Molecular modeling study, synthesis and biological evaluation of combretastatin A-4 analogues as anticancer agents and tubulin inhibitors

Article abstract of DOI:10.1039/c7md00416h

As the major structural component of microtubules, tubulin is an interesting target for the development of anticancer agents. In this study, 64 tubulin polymerization inhibitors of five-membered heterocycle-based combretastatin A-4 analogues were studied by a combination of molecular modeling techniques including 3D-QSAR, molecular docking, and molecular dynamics (MD) simulation. The CoMFA (comparative molecular field analysis) and CoMSIA (comparative molecular similarity indices analysis) models were established with desirable statistical parameters and excellent predictive ability. 20 ns MD simulations were successfully performed to confirm the detailed binding mode and validate the rationality of docking results. Combining the binding free energy calculations and 3D-QSAR results, some new heterocycle-based combretastatin A-4 analogues were designed. Three of them were synthesized and biologically evaluated. Compound 13a displayed potent antiproliferative activity (IC₅₀ value of 1.31 M against HepG2 cells, IC₅₀ value of 1.37 M against A549 cells) and inhibition of tubulin polymerization activity (IC₅₀ value of 0.86 M). Compound 13b also presented good activity against HepG2 cells (IC₅₀ value of 4.75 μM). The experimental results demonstrated that the built models were effective for the development of novel anticancer agents and tubulin inhibitors.

Full text of DOI:10.1039/c7md00416h

View Article Online
View Journal
MedChemComm
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: Y. P. Quan, L. P.
Cheng, T. C. Wang, W. Pang, F. H. Wu and J. W. Huang, Med. Chem. Commun., 2017, DOI:
1
0.1039/C7MD00416H.
Volume 7 Number 1 January 2016 Pages 1–204
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
MedChemComm
Broadening the field of opportunity for medicinal chemists
www.rsc.org/medchemcomm
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
author guidelines.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the ethical guidelines, outlined
in our author and reviewer resource centre, still apply. In no
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
Themed issue: Antibiotic Resistance
ISSN 2040-2503
rsc.li/medchemcomm

Page 1 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
MedChemComm
RESEARCH ARTICLE
Molecular modeling study, synthesis and biological evaluation of Combretastatin A-4
analogues as anticancer agents and tubulin inhibitors
†
†,
Yang Ping Quan , Li Ping Cheng *, Tian Chi Wang, Wan Pang*, Fan Hong Wu*, Jin Wen Huang
As the major structural component of microtubules, tubulin is an interesting target for the development of anticancer agents. In this study, 64 tubulin
polymerization inhibitors of five-membered heterocycle-based combretastatin A-4 analogues were studied by combination of molecular modeling
techniques including 3D-QSAR, molecular docking, and molecular dynamics (MD) simulation. The CoMFA (comparative molecular field analysis) and CoMSIA
(comparative molecular similarity indices analysis) models were established with desirable statistical parameters and excellent predictive ability. 20 ns MD
simulations were successfully performed to confirm the detailed binding mode and validate the rationality of docking results. Combined the binding free
energy calculations and 3D-QSAR results, some new heterocycle-based combretastatin A-4 analogues were designed. Three of them were synthesized and
biologically evaluated. Compound 13a displayed potent antiproliferative activity (IC50 value of 1.31 µM against HepG2 cells, IC50 value of 1.37 µM against
A549 cells) and inhibition of tubulin polymerization activity (IC50 value of 0.86 µM). Compound 13b also presented a good activity against HepG2 cells (IC50
value of 4.75 µM). The experimental results demonstrated that the built models were effective for the development of novel anticancer agents and tubulin
inhibitors.
Introduction
Microtubules are cytoskeletal filaments composed of α, β-tubulin
heterodimer assemblies. The cellular microtubule system is essential in
many important cellular processes such as cell division, formation, and
maintenance of cell shape, regulation of motility, cell signaling, secretion
combretastatins indicate that both the 3',4',5'-trimethoxy substitution
pattern on the A-ring and the cis ethylene linkage between the two aryl
10-11
rings lead to the optimal activity
(see Fig. 1). However, the
cis-configuration of CA-4 is prone to isomerize more stable and essentially
inactive trans-isomer thermodynamically. To retain the cis-configuration of
CA-4 required for bioactivity, a series of derivatives were obtained by
incorporating the olefinic double bond into some five-membered aromatic
1
-3
and intracellular transport. Microtubules are generally recognized as an
attractive target for the development of potential new anti-cancer
4
-5
agents. Among the tubulin targeting agents, combretastatin A-4 (CA-4),
6
12-19
12
derived from the bark of the South African tree Combretum caffrum, is
heterocyclic rings.
For example, Wang and co-workers utilized
one of the well-known natural tubulin-binding molecules. This cis-stilbene
strongly inhibits the polymerization of tubulin by binding to the colchicine
site on β-tubulin, prevent the polymerization of tubulin into microtubules,
1,2-substituted five-membered aromatic heterocycles such as imidazole,
oxazole, and pyrazole to mimic the cis double bonds in CA-4. The obtained
compounds have potent anti-tubulin and cytotoxic activity. Romagnoli and
7
13,14
thiazole,
15
16
and displays exceptional cytotoxicity towards a variety of cancer cell lines.
co-workers confirmed that
a
triazole,
tetrazole,
1
7-18
19
CA-4 has been entered into clinical trials, but unfortunately its efficacy in
thiophene,
or furan ring could replace the cis double bonds to
8
vivo is limited due to the drug's poor solubility. A more water-soluble
maintain the potent cytotoxicity. Although great progress has been made
in experimental research, so far theoretical studies on the mechanism of
these compounds toward tubulin as well as the structural factors
disodium phosphate of CA-4 (CA-4P) (see Fig. 1) is currently under
9
evaluation, and has shown promising results in human clinical trials.
2
0
Due to the relatively simple chemical structure and high affinity for
responsible for the cytotoxicity remain limited. Brown et al. had
performed a molecular modeling study for colchicine binding site based on
a training set of 23 compounds using Comparative Molecular Field Analysis
the colchicine binding site, a great number of CA-4 analogues had been
10-19
synthesized and evaluated in SAR studies.
Previous SAR studies on
(
CoMFA) method. The built model can accurately predict the IC50 for
*
School of Chemical and Environmental Engineering, Shanghai Institute of
Technology, Shanghai 201418, China
Corresponding authors. Telephone:
chengliping@sit.edu.cn; pangwan@sit.edu.cn; wfh@sit.edu.cn
These two authors (Yang Ping Quan and Li Ping Cheng) contributed equally to
this work and should be considered as co-first authors.
2
3
tubulin polymerization with an R of 0.88 (n=6) and those of [ H] colchicine
*
+86-21-6087-3250.
email:
2
21
displacement with an R of 0.80 (n=7). Liao et al. reported a theoretical
†
study on the binding conformations and QSAR of CA-4 analogs as inhibitors
1

