Bruker spectrometer(125 MHz) byuse of CDCl3 or DMSO-
d6 as solvent. Carbon-13 chemical shifts are referenced to the center
8.1.3. Generalprocedureforthesynthesisof3-O-acylated(e)-epigallo-
catechin penta(tert-butyldimethylsilyl) ether (3)
To a solution of compound 2 (1.0 equiv), N,N0-dicyclohex-
ylcarbodiimide (DCC, 2.2 equiv), and 4-dimethylaminopyridine
(DMAP, 0.40 equiv) in CH2Cl2 was added with various fatty acids or
carboxylic acids (2.0 equiv). The solution was stirred at room
temperature for 12 h and the insoluble urea was filtered. The filtrate
was concentrated under reduced pressure to provide crude
compound 3, which was directly used for next step.
8.1.4. Generalprocedureforthesynthesisof3-O-acylated(e)-epigallo-
catechin (4)
To a plastic beaker containing compound 3 in THF was added
aqueous HF (100 equiv). The reaction was stirred at room tempera-
ture and the reaction was monitored by TLC. After the reaction was
completed,thesolutionwasquenchedwithwaterandextractedwith
EtOAc. The combined organic layer was concentrated under reduced
pressure and the residue was purified by gravity column chroma-
tography using silica gel (EtOAc:hexanes :formic acid ¼ 1:1:0.01) to
provide compound 4.
peak of CDCl3 (d 77.0 ppm) or DMSO-d6 (d 39.5 ppm). Multiplicities
are recorded by the following abbreviations: s, singlet; d, doublet; t,
triplet; q, quartet; m, multiplet; J, coupling constant (Hz). ESI-MS
spectra were obtained from an Applied Biosystems API 300 mass
spectrometer.
8.1.1. Synthesis of (e)-epigallocatechin-3-gallate octa(tert-butyldi-
methylsilyl) ether (1) [22]
8.1.5. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl acetate (4a)
To a solution of (e)-epigallocatechin-3-gallate(10.0 g, 21.8mmol,
1.0 equiv) in DMF (50 mL) was added imidazole (14.9 g, 219 mmol,
10 equiv) followed by TBDMSCl (32.9 g, 218 mmol, 10 equiv) at 0 ꢂC.
The solution was warmed to room temperature and stirred for 12 h.
The reaction mixture was quenched with water (50 mL) and
extracted with hexanes (100 mL ꢃ 3). The combined organic layers
were washed with water (50 mL ꢃ 3), saturated aqueous NaCl
(50 mL), dried over MgSO4(s), filtered, and concentrated under
reduced pressure. The residue was purified by gravity column
chromatography using silica gel (2.0% EtOAc in hexanes) to provide
(e)-epigallocatechin-3-gallate octa(tert-butyldimethylsilyl) ether
(1, 27.3 g, 19.9 mmol) as white solids in 91% yield: 1H NMR (CDCl3,
Total yield from 2, 61%; 1H NMR (CD3OD, 500 MHz)
d 6.46 (s, 2H,
ArH), 5.94 (s, 1H, ArH), 5.90 (s,1H, ArH), 5.34 (brs, 1H, C2-H), 4.86 (s,
1H, C3-H), 2.89 (dd, J ¼ 17.2, 4.5 Hz, 1H, C4-H), 2.78 (dd, J ¼ 17.2,
2.2 Hz, 1H, C4-H), 1.92 (s, 3H, CH3C(¼O)); 13C NMR (DMSO-d6,
125 MHz)
d 21.55, 22.18, 67.68, 76.27, 94.20, 95.45, 97.26, 105.42,
128.35, 132.31, 145.50, 155.46, 156.50, 156.55, 170.38; ESI-MS m/z
349.0 (M þ H)þ.
8.1.6. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl butyrate (4b)
Total yield from 2, 51%; 1H NMR (CD3OD, 500 MHz)
d 6.49 (s, 2H,
ArH), 5.94 (s,1H, ArH), 5.91 (s, 1H, ArH), 5.34 (brs, 1H, C2-H), 4.62 (s,
1H, C3-H), 2.92 (dd, J ¼ 17.0, 4.4 Hz, 1H, C4-H), 2.76 (dd, J ¼ 17.0,
2.0 Hz, 1H, C4-H), 2.18 (t, J ¼ 7.1 Hz, 2H, eCH2C(¼O)), 1.28e1.33 (m,
2H, CH2), 0.76 (t, J ¼ 7.6 Hz, 3H, CH3); 13C NMR (DMSO-d6, 125 MHz)
500 MHz)
d
7.05 (s, 2H, ArH), 6.58 (s, 2H, ArH), 6.16 (d, J ¼ 2.2 Hz,1H,
ArH), 5.93 (d, J ¼ 2.2 Hz, 1H, ArH), 5.58 (s, 1H, C2-H), 5.01 (s, 1H, C3-
H), 2.93 (brs, 2H, C4-H), 0.86e0.98 (m, 72H, 8 ꢃ C(CH3)3), 0.08e0.20
(m, 48H, 8 ꢃ Si(CH3)2); 13C NMR (CDCl3, 125 MHz)
de4.22, e3.98,
d 13.56, 18.43, 22.18, 36.28, 67.71, 76.28, 94.21, 95.49, 97.28, 105.43,
e3.77, e3.65, e3.54, e2.77, 18.30, 18.38, 18.58 (2 ꢃ C), 18.85, 18.91,
25.84, 25.88 (2 ꢃ C), 26.26, 26.33, 26.40, 27.01, 68.08, 77.42, 101.84,
103.89, 103.98, 112.57, 115.62, 121.97, 130.09, 138.01, 143.08, 148.34,
148.57, 154.86, 155.12, 155.78, 165.19; ESI-MS m/z 1372.0 (M þ H)þ.
128.39, 132.34, 145.56, 155.49, 156.52, 156.59, 172.26; ESI-MS m/z
377.0 (M þ H)þ.
