Cytotoxicity against cholangiocarcinoma cell lines of zerumbone derivatives

Article abstract of DOI:10.1016/j.ejmech.2010.05.029

Cholangiocarcinoma (CCA) is an aggressive malignancy with a very high morbidity and mortality for which an effective treatment is lacking. In this study, seventeen zerumbone derivatives were synthesized and evaluated for in vitro cytotoxicity against chol

Full text of DOI:10.1016/j.ejmech.2010.05.029

European Journal of Medicinal Chemistry 45 (2010) 3794e3802
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Cytotoxicity against cholangiocarcinoma cell lines of zerumbone derivatives
Uraiwan Songsiang a, Siripit Pitchuanchom a, Chantana Boonyarat b, Chariya Hahnvajanawong c,
,
Chavi Yenjai a
*
a Natural Products Research Unit, Center for Innovation in Chemistry, Department of Chemistry, Faculty of Science, Khon Kaen University, Khon Kaen 40002, Thailand
b Faculty of Pharmaceutical Sciences, Khon Kaen University, Khon Kaen 40002, Thailand
c Department of Microbiology, Liver Fluke and Cholangiocarcinoma Research Center, Faculty of Medicine, Khon Kaen University, Khon Kaen 40002, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history:
Cholangiocarcinoma (CCA) is an aggressive malignancy with a very high morbidity and mortality for
which an effective treatment is lacking. In this study, seventeen zerumbone derivatives were synthesized
and evaluated for in vitro cytotoxicity against cholangiocarcinoma cell lines. 5 showed the most potent
Received 19 January 2010
Received in revised form
6 May 2010
Accepted 12 May 2010
Available online 20 May 2010
antiproliferative activity against KKU-100 cell line with an IC50 value of 16.44 mM. To investigate the
potential molecular target of the most active compound, the docking was performed using different
enzymes and receptor proteins including CDK-2, CDK-5, EGFR, and GSK-3. The docking results revealed
that 5 exhibited better binding interaction to EGFR than CDK-2, CDK-5 and GSK-3. All results indicate that
5 should be a promising candidate for treatment of cancer.
Keywords:
Zingiber zerumbet Smith
Zerumbone
Ó 2010 Elsevier Masson SAS. All rights reserved.
Sesquiterpene
Cholangiocarcinoma
Cytotoxicity
Docking
1. Introduction
factor receptor (EGFR) and protein kinases are found to play an
important role in cell cycle progression, cell proliferation and
Cholangiocarcinoma (CCA) is a primary cancer of the bile duct
epithelial cells. It is one of the most serious and highest incidence
diseases in the northeast of Thailand where the prevalence of
Opisthorchis viverrini (OV) infection is high [1]. An increasing trend
of CCA incidence has been reported in several countries in the
world [2]. This tumor continues to be associated with poor prog-
nosis [1]. At present, surgical resection is the best treatment
available and is a potentially curative therapy for CCA. However,
more than half CCA patients present with advanced, unresectable
malignancy. In addition, the response of CCA to chemotherapy and
radiotherapy is relatively poor [3]. At present, there is no effective
chemotherapy regimen for treating patients with advanced chol-
angiocarcinoma. Therefore, novel and effective therapeutic agents
and more effective medical treatment options are urgently needed.
Much attention has been focused on the targeted medical
therapy. Recently, the status and future perspectives of anti-
angiogenic and growth factor receptor-based pharmacological
approaches for the treatment of biliary tract cancer have been
reported. Among the cellular molecules, the epidermal growth
induction of apoptosis [4].
Medicinal plants which also show promising effects in treat-
ment of cancers are now attracting great attention in the world. The
rhizomes of Zingiber zerumbet Smith are employed as traditional
medicine in relieving stomach ache, as a diuretic, and when
macerated in alcohol are regarded as a tonic and depurative.
Moreover, it is also used as the spice ginger and a novel food factor
for mitigating experimental ulcerative colitis [5].
Zerumbone, a crystalline sesquiterpene, is the major component
in the rhizomes of this plant and is readily available from a wide-
spread natural source. It contains three double bonds; an isolated
one at C6 and two double bonds at C2 and C10 which are part of
a cross conjugated dienone system. It was found that the C2 double
bond appears least hindered due to being furthest from the gem
dimethyl substituents at C9. It has been reported that the X-ray
structure shows that the dienone system lies in a slightly distorted
plane which is perpendicular to that of the isolated double bond [6].
Zerumbone has been studied for various biological activities
because it acts to suppress tumor promoters [7] and to inhibit the
growth of a human leukemia cell line [8]. It is also antiproliferative
and an anti-platelet activating factor [9,10], and has anti HIV
activities [11]. Thus, it is expected that the zerumbone derivatives
* Corresponding author. Tel.: þ66 43 202222x12243; fax:þ66 43 202373.
E-mail address: chayen@kku.ac.th (C. Yenjai).
0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2010.05.029

U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
3795
might show useful biological activities, especially cytotoxicity
O
O
against CCA cell lines.
Kitayama and coworkers [12e15] have reported the isolation
and structural transformation of zerumbone as well as the regio-
and stereochemistry of its derivatives. Due to this chemical infor-
mation and as part of our research program on anti-tumor drugs
from natural products, we planned to synthesize known and new
derivatives of zerumbone and evaluate them for in vitro cytotox-
icity against CCA cell lines including KKU-100 (poorly-differenti-
ated adenocarcinoma), KKU-M139 (squamous cell carcinoma),
KKU-M156 (moderately-differentiated adenocarcinoma), KKU-
M213 (adenosquamous carcinoma) and KKU-M214 (moderately-
differentiated adenocarcinoma) cell lines [16]. It is expected that
this work will find potent compounds for the development of anti-
CCA agents. In addition, to understanding the mode of reaction, the
interaction between the most active compound and the enzymes or
receptor proteins involved with cell cycle and cell growth was
studied using molecular modeling techniques.
Y
a
b
X
1
5 X = NH2
8 Y = NHAc
9 Y = NAcBu
6 X = NHBu
7 X = NHBn
Scheme 2. Reagents and conditions: (a) NH3 or BuNH2 or BnNH2, MeCN, rt, 5 days, 56%
(5), 93% (6), 68% (7); (b) Ac2O, pyridine, 0 ꢀC, 8 h, 92% (8), 85% (9).
Epoxidation of the isolated double bond in zerumbone at C6 and
C7 using mCPBA provided epoxide 12 in 97% yield [13]. This
formation has high regioselectivity due to the high electron density
and the less steric double bond. The epoxide 12 was further treated
with an excess amount of various amines at room temperature
providing corresponding amines 13e15. As in the reaction of zer-
umbone, conjugate additions occurred at the C3 position and no
nucleophilic addition of the remaining double bond (C10) was
observed. As in a previous report, these amino derivatives were
obtained as relatively trans configuration on C2 and C3. The relative
orientation of the amino groups and epoxide oxygen showed that
amination occurred on the same face of the epoxide. The reason for
this observation was the steric repulsion of the two methyl groups
at C2 and C6 of 12 which are located on one face of the ring while
the oxide oxygen lies on the opposite face [14]. Further acetylation
of 13 using Ac2O in pyridine gave corresponding amide 16 in 66%
yield (Scheme 4).
In contrast to the reaction of 1 with various amines, the excess
KCN reacted with zerumbone at 40 ꢀC for 3 days provided conjugate
addition of cyanide ion at both C3 and C10 double bonds [15]. A
mixture of four diastereoisomeric dicyano derivatives 17 was
obtained. It was reported that in the major diastereoisomer (17a),
two cyano groups were located on the same face of the ring while
the two methyl groups at C2 and C6 lie on the opposite face
(Scheme 5) [15].
2. Results and discussion
2.1. Chemistry
Seventeen zerumbone derivatives were successfully prepared
using organic reactions. Reduction of zerumbone (1) using LiAlH4 at
0
ꢀC gave crystalline (ꢁ) zerumbol (2) [12]. Acetylation of this
compound using Ac2O/pyridine afforded 3 while epoxidation of 2
with mCPBA yielded racemic 4 (Scheme 1). The relative stereo-
chemistry of hydroxyl and epoxide groups was trans configuration.
Zerumbone was stirred with excess ammonia, butylamine and
benzylamine at room temperature to provide monoamines 5, 6 and
7, respectively. This conjugated amination yielded a single diaste-
reomer as relatively trans configuration adducts of each monoamine
(Scheme 2). As in a previous report [13], no diamine and no another
diastereomer of products was observed. Michael addition of amines
occurred selectively at the less hindered conjugated double bond
(C2eC3) of zerumbone. Acetylation of amine groups at C3 position
of 5 and 6 using Ac2O/pyridine at 0 ꢀC provided the corresponding
amides 8 and 9, respectively. The reaction of butylamine 6 with
NaBH4 afforded hydroxyamine 10 as a single diastereoisomer in 81%
yield. The reduction of amide 8 with NaBH4 at 0 ꢀC afforded a single
diastereoisomer hydroxyamide 11 in the yield of 54%. The NOESY
experiment showed the correlation of protons at C1 and C2 which
indicated the relatively cis orientation of these positions (Scheme 3).