MedChemComm
Page 2 of 14
View Article Online
DOI: 10.1039/C7MD00416H
1
2, 15-18, 22-28
toward tubulin based on docking and CoMFA methods. In this study, some
key structural factors of the compounds responsible for cytotoxicity were
reasonably presented. The results of these studies provide evidences for
further designing and synthesizing tubulin inhibitors and conducting
structure optimization. However, on the one hand, the total number of
samples in their training sets is so small (less than 35) that the reliability of
the built models seems to be questionable. On the other hand, although
molecular docking had been performed in some theoretical studies to
explore the interactions between the inhibitors and receptor proteins, the
reliability of docking results should be further verified by molecular
dynamics (MD) simulation. Up to now, few research groups had
comprehensively studied the interaction mechanism of CA-4 analogs with
tubulin receptors by MD simulation. To obtain more understanding of the
structural and chemical properties required for tubulin inhibitory activity,
more 3D-QSAR and other computational exploration analyses for
comparatively large number of newly synthesized and tested tubulin
inhibitors are necessary. In the present work, 64 five-membered
heterocycle-based CA-4 analogues reported as potent tubulin
polymerization inhibitors were collected as dataset. A series of studies
were performed by combination of molecular modeling techniques
including 3D-QSAR, molecular docking and MD simulation. The results of
CoMFA and CoMSIA studies will not only explain the conformation or
spatial orientation of CA-4 analogs but also provide useful information for
the design of potent tubulin polymerization inhibitors. Molecular docking
and MD simulations were performed to validate the 3D-QSAR models and
quantify the interactions of tubulin-ligand. Some heterocycle-based CA-4
analogues were designed based on the built 3D-QSAR contour maps.
Finally, the syntheses and biological evaluation of some novel
heterocycle-based CA-4 derivatives were further carried out. The cytotoxic
effects of the synthesized compounds against six human cancer cell lines,
A549, HCT-116, HeLa, HepG2, MGC-803, MCF-7, and the inhibitory activity
on tubulin polymerization was studied.
studies
as many as possible in the present study. However, we
encountered a problem that the data from different sources were hard to
be compared and analyzed. Furthermore, the activity values of CA-4 were
not the same in different studies. To eliminate data errors from different
sources, a normalization method of processing data was used in the
present study. Details are as follows: All data were normalized to an
experimental value for combretastatin A-4 (1) from the paper in which the
data was taken. The active value of CA-4 is set to 1, corrected IC50 (IC50c) =
IC50 (compound) / IC50 (CA-4). If one compound has higher activity than
CA-4, its IC50c are less than 1, and vice versa. Then the IC50c values are
converted to the negative logarithm of this values (pIC50c value), which are
used as dependent variables in the 3D-QSAR models. The chemical
structures are divided into skeleton templates a, b, c, d, and pIC50c values
are summarized in Supplementary Material (Table 1S). The whole dataset
was randomly divided into a training set of 51 compounds (79.7%) and a
test set of 13 compounds (20.3%), respectively. The selection of training
and test sets was done by considering structural diversity and the range of
biological activities of test compounds similar to that of the training set.
The model helped us comprehend the structure-activity relationships of
this class of compounds and facilitate the design of novel inhibitors with
good activities. In this study, we used the ligand-based alignment rule. The
alignment of training set compounds is shown in Fig. 2.
Figure 2. Alignment of the training set and compound 13 used as a template for
alignment, with the common substructure shown in bold (left). Molecules are
colored in white for common C, red for O, yellow for S, cyan for H, green for F
and Cl, respectively.
Molecular modeling and statistical analysis
2
9
According to the partial least squares (PLS) method, some statistical
Figure 1. The structures of CA-4 and CA-4P.
parameters were used to evaluate and analysis the stand or fall of these
2
2
models, including q , r , standard error of estimate (SEE) and F-statistic
Results and Discussion
2
values, r pred. Table 1 lists the statistical parameter results of the CoMFA
and CoMSIA analyses. From Table 1, the optimal CoMFA and CoMSIA
Data sets
2
models show good predictive power with the cross-validated q of 0.702
2
To construct a global predictive combretastatin model for the inhibition of
tubulin polymerization, various five-membered heterocycle-based
combretastatin A-4 analogues were collected from some different
and 0.700. The correlation coefficient r , SEE and F-statistic values are
0.988, 0.054, and 505.339 for CoMFA model and 0.986, 0.060, and 359.657
2
for CoMSIA model. The values of r pred representing the predictive ability
2

