Syntheses and QSAR studies of benzylimidazole derivatives and benzylcarbazole as potential aromatase inhibitors

Article abstract of DOI:10.14233/ajchem.2014.16016

In this study, in order to explore new structures and chemical entities as aromatase inhibitors, benzylcarbazole, 12 benzylimidazole derivatives with different substituents on both phenyl and imidazole rings were synthesized and their aromatase inhibitory were evaluated with fluorescent substrate detection method. The results showed that the compounds with carboxyl and ester groups in phenyl ring show better inhibitory activity. The introduction of alkyl groups in imidazole may improve the aromatase inhibitory activity. Most of the compounds were more potent than aminoglutethimide and tamoxifen. 2-[2-{(2-ethyl-4-methyl- 1H-imidazol-1-yl)methyl}phenyl]acetic acid and benzylcarbazole have the highest bioactivities with IC50 values of 6.19 μM and 2.72 μM respectively. A meaningful QSAR model with LOF = 0.00359, R² = 0.9914, Adj-R ² = 0.9853, R²cv = 0.9639, F = 161.8, was constructed with genetic functional algorithm using discovery studio 2.1 package.

Full text of DOI:10.14233/ajchem.2014.16016

Asian Journal of Chemistry; Vol. 26, No. 8 (2014), 2381-2388
ASIAN JOURNAL OF CHEMISTRY
http://dx.doi.org/10.14233/ajchem.2014.16016
Syntheses and QSAR Studies of Benzylimidazole Derivatives
and Benzylcarbazole as Potential Aromatase Inhibitors
1,*
1
1
1
1
2
3
1
1
Y. DAI , Y. XIAO , Q. WANG , S. WEI , X. ZHANG , Z. MA , H. ZHENG , M. HOU and T. ZHANG
1Key Laboratory of Industrial Fermentation Microbiology (Tianjin University of Science & Technology), Ministry of Education, College of
Bioengineering, Tianjin University of Science and Technology, Tianjin 300457, P.R. China
2Department of Ocean Science and Engineering, Zhejiang University, Hangzhou 310058, P.R. China
3School of Life Science and Technology, China Pharmaceutical University, Nanjing 210009, P.R. China
*Corresponding author: Fax: +86 22 60602298; Tel: +86 22 60601265; E-mail: yjdai@126.com
Received: 8 July 2013;
Accepted: 1 October 2013;
Published online: 15 April 2014;
AJC-15031
In this study, in order to explore new structures and chemical entities as aromatase inhibitors, benzylcarbazole, 12 benzylimidazole
derivatives with different substituents on both phenyl and imidazole rings were synthesized and their aromatase inhibitory were evaluated
with fluorescent substrate detection method. The results showed that the compounds with carboxyl and ester groups in phenyl ring show
better inhibitory activity. The introduction of alkyl groups in imidazole may improve the aromatase inhibitory activity. Most of the
compounds were more potent than aminoglutethimide and tamoxifen. 2-[2-{(2-ethyl-4-methyl-1H-imidazol-1-yl)methyl}phenyl]acetic
acid and benzylcarbazole have the highest bioactivities with IC50 values of 6.19 µM and 2.72 µM respectively. A meaningful QSAR model
with LOF = 0.00359, R2 = 0.9914, Adj-R2 = 0.9853, R2cv = 0.9639, F = 161.8, was constructed with genetic functional algorithm using
discovery studio 2.1 package.
Keywords: Benzylimidazole, Benzylcarbazole, Aromatase inhibitors, QSAR.
of enzyme present in the ovaries of premenopausal women, in
the placenta of pregnant women and in the peripheral adipose
tissues of postmenopausal women and of men7.
INTRODUCTION
Cancer is the leading cause of death among women
between the ages of 30 and 54, with breast and uterine cancers
comprising 10 % and 28 %, respectively, of all cancers in
females per year1. Excluding skin cancers, breast cancer is the
most common malignancy among women, accounting for
nearly 1 in 3 cancers diagnosed among women in the United
States and it is the second leading cause of cancer death among
women2. Although the rate of postmenopausal breast cancer
were ten fold higher in Americans than in Chinese, whereas
the rates for Chinese-Americans were intermediate3, compared
with 2000 years. There are 470 thousands more new breast
cancer cases and 130 thousands more deaths from breast cancer
in 2005 in China4. Approximately 2 in 3 of postmenopausal
breast cancer patients have estrogen-dependent breast cancer,
which includes estrogen receptors and requires estrogen for
tumor growth. Estradiol is the most potent endogenous estrogen.
Consequently, inhibition of estrogen biosynthesis by means
of selective aromatase inhibitors is a potentially useful thera-
peutic option in hormone-sensitive breast cancer5. Estradiol is
biosynthesized from androgens by the cytochrome P450
enzyme called "aromatase"(CYP19)6, with the highest levels
Furthermore, expression of aromatase is the highest in or
near breast tumor sites8,9. Thus, aromatase is a key drug target
for the treatment of breast cancer. Aromatase inhibitors have
emerged over the last many years as modulators of the growth-
stimulatory effects of estrogens in estrogen-dependent breast
cancer10.Aromatase inhibitors are divided into two categories:
steroidal aromatase inhibitors and nonsteroidal aromatase
inhibitors. Steroidal inhibitors have been developed to now,
such as exemestane, build upon the basic androstenedione
nucleus and incorporate substituents at varying positions on the
steroid11. Most of aromatase inhibitors belong to nonsteroidal
aromatase inhibitors, such as tamoxifen (which has been used
as the standard anti-estrogen in treating such tumors12),
aminoglutethimide, anastrozole, letrozole, etc. (Scheme-I).