After treatment of 1 with dimethylamine in the presence of
acetonitrile at room temperature, followed by stirring with excess
KCN at 15 ꢀC, the nitrile derivative 20 was detected [13]. This can be
explained as that conjugate addition of dimethylamine at C3 gave
intermediate 18, while conjugate addition of the cyanide ion at C10
yielded intermediate 19. After the easy elimination of the dime-
thylamino group, cyano 20 was observed as a sole product (Scheme
6). The 1H NMR spectrum showed an olefinic proton at
(H3) and a singlet signal at 1.79 (CH3-12) which indicates the
presence of -unsaturated ketone at C2 and C3. Two doublet
d 5.95e6.50
d
O
1
OH
OAc
a,b
12
2
6
11
10
signals of olefinic protons at H10 and H11 disappeared.
a
b
15
14
3
4
5
7
2.2. Biological activity [17]
8
13
1
2
3
Zerumbone (1) and its derivatives (2e17 and 20) were tested
for their cytotoxicity against CCA cell lines; KKU-100, KKU-M139,
c
OH
NOESY
OH
NOESY
OH
O
H
H
H
H
a
NHBu
NHAc
or
6 or 8
4
Scheme 1. Reagents and conditions: (a) LiAlH4, THF, 0 ꢀC, 1 h, 88%; (b) Ac2O, pyridine,
reflux, 30 min, 74%; (c) mCPBA, EtOAc, rt, 24 h, 15%. Note: The stereochemistry shown
in all schemes are relative configuration.
10
11
Scheme 3. Reagents and conditions: (a) NaBH4, MeOH, 0 ꢀC, 3 h, 81% (10), 54% (11).

3796
U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
O
O
O
against KKU-100 cell line. In order to understand the modes of
action of 5, molecular modeling based on the docking approach was
used to find possible targets of cholangiocarcinoma activity.
Excessive activity of cyclin-dependent kinases (CDKs) has been
observed as one of the mechanisms underlying pathological
hyperproliferation [18]. Consequently, inhibition of CDKs could be
a therapeutic option for treatment of cancer [19]. In addition, since
the epidermal growth factor receptor (EGFR) is found to play an
important role in cell cycle progression, cell proliferation and
induction of apoptosis, it has become an important target in cancer
therapy [20]. In cholangiocarcinoma cells, the expression of CDK-2,
CDK-5 and EGFR in tumor tissues were found in both Thai intra-
hepatic cholangiocarcinoma (ICCs) and Japanese ICCs. Several
reports indicate that the EGFR is frequently overexpressed in CCA
[21]. Therefore, inhibition of CDK-2, CDK-5 or EGFR could provide
control of the cell proliferation of certain cancers. Moreover, the
a
b
c
X
Y
1
O
O
O
12
13 X = NH2
14 X = NHBu
15 X = NHBn
16 Y = NHAc
Scheme 4. Reagents and conditions: (a) mCPBA, EtOAc, rt, 24 h, 97%; (b) NH3 or BuNH2
or BnNH2, MeCN, rt, 5 days, 51% (13), 15% (14), 18% (15); (c) Ac2O, pyridine, 0 ꢀC, 10 h,
66%.
KKU-M156, KKU-M213 and KKU-M214 and their activities are
shown in Table 1. Among these cell lines, KKU-100 was the least
sensitive cell line which showed the highest IC50 value of 25.21 mM
to ellipticine. It was found that zerumbone and alcohol derivatives
showed no cytotoxic activity against all cell lines. Fortunately,
primary amine 5 exhibited cytotoxic activity against KKU-100 cell
expression of GSK-3
b
which plays an important in regulation of
-cat-
tumor cell proliferation through the activation of the Wnt/
b
line with an IC50 value of 16.44
m
M which is about 1.5-fold higher
enin pathway was found in 16/20 (80%) of Thai ICCs and in 16/20
(80%) of Japanese ICCs [22]. Therefore, in the course of our molec-
ular modeling, we selected CDK-2, CDK-5, EGFR and GSK-3 for the
elucidation of the possible characteristics of zerumbone derivatives
in cholangiocarcinoma cells. The template of each protein target
was constructed and validated by redocking with the native ligand.
The chemical structures of these ligands are shown in Table 2 which
summarizes the data and source of these structures [23e26]. The
3D-configuration of four inhibitors obtained from redocking with
the developed protein templates were compared with the crystal-
lographic poses. The results showed a good match of the docked
and the crystallographic binding orientation with RMSD of less
than 2.0 A (Table 3). The validation results show that the con-
structed template is a good model system for predicting ligand
binding orientation and binding affinity.
The docking results using AutoDock 4 indicate that 5 interacted
with CDK-2 through the 2 H-bond at Asp145 with binding energy of
ꢂ6.18 kcal/mol. In contrast, for CDK-5 and GSK-3 complexes it is
seen that hydrogen bonds were not formed (Table 4).
The best binding interactions of 5 was formed in the complex
with EGFR with the lowest binding energy of ꢂ7.24 kcal/mol. The
binding energy of zerumbone, which showed no cytotoxic activity
toxicity than the ellipticine standard (IC50 ¼ 25.21
mM). This
compound also showed cytotoxicity against KKU-M139, KKU-
M156, KKU-M213 and KKU-M214 cell lines with IC50 values
ranging from 51 to 69
against KKU-M213 and KKU-M214 cell lines with IC50 values of
36.61 and 41.55 M, respectively. In contrast, benzylamine 7
mM. Butylamine 6 possessed cytotoxicity
m
showed no cytotoxicity against these five cell lines. It is interesting
to note that the more polar primary amine 5 may play an
important role in cytotoxicity.
Zerumbol (2) showed no cytotoxicity against all CCA cell lines.
But, when hydroxyamine 10 was synthesized and evaluated for
cytotoxicity, it was found that this derivative displayed interesting
results by showing cytotoxic activity against all cell lines with IC50
ꢀ
values ranging from 16 to 35
Epoxide 12 demonstrated weak cytotoxicity to KKU-M214 cell line
with an IC50 value of 52.61 M. From the cytotoxicity results of amine
mM.
m
5 and epoxide 12, we hoped that epoxyamine 13 would exhibit good
results on cytotoxicity testing. However, this compound displayed
strong cytotoxicity only to the KKU-M214 cell line with an IC50 value
of 18.30
mM. Epoxyamine 14 displayed good results to KKU-100 cell
line by showing an IC50 value of 16.52
mM. In addition, this compound
exhibited activity with IC50 values of 48.20, 37.63 and 51.00
against KKU-M156, KKU-M213 and KKU-M214, respectively.
mM
against KKU-100 at 75
mM, with EGFR was also calculated. The
docking result showed that the binding energy of zerumbone was
higher than the standard known inhibitor with the binding energy
of ꢂ6.22 kcal/mol, as a result, the affinity of zerumbone decrease.
Compound 20 showed strong cytotoxicity against the least
sensitive cell line (KKU-100) with an IC50 value of 25.72 M which
m
is nearly equal to the ellipticine standard. This derivative also dis-
played moderate cytotoxicity against the other cell lines with IC50
values ranging from 37 to 56 mM. It was suggested that the nitrile
group at the C10 position was favorable for CCA cell lines. All results
show convincingly that amine, hydroxylamine, epoxyamine and
nitrile groups are essential for cytotoxicity.
Thus, zerumbone showed no antiproliferation activity at 75 mM.
These patterns correlated with antipoliferative activities. For the
binding mode of compound 5, the core structure, sesquiterpene, is
tightly bound to the pocket binding site with a hydrophobic
interaction. The oxygen atom of the carbonyl group in 5 forms
ꢀ
a hydrogen bond with the HeN of Lys721 (bond length 1.9 A). The
interaction between the primary amine and the oxygen atom of the
ꢀ
Asp831 residue has a 1.7 A bond length (Fig. 1). The docking result
2.3. Molecular modeling simulation
suggests that 5 might inhibit the proliferation of cancer via EGFR
inhibition.
From the results of the bioactivity testing, 5 containing an amine
group exhibits the most potent antiproliferative activity, especially
3. Conclusion
O
A series of zerumbone derivatives were synthesized and their in
vitro cytotoxicity against cholangiocarcinoma cell lines were eval-
uated. Compounds 5, 10, 14 and 20 exhibited cytotoxic activity
against all CCA cell lines to different extents, indicating their broad
spectrum of anti-CCA effects. Our results indicate that these five
CCA cell lines, with different histological types, respond differently
to different compounds. This may be due to the variation in func-
tional groups present in the compounds. The chemical structure
NC
CN
a
+
17b-d
1
17a
Scheme 5. Reagents and conditions: (a) excess KCN, MeCN-H2O, 40 ꢀC,
42% (17a).
3 days,

U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
3797
O
O
O
NMe2
NC
NMe2
NC
a
b
1
18
19
20
Scheme 6. Reagents and conditions: (a) Me2NH, MeCN, rt, 5 days; (b) KCN, MeCN-H2O, 15 ꢀC, 2 days, 30% in two steps.
diversity of the four compounds reflects the biological activities.