8.1.7. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl hexanoate (4c)
8.1.2. Synthesis of (e)-epigallocatechin penta(tert-butyldimethylsilyl)
ether (2) [22]
To a THF solution (100 mL) containing (e)-epigallocatechin-3-
gallate octa(tert-butyldimethylsilyl) ether (1, 26.0 g, 18.9 mmol,
Total yield from 2, 56%; 1H NMR (CD3OD, 500 MHz)
d 6.47 (s, 2H,
ArH), 5.94 (s, 1H, ArH), 5.91 (s, 1H, ArH), 5.33e5.34 (m, 1H, C2-H),
4.63 (s, 1H, C3-H), 2.88 (dd, J ¼ 17.2, 4.6 Hz, 1H, C4-H), 2.79 (dd,
J ¼ 17.2, 2.4 Hz, 1H, C4-H), 2.21 (t, J ¼ 7.2 Hz, 2H, eCH2C(¼O)),

S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
6073
1.08e1.28 (m, 6H, 3 ꢃ CH2), 0.82 (t, J ¼ 7.2 Hz, 3H, CH3); 13C NMR
(brs, 1H, C2-H), 4.88 (s, 1H, C3-H), 2.92 (dd, J ¼ 17.5, 4.5 Hz, 1H, C4-
H), 2.78 (dd, J ¼ 17.5, 4.5 Hz,1H, C4-H), 2.20 (t, J ¼ 7.2 Hz, 2H, eCH2C
(DMSO-d6, 125 MHz) d 14.07, 22.18, 23.98, 30.95, 31.44, 34.68, 67.73,
76.29, 94.24, 95.47, 97.26, 105.42, 128.38, 132.33, 145.55, 155.48,
(¼O)), 1.11e1.45 (m, 30H, 15 ꢃ CH2), 0.89 (t, J ¼ 7.2 Hz, 3H, CH3); 13
C
156.51, 156.58, 172.25; ESI-MS m/z 405.0 (M þ H)þ.
NMR (DMSO-d6, 125 MHz) d 14.01, 22.15, 24.43, 25.49, 28.35, 28.70,
28.76, 28.82, 28.95e29.11 (8 ꢃ CH2), 31.36, 33.67, 67.74, 76.33,
94.22, 95.45, 97.29, 105.47, 128.40, 132.46, 145.61, 155.44, 156.52,
156.62, 172.21; ESI-MS m/z 573.0 (M þ H)þ.
8.1.8. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl octanoate (4d)
Total yield from 2, 57%; 1H NMR (CDCl3, 500 MHz)
d 6.47 (s, 2H,
ArH), 5.92 (s, 1H, ArH), 5.88 (s, 1H, ArH), 5.30 (s, 1H, C2-H), 4.80 (s,
8.1.14. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
1H, C3-H), 2.78e2.81 (m, 2H, C4-H), 2.20 (t, J ¼ 7.2 Hz, 2H, eCH2C
3-yl icosanoate (4j)
(¼O)), 1.17e1.29 (m, 10H, 5 ꢃ CH2), 0.88 (t, J ¼ 7.3 Hz, 3H, CH3); 13
C
Total yield from 2, 51%; 1H NMR (CD3OD, 500 MHz)
d 6.46 (s, 2H,
NMR (DMSO-d6, 125 MHz)
d
14.01, 22.16, 23.57, 24.45, 28.99, 29.23,
ArH), 5.94 (s,1H, ArH), 5.91 (s,1H, ArH), 5.32 (s,1H, C2-H), 4.87 (s,1H,
C3-H), 2.73e2.90 (m, 2H, C4-H), 2.14 (t, J ¼ 7.2 Hz, 2H, eCH2C(]O)),
1.19e1.38 (m, 34H, 17 ꢃ CH2), 0.89 (t, J ¼ 6.8 Hz, 3H, CH3); 13C NMR
31.36, 34.51, 67.72, 76.28, 94.23, 95.46, 97.24, 105.41, 128.36, 132.39,
145.57, 155.49, 156.52, 156.58, 172.23; ESI-MS m/z 433.0 (M þ H)þ.
(DMSO-d6, 125 MHz)
d 13.92, 22.17, 24.42, 25.47, 28.36, 28.72,
8.1.9. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
28.95e29.11 (12ꢃ CH2), 31.37, 33.63, 67.73, 76.32, 94.21, 95.43, 97.28,
3-yl decanoate (4e)
105.43, 128.39, 132.42, 145.60, 155.46, 156.51, 156.61, 172.24; ESI-MS
Total yield from 2, 55%; 1H NMR (CDCl3, 500 MHz)
d
6.45 (s, 2H,
m/z 601.0 (M þ H) þ
.
ArH), 5.96 (s, 1H, ArH), 5.91 (s, 1H, ArH), 5.34 (s, 1H, C2-H), 4.83 (s,
1H, C3-H), 2.76e2.85 (m, 2H, C4-H), 2.13 (t, J ¼ 7.2 Hz, 2H, eCH2C
8.1.15. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
(¼O)), 1.13e1.31 (m, 14H, 7 ꢃ CH2), 0.91 (t, J ¼ 7.4 Hz, 3H, CH3); 13
C
3-yl docosanoate (4k)
NMR (DMSO-d6, 125 MHz)
d
14.03, 22.15, 23.54, 24.43, 28.37, 28.76,
Total yield from 2, 56%; 1H NMR (DMSO-d6, 500 MHz)
d 9.24 (s,
28.91, 29.06, 31.34, 34.50, 67.68, 76.29, 94.21, 95.45, 97.23, 105.45,
128.39, 132.41, 145.63, 155.47, 156.52, 156.64, 172.32; ESI-MS m/z
461.0 (M þ H)þ.