Now anastrozole and letrozole are approved aromatase
inhibitors for the treatment of metastatic estrogen-dependent
breast cancer13-17. In spite of their clinical success, negative
side effects and partial selectivity of existing aromatase inhibi-
tors are significant issues which must be addressed.Aromatase
inhibitors are associated with osteoporosis, reproductive

2382 Dai et al.
Asian J. Chem.
O
NC
CN
H
N
N
O
NH
H
H
N
CN
CN
N
NH
O
2
N
N
O
N
O
Aminoglutethimide
Exemestane
Anastrozole
Letrozole
Tamoxifen
Scheme-I: Commonly used drugs for the treatment of breast cancer
problems and androgenic side effects. These factors in combi-
nation necessitate the development of more selective new
aromatase inhibitors for the treatment of breast cancer and up
to now the search for potent and selective aromatase inhibitors
still remains an attractive subject18, Moreover, alternative
strategies such as development of single agents against multiple
drug targets19,20 are being evaluated from different research
groups to deal with the complexity and multiplicity of factors
involved in the development of hormone-dependent breast
cancer. Various techniques were applied to search for the new
aromatase inhibitors, including using the existing drugs as
templates21, structural modification of natural products22, high-
throughput docking23 or structure-guided drug design24, etc.
It was found that most of the non-steroidal aromatase inhi-
bitors are composed of 3 or 4 aromatic rings25. However, some
compounds with simple structure and comparatively fewer
aromatic rings also behave good aromatase inhibitory activities.
It has been reported that inhibitory activities (IC50) of CHEMBL
349822 (with 2 aromatic rings) toward human aromatase
enzyme was 0.368 µM26. These aromatase inhibitors have
advantage of higher activity, relatively simple structure and
easier to synthesize and then we were interested in one group
of these aromatase inhibitors (Scheme-II). In particular, the
stereochemically defined functionalized phenyl ring is essential
for potent activity, so researchers added some substituents to
the phenyl ring to improve the activity of these aromatase
inhibitors26-28. In this research, we intend to change the structure
of imidazole ring (such as adding some substituents to the
imidazole ring, or using carbazole ring instead) in order to
obtain the new aromatase inhibitors with high activity and 13
compounds were obtained (Scheme-III).
routinely used in modern drug design to help in understanding
drug-receptor interaction. Relatively large number of quanti-
tative structure-activity relationship (QSAR) studies involving
inhibitors of dipeptidyl peptidase IV, glycogen phosphorylase,
acetylcholinesterase, HIV protease, estrogens and so on have
been reported32-37. Also there are some 2D-3D hybrid QSAR
studies on the steroid and non-steroid aromatase inhibitors have
been conducted with molecular shape analysis descriptors
along with thermodynamic and structural descriptors and also
with selected topological parameters on structurally diverse
datasets of aromatase inhibitors38. Since Ghosh et al.39 published
the crystal structure of CYP19 (Fig. 1), various ligand-receptor
interaction investigations between CYP19 and the small mole-
cule inhibitors using docking process were reported. These
computational techniques have been proved particularly helpful
in the design of novel, more potent inhibitors by revealing the
mechanism of drug-receptor interaction40.
Molecular modeling has become a useful tool for the
characterization of structure-function relationships for a
receptor and for the identification of structural motives of both
ligands and receptors, which may play important roles in the
binding processes30. Quantitative structure-activity relationship
(QSAR) is an important tool to keep the number of synthesized
and tested compounds at a minimum in the process of develop-
ment of new drugs31. Nowadays, three-dimensional (3D) quant-
itative structure activity relationship (3D-QSAR) techniques,
such as comparative molecular field analysis (CoMFA) are
Fig. 1
In this study, 13 compounds which contain 12 benzyl
imidazoles derivatives and benzyl carbazole were synthesized
and their aromatase inhibitory activities were determined with
molecular fluorescence analysis method41.A QSAR model was
built on the basis of the synthesized compounds and some
CN
N
N
N
N
N
CN
N
N
O N
N
CHEMBL 349822
2
Br
Scheme-II: Some aromatase inhibitors with the structure of benzyl-imidazoles 26, 29

Vol. 26, No. 8 (2014)
Syntheses and QSAR Studies of Benzylimidazole Derivatives and Benzylcarbazole 2383
O
O
N
O
NH2
O
OH
N
Cl
O
N
N
N
N
O
1
2
3
4
C
CONH2
COOCH3
COOCH3
N
N
N
N
N
N
N
COOC2H
5
N
8
7
5
6
O
O
O
O
N
N
N
N
N
N
N
N
12
10
11
9
N
N
N
13
14
Scheme-III: Aromatase inhibitors synthesized in this paper
reported benzyl imidazole derivatives, also molecular docking
for some ligands with higher bioactivities to the CYP19 binding
site was conducted.