Page 3 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
Table 1. Statistical results for the CoMFA and CoMSIA models
CoMFA and CoMSIA contour maps
Statistical parameters
CoMFA
CoMSIA
0.700
8
To visualize the field effects on the target compounds in 3D space, the
contour maps generated by CoMFA and CoMSIA were analyzed by
superimposing them onto the most active molecule 13. These contour
maps are significant for new drugs design, as they show regions in 3D
space where modifications of the molecular fields strongly correlated with
variation in biological activity.
2
q
0.702
NOC
7
0.988
0.711
0.054
505.339
-
2
r
0.986
0.623
0.060
359.657
-
2
r pred
The electrostatic field contour maps of CoMFA and CoMSIA are
shown in Fig. 4A and 4B. The electrostatic field is indicated by blue and red
contours, where the blue regions denote that the electropositive groups
are favorable to the activity and the red regions indicate that the
electronegative groups are favorable to the activity. As shown in Fig. 4A, it
had been recognized that the furan ring and the portion of the benzene
ring connected to the C-2 position of furan are encompassed by a large
blue colored map. This observation demonstrates that these positions are
suitable for substitution with electropositive groups. This is reflected in the
SEE
F
fraction
Steric
0.533
0.467
-
0.099
0.273
0.171
0.227
0.227
Electrostatic
H-acceptor
H-donor
Hydrophobic
3 2 2
inhibitory order of 27 (CH ) > 30 (CH I) > 28 (CH Cl), and 33 (Cl) > 34 (F). A
large red contour located at the 5-position of benzo[b]furan ring indicates
that this position is suitable for substitution with electronegative group.
However, in Fig. 4B, CoMSIA contour map shows that no corresponding
red contour presents to the 5-position of benzo[b]furan ring. In Fig. 4B, a
large red contour appears on the 4-position of furan ring. Generally
speaking, CoMSIA descriptors are more convincing. It suggests that the
substituent at the 5-position of benzo[b]furan may not obtain the desired
bioactivity. In addition, two pieces of region of red contour near the
trimethoxyphenyl ring indicates that some electronegative groups can be
introduced into this region to produce better biological activity.
-
-
for the CoMFA and CoMSIA models are 0.711 and 0.623, respectively. For
the CoMFA model, the contributions of the steric and electrostatic fields
are 53.3 and 46.7%. The main contribution in the model was steric field.
For the optimal CoMSIA model, the steric, electrostatic, hydrophobic,
H-Bond donor and acceptor field contribution was 9.9, 27.3, 22.7, 23.0,
and 17.1%, respectively. The electrostatic, H-bond donor and hydrophobic
fields were found to be the important contributions in the optimal CoMSIA
model. The experimental pIC50c, predicted pIC50c, and the deviation
between predicted and experimental values are shown in Supplementary
Material (Table 2S). The relationship between actual and predicted pIC50
c
values of the training and test set is illustrated in Fig. 3A and 3B for CoMFA
and CoMSIA models, in which most points are evenly distributed along the
line Y = X, suggesting that the 3D-QSAR models are of good quality.
Figure 4. CoMFA (A) and CoMSIA (B) contour maps. Electrostatic fields:
electropositive (blue) and electronegative (red).
As shown in Fig. 5, the sterically favorable regions are represented
in green and unfavorable regions in yellow, respectively. As shown in Fig.
5
3 6
A, a large green contour surrounding the C ~C moiety of the benzene
ring which is connected to the C-2 position of furan ring of template
molecule 13 indicates the importance of the presence of a bulky group in
this region for biological activity. In comparison with molecules of
template a, c, d, most molecules in template b are substituted by bulky
Figure 3. Plots of predicted versus experimental pIC50c values for all the
molecules based on CoMFA (A) and CoMSIA models (B).
7
groups at the position of R substituent, which exhibit excellent biological
3