The presence of amine (5), hydroxylamine (10), epoxyamine (14),
and nitrile (20) groups is believed to play an important role in
potent anticancer activities. Among the tested compounds, 5,
which contained an amine group, exhibited higher potency. The
docking result indicates that 5 may inhibit the proliferation of
cancer through EGFR inhibition. This preliminary investigation is
a good basis for further medicinal chemistry study.
polarity of elution solvents system. The eluents were collected and
monitored by TLC resulting in 15 fractions (F1eF15). The solid in F6
was separated by filtration, recrystallized from CH2Cl2-hexane to
afford zerumbone (1, 15.0 g).
4.3. Preparation of zerumbone derivatives
4.3.1. Preparation of zerumbol (2)
To a solution of compound 1 (503 mg, 2.30 mmol) in dry THF
(5.0 mL) was added LiAlH4 (348 mg, 9.2 mmol) at 0 ꢀC and the
solution was stirred for 1 h. The reaction mixture was quenched
with 10% HCl (10 mL) and extracted with EtOAc (3 times). The
organic layer was combined, washed with water, brine, dried over
anhydrous Na2SO4 and evaporated to dryness. Purification of crude
oil by column chromatography (silica gel, 10% EtOAc:hexane) gave
a white solid of compound 2.
4. Experimental
4.1. General
NMR spectra were recorded on a Varian Mercury plus spec-
trometer operating at 400 MHz (1H) and at 100 MHz (13C). IR
spectra were recorded as KBr disks or thin films, using Perkin Elmer
Spectrum One FT-IR spectrophotometer. Mass spectra were deter-
mined on Micromass Q-TOF 2 hybrid quadrupole time-of-flight (Q-
TOF) mass spectrometer with a Z-spray ES source (Micromass,
Manchester, UK). Melting points were determined on a SANYO
Gallenkamp melting point apparatus and were uncorrected. Thin
layer chromatography (TLC) was carried out on MERCK silica gel 60
F254 TLC aluminum sheet. Column chromatography was done with
silica gel 0.063e0.200 mm or less than 0.063 mm. Preparative layer
chromatography (PLC) was carried out on glass supported silica gel
plates using silica gel 60 PF254 for preparative layer chromatog-
raphy. All solvents were routinely distilled prior to use.
4.3.2. (ꢁ)-(2E,6E,10E)-2,6,9,9-tetramethylcycloundeca-2,6,10-
trienol (2)
Yield 447 mg, 88%; mp 72e73 ꢀC; IR nmax 3290, 2955, 1445, 1297,
1076, 969 cmꢂ1; 1H NMR (CDCl3)
d
5.56 (1H, dd, J ¼ 16.1, 7.2 Hz, H-
11), 5.25 (1H, d, J ¼ 16.1 Hz, H-10), 5.22 (1H, t, J ¼ 7.5 Hz, H-3), 4.83
(1H, dd, J ¼ 10.1, 4.0 Hz, H-7), 4.63 (1H, d, J ¼ 7.2 Hz, H-1), 2.16e2.24
(1H, m, H-4), 2.04e2.14 (3H, m, H-4 and 2H-5), 2.01 (1H, dd,
J ¼ 13.7, 2.8 Hz, H-8), 1.80 (1H, dd, J ¼ 13.7, 3.8 Hz, H-8), 1.67 (3H, s,
CH3-12), 1.44 (3H, s, CH3-13), 1.08 (3H, s, CH3-14 or 15), 1.06 (3H, s,
CH3-15 or 14); 13C NMR (CDCl3)
d 141.8 (C2), 139.3 (C10), 133.0 (C6),
131.3 (C11), 124.9 (C7), 124.8 (C3), 78.6 (C1), 41.9 (C8), 39.1 (C5), 37.1
(C9), 29.4 (CH3-14 or 15), 24.2 (CH3-15 or 14), 22.9 (C4), 15.0 (CH3-
13), 12.7 (CH3-12); HRMS m/z calcd mass for C15H24O 243.1725
[M þ Na]þ, found 243.1725.
4.2. Extraction and purification of zerumbone (1)
Dried rhizomes of Z. zerumbet Smith (4.0 kg) were ground into
powder and then extracted with EtOAc (3 ꢃ 5 L) at room temper-
ature. Removal of solvent under reduced pressure gave crude EtOAc
extract (32.8 g) which was further subjected to column chroma-
tography (CC) on silica gel 60 (500 g) and subsequently eluted with
three solvents (hexane, EtOAc and MeOH) by gradually increasing
4.3.3. Preparation of compound 3
Compound 2 (35 mg, 0.16 mmol) was refluxed with acetic
anhydride (1.0 ml) in the presence of pyridine (1.0 ml) for 30 min.
The water was added and the mixture was extracted with CH2Cl2 (3
Table 1
Cytotoxicity of all compounds against cholangiocarcinoma cell lines.a
Compound
IC50(mM)
KKU-100
KKU-M139
KKU-M156
KKU-M213
KKU-M214
Zerumbone
NR
NR
NR
NR
NR
5
6
10
12
13
14
17aed
20
Other
Ellipticine
16.44 ꢁ 0.59
63.26 ꢁ 5.48
69.04 ꢁ 5.40
51.88 ꢁ 2.25
36.61 ꢁ 0.65
27.19 ꢁ 1.29
NR
59.65 ꢁ 7.39
41.55 ꢁ 0.58
16.05 ꢁ 6.06
52.61 ꢁ 3.88
18.30 ꢁ 0.39
51.00 ꢁ 0.78
66.52 ꢁ 3.63
54.50 ꢁ 1.5
NR
NR
NR
NR
35.33 ꢁ 5.89
NR
NR
19.63 ꢁ 1.26
26.34 ꢁ 4.0
NR
NR
NR
NR
NR
NR
NR
16.52 ꢁ 0.84
48.20 ꢁ 1.85
37.63 ꢁ 2.4
NR
NR
NR
25.72 ꢁ 2.60
NR
25.21 ꢁ 0.2
55.88 ꢁ 4.4
37.30 ꢁ 3.58
NR
9.34 ꢁ 1.66
45.12 ꢁ 0.89
NR
1.62 ꢁ 0.08
NR
4.5 ꢁ 0.28
1.01 ꢁ 0.02
NR ¼ no response at > 75
mM.
a
Data shown are from triplicate experiments.

3798
U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
Table 2
Table 4
Crystal ligand structures used in this study.
Geometric filters for training set of compound 5.
ꢀ
Ligand
PDB
Resolution A
Kinase
targets
Number of
H-bond
Geometric filters
Ebinding
(kcal/mol)
Amino acid
residue
H-bonding
atom
Distance
(A)
ꢀ
CDK-2
2
Asp145
Asp145
e
O (C]O)
N (NH2)
e
2.211
2.161
e
ꢂ6.18
CDK-5
EGFR
0
2
ꢂ6.06
ꢂ7.24
1DI8
2.20
Lys721
Asp831
e
O (C]O)
N (NH2)
e
1.951
1.718
e
GSK-3
0
ꢂ5.48
1.08 (3H, s, CH3-14 or 15), 1.07 (3H, s, CH3-15 or 14); 13C NMR
(CDCl3) 170.4 (C]O of OAc), 140.8 (C2), 138.7 (C6), 133.2 (C11),
d
127.7 (C7), 127.1 (C10), 124.7 (C3), 79.9 (C1), 41.1 (C8), 39.2 (C5), 37.3
(C9), 29.5 (CH3-14 or 15), 23.8 (CH3-15 or 14), 23.0 (C4), 21.2 (CH3 of
OAc), 15.0 (CH3-13), 13.2 (CH3-12); HRMS m/z calcd mass for
C17H26O2 285.1825 [M þ Na]þ, found 285.1835.
1UNG
2.30
4.3.5. Preparation of compound 4
To a solution of compound 2 (39 mg, 0.18 mmol) in EtOAc
(0.5 ml) was added mCPBA (37.4 mg, 0.22 mmol) at 0 ꢀC. The
solution was stirred for 1 h and then allowed to warm at room
temperature for 24 h. The water was added and the mixture was
extracted with EtOAc (3 times). The organic layer was combined,
washed with water, brine and dried over anhydrous Na2SO4.
Evaporation of solvent gave a crude oil, which was purified by PLC
(25% EtOAc:hexane) to give a white solid of 4.
1M17
2.60
1UV5
2.80
times). The organic layer was combined, washed with water, brine
and dried over anhydrous Na2SO4. Evaporation of solvent gave
a crude oil, which was purified by PLC (40% CH2Cl2:Hexane) to give
a white solid of compound 3.