to give the phenyl acetate derivative (compound 1) in 85.67 %
yield42, which was reacted further with imidazole, carbazole
in CHCl3 under K2CO3 basic condition with 10 h reflux to
form compound 9 in 73.22 % yield. In a similar fashion, comp-
ound 2 was obtained from compound 1 via the K2CO3 in CHCl3
by reaction with 2-ethyl-4-methyl-1H-imidazole. The primary
ester group in compound 2 was converted to amide 4 via nucleo-
philic substitution with NH3 using ammonia water at room
temperature to give a yield of 53.67 %. Compound 3was obtained
by hydrolysis of ester bond of compound 2. Similarly, compound
1 was also converted to compound 5 in 67.55 % yield by
reaction with 4-imidazole carboxylate. The substitution of the
chlorine group of compound 1 with 2-methyl imidazole led to
the formation of compound 6 and compound 7 was obtained
by ammonia substitution of the ester group in compound 6. The
nitrile group of compound 8 was formed by the dehydration
reaction of compound 7 with POCl3 in ethyl acetate in 52.82 %
yield.
EXPERIMENTAL
It is well known that benzyl imidazole skeleton was easily
obtained by the reaction of benzyl halide and imidazole at
room temperature and N-substituted imidazole was easily
obtained by the reaction of halogenated hydrocarbons and
imidazole or carbazole. In order to avoid the substitution
reaction taken place at the para-position of the benzene ring,
3-isochromanone was adopted as the source regent to prepare
the ortho-substituted benzyl halide (compound 1).All reaction
for the object compounds are not only easy to operate, but
also have less byproducts, which makes it easy for the purifi-
cation of the products.
Synthesis of compounds 2-9: The lactonic ring of 3-
isochromanone was opened with SOCl2 under basic conditions

2384 Dai et al.
Asian J. Chem.
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
(d, 2H, ArH + imidazoleH), δ7.338 (m, 3H, ArH), δ7.691 (s,
1H, ArH), δ7.812 (s, 1H, imidazoleH).
compound 2: δ1.105 (t, 3H, J = 7.6 Hz, CH2CH3), δ2.063 (s,
3H, imidazole-CH3), δ2.490 (m, 2H, CH2CH3), δ3.623 (s, 3H,
OCH3), δ3.802 (s, 2H, CH2CO), δ5.086 (s, 2H, imidazole-
CH2), δ6.41 (s, 1H, imidazoleH), δ6.632 (m, 2H,ArH), δ7.237
(m, 2H, ArH).
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 11: δ2.349 (s, 3H, imidazole-CH3), δ5.053 (s, 2H,
ArCH2-imidazole), δ6.847 (d, 1H, J = 1.2 Hz, imidazoleH),
δ6.966 (d, 1H, J = 1.2 Hz, imidazoleH), δ7.065 (d, 2H, J =
6.4 Hz, ArH), δ7.333 (m, 3H, ArH).
1H NMR (D2O, ppm, 400 MHz ) of compound 3: δ1.100
(t, 3H, J = 7.6 Hz, CH2CH3), δ2.083 (s, 3H, imidazole-CH3),
δ2.536 (m, 2H, CH2CH3), δ3.503 (s, 2H, CH2CO), δ4.703 (s,
2H, imidazole-CH2), δ6.556(s, 1H, imidazoleH), d7.216 (m,
4H, ArH).
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 12: δ1.238 (t, 3H, J = 7.6 Hz, imidazole-CH2CH3),
δ2.181 (s, 3H, imidazole-CH3), δ2.595 (m, 2H, imidazole-
CH2CH3), δ4.965 (s, 2H, ArCH2-imidazole), δ6.508 (s, 1H,
imidazoleH), δ7.041 (d, 2H, J = 6.8 Hz, ArH), δ7.304 (m, 3H,
ArH).
1H NMR (DMSO, ppm, relative to TMS, 400 MHz) of
compound 4: δ1.105 (t, 3H, J = 7.6 Hz, CH2CH3), δ2.058 (s,
3H, imidazole-CH3), δ2.504 (m, 2H, CH2CH3), δ3.497 (s, 2H,
CH2CO), δ5.164 (s, 2H, imidazole-CH2), δ6.514 (d, 1H, J =
7.2 Hz, imidazoleH), δ6.643 (d, 1H, J = 0.8 Hz ArH), δ6.985
(s, 1H, NH), δ7.234 (m, 3H, ArH + NH), δ7.538(s, 1H, ArH).
1H NMR (DMSO, ppm, 400 MHz) of compound 5:
δ1.169 (t, 3H, OCH2CH3), δ3.628 (s, 3H, COOCH3), δ3.869
(s, 2H,COOCH2Ar), δ4.171 (m, 2H, imidazole-COOCH2),
δ5.555 (s, 2H, imidazole-CH2Ar), δ6.559 (m, 1H, imidazoleH),
δ7.247 (m, 3H, ArH), δ7.740 (δ, 2H, J = 0.8 Hz,ArH), δ7.990
(s, 1H, imidazoleH).