MedChemComm
Page 4 of 14
View Article Online
DOI: 10.1039/C7MD00416H
activity. A yellow contour located at the moiety of benzo[b]furan ring
indicates that the presence of bulky substitution in this region will
decrease biological activity. For example, the inhibitory order of some
compounds is: 26 > 27-30. As shown in Fig. 5B, a green contour located at
the 6-position of benzo[b]furan ring suggests that compounds with bulky
substitution would increase biological activity. For example, the good
activity of 52 compared with 48-51 appears to be highly dependent upon
the presence of a methoxy substituted H atom at this position. In addition,
a large yellow contour surrounding trimethoxyphenyl ring suggests that
this position is not suitable for substitution with bulky groups. A large
green contour near the -2-(4-methoxy phenyl) suggests that this position is
suitable for substitution with bulky groups, which is similar to the result
from CoMFA steric contour map.
heterocycle at the furan ring position, which reveals the importance of the
H-bond donor substituent. large cyan contour map covers the
A
meta-position and para-position at the benzene ring of
-2-(4-methoxyphenyl). The inhibitory order of 5, 6, and 7 (7 < 6 < 5) shows
that H-Bond donors in this region are beneficial to bioactivity. A purple
contour map which was observed above the trimethoxyphenyl indicates
that the H-Bond donors in these regions are unfavorable for the inhibitory
activity. In addition, other medium-sized cyan contours are far from the
molecular, and they are neglected.
As shown in Fig. 6C, the hydrogen bond acceptor field is represented
by magenta and red contours, in which magenta contours denote regions
where hydrogen bond acceptor group would be beneficial to the
bioactivity, whereas red contours represent hydrogen bond acceptor
group would decrease the bioactivity. A magenta contour map covering
the connection part and the 4-position of the benzo[b]furan ring and a bit
smaller magenta contour near the C-3 position of -2-(4-methoxyphenyl)
indicate the H-bond acceptor groups could have a positive influence on the
inhibitory activity. Most of the derivatives involved in this study possess
groups containing H-bond acceptors (such as carbonyl, hydroxy, imino, or
five-membered heterocycles containing N or O atoms) at this site, which
reveals the importance of the H-bond acceptor substituents. A red contour
map covering the 1,2-position of benzo[b]furan ring implies that H-bond
acceptor groups maybe decrease the inhibitory activity.
Figure 5. CoMFA (A) and CoMSIA (B) contour maps. Steric fields: favored
(
green) and disfavored (yellow).
The CoMSIA hydrophobic contour maps are depicted in Fig. 6A.
Yellow and white contours highlight areas where hydrophobic and
hydrophilic properties are favored. As shown in Fig. 6A, a white
polyhedron covering carbonyl group indicates that this position is suitable
for substitution with hydrophilic group. Most of the derivatives involved in
this study possess hydrophilic groups (such as carbonyl, hydroxy, imino, or
five-membered heterocycles containing N or O atoms) at this site, which
reveals the importance of the hydrophilic substituent. For instance, the
inhibitory order of 42, 47, and 1 is 42 > 47 > 1. One medium-sized region of
yellow contour located at the 4-position of benzo[b]furan ring shows the
importance of hydrophobic substitute in imparting better biological
activity. In addition, the biggest yellow maps are far from the molecular
skeleton and they are ignored.
Figure 6. CoMSIA contour maps. (A) hydrophobic field: favored (yellow) and
disfavored (white). (B) Hydrogen bond donor field: favored (cyan) and
disfavored (purple). (C) Hydrogen bond acceptor field: favored (magenta) and
disfavored (red).
Molecular design of new tubulin inhibitors
According to the information derived from contour maps generated by
3
D-QSAR models, some information about the template modification were
presented for region A, B, C and D on inhibitory activities (Fig. 7). Based on
the summarized structure-activity relationships, six new compounds 13a-f
were designed by introduction of F atom or other substituent groups into
the regions. The most active molecule 13 was taken as a template
molecule. The structures and predicted pIC50c values of 13a-f based on
CoMFA and CoMSIA models are listed in Table 2. Table 2 shows that the
pIC50c values of all designed new molecules are higher than those of CA-4,
suggesting their good inhibitory activity.
The CoMSIA hydrogen bond donor and accepter contour are shown
in Fig. 6B and 6C. As shown in Fig. 6B, the purple contours represent the
position where hydrogen bond donor disfavors the biological activity and
the cyan contours show that the presence of donor groups in this region
should produce better biological activity. A big cyan (H-bond donor
4 5
favorable) contour map next to the C ~C moiety of furan suggests that the
electronegative H-Bond donors in these regions are favorable for the
inhibitory activity. Most of the derivatives involved in this study possess
strong electronegative atoms (such as O and N) in the five-membered
4