4.3.4. (ꢁ)-(2E,6E,10E)-2,6,9,9-tetramethylcycloundeca-2,6,10-
trienyl acetate (3)
Yield 31 mg, 74%; mp 70e71 ꢀC; IR nmax 2955, 1737, 1368,
1240 cmꢂ1; 1H NMR (CDCl3)
d
5.54 (1H, dd, J ¼ 15.8, 7.4 Hz, H-11),
5.48 (1H, d, J ¼ 7.4 Hz, H-1), 5.31 (1H, d, J ¼ 15.8 Hz, H-10), 5.26 (1H,
t, J ¼ 7.7 Hz, H-3), 4.84 (1H, dd, J ¼ 10.6, 4.2 Hz, H-7), 2.12e2.20 (3H,
m, 2H-4 and H-8), 2.02e2.10 (2H, m, H-5), 2.06 (3H, s, CH3 of OAc),
1.78e1.84 (1H, m, H-8), 1.64 (3H, s, CH3-12), 1.45 (3H, s, CH3-13),
Table 3
Docking of crystal inhibitors.
Ligand Macromolecule Members in the Ebinding
RMSD from
highest cluster
(kcal/mol) crystal orientation
ꢀ
(A)
I
II
III
IV
CDK-2
CDK-5
EGFR
87
93
67
ꢂ7.32
ꢂ6.53
ꢂ7.69
ꢂ9.14
0.71
0.92
1.67
0.56
GSK-3
150
Fig. 1. Binding orientation of compound 5 on EGFR kinase.

U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
3799
4.3.6. (ꢁ)-[6E,10E]-2,3-epoxy-2,6,9,9-tetramethylcycloundeca-
6,10-dienol (4)
(1H, dd, J ¼ 7.5, 2.7 Hz, H-3), 2.10 (1H, t, J ¼ 12.3 Hz, H-8),
1.92e1.99 (1H, m, H-5), 1.61 (NH), 1.75 (1H, t, J ¼ 12.0 Hz, H-5), 1.68
(1H, dd, J ¼ 12.0, 4.5 Hz, H-8), 1.31 (3H, s, CH3-13), 1.08 (3H, s, CH3-
14 or 15), 1.01e1.05 (1H, m, H-4), 0.96 (3H, d, J ¼ 6.6 Hz, CH3-12),
0.86e0.94 (1H, m, H-4), 0.77 (3H, s, CH3-15 or 14); 13C NMR
Yield 41 mg, 15%; mp 84e85 ꢀC; IR nmax 3418, 2959, 1388, 1268,
1059 cmꢂ1; 1H NMR (CDCl3)
d
5.30 (1H, dd, J ¼ 16.0, 6.9 Hz, H-11),
5.23 (1H, d, J ¼ 16.0 Hz, H-10), 4.95 (1H, dd, J ¼ 10.7, 3.7 Hz, H-7),
3.57 (1H, d, J ¼ 7.5 Hz, H-1), 2.46 (1H, dd, J ¼ 9.9, 4.3 Hz, H-3),
2.20e2.32 (2H, m, H-4 and H-5), 2.00e2.12 (2H, m, H-5 and H-8),
1.82 (1H, br d, J ¼ 13.7 Hz, H-8), 1.55 (3H, s, CH3-13), 1.36e1.44 (1H,
m, H-4), 1.32 (3H, s, CH3-12), 1.12 (3H, s, CH3-14 or 15), 1.08 (3H, s,
(CDCl3)
d 203.1 (C1), 151.4 (C10), 140.0 (C-Ar), 137.9 (C-Ar), 137.9
(C6), 128.6 (C-Ar), 128.4 (C-Ar), 127.9 (C11), 127.3 (C-Ar), 121.9 (C7),
56.0 (C3), 50.5 (C10), 47.4 (C2), 41.3 (C8), 39.7 (C9), 37.5 (C5), 29.0
(CH3-14 or 15), 28.9 (C4), 22.7 (CH3-15 or 14), 16.4 (CH3-13), 6.0
(CH3-12); HRMS m/z calcd mass for C22H31NO 348.2303
[M þ Na]þ, found 348.2303.
CH3-15 or 14); 13C NMR (CDCl3)
d 142.7 (C10), 132.0 (C6), 125.7 (C7),
125.2 (C11), 81.6 (C1), 65.6 (C2), 60.5 (C3), 40.2 (C8), 36.4 (C9), 36.1
(C5), 30.9 (CH3-14 or 15), 29.8 (CH3-15 or 14), 24.5 (C4), 15.0 (CH3-
13), 10.6 (CH3-12); HRMS m/z calcd mass for C15H24O2 237.1849
[M þ H]þ, found 237.1848.
4.3.11. Preparation of compounds 8 and 9
Compound 5 (or 6) (0.34 mmol) reacted with acetic anhydride
(1.0 mL) in the presence of pyridine (1.0 mL) at 0 ꢀC for 8 h. The
water was added and the mixture was extracted with EtOAc (3
times). The organic layer was combined, washed with water, brine
and dried over anhydrous Na2SO4. Evaporation of solvent gave
a crude oil, which was purified by PLC (50% EtOAc:hexane) to give
a white solid of 8 or a yellow solid of 9, respectively.
4.3.7. Preparation of amines 5, 6 and 7
A mixture of compound 1 (50 mg, 0.23 mmol) in MeCN
(1.0 mL) and ammonia (30% in water, 2.0 mL) was stirred for 5
days at room temperature. The solution was concentrated on
a rotary evaporator under reduced pressure. The water was added
and the mixture was extracted with EtOAc (3 times). The organic
layer was combined, washed with water, brine and dried over
anhydrous Na2SO4. Evaporation of solvent gave a crude oil, which
was purified by PLC (10% MeOH:EtOAc) to give a white solid of
compound 5. The reaction of compound 1 with n-butylamine or
benzylamine (excess, 4.0 mL) gave a yellow oil of 6 and a white
solid of 7, respectively.
4.3.12. (ꢁ)-[6E,10E]-3-(N-Acetylamino)-2,6,9,9-tetramethyl-
cycloundeca-6,10-dienone (8)
Yield 92 mg, 92%; mp 199e200 ꢀC; IR nmax 3300, 1694, 1633,
1547 cmꢂ1
;
1H NMR (CDCl3)
d
6.40 (1H, d, J ¼ 16.2 Hz, H-11), 6.14
(1H, d, J ¼ 16.2 Hz, H-10), 5.15 (1H, dd, J ¼ 11.7, 4.2 Hz, H-7), 4.47
(1H, br d, J ¼ 8.3 Hz, H-3), 2.90 (1H, dq, J ¼ 6.7, 2.6 Hz, H-2), 2.20
(1H, t, J ¼ 12.3 Hz, H-8), 1.98 (3H, s, CH3 of NHAc), 1.90e1.96 (1H, m,
H-5),1.85 (1H, dd, J ¼ 13.0, 4.4 Hz, H-8),1.70 (1H, t, J ¼ 12.3 Hz, H-5),
1.36 (3H, s, CH3-13), 1.19 (3H, s, CH3-14 or 15), 1.13 (3H, s, CH3-15 or
14), 0.95e1.00 (2H, m, H-4), 0.95 (3H, d, J ¼ 6.7 Hz, CH3-12); 13C
4.3.8. (ꢁ)-[6E,10E]-3-amino-2,6,9,9-tetramethylcycloundeca-6,10-
dienone (5)
Yield 32 mg, 56%; mp 90e91 ꢀC; IR nmax 3373, 3304, 1677, 1625,
1451, 995 cmꢂ1; 1H NMR (CD3OD)
d
6.13 (1H, d, J ¼ 16.1 Hz, H-10),
NMR (CDCl3) d 202.6 (C1), 170.7 (C]O of NHAc), 152.1 (C10), 136.5
6.08 (1H, d, J ¼ 16.1 Hz, H-11), 5.04 (1H, dd, J ¼ 11.7, 4.4 Hz, H-7),
3.23e3.35 (1H, m, H-3), 2.65 (1H, dq, J ¼ 6.6, 2.9 Hz, H-2), 2.20 (1H,
t, J ¼ 12.4 Hz, H-8), 1.98e2.06 (1H, m, H-5), 1.78e1.90 (2H, m, H-5
and H-8), 1.30 (3H, s, CH3-13), 1.10 (3H, s, CH3-14 or 15), 1.08 (3H, s,
CH3-15 or 14), 0.98e1.02 (2H, m, H-4), 0.90 (3H, d, J ¼ 6.6 Hz, CH3-
(C6),128.4 (C11),123.2 (C7), 50.8 (C2), 50.7 (C3), 41.5 (C8), 40.0 (C9),
37.0 (C5), 28.8 (CH3-14 or 15), 27.4 (C4), 22.9 (CH3 of NHAc), 22.8
(CH3-15 or 14), 16.3 (CH3-13), 6.6 (CH3-12); HRMS m/z calcd mass
for C17H27NO2 300.1939 [M þ Na]þ, found 300.1939.
12); 13C NMR (CD3OD)
d
202.6 (C1), 152.5 (C10), 137.7 (C6), 127.8
4.3.13. (ꢁ)-[6E,10E]-3-(N-Acetylbutylamino)-2,6,9,9-
tetramethylcycloundeca-6,10-dienone (9)
(C11), 122.2 (C7), 54.4 (C2), 52.7 (C3), 41.2 (C8), 39.9 (C9), 37.7 (C5),
29.9 (C4), 28.7 (CH3-14 or 15), 22.7 (CH3-15 or 14),16.2 (CH3-13), 5.7
(CH3-12).