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 13: δ5.129 (s, 2H, ArCH2-imidazole), δ6.9261 (s,
1H, imidazoleH), δ7.173 (m, 3H, Ar + imidazoleH), δ7.370
(m, 3H, ArH), δ7.560 (s, 1H, imidazoleH).
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 14: δ5.557 (s, 2H, ArCH2-carbazole), δ7.188 (d,
2H, J = 6 Hz,ArH), δ7.291 (m, 5H,ArH + carbazoleH), δ7.404
(d, 2H, J = 8 Hz, carbazoleH), δ7.469 (m, 2H, carbazole H),
δ8.175 (d, 2H, J = 8 Hz, carbazoleH).
Inhibitory activities: The inhibitory activities of the target
compounds were evaluated by a modification of the method
of Luo et al.41.
1H NMR (DMSO, ppm, 400 MHz) of compound 6:
δ2.169 (s, 3H, imidazole-CH3), δ3.627 (s, 3H, COOCH3),
δ3.824 (s, 2H, ArCH2CO), δ5.176 (s, 2H, ArCH2-imidazole),
δ6.554 (d, 1H, J = 6.8 Hz, imidazoleH), δ6.813 (d, 1H, J =
1.2 Hz, imidazoleH), δ6.982 (d, 1H, ArH), δ7.247 (m, 3H,
ArH).
Molecular docking: The pdb data about the crystal struc-
ture of CYP19 complexing with androstenedione (EC:
1.14.14.1, 3EQM.pdb)39 was obtained from the RCSB protein
data bank (http://www.pdb.org). The molecular docking of two
typical compounds with the highest inhibitory activities in
benzyl imidazoles and benzyl carbazole synthesized (comp-
ounds 3 and 14) was carried out using the CDocker protocol
in accelrys discovery studio 2.1 software package (DS 2.1)44.
Initially both the ligands and the enzyme (aromatase) were
pretreated. The 3D structures of compounds 3 and 14 were
generated with Cambridge Chem Bio Office 200845 and
optimized with AM1 method. For enzyme preparation, the
hydrogen atoms were added with the pH of the protein in the
range of 6.5-8.5.
Model development: In order to give a systematic evalua-
tion on benzyl imidazoles and benzyl carbazole as aromatase
inhibitors and to explore more potent and selective aromatase
inhibitors, a QSAR model was built using some 2D and 3D
physiochemical properties as candidates of the descriptors
which values were obtained from the calculation using the
protocol of calculate molecular properties in DS 2.1. The
compounds physiochemical properties include 2D (AlogP,
Molecular_Weight, Num_H_Acceptors, Num_H_Donors,
Num_Rotatable Bonds, Molecular_Surface Area, topological
descriptors such as BIC, SC_1, CIC, E_ADJ_equ, IAC_Total,
IC and SIC, etc.) and 3D (Dipole, Jurs descriptors, shadow
indices and Molecular_Volume, etc.) parameters. pIC50 (-lgIC50
of the compounds was taken as the dependent variable. For
the development of the QSAR model for the synthesized
compounds 2-14, the statistical techniques of genetic function
algorithm and partial least squares (PLS) in Accelrys DS 2.1
were employed. In this study, all the 12 synthesized benzyl
imidazoles derivatives and benzyl carbazole with definite IC50
values were selected as the model dataset. More than 120
1H NMR (DMSO, ppm, 400 MHz) of compound 7:
δ2.177 (s, 3H, imidazole-CH3), δ3.516 (s, 2H, ArCH2CO),
δ5.241 (s, 2H, ArCH2-imidazole), δ6.483 (d, 1H, J = 7.2 Hz,
imidazoleH), δ6.804 (d, 1H, 1.2 Hz, imidazoleH), δ6.993
(d, 2H, J = 1.2Hz, ArH), δ7.231(m, 3H, ArH + NH2), δ7.543
(s, 1H, ArH).
1H NMR (DMSO, ppm, 400 MHz) of compound 8:
δ2.210 (s, 3H, imidazole-CH3), δ4.170 (s, 2H, CNCH2), δ5.271
(s, 2H, ArCH2-imidazole), δ6.548 (d, 1H, J = 7.6 Hz,
imidazoleH), δ6.887 (s, 1H, imidazoleH), δ7.067 (s, 1H,ArH),
δ7.321 (m, 2H, ArH), δ7.475 (d, 1H, J = 7.2 Hz, ArH).
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 9: δ3.621 (s, 2H, ArCH2CO), δ3.675 (s, 3H,
COCH3), δ5.206 (s, 2H, ArCH2-imidazole), δ6.864 (s, 1H,
imidazole H), δ7.013 (d, 1H, J = 5.6 Hz, imidazole H), δ7.083
(s, 1H, ArH), δ7.292 (m, 3H, ArH), δ7.491 (s, 1H, imidazole
H).