1H, OH), 8.99 (s, 1H, OH), 8.75 (s, 2H, OH), 7.98 (s, 1H, OH), 6.34 (s,
2H, ArH), 5.92 (s, 1H, ArH), 5.75 (s, 1H, ArH), 5.20 (s, 1H, C2-H), 4.85
(s, 1H, C3-H), 2.76e2.83 (m, 2H, C4-H), 2.16 (t, J ¼ 7.0 Hz, 2H, eCH2C
(¼O)), 1.15e1.38 (m, 38H, 19 ꢃ CH2), 0.85 (t, J ¼ 6.6 Hz, 3H, CH3); 13
C
8.1.10. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
NMR (DMSO-d6, 125 MHz) d 13.96, 22.16, 24.44, 25.48, 28.39, 28.79,
3-yl dodecanoate (4f)
28.96e29.12 (14 ꢃ CH2), 31.37, 33.62, 67.70, 76.30, 94.22, 95.45,
97.25, 105.46, 128.39, 132.41, 145.62, 155.47, 156.50, 156.59, 172.23;
ESI-MS m/z 627.0 (M ꢁ H)ꢁ.
Total yield from 2, 61%; 1H NMR (DMSO-d6, 500 MHz)
d 9.24 (s,
1H, OH), 8.99 (s, 1H, OH), 8.75 (s, 2H, OH), 7.98 (s, 1H, OH), 6.34 (s,
2H, ArH), 5.92 (s, 1H, ArH), 5.75 (s, 1H, ArH), 5.20 (s, 1H, C2-H), 4.85
(s, 1H, C3-H), 2.75e2.85 (m, 2H, C4-H), 2.12 (t, J ¼ 7.2 Hz, 2H, eCH2C
8.1.16. (E)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)
(¼O)), 1.09e1.35 (m, 18H, 9 ꢃ CH2), 0.85 (t, J ¼ 6.6 Hz, 3H, CH3); 13
C
chroman-3-yl) dec-2-enoate (5)
NMR (DMSO-d6, 125 MHz)
d
14.01, 22.17, 23.58, 24.44, 28.35, 28.73,
Total yield from 2, 55%; 1H NMR (CD3OD, 500 MHz)
d 6.85e6.88
28.80, 28.91, 29.05, 29.08, 31.37, 34.51, 67.71, 76.30, 94.23, 95.46,
97.26, 105.45, 128.39, 132.41, 145.63, 155.48, 156.50, 156.59, 172.27;
ESI-MS m/z 489.0 (M þ H)þ.
(m, 1H, C]CH), 6.46 (s, 2H, ArH), 5.95 (s, 1H, ArH), 5.91 (s, 1H, ArH),
5.74 (d, J ¼ 15.6 Hz, 1H, C]CH), 5.37 (m, 1H, C2-H), 4.90 (s, 1H, C3-
H), 2.75e2.90 (m, 2H, C4-H), 2.14e2.15 (m, 2H, C]CCH2), 1.28e1.41
(m, 10H, 5 ꢃ CH2), 0.88 (t, J ¼ 6.8 Hz, 3H, CH3); 13C NMR (DMSO-d6,
8.1.11. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
125 MHz) d 14.12, 23.12, 22.54, 28.58, 29.74, 30.01, 31.24, 33.56,
3-yl tetradecanoate (4g)
68.18, 76.71, 94.60, 95.84, 97.88, 105.76, 121.38, 128.81, 132.76,
146.00, 150.19, 155.86, 156.74, 157.21, 168.54; ESI-MS m/z 459.0
(M þ H)þ.
Total yield from 2, 58%; 1H NMR (CD3OD, 500 MHz)
d 6.47 (s, 2H,
ArH), 5.94 (s, 1H, ArH), 5.90 (s, 1H, ArH), 5.33 (m, 1H, C2-H), 4.86 (s,
1H, C3-H), 2.79e2.89 (m, 2H, C4-H), 2.18e2.21 (m, 2H, eCH2C(¼O)),
1.20e1.45 (m, 22H, 11 ꢃ CH2), 0.90 (t, J ¼ 6.8 Hz, 3H, CH3); 13C NMR
8.1.17. (Z)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)
(DMSO-d6,125 MHz)
d
14.00, 22.15, 23.53, 24.41, 28.36, 28.72, 28.81,
chroman-3-yl) hexadec-9-enoate (6)
28.86, 28.93, 28.97, 29.07, 29.11, 31.32, 34.50, 67.69, 76.33, 94.24,
95.47, 97.21, 105.46, 128.38, 132.42, 145.61, 155.49, 156.47, 156.56,
172.28; ESI-MS m/z 517.0 (M þ H)þ.
Total yield from 2, 52%; 1H NMR (CD3OD, 500 MHz)
d 6.48 (s, 2H,
ArH), 5.96 (d, J ¼ 2.3 Hz, 1H, ArH), 5.93 (d, J ¼ 2.3 Hz, 1H, ArH),
5.32e5.39 (m, 3H, eCH ¼ CHe þ C2-H), 4.89 (s, 1H, C3-H), 2.91 (dd,
J¼ 17.4,4.6Hz,1H, C4-H), 2.79(dd,J¼17.4, 2.1Hz,1H, C4-H), 2.17e2.20
(m, 2H, eCH2C(¼O)), 1.98e2.04 (m, 4H, eCH2CH ¼ CHCH2e),
1.13e1.33 (m, 18H, 9 ꢃ CH2), 0.89 (t, J ¼ 6.8 Hz, 3H, CH3); 13C NMR
8.1.12. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl palmitate (4h)
Total yield from 2, 60%; 1H NMR (CD3OD, 500 MHz)
d
6.46 (s, 2H,
(DMSO-d6,125MHz)
d
14.39,22.55, 24.85,25.91, 27.07(2ꢃ CH2), 28.75
ArH), 5.94 (s, 1H, ArH), 5.91 (s, 1H, ArH), 5.33 (s, 1H, C2-H), 4.88 (s,
1H, C3-H), 2.90 (dd, J ¼ 17.4, 4.5 Hz, 1H, C4-H), 2.77 (dd, J ¼ 17.4,
1.5 Hz, 1H, C4-H), 2.19 (t, J ¼ 7.3 Hz, 2H, eCH2C(¼O)), 1.11e1.33 (m,
26H, 13 ꢃ CH2), 0.90 (t, J ¼ 7.1 Hz, 3H, CH3); 13C NMR (DMSO-d6,
(2 ꢃ CH2), 28.92, 29.00, 29.57 (2 ꢃ CH2), 31.61, 34.05, 68.15, 76.73,
94.67, 95.90, 97.70, 105.90, 128.83, 130.02, 130.19, 132.85, 146.06,
155.90, 156.92, 157.02, 172.67; IR (diffuse reflectance) 3364 (br), 2928
(s), 2977 (2854), 2092 (s),1705 (s),1607 (s),1518 (m),1464 (s),1307 (s),
1251 (s), 1184 (s), 1141 (s), 1094 (s), 1032 (s), 1016 (s), 889 (w), 822 (s),
784 (m), 728 (m), 700 (s), 628 (w), 538 (w), 460 (w) cmꢁ1; ESI-MS m/z
543.1 (M þ H)þ.