Yield 97 mg, 85%; mp 72e73 ꢀC; IR nmax 3554, 3476, 1693,
1619 cmꢂ1
;
1H NMR (CDCl3)
d
6.62 (1H, d, J ¼ 16.2 Hz, H-11), 6.16
(1H, d, J ¼ 16.2 Hz, H-10), 5.14e5.20 (2H, m, H-3 and H-7),
3.28e3.60 (1H, m, H-10), 3.14e3.23 (1H, m, H-10), 2.80 (1H, dq,
J ¼ 6.7, 2.2 Hz, H-2), 2.22e2.32 (1H, m, H-8), 2.18 (3H, s, CH3 of
NHAc), 1.95 (1H, dd, J ¼ 12.5, 6.5 Hz, H-5), 1.87 (1H, dd, J ¼ 13.2,
4.3 Hz, H-8),1.56e1.66 (2H, m, H-4 and H-5),1.46e1.56 (2H, m, H20),
1.41 (3H, s, CH3-13),1.28e1.36 (2H, tq, J ¼ 14.7, 7.3 Hz, H30),1.26 (3H,
s, CH3-14 or 15), 1.15 (3H, s, CH3-15 or 14), 0.99 (3H, d, J ¼ 6.7 Hz,
CH3-12), 0.94 (3H, t, J ¼ 7.3 Hz, H-40), 0.80e0.90 (1H, m, H-4); 13C
4.3.9. (ꢁ)-[6E,10E]-3-butylamino-2,6,9,9-tetramethylcycloundeca-
6,10-dienone (6)
Yield 62 mg, 93%; IR nmax 3332, 1693, 1674, 1626, 1455,
998 cmꢂ1; 1H NMR (CDCl3)
d
6.12 (1H, d, J ¼ 16.1 Hz, H-10), 5.58 (1H,
d, J ¼ 16.1 Hz, H-11), 5.00 (1H, dd, J ¼ 11.7, 4.6 Hz, H-7), 2.97e2.99
(1H, m, H-3), 2.84e2.92 (1H, m, H-2), 2.60e2.66 (1H, m, H-10),
2.47e2.55 (1H, m, H-10), 2.12 (1H, t, J ¼ 12.3 Hz, H-8),1.95e2.02 (1H,
m, H-5), 1.78e1.81 (2H, m, H-5 and H-8), 1.40e1.50 (2H, m, H-20),
1.25e1.33 (2H, m, H-30), 1.28 (3H, s, CH3-13), 1.10 (3H, s, CH3-14 or
15), 1.08 (3H, s, CH3-15 or 14), 0.90e1.00 (2H, m, H-4), 0.85 (3H, t,
J ¼ 7.3 Hz, H-40), 0.80 (3H, d, J ¼ 6.6 Hz, CH3-12); 13C NMR (CDCl3)
NMR (CDCl3)
d 202.5 (C1), 171.8 (C]O of NHAc), 151.7 (C10), 136.2
(C6),128.8 (C11),123.7 (C7), 54.1 (C3), 51.7 (C2), 45.9 (C10), 41.7 (C8),
40.2 (C9), 37.1 (C5), 33.3 (C20), 29.0 (CH3-14 or 15), 23.1 (CH3-15 or
14), 23.0 (C4), 22.3 (CH3 of NHAc), 20.2 (C30), 16.3 (CH3-13), 13.7
(C40), 8.3 (CH3-12); HRMS m/z calcd mass for C21H35NO2 356.2565
[M þ Na]þ, found 356.2566.
d
203.2 (C1), 152.0 (C10), 138.1 (C6), 127.7 (C11), 122.0 (C7), 58.5
(C3), 48.6 (C2), 46.8 (C10), 41.3 (C8), 40.0 (C9), 38.3 (C5), 31.9 (C20),
28.8 (CH3-14 or 15), 28.7 (C4), 22.7 (CH3-15 or 14), 20.4 (C30), 16.2
(CH3-13), 13.9 (C40), 5.9 (CH3-12).
4.3.14. Preparation of compounds 10 and 11
To a solution of compound 6 (or 8) (0.10 mmol) in MeOH
(1.0 mL) was added NaBH4 (0.15 mmol) at 0 ꢀC and the solution
was stirred for 3 h. The reaction mixture was quenched with 10%
HCl (10 mL) and extracted with EtOAc (3 times). The organic layer
was combined, washed with water, brine and dried over anhy-
drous Na2SO4. Evaporation of solvent gave a crude oil, which was
purified by PLC (100% EtOAc) to give a white solid of compound 10
(or 11).
4.3.10. (ꢁ)-[6E,10E]-3-benzylamino-2,6,9,9-
tetramethylcycloundeca-6,10-dienone (7)
Yield 51 mg, 68%; mp 80e82 ꢀC; IR nmax 3421, 2922, 1687, 1624,
1042, 744, 701 cmꢂ1 1H NMR (CDCl3)
; d 7.26e7.60 (5H, m, Ar-H),
6.05 (1H, d, J ¼ 16.1 Hz, H-10), 5.65 (1H, d, J ¼ 16.1 Hz, H-11), 4.54
(1H, dd, J ¼ 11.8, 4.2 Hz, H-7), 4.02 (1H, d, J ¼ 13.6 Hz, H-10), 3.77
(1H, d, J ¼ 13.6 Hz, H-10), 3.00 (1H, dq, J ¼ 6.6, 2.7 Hz, H-2), 2.92

3800
U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
4.3.15. (ꢁ)-[6E,10E]-3-butylamino-2,6,9,9-
tetramethylcycloundeca-6,10-dienol (10)
and H-5), 0.90 (3H, d, J ¼ 6.7 Hz, CH3-12); 13C NMR (CDCl3)
d 203.1
(C1), 150.8 (C10), 127.7 (C11), 62.0 (C6), 60.7 (C7), 54.3 (C3), 53.2
(C2), 40.2 (C8), 36.6 (C5), 36.1 (C9), 29.4 (CH3-14 or 15), 27.4 (C4),
23.1 (CH3-15 or 14), 16.6 (CH3-13), 5.6 (CH3-12); HRMS m/z calcd
mass for C15H26NO2 252.1964 [M þ H]þ, found 252.1962.
Yield 81%; IR nmax 3428, 2959, 1563, 1413, 1022 cmꢂ1 1H NMR
;
(CDCl3)
d
5.26 (1H, br d, J ¼ 15.3 Hz, H-10), 4.98e5.10 (2H, m, H-7
and H-11), 3.98 (1H, br s, H-1), 2.58e2.70 (1H, m, H-3), 2.40e2.52
(2H, m, H-10), 2.02e2.12 (1H, m, H-8), 1.80e1.94 (3H, m, H-2 and
2H-5), 1.68 (1H, t, J ¼ 13.0 Hz, H-8), 1.54 (3H, s, CH3-13), 1.38e1.46
(2H, m, H-20), 1.20e1.30 (4H, m, 2H-4 and 2H-30), 1.00 (6H, s, 2CH3-
14, 15), 0.88 (3H, d, J ¼ 6.9 Hz, CH3-12), 0.84 (3H, t, J ¼ 7.2 Hz, H-40);
4.3.21. (ꢁ)-[10E]-3-butylamino-6,7-epoxy-2,6,9,9-
tetramethylcycloundeca-10-enone (14)
Yield 15%; mp 76e77 ꢀC; IR nmax 3434,1688,1628,1114, 917 cmꢂ1
;
13C NMR (CDCl3)
d
145.1 (C10), 140.2 (C6), 128.8 (C11), 128.4 (C7),
1H NMR (CDCl3)
d
6.30 (1H, d, J ¼ 16.0 Hz, H-11), 6.25 (1H, d,
79.9 (C1), 60.2 (C3), 50.8 (C2), 44.4 (C10), 44.1 (C8), 42.1 (C9), 39.0
(C5), 34.7 (C4), 31.9 (CH3-14 or 15), 29.4 (CH3-15 or 14), 27.8 (C20),
24.3 (C30), 23.0 (CH3-13), 17.6 (C40), 11.7 (CH3-12); HRMS m/z calcd
mass for C19H35NO 294.2797 [M þ H]þ, found 294.2798.