Synthesis of compounds 10-14: Benzyl chloride was
easily reacted with imidazoles at ambient temperature in CHCl3
under stirring for 24 h. Compounds 10-13 was obtained in the
yields of 79.16, 75.56, 75.35 and 73.74 %, respectively. Carbazole
derivatives are important pharmaceutical intermediates and
recently are concerned as selective androgen receptor modu-
lators43, however, their aromatase inhibitory activities have not
attracted much attention, thus in this article, compound 14 was
synthesized by the reaction of carbazole with benzyl chloride
in 56.61 % yield.
1H NMR (CDCl3, ppm, relative to TMS, 400 MHz) of
compound 10: δ1.332 (t, 3H, J = 7.6 Hz, OCH2CH3), δ4.292
(m, 2H, OCH2CH3), δ5.539 (s, 2H,ArCH2-imidazole), δ7.195

Vol. 26, No. 8 (2014)
Syntheses and QSAR Studies of Benzylimidazole Derivatives and Benzylcarbazole 2385
physiochemical properties such as AlogP, Molecular_Weight,
Dipole and Jurs descriptors obtained from DS 2.1 as the initial
independent descriptors to construct the QSAR model. For
the system of a small quantity of samples with large parameters,
the number of parameters should be reduced in order to make
the obtained model with statistical meaning. In this study,
genetic functional algorithm was employed to cut down the
number of parameters and optimize the QSAR model. It solves
problems in an analogous way of the organism's evolution.
This method generates a series of potential solutions to a problem
(the population of organisms) and then these solutions are
modified and tested repeatedly until an approximate optimal
solution is found38,46. During the optimizing process of genetic
functional algorithm, Friedman lack-of-fit (LOF) was used as
a scoring function47 to control the model size to resist over-
fitting, which is a problem often encountered in constructing
statistical models. The smoothing factor of LOF was set to
0.5. For a given smoothing factor, the optimization of a QSAR
model was considered to be obtained when descriptors usage
in genetic functional algorithm became constant and
independent of the increasing number of genetic functional
algorithm crossover operations. All the descriptors in the
QSAR trial descriptor pool were used as linear terms during
genetic functional algorithm to generate QSAR models. The
obtained model was further optimized using G/PLS protocol
in DS 2.5 and the correlation coefficient R2 and the adjusted
R2(Adj-R2), were taken as objective functions to select an equa-
tion. Leave-one-out cross-validation (R2cv) was employed to
validate the predictivity of generated QSAR model equation.
RESULTS AND DISCUSSION
In order to examine the effect of substituents in the imida-
zoles on the aromatase ihibitory activity of benzyl imidazoles,
imidazoles with substituents of 2-methyl, 2-ethyl, 2-ethyl-4-
methyl, 4-carboxylate and benzo (carbazole) were synthesized.
Considering the introduction of some substituents in the phenyl
ring may improve the activity, infrequently used groups in
aromatase inhibitors, carboxymethyl ester and its derivatives
such as carboxymethyl, acetonitrile groups, carboxylation
formamide were also considered, the structure of all the synthe-
sized compounds are shown in Schemes IV and V.
The synthetic routes of compounds 1-9 are outlined in
Scheme-IV. In order to check if imidazole was the optimal
activity spacer for aromatase inhibition, benzyl imidazoles with
different substitutional groups of compounds 10-13 were synthe-
sized. The synthetic routes were illustrated in Scheme-V.
Scheme-IV: Synthesis of methyl 2-(2-(chloromethyl) phenyl)acetate and derivatives. Reagents and conditions: (a) SOCl2, CH3OH, -5 °C to 0 °C, 2h,
85.67 %; (b) K2CO3, CHCl3, reflux, 10h, 73.22 %; (c) K2CO3, CHCl3, reflux, 3h, 66.42 %; (d) 5 % NaOH, CH3CH2OH, reflux, 10h, 82.78 %; (e)
NH4OH, CH3CH2OH, rt, 72h, 63.51 %; (f) K2CO3, CHCl3, reflux, 3h, 67.55 %; (g) K2CO3, CHCl3, reflux, 5h, 75.80 %; (h) NH4OH, CH3CH2OH,
rt, 72h, 40.13 % and (i) POCl3, K2CO3, EtOAc, reflux, 12h, 52.82 %

2386 Dai et al.
Asian J. Chem.