125 MHz)
d 14.03, 22.11, 23.56, 24.40, 28.38, 28.73, 28.81, 28.87,
28.92e29.12 (6 ꢃ CH2), 31.33, 34.55, 67.66, 76.38, 94.23, 95.48,
97.20, 105.47, 128.35, 132.43, 145.60, 155.52, 156.48, 156.57, 172.29;
ESI-MS m/z 545.0 (M þ H)þ.
8.1.18. (Z)-((2R,3R)-5,7-dihydroxy-2-(3,4,5-trihydroxyphenyl)
8.1.13. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
chroman-3-yl) octadec-9-enoate (7)
3-yl stearate (4i)
Total yield from 2, 68%; 1H NMR (CD3OD, 500 MHz)
d 6.48 (s, 2H,
Total yield from 2, 65%; 1H NMR (CD3OD, 500 MHz)
d
6.47 (s, 2H,
ArH), 5.96 (s, 1H, ArH), 5.93 (s, 1H, ArH), 5.32e5.35 (m, 3H,
CH ¼ CHþC2-H), 4.89 (s,1H, C3-H), 2.82e2.93 (m, 2H, C4-H), 2.16e2.18
ArH), 5.94 (d, J ¼ 2.2 Hz, 1H, ArH), 5.91 (d, J ¼ 2.2 Hz, 1H, ArH), 5.33

6074
S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
(m, 2H, eCH2C(¼O)), 2.00e2.02 (m, 4H, eCH2CH ¼ CHCH2e),
1.20e1.33 (m, 22H, 11 ꢃ CH2), 0.89 (t, J ¼ 6.8 Hz, 3H, CH3); 13C NMR
8.1.24. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-hex-5-ynoate (13)
(DMSO-d6, 125 MHz)
d
14.48, 22.66, 23.76, 26.10, 28.14, 29.43, 30.21,
Total yield from 2, 48%; 1H NMR (CD3OD, 500 MHz)
d 6.46 (s,
30.26, 30.33, 30.37, 30.48, 30.64, 30.81, 30.86, 33.08, 34.95, 68.23, 76.75,
94.69, 95.74, 97.95,106.01,128.67,130.79,130.88,132.74,146.02,155.93,
156.91, 157.12, 172.56; ESI-MS m/z 571.0 (M þ H)þ.
2H, ArH), 5.94 (s, 1H, ArH), 5.90 (s, 1H, ArH), 5.35 (m, 1H, C2-H),
4.88 (s, 1H, C3-H), 2.79e2.90 (m, 2H, C4-H), 2.31e2.33 (m, 2H,
ChCCH2), 2.16 (s, 1H, ChCH), 2.01e2.06 (m, 2H, eCH2C(¼O)),
1.62e1.65 (m, 2H, CH2); 13C NMR (DMSO-d6, 125 MHz)
d 20.72,
8.1.19. (E)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)
22.55, 23.91, 33.20, 67.98, 71.32, 76.28, 84.71, 94.55, 95.94, 97.47,
105.68, 128.74, 132.75, 145.87, 155.06, 156.90, 157.26, 173.44; ESI-
MS m/z 401.0 (M þ H)þ.
chroman-3-yl)-3,7-dimethylocta-2,6-dienoate (8)
Total yield from 2, 50%; 1H NMR (CD3OD, 500 MHz)
d 6.66 (s, 2H,
ArH), 6.01 (s, 1H, ArH), 5.95 (m, 2H, ArH þ C]CH), 5.29e5.33 (m,
2H, C]CH þ C2-H), 4.87 (s, 1H, 1H, C3-H), 2.79e2.82 (m, 2H, C4-H),
2.00e2.25 (m, 4H, 2 ꢃ CH2), 1.55e1.58 (m, 9H, 3 ꢃ CH3); 13C NMR
8.1.25. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-4-tert-butylcyclohexanecarboxylate (14)
(DMSO-d6, 125 MHz)
d
18.54, 18.62, 22.42, 24.60, 26.15, 39.12, 68.21,
Total yield from 2, 52%; 1H NMR (CD3OD, 500 MHz)
d 6.47 (s, 2H,
76.64, 94.55, 95.81, 97.60, 105.91, 114.23, 123.55, 128.82, 132.06,
132.76, 146.04, 155.91, 156.90, 157.12, 161.23, 166.02; ESI-MS m/z
457.0 (M þ H)þ.
ArH), 5.96 (s, 1H, ArH), 5.93 (s, 1H, ArH), 5.32 (m, 1H, C2-H), 4.90 (m,
1H, C3-H), 2.76e2.89 (m, 2H, C4-H), 2.02e2.08 (m, 1H, CH),
1.70e1.85 (m, 4H, 2 ꢃ CH2),1.36e1.40 (m,1H, CH),1.29e1.34 (m, 4H,
2 ꢃ CH2), 0.82 (s, 9H, 3 ꢃ CH3); 13C NMR (DMSO-d6, 125 MHz)
8.1.20. (9Z,12Z,15Z)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxy-
d 22.43, 25.54, 27.62, 29.88, 32.52, 43.62, 48.23, 67.94, 76.38, 94.75,
phenyl)chroman-3-yl)octadeca-9,12,15-trienoate (9)
95.84, 97.37, 105.55, 128.65, 132.87, 145.98, 155.16, 157.01, 157.28,
Total yield from 2, 22%; 1H NMR (CD3OD, 500 MHz)
d
6.47 (s, 2H,
175.42; ESI-MS m/z 473.0 (M þ H)þ.