J ¼ 16.0 Hz, H-10), 2.92e3.00 (2H, m, H-2 and H-7), 2.66e2.74 (2H, m,
H-3 and H-10), 2.44e2.52 (1H, m, H-10), 2.05e2.09 (1H, m, H-5),1.90
(1H, d, J ¼ 13.8 Hz, H-8),1.42e1.50 (2H, m, H-20),1.34e1.41 (3H, m, H-
8 and 2H-30),1.26 (3H, s, CH3-14 or 15),1.15 (3H, s, CH3-15 or 14),1.10
(3H, s, CH3-13), 1.00e1.10 (2H, m, H-4), 1.00e1.05 (1H, m, H-5), 0.93
(3H, t, J ¼ 7.2 Hz, H-40), 0.90 (3H, d, J ¼ 6.4 Hz, CH3-12); 13C NMR
4.3.16. (ꢁ)-[6E,10E]-3-(N-Acetylamino)-2,6,9,9-
tetramethylcycloundeca-6,10-dienol (11)
(CDCl3) d 203.8 (C1),150.3 (C10),127.8 (C11), 62.0 (C6), 60.7 (C7), 60.7
Yield 54%; mp 84e85 ꢀC; IR nmax 3383, 1632, 1290, 1023,
(C3), 47.8 (C2), 47.0 (C10), 40.5 (C8), 37.2 (C5), 36.2 (C9), 32.4 (C20),
29.6 (CH3-14 or 15), 26.2 (C4), 23.3 (CH3-15 or 14), 20.4 (C30), 16.8
(CH3-13), 14.0 (C40), 5.8 (CH3-12); HRMS m/z calcd mass for
C19H34NO2 308.2590 [M þ H]þ, found 308.2593.
978 cmꢂ1; 1H NMR (CDCl3)
d 5.23e5.36 (2H, m, H-10 and H-11), 5.13
(1H, d, J ¼ 8.0 Hz, H-7), 4.19e4.23 (1H, m, H-1), 2.88e2.91 (1H, m,
H-3), 1.90e2.02 (2H, m, H-2 and H-8), 1.89 (3H, s, CH3 of NHAc),
1.77e1.83 (1H, m, H-8), 1.58e1.63 (2H, m, H-5), 1.57 (3H, s, CH3-13),
1.11e1.28 (2H, m, H-4), 1.12 (3H, s, CH3-14 or 15), 1.03 (3H, s, CH3-15
4.3.22. (ꢁ)-[10E]-3-benzylamino-6,7-epoxy-2,6,9,9-
tetramethylcycloundec-10-enone (15)
or 14), 0.91 (3H, d, J ¼ 7.3 Hz, CH3-12); 13C NMR (CDCl3)
d 170.4 (C]
O of NHAc),138.9 (C10), 136.2 (C6), 126.4 (C11), 124.1 (C7), 74.7 (C1),
48.1 (C3), 42.9 (C8), 41.3 (C9), 38.2 (C2), 35.4 (C5), 27.8 (CH3-14, 15),
23.0 (C4), 22.9 (CH3 of NHAc), 18.0 (CH3-13), 8.9 (CH3-12); HRMS m/
z calcd mass for C17H29NO2 302.2096 [M þ Na]þ, found 302.2096.
Yield 18%; mp 80e81 ꢀC; IR nmax 3350, 1686, 1626, 1113, 917, 749,
704 cmꢂ1 1H NMR (CDCl3)
; d 7.28e7.42 (5H, m, Ar-H), 6.16 (1H, d,
J ¼ 16.0 Hz, H-10), 5.84 (1H, d, J ¼ 16.0 Hz, H-11), 4.02 (1H, d,
J ¼ 13.4 Hz, H-10), 3.72 (1H, d, J ¼ 13.4 Hz, H-10), 3.00 (1H, dq, J ¼ 6.6,
3.2 Hz, H-2), 2.80 (1H, dd, J ¼ 9.3, 3.2 Hz, H-3), 2.14 (1H, dd, J ¼ 11.2,
1.6 Hz, H-7), 2.00 (1H, dd, J ¼ 8.6, 8.2 Hz, H-5), 1.74 (1H, d,
J ¼ 13.5 Hz, H-8), 1.20e1.28 (1H, m, H-8), 1.08e1.14 (2H, m, H-4),
1.06 (3H, s, CH3-14 or 15), 1.04 (3H, s, CH3-13), 0.96 (3H, d,
J ¼ 6.6 Hz, CH3-12), 0.90 (3H, s, CH3-15 or 14); 13C NMR (CDCl3)
4.3.17. Preparation of compound 12
To a solution of 1 (39 mg, 0.18 mmol) in EtOAc (0.5 mL) was
added mCPBA (37.4 mg, 0.22 mmol) at 0 ꢀC, the solution was stirred
for 1 h and then allowed to warm at room temperature for 24 h. The
water was added and the mixture was extracted with EtOAc (3
times). The organic layer was combined, washed with water, brine
and dried over anhydrous Na2SO4. Evaporation of solvent gave
a crude oil, which was purified by PLC (20% EtOAc:hexane) to give
a crytalline solid of compound 12.
d
203.8 (C1), 150.1 (C10), 139.7 (C-Ar), 128.7 (C-Ar), 128.4 (C-Ar),
127.7 (C11), 127.5 (C-Ar), 61.7 (C6), 60.4 (C7), 57.2 (C3), 50.3 (C10),
46.4 (C2), 40.7 (C8), 36.5 (C5), 36.0 (C9), 29.6 (CH3-14 or 15), 25.9
(C4), 23.0 (CH3-15 or 14), 16.8 (CH3-13), 5.7 (CH3-12); HRMS m/z
calcd mass for C22H32NO2 342.2433 [M þ H]þ, found 342.2433.
4.3.18. (ꢁ)-[2E,10E]-6,7-epoxy-2,6,9,9-tetramethylcycloundeca-
2,10-dienone (12)
4.3.23. Preparation of compound 16
Acetylation reaction of compound 13 (23 mg 0.08 mmol) was
similar as described above.
Yield 41 mg, 97%; mp 84e85 ꢀC; IR nmax 2963, 1656, 1262, 1119,
971 cmꢂ1; 1H NMR (CDCl3)
d 6.02e6.10 (3H, m, H-3, H-10 and H-11),
2.70 (1H, dd, J ¼ 11.2, 1.5 Hz, H-7), 2.34e2.45 (2H, m, H-4),
2.21e2.27 (1H, m, H-5), 1.89 (1H, d, J ¼ 14.0 Hz, H-8), 1.80 (3H, s,
CH3-12), 1.41 (1H, dd, J ¼ 14.0, 11.2 Hz, H-8), 1.26e1.34 (1H, m, H-5),
1.25 (3H, s, CH3-14 or 15), 1.18 (3H, s, CH3-13), 1.03 (3H, s, CH3-15 or
4.3.24. (ꢁ)-[10E]-3-(N-Acetylamino)-6,7-epoxy-2,6,9,9-
tetramethylcycloundeca-10-enone (16)
Yield 15.6 mg, 66%; mp 205e206 ꢀC; IR nmax 3243, 3072, 1699,
1632, 1551, 1107, 916 cmꢂ1 1H NMR (CDCl3)
; d 6.59 (1H, d,
14); 13C NMR (CDCl3)
d
202.8 (C1),159.4 (C10),147.7 (C3),139.4 (C2),
J ¼ 16.1 Hz, H-11), 6.29 (1H, d, J ¼ 16.1 Hz, H-10), 4.47 (1H, dt, J ¼ 7.5,
3.1 Hz, H-3), 2.96 (1H, dq, J ¼ 6.7, 3.1 Hz, H-2), 2.85 (1H, dd, J ¼ 11.2,
1.7 Hz, H-7), 1.98e2.03 (1H, m, H-5), 2.01 (3H, s, CH3 of NHAc),
1.90e1.95 (1H, m, H-8), 1.36e1.41 (1H, dd, J ¼ 11.2, 2.5 Hz, H-8), 1.34
(3H, s, CH3-13), 1.19e1.28 (1H, m, H-4), 1.15 (3H, s, CH3-14 or 15),
1.12 (3H, s, CH3-15 or 14), 1.00e1.06 (1H, m, H-4), 0.95 (3H, d,
J ¼ 6.7 Hz, CH3-12), 0.95e0.98 (1H, m, H-5); 13C NMR (CDCl3)
128.2 (C11), 62.7 (C7), 61.3 (C6), 42.6 (C8), 38.1 (C5), 35.9 (C9), 29.7
(CH3-14 or 15), 24.6 (C4), 24.0 (CH3-15 or 14), 15.6 (CH3-13), 12.0
(CH3-12); HRMS m/z calcd mass for C15H22O2 257.1517 [M þ Na]þ,
found 257.1517.
4.3.19. Preparation of compounds 13e15
The synthesis of compounds 13e15 was similar as described in
the preparation of compounds 5e7.
d 202.6 (C1), 170.1 (C]O of NHAc), 150.6 (C10), 128.4 (C11), 61.1
(C6), 60.3 (C7), 52.0 (C3), 49.7 (C2), 40.8 (C8), 36.3 (C9), 36.0 (C5),
29.5 (CH3-14 or 15), 24.9 (C4), 23.4 (CH3-13), 23.3 (CH3 of NHAc),
16.8 (CH3-15 or 14), 6.5 (CH3-12); HRMS m/z calcd mass for
C17H27NO3 316.1889 [M þ Na]þ, found 316.1889.