O
Cl
O
N
N
HN
NH
O
N
NH
N
O
e
a
N
N
N
11
10
NH
d
HN
c
b
N
N
N
N
N
13
N
12
14
Scheme-V: Syntheses of benzylimidazole, benzylcarbazole and derivatives. Reagents and conditions: (a) K2CO3, CHCl3, reflux, 6h, 75.56 %; (b) K2CO3,
CHCl3, reflux, 8h, 73.74 %; (c) K2CO3, DMF, rt to 40 °C, 8h, 56.41 %; (d) K2CO3, CHCl3, reflux, 8h, 75.35 % and (e) K2CO3, CHCl3, reflux, 5h,
79.16 %
Inhibitory activities: The in inhibitory activities of the
target compounds were evaluated by a modification of the
method of Luo et al.41 and the results are listed in Table-1. It
was found that if no substituent is existed in phenyl ring, the
compounds have less even no inhibitory activity, such as com-
pound 10 (IC50 = 173.72 µM ), 11 (IC50 = 165.09 µM) and 13
(IC50 = 140.96 µM). On the other hand, the introduction of some
polar groups such as -CH2CONH2 (compound 4) and -CH2CN
(compound 8) also have less effect for the improvement of
aromatase activity (IC50 > 155.44 µM). However, the compounds
with carboxyl and ester groups in phenyl ring show better
inhibitory activity (the IC50 values compounds 2 and 3 are
63.52 µM and 6.19 µM respectively). The introduction of alkyl
groups in imidazole may improve the aromatase inhibitory
activity. The inhibitory activity order is that compounds with
2-ethyl-4-methyl-1H-imidazole > compounds with 4-methyl-
1H-imidazole > compounds with 1H-imidazole. For example,
compound 12 (IC50 = 42.94 µM) > compound 11(IC50 = 165.09
µM) > compound 13 (IC50 = 140.96 µM) and compound 6
(IC50 = 6.96 µM) > compound 2 (IC50 = 63.52 µM) > compound
9 (IC50 = 173.72 µM). However, the introduction of ester in
imidazole ring (compounds 5 and 6) has little effect on the
improvement of aromatase inhibitory activity.
Compared with the benzylimidazoles, the carbazole deri-
vative (compound 14) is more potent as aromatase inhibitors.
Compound 14 displayed the most potent inhibitory activity
among all the compounds, with an IC50 of 2.72 µM.
Molecular docking and QSAR: The molecular docking
of two typical compounds with the highest inhibitory activities
in synthesized benzyl imidazoles and benzyl carbazole (comp-
ounds 3 and 14) was carried out using the CDocker protocol
in accelrys discovery studio 2.1 software package42. The poses
TABLE-1
OBSERVED AND PREDICTED AROMATASE INHIBITORY ACTIVITIES, PHYSIOCHEMICAL PROPERTIES OF
DIFFERENT STEROIDAL COMPOUNDS USED FOR THE CONSTRUCTION OF QSAR MODELS
IC50 (Obsa)
(µM)
pIC50 (Predb)
(M)
Compounds
pIC50 (Obsa)
Residual E_ADJ_equ SC_1 Jurs_TASA Shadow_XZ
CIC
63.53
6.19
155.60
25.82
6.97
10.47
189.23
173.78
173.78
194.98
42.95
140.93
2.72
4.197
5.208
3.808
4.588
5.157
4.98
3.723
3.76
3.76
4.207
5.12
-0.01
0.088
0.012
-0.101
-0.002
-0.005
-0.095
0.083
0.081
-0.06
325.212
310.764
310.764
354.413
282.193
268.078
240.215
254.084
254.084
186.117
226.477
160
21
20
20
23
19
18
17
18
18
14
16
13
23
478.76
401.828
444.491
395.107
349.421
355.493
397.365
388.828
384.577
353.371
429.07
59.977
57.194
66.052
67.836
50.184
47
49.001
52.13
60.1
47.043
54.279
43.889
57.057
1
2
0.947
0.947
0.818
1.084
1.265
1.219
1.176
0.868
1.231
0.717
1.801
1.683
3
3.796
4.689
5.159
4.985
3.818
3.677
3.679
3.77
4
5
6
7
8
9
10
3.71
11
4.367
3.851
5.565
4.480
5.490
4.4
3.824
5.551
-0.033
0.027
0.014
12
332.011
482.628
13
14
384
Aminoglutethimide
Tamoxifen
33.11
3.24
a Obs, observed, b Pred, predicted

Vol. 26, No. 8 (2014)
Syntheses and QSAR Studies of Benzylimidazole Derivatives and Benzylcarbazole 2387
TABLE-2
of compounds 3 and 14 within the important amino acid
residues of human placental aromatase are shown in Figs. 2
and 3 (the molecule in pink stick is haem and that in gray stick
is compound 3. The green dash line refers to hydrogen bond).
The results showed that the oxygens in carbonyl of compound
3 coordinated with the Fe(III) in haem molecule. The distances
of O-Fe are 2.304 Å and 2.645 Å respectively. There are also
hydrogen bonds formed between the hydroxyl group in
compound 3 and carbonyl group in ALA306, hydroxyl group
in THR310. However, there are no hydrogen bond formed
between compound 14 and CYP19. It may be the hydrophobic
interaction plays the main role for their interaction. The highest
bioactivity of compound 14 may be attributed to the strong
hydrophobic interaction between them.
INITIAL PARAMETERS FOR GFA OPTIMIZATION
Population
Maximum
generations
Scoring
function
Scaled
LOF
smoothness
parameter
0.50
Mutation
probability
100
5000
Friedman
LOF
0.100
Under this condition, the model with the Friedman LOF
value was acquired as follows:
pIC50 = 17.2938 + 0.124234 × E_ADJ_equ- 1.59464
× SC_1- 0.00726719 × Jurs_TASA- 0.114515
× Shadow_XZ + 2.86407 × < 1.23168 - CIC >
(1)
The sample number is 13 Friedman LOF = 0.00359, R2 =
0.9914,Adj-R2 = 0.9853, R2cv = 0.9639, F = 161.8. The observed
and predicted pIC50 results and the values of physiochemical
properties of the 13 ligands for Eqn. (1) are listed in Table-1.