ArH), 5.95 (s, 1H, ArH), 5.91 (s, 1H, ArH), 5.28e5.35 (m, 7H,
3 ꢃ eCH ¼ CHe þ C2-H), 4.89 (s, 1H, C3-H), 2.77e2.81 (m, 6H,
3 ꢃ CH2), 2.12e2.22 (m, 2H, CH2), 2.01e2.12 (m, 2H, CH2), 1.95e2.00
(m, 2H, CH2), 1.29e1.37 (m, 10H, 5 ꢃ CH2), 0.89 (t, J ¼ 6.9 Hz, 3H,
8.1.26. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-4-methylbenzoate (15)
Total yield from 2, 60%; 1H NMR (CD3OD, 500 MHz)
d 7.76 (d,
CH3); 13C NMR (DMSO-d6, 125 MHz)
d
14.46, 22.39, 23.63, 25.71,
J ¼ 8.2 Hz, 2H, ArH), 7.19 (d, J ¼ 8.2 Hz, 2H, ArH), 6.51 (s, 2H, ArH),
5.97 (s, 1H, ArH), 5.95 (s, 1H, ArH), 5.33e5.35 (m, 1H, C2-H), 4.91 (s,
1H, C3-H), 2.90 (dd, J ¼ 17.4, 4.6 Hz, 1H, CH), 2.80 (dd, J ¼ 17.4,
2.1 Hz, 1H, CH), 2.36 (s, 3H, ArCH3); 13C NMR (DMSO-d6, 125 MHz)
26.56, 27.90, 28.19, 30.20, 30.46, 32.66, 34.83, 68.12, 76.68, 94.55,
95.84, 97.72, 105.91, 128.81, 128.95, 129.08, 129.24, 129.26, 130.60,
131.14, 132.74, 146.16, 155.83, 156.91, 157.10, 172.60; ESI-MS m/z
567.5 (M þ H)þ.
d
21.12, 22.57, 67.84, 76.43, 94.80, 95.72, 97.27, 105.48, 128.62,
128.33, 129.39, 129.62, 133.66, 142.86, 146.92, 155.06, 157.04, 157.15,
8.1.21. (2E,4E,6E,8E)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxy-
phenyl)chroman-3-yl)-3,7-dimethyl-9-(2,6,6-trimethylcyclohex-1-
enyl)nona-2,4,6,8-tetraenoate (10)
170.21; ESI-MS m/z 425.0 (M þ H)þ.
8.1.27. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
Total yield from 2, 15%; 1H NMR (CD3OD, 500 MHz)
d
7.64e7.67
3-yl-4-chlorobenzoate (16)
(m, 1H, C]CH), 6.95e7.02 (m, 1H, C]CH), 6.48 (s, 2H, ArH),
6.08e6.24 (m, 4H, C]CH), 5.97 (s, 1H, ArH), 5.94 (s, 1H, ArH), 5.32
(s,1H, C2-H), 4.92 (s,1H, C3-H), 2.73e2.95 (m, 2H, C4-H), 2.18 (s, 3H,
CH3), 2.01e2.03 (m, 2H, CH2), 1.92 (s, 3H, CH3), 1.63 (s, 3H, CH3),
1.29e1.50 (m, 4H), 1.04 (s, 3H, CH3), 0.96 (s, 3H, CH3); 13C NMR
Total yield from 2, 55%; 1H NMR (CD3OD, 500 MHz)
d 7.84 (d,
J ¼ 8.6 Hz, 2H, ArH), 7.41 (d, J ¼ 8.6 Hz, 2H, ArH), 6.50 (s, 2H, ArH),
5.97 (s, 1H, ArH), 5.96 (s, 1H, ArH), 5.33e5.35 (m, 1H, C2-H), 4.97 (s,
1H, C3-H), 2.95 (dd, J ¼ 17.2, 4.5 Hz, 1H, C4-H), 2.85 (dd, J ¼ 17.2,
2.2 Hz, 1H, C4-H); 13C NMR (DMSO-d6, 125 MHz)
d 22.67, 67.75,
(DMSO-d6, 125 MHz)
d
12.85, 13.91, 20.31, 21.91, 22.56, 29.40, 33.98,
76.45, 94.82, 95.49, 97.20, 105.34, 128.62, 128.69, 129.74, 131.23,
132.62, 138.04, 146.83, 155.07, 157.11, 157.15, 170.37; ESI-MS (for
35Cl) m/z 445.0 (M þ H)þ.
35.25, 40.77, 68.23, 76.97, 94.76, 95.91, 97.88, 105.92, 119.83, 128.85,
129.49,130.68,130.94,132.21,132.83,136.52,138.96,139.06,140.45,
146.26, 154.28, 156.14, 157.02, 157.22, 165.78; ESI-MS m/z 589.0
(M þ H)þ.
8.1.28. 5-(((2R,3R)-5,7-Diacetoxy-2-(3,4,5-triacetoxyphenyl)
chroman-3-yloxy)carbonyl)benzene-1,2,3-triyl triacetate (17) [34]
Prepared from the reaction of EGCG with excess acetyl anhy-
8.1.22. (E)-((2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)
chroman-3-yl)-2-methylbut-2-enoate (11)
dride in pyridine. Yield, 65%; 1H NMR (CDCl3, 500 MHz)
d 7.11 (s, 2H,
Total yield from 2, 57%; 1H NMR (CD3OD, 500 MHz)
d
6.73e6.74
ArH), 7.26 (s, 2H, ArH), 6.76 (s, 1H, ArH), 6.64 (s, 1H, ArH), 5.66 (brs,
(m, 1H, C]CH), 6.55 (s, 2H, ArH), 5.96 (s, 1H, ArH), 5.93 (s, 1H, ArH),
5.34 (s, 1H, C2-H), 4.92 (s, 1H, C3-H), 2.80e2.90 (m, 2H, C4-H),
1H, C2-H), 5.21 (s, 1H, C3-H), 3.00e3.15 (m, 2H, C4-H), 2.22e2.45
(m, 24H, 3 ꢃ (C]O)CH3); 13C NMR (DMSO-d6, 125 MHz)
d 20.29,
1.70e1.73 (m, 6H, 2 ꢃ CH3); 13C NMR (DMSO-d6, 125 MHz)
d
12.86,
20.40, 20.59, 20.72, 20.84, 20.93, 26.05, 68.18, 76.91, 108.26, 109.18,
118.38, 118.72, 122.27, 122.53, 134.53, 135.25, 139.11, 143.47, 143.57,
149.82, 149.90, 154.91, 163.72, 166.39, 166.94, 167.63, 167.82, 168.64,
168.83; ESI-MS m/z 715.0 (M þ H)þ.