4.3.20. (ꢁ)-[10E]-3-amino-6,7-epoxy-2,6,9,9-
tetramethylcycloundeca-10-enone (13)
Yield 51%; mp 126e127 ꢀC; IR nmax 3382, 1693, 1631, 1009,
914 cmꢂ1
;
1H NMR (CDCl3)
d
6.29 (1H, d, J ¼ 16.0 Hz, H-11), 6.20
4.3.25. Preparation of compounds 17a-d
(1H, d, J ¼ 16.0 Hz, H-10), 3.12e3.17 (1H, m, H-3), 2.62e2.68 (2H, m,
H-2 and H-7), 2.00e2.06 (1H, m, H-5), 1.84 (1H, t, J ¼ 12.5 Hz, H-8),
1.30 (1H, dd, J ¼ 11.3, 2.5 Hz, H-8), 1.20 (3H, s, CH3-14 or 15), 1.06
(3H, s, CH3-15 or 14), 1.04 (3H, s, CH3-13), 0.94e0.98 (3H, m, 2H-4
To a solution of compound 1 (66 mg, 0.30 mmol) in MeCN
(1.0 mL) was added the solution of KCN (63 mg, 0.97 mmol) in H2O
(1.0 mL), the reaction mixture was stirred for 3 days at 40 ꢀC. The
reaction mixture was quenched with H2O (10 mL) and extracted

U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
3801
with EtOAc (3 times). The organic layer was combined, washed
with water, brine and dried over anhydrous Na2SO4 and evaporated
to dryness. Purification of crude oil by PLC (30% EtOAc:Hexane)
gave a white solid of a major diastereomer 17a.
starting cells)] ꢃ 100. Dose-response curves were plotted and
growth inhibition of 50% (IC50) was determined at compound
concentration which results in 50% reduction of total protein
increase in control cells.
4.3.26. (ꢁ)-[6E]-3,10-dicyano-2,6,9,9-tetramethylcycloundeca-6-
enone (17a)
4.5. Investigation the target protein of compounds 1 and 5 by
docking study
Yield 35 mg, 42%; mp 115e116 ꢀC; IR nmax 2962, 2234,
1726 cmꢂ1
;
1H NMR (CDCl3)
d
5.11 (1H, dd, J ¼ 10.5, 4.0 Hz, H-7),
4.5.1. Template preparation
3.36 (1H, dd, J ¼ 19.4, 5.2 Hz, H-11), 3.25e3.28 (1H, m, H-10),
2.83e2.86 (2H, m, H-2 and H-3), 2.50 (1H, dd, J ¼ 19.4, 3.2 Hz, H-11),
2.12e2.22 (3H, m, 2H-8 and H-5), 1.84e1.92 (3H, m, 2H-4 and H-5),
1.45 (3H, s, CH3-13), 1.26 (3H, s, CH3-14 or 15), 1.22 (3H, d, J ¼ 6.8 Hz,
A crystallographic structure of cyclin-dependent kinases 2
(CDK-2) [23], cyclin-dependent kinases 5 (CDK-5) [24], epidermal
growth factor receptor (EGFR) [25] and glycogen synthase kinase 3
(GSK-3) [26] bound to the selective inhibitor 4-[3-hydroxyanilino]-
6,7-dimethoxyquinazoline (ligand I) for CDK-2, 6-phenyl[5H]pyr-
rolo[2,3-b]pyrazine (ligand II) for CDK-5, [6,7-bis(2-methoxy-
ethoxy)quinazoline-4-yl]-(3-ethynylphenyl)amine (ligand III) for
EGFR and 6-bromoinidirubin-30-oxime (ligand IV) for GSK-3 were
selected as the docking template. These structures were obtained
from the Protein Data Bank (http://www.rcsb.org/pdb/) (PDB
codes: 1DI8 for CDK-2, 1UNG for CDK-5, 1M17 for EGFR and 1UV5
for GSK-3, Table 2). In order to prepare the target protein as
a template, ligand and crystallographic water were removed.
Hydrogens and Gasteiger charges were assigned by using Auto-
DockTools (ADT) [27]. Atomic salvation parameters, based on the
Stouten model, and fragmental volumes were added in accordance
with the AutoDock force field [28]. Cubic affinity grid maps for each
atom type in the ligand set (plus an electrostatics map) centered on
CH3-12), 1.16 (3H, s, CH3-15 or 14); 13C NMR (CDCl3)
d 202.2 (C1),
136.1 (C6), 122.1 (C7), 121.9 (CN at C10), 121.6 (CN at C3), 45.9 (C3),
41.3 (C8), 41.1 (C11), 38.9 (C5), 35.9 (C9), 33.5 (C10), 33.3 (C2), 31.9
(CH3-14 or 15), 24.5 (C4), 21.5 (CH3-15 or 14), 16.3 (CH3-13), 11.2
(CH3-12); HRMS m/z calcd mass for C17H24N2O 295.1786 [M þ Na]þ,
found 295.1787.
4.3.27. Preparation of compound 20
Dimethylamine derivative 18 was prepared in the same proce-
dure as described above and used as the starting material for the
further step. A mixture of 18 (50 mg, 0.19 mmol) and KCN
(210.3 mg, 3.2 mmol) in MeCN:H2O (1:0.5 mL) was stirred for 2
days at 15 ꢀC. The solution was concentrated on a rotary evaporator
under reduced pressure. The water was added and the mixture was
extracted with EtOAc (3 times). The organic layer was combined,
washed with water, brine and dried over anhydrous Na2SO4.
Evaporation of solvent gave a crude oil, which was purified by PLC
(50% EtOAc:hexane) to give a white solid of compound 20.
ꢀ
ꢀ
the cavity with dimensions of 45 ꢃ 45 ꢃ 45 A and 0.375 A spacing
between grid points were computed using AutoGrid 4.0 for each
protein model. LennardeJones parameters 12-10 and 12-6 were
used for modeling H-bonds and Van Der Waals interactions,
respectively.
4.3.28. [2E,6E]-10-cyano-2,6,9,9-tetramethylcycloundeca-2,6-
dienone (20)
4.5.2. Template validation
Yield 32 mg, 60% in the final step; mp 104e105 ꢀC; IR nmax 2966,
To validate the constructed templates, the crystal structures of
native ligands including ligands I, II, III and IV were used to redock
with its template, CDK-2, CDK-5, EGFR and GSK-3, respectively.
To prepare the inhibitors to dock to the constructed template, all
hydrogens were added and Gasteiger charges were assigned for all
inhibitors. ADT was used to merge nonpolar hydrogens and define
which bonds were rotatable for each ligand.
2232, 1672, 1461, 1267, 817 cmꢂ1; 1H NMR (CDCl3)
d 5.95e6.05 (1H,
m, H-3), 5.11 (1H, t, J ¼ 5.9 Hz, H-7), 2.88e3.07 (1H, br s, H-11),
2.72e2.80 (1H, m, H-10), 2.53e2.66 (1H, m, H-11), 2.42e2.52 (1H,
m, H-5), 2.24e2.42 (3H, m, 2H-4 and H-5), 2.00e2.22 (2H, m, H-8),
1.79 (3H, s, CH3-12), 1.57 (3H, s, CH3-13), 1.19 (3H, s, CH3-14 or 15),
1.07 (3H, s, CH3-15 or 14); 13C NMR (CDCl3)
d 202.9 (C1), 143.7 (C3),
137.0 (C6), 134.9 (C2), 122.8 (C7), 120.2 (CN at C10), 39.9 (C8), 39.6
(C4), 39.0 (C11), 37.3 (C9), 36.6 (C10), 28.3 (CH3-14 or 15), 25.8 (CH3-
15 or 14), 24.6 (C5), 15.8 (CH3-13), 12.5 (CH3-12); HRMS m/z calcd
mass for C16H23NO 268.1677 [M þ Na]þ, found 268.1677.
AutoDock 4 was employed to perform the docking calculation.
Each ligand was docked to the developed CDK-2, CDK-5, EGFR and
GSK-3 templates by using a Lamarkian genetic algorithm search
(LGA). Due to the stochastic nature of genetic search algorithms,
each ligand was docked in 150 trials. Docking trial for each ligand
was initiated with a randomly generated population of 150 ligand
orientations and the highest affinity configuration was selected
after 2 million energy evaluations had been performed. Standard
AutoDock parameters were used for the genetic algorithm: 2%
point mutation; 80% cross over rate; 6% local search rate.
The resulting ligand configurations, from 150 trials within a 2.0
RMSD (root-mean-square deviation) tolerance of each other, were
grouped together in clusters. The results of the docking experi-
ments were evaluated by calculating the positional RMSD of the
corresponding atoms of each conformation. The final docked
structure, RMSD from the bound crystal structure, docked energy
and predicted free energy of binding were all used to analyze its
interaction with the active site.