In our study, R2,Adj-R2, R2cv and F-value were used to evaluate
the regression model. Eqn. (1) can explain 98.33 % of the
variance (Adj-R2) while it could predict 97.06 % of the variance
(R2cv). F > F (a = 0.05) = 4 shows that the model is in the
confidence interval of 95 %. The difference between R2 and
Pred-R2 values is not very high (less than 0.3)34. It can be seen
from Eq. (1) that SC_1, Jurs_TASA, Shadow_XZ and CIC have
negative contribution to the bioactivity of the ligands, however,
only E_ADJ_equ, has the positive effect on the bioactivity of
the ligands. The plot of the observed pIC50 vs. the predicted
data for compounds 2-14 is shown in Fig. 4. The high corre-
lation coefficient (R2 = 0.9914) for the regression of predicted
versus measured pIC50 for compounds 2-14 and all the points
locating near the y = x axis (the solid line) indicated that the
predicted data by this model is in accordance well with the
experimental results.
Fig. 2
5.5
5.0
4.5
4.0
3.5
3.5
4.0
4.5
5.0
5.5
pIC (obs)
50
Fig. 4
Fig. 3
The standardized regression coefficient for each variable
is 11.84322, 7.249866, 0.520296, 0.031171 and 0.805255
respectively. Therefore, the relative importance of the descrip-
tors according to their standardized regression coefficients is
in the following order:
The QSAR model was optimized and descriptors were
reduced with the genetic function approximation. Some initial
parameters for genetic function approximation are listed in
Table-2.

2388 Dai et al.
Asian J. Chem.
19. L.L. Woo, T. Jackson, A. Putey, G. Cozier, P. Leonard, K.R. Acharya,
S.K. Chander, A. Purohit, M.J. Reed and B.V.L. Potter, J. Med. Chem.,
53, 2155 (2010).
E_ADJ_equ>SC_1>><1.23168-CIC>>
Jurs_TASA > Shadow_XZ
20. A. Stefanachi, A.D. Favia, O. Nicolotti, F. Leonetti, L. Pisani, M. Catto,
C. Zimmer, R.W. Hartmann and A. Carotti, J. Med. Chem., 54, 1613
(2011).
21. P.M. Wood, L. Woo, M.P. Thomas, M.F. Mahon, A. Purohit and B.V.L.
Potter, Chem. Med. Chem., 6, 1423 (2011).
It was found that E_ADJ_equ and SC_1 play the key role
for the bioactivity of the ligands.
Conclusion
Benzyl carbazole, 12 benzyl imidazole derivatives with
different substituents on both phenyl, imidazole rings were
synthesized and their aromatase inhibitory were evaluated with
fluorescent substrate detection method. The results showed
that the compounds with carboxyl and ester groups in phenyl
ring show better inhibitory activity. The introduction of alkyl
groups in imidazole may improve the aromatase inhibitory
activity. The obtained QSAR model with genetic functional
algorithm gives comparatively better correlation coefficient
and cross-validation correlation coefficient.
22. K. Bonfield, E. Amato, T. Bankemper, H. Agard, J. Steller, J.M. Keeler,
D. Roy, A. McCallum, S. Paula and L. Ma, Bioorg. Med. Chem., 20,
2603 (2012).
23. F. Caporuscio, G. Rastelli, C. Imbriano and A. Del Rio, J. Med. Chem.,
54, 4006 (2011).
24. D. Ghosh, J. Lo, D. Morton, D. Valette, J. Xi, J. Griswold, S. Hubbell,
C. Egbuta, W. Jiang, J. An and H.M.L. Davies, J. Med. Chem., 55,
8464 (2012).
25. T. Liu, Y. Lin, X. Wen, R.N. Jorissen and M.K. Gilson, Nucleic Acids
Res., 35(Database), D198 (2007).
26. L. Roumen, J.W. Peeters, J.M. Emmen, I.P.E. Beugels, E.M.G. Custers,
M. de Gooyer, R. Plate, K. Pieterse, P.A.J. Hilbers, J.F.M. Smits, J.A.J.
Vekemans, D. Leysen, H.C.J. Ottenheijm, H.M. Janssen and J.J.R.
Hermans, J. Med. Chem., 53, 1712 (2010).
ACKNOWLEDGEMENTS
27. J. Doiron, A.H. Soultan, R. Richard, M.M. Touré, N. Picot, R. Rich-
ard, M. Cuperlovic-Culf, G.A. Robichaud and M. Touaibia, Eur. J.
Med. Chem., 46, 4010 (2011).
28. C.P. Owen, S. Dhanani, C.H. Patel, I. Shahid and S. Ahmed, Bioorg.
Med. Chem. Lett., 16, 4011 (2006).
This work was supported by the National Natural Science
Foundation (Grant No. 21272171 and 31271809).
REFERENCES
29. P.P. Roy and K. Roy, J. Mol. Model., 16, 1597 (2010).
30. B. Balogh, B. Jójárt, Z. Wágner, P. Kovacs, G. Mate, K. Gyires, Z.
Zadori, G. Falkay, A. Marki and B. Viskolcz, Neurochem. Int., 51, 268
(2007).