15.32, 22.43, 67.28, 76.26, 94.54, 95.83, 97.88, 105.75, 128.40,
128.68, 132.65, 138.67, 145.72, 155.96, 157.21, 157.30, 166.20; ESI-MS
m/z 389.0 (M þ H)þ.
8.1.23. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
8.1.29. (2R,3R)-5,7-Dimethoxy-2-(3,4,5-trimethoxyphenyl)
chroman-3-yl-3,4,5-trimethoxybenzoate (18) [35]
3-yl-3-methylbut-2-enoate (12)
Total yield from 2, 51%; 1H NMR (CD3OD, 500 MHz)
d
6.55 (s, 2H,
Prepared from the reaction of EGCG with excess dimethylsulfate
ArH), 6.46 (s, 1H, C]CH), 5.97 (s, 1H, ArH), 5.91 (s, 1H, ArH), 5.36 (s,
1H, C2-H), 4.94 (s, 1H, C3-H), 2.83e2.94 (m, 2H, C4-H), 1.58 (s, 3H,
in the presence of K2CO3 using acetone as the solvent. Yield, 43%; 1H
NMR (CD3OD, 500 MHz) d 7.11 (s, 2H, ArH), 6.78 (s, 2H, ArH), 6.22 (s,
CH3), 1.57 (s, 3H, CH3); 13C NMR (DMSO-d6, 125 MHz)
d
20.41, 22.38,
1H, ArH), 6.16 (s, 1H, ArH), 5.63 (m, 1H, C2-H), 5.12 (s, 1H, C3-H),
26.45, 68.15, 76.31, 94.59, 95.88, 97.50, 105.76, 116.12, 128.85,
132.60, 145.96, 155.88, 156.92, 157.02, 157.20, 166.20; ESI-MS m/z
389.0 (M þ H)þ.
3.65e3.82 (m, 24H, 3 ꢃ CH3), 2.90e3.12 (m, 2H, C4-H); 13C NMR
(DMSO-d6,125 MHz) d 22.38, 55.38, 56.14, 56.52, 56.88, 57.02, 61.54,
68.50, 76.81, 92.17, 92.95, 99.66, 100.88, 106.75, 125.90, 129.45,

S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
6075
138.19, 143.01, 153.89, 154.24, 156.38, 159.28, 160.75, 164.09, ESI-MS
DHT on the TLC plate were exposed by spraying with PMA (12 g in
250 mL of 95% ethanol). The spots on the TLC were cut separately
and placed into counting vials, followed by the addition of 0.50 mL
of ethyl acetate to extract [3H]T or [3H]DHT. Five mL of Ready
OrganicTM liquid scintillation cocktail was added to vials and the
radioactivity was measured using an LS6500 multi-purpose scin-
tillation counter (Beckman Instruments, Inc.). The concentration of
test compound inhibiting the conversion of [3H]testosterone to [3H]
DHT by 50% (IC50) was determined by interpolation between
appropriate data points.
m/z 571.0 (M þ H)þ.
8.1.30. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-p-tolylcarbamate (19)
From the reaction of 2 with p-tolyl isocyanate followed by
desilylation with HF. Total yield from 2, 44%; 1H NMR (DMSO-d6,
500 MHz) d 9.37 (s,1H), 9.27 (s, 1H), 8.99 (s,1H), 8.73 (s, 2H), 8.03 (s,
1H), 7.27 (brs, 2H, ArH), 7.00 (d, J ¼ 8.2 Hz, 2H, ArH), 6.41 (s, 2H,
ArH), 5.93 (s, 1H, ArH), 5.77 (s, 1H, ArH), 5.28 (s, 1H, C2-H), 4.93 (s,
1H, C3-H), 2.85e2.95 (m, 2H, C4-H), 2.19 (s, 3H, CH3); 13C NMR
(DMSO-d6,125 MHz)
d
20.37, 25.97, 67.78, 76.42, 94.46, 95.54, 97.66,
8.2.2. In vivo anti-cancer activity studies
105.63, 118.41, 128.57, 129.03, 131.16, 132.49, 136.56, 145.72, 153.07,
PC-3 (1 ꢃ 106 cell/mouse) tumor cells were subcutaneously
injected into the right flank of 5 week old male NOD/SCID mice
(BioLASCO, Taiwan). Tumor volume was measured with a digital
caliper once tumor was palpable (within 10 to 15 days after
implantation). The tumor-bearing animals were treated when the
size of tumor have reached to an average volume of w50 to
100 mm3. Twenty mice were divided into 4 groups and treated
intraperitoneally with compound 6 (20 mg/kg), EGCG (20 mg/kg or
80 mg/kg) and vehicle control (15% PEG 400 and 5% EtOH in H2O)
with one dose per day for 14 days. The test compound 6 and EGCG
were formulated in a solution of 15% PEG 400, 5% EtOH, and 80%
H2O. Body weights of the mice were measured every two days and
the size of tumor was measured by a digital caliper every two days.
The tumor size plotted in Fig. 4 represents the mean þ/ꢁ S.E. of data
from 5 mice.
155.53, 156.58, 156.61; ESI-MS m/z 438.0 (M ꢁ H)ꢁ.