4.4. Cytotoxicity assay
Cytotoxic activity of the compounds was determined by the
sulforhodamine B (SRB) assay [17]. Briefly, 190 mL/well of cell
suspensions (0.5e1.0 ꢃ 105 cells/mL) were seeded in 96-well plates
and incubated at 37 ꢀC for 24 h. Then 10 mL/well of each concen-
tration of the compounds was added in triplicate to obtain final
concentration of 0.025, 0.16, 0.8, 4 and 20 mg/well and 0.1% DMSO
was used as the solvent-control wells. The plates were incubated
for 1 h (starting cells) and 72 h at 37 ꢀC. At the end of each exposure
time, the medium was removed. The cells were fixed with 20% (w/
v) trichloroacetic acid (TCA, Sigma, St. Louis, MO, USA) at 4 ꢀC for 1 h
and stained with 0.4% (w/v) SRB (Sigma) dissolved in 1% acetic acid
(Sigma) at room temperature for 30 min. After five times washing
with 1% acetic acid, protein-bound dye was solubilized with 10 mM
Tris base, pH10 (Sigma) and the absorbance (OD) at 510 nm was
determined with an ELISA plate reader (ELX-800; Bio-Tek Instru-
ments, Inc.). Percentage cell viability was calculated as: [(OD
treated cells on day 3eOD starting cells)/(OD control on day 3eOD
5. Molecular docking of 1 and 5 to the constructed protein
templates
To prepare 1 and 5 for docking with four constructed templates,
all hydrogens were added and Gasteiger charges were assigned

3802
U. Songsiang et al. / European Journal of Medicinal Chemistry 45 (2010) 3794e3802
[12] T. Kitayama, A. Furuya, C. Moriyama, T. Masuda, S. Fushimi, Y. Yonekura,
H. Kubo, Y. Kawai, S. Sawada, Tetrahedron: Asymmetry 17 (2006) 2311e2316.
[13] T. Kitayama, T. Yokoi, Y. Kawai, R.K. Hill, M. Morita, T. Okamoto, Y. Yamamoto,
V.V. Fokin, K.B. Sharpless, S. Sawada, Tetrahedron 59 (2003) 4857e4866.
[14] T. Kitayama, K. Yamamoto, R. Utsumi, M. Takatani, R.K. Hill, Y. Kawai,
S. Sawada, Biosci. Biotechnol. Biochem. 65 (2001) 2193e2199.
[15] T. Kitayama, T. Okamoto, R.K. Hill, Y. Kawai, S. Takahashi, S. Yonemori,
Y. Yamamoto, K. Ohe, S. Uemura, S. Sawada, J. Org. Chem. 64 (1999)
2667e2672.
[29]. The ADT was used to merge nonpolar hydrogens and define
which bonds were rotatable. 1 and 5 was docked to the four protein
templates by using a Lamarkian genetic algorithm search. Due to
the stochastic nature of genetic search algorithms, each ligand was
docked in 150 trials. Each docking trial for ligand was initiated with
a randomly generated population of 150 ligand orientations and the
highest affinity orientation was selected after 2 million energy
evaluations had been performed. Standard AutoDock parameters
were used for the genetic algorithm: 2% point mutation; 80% cross
over rate; 6% local search rate [28]. The resulting ligand orientations
[16] (a) N. Tepsiri, L. Chaturat, B. Sripa, W. Namwat, S. Wongkham,
V. Bhudhisawasdi, W. Tassaneeyakul, World J. Gastroenterol. 11 (2005)
2748e2753;
(b) B. Sripa, S. Leungwattanawanit, T. Nitta, C. Wongkham, V. Bhudhisawasdi,
A. Puapairoj, C. Sripa, M. Miwa, World J. of Gastroenterol 11 (2005)
3392e3397;
(c) P. Thuwajit, W. Chawengrattanachot, C. Thuwajit, B. Sripa, F.E.B. May,
B.R. Westley, N. Tepsiri, A. Paupairoj, S. Chauin, Hepatol. Res. 37 (2007)
295e304.
ꢀ
from 150 trials with a 2.0 A RMSD tolerance of each other were
grouped together as clusters. The final docked structure, docked
energy and predicted free energy of binding were used for the
analysis of its interaction with the active site.
[17] P. Skehan, R. Storeng, D. Scdiero, A. Monks, J. McMahon, D. Vistica, J.T. Warren,
H. Bokesch, S. Kenney, M.R. Boyd, J. Natl, Cancer Inst. 82 (1990) 1107e1112.
[18] H. Lin, J.L. Juang, P.S. Wang, J. Biol. Chem. 279 (2004) 29302e29307.
[19] Y.M. Ahn, L. Vogeti, C.J. Liu, H.K.R. Santhapuram, J.M. White, V. Vasandani, L.A.
Mitscher, G.H. Lushington, P.R. Hanson, D.R. Powell, R.H. Himes, K.F. Roby,
Q. Ye, G.I. Georg, 15 (2007) 702e13.
[20] G.Y. Gwak, J.H. Yoon, C.M. Shin, Y.J. Ahn, J.K. Chung, Y.A. Kim, T.Y. Kim, H.
S. Lee, J. Cancer Res. Clin. Oncol 131 (2005) 649e652.
[21] N. Jinawath, Y. Chamgramol, Y. Furukawa, K. Obama, T. Tsunoda, B. Sripa,
C. Pairojkul, Y. Nakamura, Hepatology 44 (2006) 1025e1038.
[22] K. Lim, C. Han, L. Xu, K. Isse, A.J. Demetris, T. Wu, Cancer Res. 68 (2008)
553e560.
Acknowledgements
The authors are grateful to Associate Professor Dr. Banchob Sripa
for providing cholangiocarcinoma cell lines. Thai Research Fund
(DBG5180066) and The Center for Innovation in Chemistry (PERCH-
CIC) are gratefully acknowledged.
References
[23] L. Shewchuk, A. Hassell, B. Wisely, W. Rocque, W. Holmes, J. Veal, L.F. Kuyper,
J. Med. Chem. 43 (2000) 133e138.
[24] M. Mapelli, L. Massimiliano, C. Crovace, M.A. Seeliger, L.H. Tsai, L. Meijer,
A. Musachio, J. Med. Chem. 48 (2005) 671e679.
[1] B. Sripa, C. Pairojkul, Curr. Opin. Gastroenterol. 24 (2008) 349e356.
[2] K.A. McGlynn, R.E. Tarone, H.B. El-Serag, Cancer Epidemiol. Biomarkers Prev.
15 (2006) 1198e1203.
[3] L.K. Sampson, S.M. Vickers, W. Ying, J.O. Phillips, Cancer Res. 57 (1997)
1743e1749.
[4] M. Höpfner, D. Schuppan, H. Scherübl, World Gastroenterol. 14 (2008)
7021e7032.
[5] N.R. Fransworth, N. Bunyapraphatsara, Thai Medicinal Plants. Prachachon,
Bangkok, Thailand, 1992, pp. 261e3.
[6] S.R. Hall, S. Nimgirawath, C.L. Roston, A. Sittatrakul, S. Thadaniti,
N. Thirasasana, A.H. White, Aust. J. Chem. 34 (1981) 2243e2247.
[7] A. Murakami, D. Takahashi, T. Kinoshita, K. Koshimizu, H.W. Kim, A. Yoshihiro,
Y. Nakamura, S. Jiwajinda, J. Terao, H. Ohigashi, Carcinogenesis 23 (2002) 795e802.
[8] G.C. Huang, T.Y. Chien, C.C. Wang, Planta. Med. 71 (2005) 219e224.
[9] S. Alwi, S. Sakinah, M. Nallappan, A. Pihie, L. Hawariah, Malaysian J. Biochem.
Mol. Biol. 15 (2007) 19e23.
[10] I. Jantan, I.A.A. Rafi, J. Jalil, Phytomedicine 12 (2005) 88e92.
[11] J.R. Dai, J.H. Cardellina, J.B. McMahon, M.R. Boyd, Nat. Prod. Lett. 10 (1997)
115e118.
[25] J. Stamos, M.X. Sliwkowdki, C. Eigenbrot, J. Biol. Chem. 277 (2002)
46265e46272.
[26] L. Meijer, A.L. Skaltsounis, P. Magiatis, P. Polychronopoulos, M. Knockaert,
M. Leost, X.P. Ryan, C.A. Vonica, A. Brivanlou, R. Dajani, C. Crovace,
C. Tarricone, A. Musacchio, S.M. Roe, L. Pearl, P. Greengard, Chem. Biol. 10
(2003) 1255e1266.
[27] S.J. Weiner, P.A. Kollman, D.A. Case, U.C. Singh, C. Ghio, G. Alahona, S. Profeta,
P. Weiner, J. Am. Chem. Soc. 106 (1984) 765e784.
[28] [a] G.M. Morris, D.S. Goodsell, R.S. Halliday, R. Huey, W.E. Hart, R.K. Belew, A.
J. Olson, J. Comput. Chem. 19 (1998) 1639e1662;
(b) P.F.W. Stouten, C. Froemmel, H. Nakamura, C. Sander, Mol. Simulation 10
(1993) 97e120;
(c) G.M. Morris, D.S. Goodsell, R. Huey, A.J. Olson, J. Comput. Aided. Mol. Des.
10 (1996) 293e304.
[29] J. Gasteiger, M. Marsili, Tetrahedron 36 (1980) 3219e3228.