31. Y.C. Martin, Quantitative Drug Design: A Critical Introduction, M.
Dekker, New York (1978).
32. R.R. Pissurlenkar, M.S. Shaikh and E.C. Coutinho, J. Mol. Model.,
13, 1047 (2007).
33. P. Zhou and Z. Li, Sci. China B, 50, 568 (2007).
34. A. Liu, H. Guang, L. Zhu, G.H. Du, S.M.Y. Lee and Y.T. Wang, Sci.
China C Life Sci., 50, 726 (2007).
35. J.T. Leonard and K. Roy, Bioorg. Med. Chem., 14, 1039 (2006).
36. X. Wang, Q. Xiao and S. Cui, Sci. China B, 48, 156 (2005).
37. I.L. Ruiz, M. Urbano-Cuadrado and M.Á. Gómez-Nieto, J. Math.
Chem., 43, 1549 (2008).
38. Y. Dai, Q. Wang, X. Zhang, S. Jia, H. Zheng, D. Feng and P. Yu, Eur. J.
Med. Chem., 45, 5612 (2010).
39. D. Ghosh, J. Griswold, M. Erman and W. Pangborn, Nature, 457, 219
(2009).
40. G.F. Yang, H.T. Lu, Y. Xiong and C.-G. Zhan, Bioorg. Med. Chem., 14,
1462 (2006).
41. L. Luo, L. Shen, F. Sun, Y. Dai, H. Zheng, Z. Ma, Y. Xu and Z. Guo,
Anal. Methods, 4, 230 (2012).
42. Discovery Studio. Accelrys Software Inc, San Diego, CA.
43. D.J. Ritchie, US Patent 6048998(2000).
44. C.P. Miller, P. Bhaket, N. Muthukaman, C.R. Lyttle, M. Shomali, K.
Gallacher, C. Slocum and G. Hattersley, Bioorg. Med. Chem. Lett., 20,
7516 (2010).
45. Chem. Bio. Office UltraVersion 11.0, Cambridge Soft. Inc., Cambridge,
USA.
1. R.W. Brueggemeier, J.C. Hackett and E.S. Diaz-Cruz, Endocr. Rev.,
26, 331 (2005).
2. C. DeSantis, R. Siegel, P. Bandi and A. Jemal, CA Cancer J. Clin., 61,
408 (2011).
3. H. Yu, R. Harris, Y. -T. Gao, R. Gao and E. Wynder, Int. J. Epidemiol.,
20, 76 (1991).
4. L. Yang, L.D. Li, Y.D. Chen and D.M. Parkin, Zhonghua Zhong Liu Za
Zhi, 28, 438 (2006).
5. R.J. Santen, Steroids, 50, 575 (1987).
6. M.S. Mahendroo, C.R. Mendelson and E.R. Simpson, J. Biol. Chem.,
268, 19463 (1993).
7. E.R. Simpson, M.S. Mahendroo, G.D. Means, M.W. Kilgore, M.M.
Hinshelwood, S. Graham-Lorence, B. Amarneh, Y. Ito, C.R. Fisher
and M.D. Michael, Endocr. Rev., 15, 342 (1994).
8. H. Lefebvre, V. Contesse, C. Delarue, C. Soubrane, A. Legrand, J.M.
Kuhn, L.M. Wolf and H. Vaudry, J. Clin. Endocrinol. Metab., 77, 1662
(1993).
9. W.R. Miller and J. O'Neill, Steroids, 50, 537 (1987).
10. J. McNulty, J. Nair, N. Vurgun, B.R. DiFrancesco, C.E. Brown, B.
Tsoi, D.J. Crankshaw and A.C. Holloway, Bioorg. Med. Chem. Lett.,
22, 718 (2012).
11. T.R. Evans, E.D. Salle, G. Ornati, M. Lassus, M.S. Benedetti, E.
Pianezzola and R.C. Coombes, Cancer Res., 52, 5933 (1992).
12. D.P. McDonnell and S.E. Wardell, Curr. Opin. Pharmacol., 10, 620
(2010).
13. A. Buzdar and A. Howell, Clin. Cancer Res., 7, 2620 (2001).
14. P.E. Goss and K. Strasser, J. Clin. Oncol., 19, 881 (2001).
15. A. Hamilton and M. Volm, Oncology (Williston Park), 15, 965-972;
discussion 972, 977-969 (2001).
16. R. Murray, Cancer Chemother. Pharmacol., 48, 259 (2001).
17. J.N. Ingle, Clin. Cancer Res., 7, 4392s (2001).
18. S. Gobbi, C. Zimmer, F. Belluti, A. Rampa, R.W. Hartmann, M.
Recanatini and A. Bisi, J. Med. Chem., 53, 5347 (2010).
46. Y. Dai, N. Chen, Q. Wang, H. Zheng, X. Zhang, S. Jia, L. Dong and D.
Feng, Iran. J. Pharm. Res., 11, 807 (2012).
47. J.H. Friedman, Ann. Statistics, 19, 1 (1991).