8.1.31. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-4-chlorophenylcarbamate (20)
From the reaction of 2 with 4-chlorophenyl isocyanate followed
by desilylation with HF. Total yield from 2, 67%; 1H NMR (CD3OD,
500 MHz)
d
7.29 (brs, 2H, ArH), 7.18 (d, J ¼ 8.8 Hz, 2H, ArH), 6.53 (s,
2H, ArH), 5.95 (s, 1H, ArH), 5.90 (s, 1H, ArH), 5.37 (s, 1H, C2-H), 4.84
(s, 1H, C3-H), 2.80e2.93 (m, 2H, C4-H); 13C NMR (DMSO-d6,
125 MHz)
d 25.91, 68.11, 76.32, 94.44, 95.54, 97.54, 105.54, 119.82,
125.98, 128.46, 128.52, 132.48, 138.12, 145.72, 152.95, 155.47, 156.59,
156.63; ESI-MS (for 35Cl) m/z 458.0 (M ꢁ H)ꢁ.
8.1.32. (2R,3R)-5,7-Dihydroxy-2-(3,4,5-trihydroxyphenyl)chroman-
3-yl-naphthalen-1-ylcarbamate (21)
From the reaction of 2 with 1-naphthyl isocyanate followed by
8.3. Computational method
desilylation with HF. Total yield from 2, 55%; 1H NMR (DMSO-d6,
500 MHz) d 9.37 (s, 1H), 9.31 (s, 1H), 9.02 (s, 1H), 8.78 (s, 2H), 8.10 (s,
The HipHop module in the CATALYST 4.11 package (Accelrys Inc.,
1H), 7.75e7.87 (m, 1H, ArH), 7.70e7.75 (m, 2H, ArH), 7.32e7.48 (m,
4H), 6.50 (s, 2H, ArH), 5.96 (s, 1H, ArH), 5.80 (s, 1H, ArH), 5.34 (s, 1H,
C2-H), 4.95 (s, 1H, C3-H), 2.65e2.98 (m, 2H, C4-H); 13C NMR
San Diego, CA) was used to identify the common features for 5a-
reductase inhibitors [24]. The HipHop algorithm was forced to
determine pharmacophore models that contain 0e5 hydrogen-
bond acceptor, 0e5 hydrogen-bond donor, 0e3 positive charge,
0e3 negative charge, 0e3 hydrophobic group and 0e5 excluded
volume features. All parameters were set default, except for the
Principal value and MaxOmitFeat value. Finally, the top 10 scoring
hypotheses composed of these pharmacophore features for each
set were exported. CATALYST produced 10 hypotheses, and all
include ten common features: four hydrogen-bond donor, one
hydrophobic group and five excluded volume. Among these 10
hypotheses, Hypol is the most reasonable (Fig. 3) as characterized.
(DMSO-d6, 125 MHz)
d 25.93, 67.92, 76.67, 94.46, 95.48, 97.71,
105.74, 122.36, 123.43, 123.66, 125.42, 125.94, 126.77, 127.78, 128.67,
128.70, 132.53, 133.65, 133.75, 145.75, 154.58, 155.66, 156.59,
156.62; ESI-MS m/z 474.0 (M ꢁ H)ꢁ.
8.2. Biological procedures
[3H]Testosterone (80.4 Ci/mmol, [3H]T) was purchased from New
England Biochemical (Boston, MA). Testosterone (T), dihy-
drotestosterone (DHT), dithiothreitol and finasteride were purchased
from Sigma. Phosphomolybdic acid (PMA), nicotinamide adenine
dinucleotide phosphate (NADPH), and analytical grade organic
solvents were purchased from Merck. Ready OrganicTM liquid scin-
tillation cocktail was purchased from Beckman Coulter, Inc.
Acknowledgements
We are grateful to the Ministry of Economic Affairs of Republic
of China for financial support.
8.2.1. Assay of 5a-reductase
References
Microsomes were isolated from rat liver as described by Sun
et al. [23]. The enzymatic assay was based on the measurement of
DHT production from testosterone in the presence of microsomes
[1] E. Haslam, Plant Polyphenols-Vegetable Tannins Revisited. Cambridge
University Press, Cambridge, 1989.
[2] For the review of biological activity and beneficial effects of tea and EGCG, see
(a) H. Fujiki, M. Suganuma, S. Okabe, N. Sueoka, A. Komori, E. Sueoka, T. Kozu,
Y. Tada, K. Suga, K. Imai, K. Nakachi, Mutat. Res. 402 (1998) 307e310;
(b) I.E. Dreosti, Nutr. Rev. 54 (1996) S51eS58;
from rat liver containing both the type 1 and type 2 5
a
-reductase.
The assay mixture, in a final volume of 200
mL, contained 35 nM
[3H]testosterone, 1.0 mM NADPH, 1 mM dithiothreitol, and 40 mM
potassium phosphate, pH 6.5, with or without a test compound.
The reaction was started by the addition of 500 ng rat liver
microsome. After incubation at 37 ꢂC for 30 min, the reaction was
(c) D.S. Wheeler, W.J. Wheeler, Drug Dev. Res. 61 (2004) 45e65;
(d) J. Gupta, Y.H. Siddique, T. Beg, G. Ara, M. Afzal, Int. J. Pharmacol. 4 (2008)
314e338;
(e) D.G. Nagle, D. Ferreira, Y.-D. Zhou, Phytochemistry 67 (2006) 1849e1855.
[3] F. Yang, J.Y. Chung, G.Y. Yang, C. Li, X. Meng, M.J. Lee, BioFactors 13 (2000) 73e79.
[4] S. Shankar, S. Ganapathy, R.K. Srivastava, Front. Biosci. 12 (2007) 4881e4899.
[5] S. Bettuzzi, M. Brausi, F. Rizzi, G. Castagnetti, G. Peracchia, A. Corti, Cancer Res.
66 (2006) 1234e1240.
[6] J.P. Deslypere, M. Young, J.D. Wilson, M.J. McPhaul, Mol. Cell Endocrinol.
88 (1992) 15e22.
[7] A.S. Wright, R.C. Douglas, L.N. Thomas, C.B. Lazier, R.S. Rittmaster, Endocrinology
140 (1999) 4509e4515.
stopped by the addition of 200
testosterone and 500 ng DHT (both radio-inactive). The reaction
mixture was vortexed for 1.0 min and 50 L upper organic layer was
collected and separated by TLC (silica gel 60 F-254 plate,
6 ꢃ 1.5 cm). These plates were developed in a solvent system
(cyclohexane:ethyl acetate ¼ 50:50, v/v) and the testosterone and
mL ethyl acetate containing 500 ng
m

6076
S.F. Lin et al. / European Journal of Medicinal Chemistry 45 (2010) 6068e6076
[8] H. Bonkhoff, U. Stein, G. Aumuller, K. Remberger, Prostate 29 (1996) 261e267.
[9] F.W. George, L. Johnson, J.D. Wilson, Endocrinology 125 (1989) 2434e2438.
[10] J. Imperato-McGinley, T. Gautier, K. Zirinsky, T. Hom, O. Palomo, E. Stein,
E.D. Vaughan, J.A. Markisz, E. Ramirez de Arellano, E. Kazam, J. Clin. Endo-
crinol. Metab. 125 (1989) 2434e2438.
[11] B. Kenny, S. Ballard, J. Blagg, D. Fox, J. Med. Chem. 40 (1997) 1293e1315.
[12] B. Djavan, M. Remzi, B. Erne, M. Marberger, Drugs Today 38 (2002) 867e876.
[13] (a) S. Khandpur, M. Suman, B.S. Reddy, J. Dermatol. 29 (2002) 489e498;
(b) R. Hoffmann, Dermatology 206 (2003) 85e95.
[14] (a) J.C. Shaw, Am. J. Clin. Dermatol. 3 (2002) 571e578;
(b) A. Cilotti, G. Danza, M. Serio, J. Endocrinol. Invest. 24 (2001) 199e203.
[15] (a) R. Azziz, E. Carmina, M.E. Sawaya, Endocr. Rev. 21 (2000) 347e362;
(b) L. Falsetti, A. Cambera, Fertil. Steril. 72 (1999) 41e46.
[16] (a) M.C. Bosland, J. Natl. Cancer Inst. Monogr. 27 (2000) 39e66;
(b) L.N. Thomas, C.B. Lazier, R. Gupta, R.W. Norman, D.A. Troyer, S.P. O’Brien,
R.S. Rittmaster, Prostate 63 (2005) 231e239.
[17] I.M. Thompson, P.J. Goodman, C.M. Tangen, M.S. Lucia, G.J. Miller, L.G. Ford,
M.M. Lieber, R.D. Cespedes, J.N. Atkins, S.M. Lippman, S.M. Carlin, A. Ryan,
C.M. Szczepanek, J.J. Crowley, C.A. Coltman Jr., N. Engl. J. Med. 349 (2003)
215e224.
[18] S. Liao, R.A. Hiipakka, Biochem. Biophys. Res. Commun. 214 (1995) 833e838.
[19] N. Khan, H. Mukhtar, Cancer Lett. 269 (2008) 269e280.
[20] S. Liao, Y. Umekira, J. Guo, J.M. Kokontis, R.A. Hiipakka, Cancer Lett. 96 (1995)
239e243.
[24] CATALYST V4.11. Accelrys Inc., San Diego, CA., 2005.
[25] W.D. Tilley, C.M. Wilson, M. Marcelli, M.J. McPhaul, Cancer Res. 50 (1990)
5382e5386.
[26] A. Kumagai, Y. Nagaoka, T. Obayashi, Y. Terashima, H. Tokuda, Y. Hara,
T. Mukainaka, H. Nishino, H. Kuwajima, S. Uesato, Bioorg. Med. Chem. 11
(2003) 5143e5148.
[27] (a) S.B. Wan, Q.P. Dou, T.H. Chan, Tetrahedron 62 (2006) 5897e5904;
(b) N.T. Zaveri, Org. Lett. 3 (2001) 843e846;
(c) S.B. Wan, K.R. Landis-Piwowar, D.J. Kuhn, D. Chen, Q.P. Dou, T.H. Chan,
Bioorg. Med. Chem. 13 (2005) 2177e2185.
[28] T.W. Greene, P.G.M. Wuts, Protective Groups in Organic Synthesis, third ed..
John Wiley, New York, 1999.
[29] (a) J.F.W. Burger, J.P. Styenberg, D.A. Young, E.V. Brandt, J. Chem. Soc. Perkin
Trans. 1 (1989) 671e681;
(b) J.P. Styenberg, J.F.W. Burger, D.A. Young, E.V. Brandt, J.A. Steenkamp,
D. Ferreira, J. Chem. Soc. Perkin Trans. 1 (1988) 3323e3329.
[30] Q.Y. Zhu, A. Zhang, D. Tsang, Y. Huang, Z.-Y. Chen, J. Agric. Food Chem. 45
(1997) 4624e4628.
[31] M. Hahn, J. Med. Chem. 38 (1995) 2080e2090.
[32] M. Hahn, D. Rogers, J. Med. Chem. 38 (1995) 2091e2102.
[33] We have tried a higher dose of 6 to 40 mg/kg to seek better in vivo activity.
However, a significant weight loss of the mice was observed (>15%) at day 5
and the dosing was thus terminated. It is hardly to attribute the weight loss to
the toxicity of 6 since EGCG is also reported to have an anti-obesity effect. The
activity of weight loss might be enhanced by the long-chain fatty acid in 6 in
comparison with EGCG.
[21] R.A. Hiipakka, H.-Z. Zhang, W. Dai, Q. Dai, S. Liao, Biochem. Pharmacol. 63
(2002) 1165e1176.
[22] J.C. Anderson, C. Headley, P.D. Stapleton, P.W. Taylor, Bioorg. Med. Chem. Lett.
15 (2005) 2633e2635.
[34] S. Mori, S. Miyake, T. Kobe, T. Nakaya, S.D. Fuller, N. Kato, K. Kaihatsu, Bioorg.
Med. Chem. Lett. 18 (2008) 4129e4252.
[23] Z.-Y. Sun, Z.-H. Tu, Meth. Find. Exp. Clin. Pharmacol. 20 (1998) 283e287.
[35] F. Hashimoto, G. Nonaka, I. Nishioka, Chem. Pharm. Bull. 37 (1989) 77e85.