Page 5 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
conducted for two representative inhibitors, CA-4 and 13, as well the new
designed molecules, 13a-f. All MD simulations were performed using the
AMBER 12 package. The overall convergence of MD simulations and
system equilibration were monitored by root-mean-square deviation
α
(RMSD) of tubulin-inhibitor backbones atoms (C, C , N, and O). The RMSD
value (Å) with respect to simulation time (ns) is shown in Fig. 1S (See
Supplementary Material). It can be seen that all complexes are mainly
stable after 6 ns simulation time run. In addition, the initial structure of
1
3a-tubulin (PDB ID: 3UT5) complex and the equilibrium structure after 20
ns MD simulation superimposed in the Fig. 8. Fig. 8 shows that there
seems to be no significant difference between initial structure (in green)
and the lowest energy structure (in magenta) except for a slight drift and
rotation of bonds. It can be inferred that the binding pocket and the
conformation of the ligand are stable, and the docking results are reliable.
Figure 7. Structure-activity relationships derived from 3D-QSAR studies.
Table 2. Structures and predicted pIC50c values of designed molecules
pIC50c (Pred)
Compound
Structure
CoMFA
CoMSIA
CA-4
0.000
0.037
1
3a
0.351
0.285
0.291
0.308
1
3b
O
O
O
Figure 8. The superimposed the initial structure (green ribbon) and the
equilibrium structure (magenta ribbon) of 13a-3UT5 complexes, initial
conformation (green), equilibrium conformation (magenta).
O
1
3c
0.348
0.425
0.503
0.576
0.282
0.346
0.206
0.182
O
O
O
N
Based on the MD simulations, the binding free energy calculations
for seven inhibitors of the last 2 ns trajectory were carried out and are
listed in Table 3. The contributions of different energy components are
also reported in Table 3. The MM/GBSA and MM/PBSA were both used in
free energy calculations. For all inhibitors, the ꢀGgas and the ꢀGsolv is
negative and positive, respectively, indicating that the former is favorable
energetic contribution and the latter is unfavorable energetic contribution
to the total binding free energy. It can also be seen that the major
contribution to the overall binding free energy is the van der Waals energy
of ꢀGgas. The binding free energy for the inhibitors CA-4, 13, 13a, 13b, 13c,
1
3d
3e
1
1
3d, 13e, and 13f via MM/GBSA was predicted to be -34.32, -57.59, -57.52,
1
3f
-
1
-
54.41, -55.78, -50.77, -55.56, and -40.52 kcal·mol and via MM/PBSA was
predicted to be -21.54, -44.02, -43.45, -42.03, -40.99, -39.79, -36.23, and
-
1
-
27.82 kcal·mol , respectively. It is obviously that the binding free energies
of 13a-13f are far more negative than that of CA-4 based on both
MM/GBSA and MM/PBSA method. Among the six new designed
compounds, 13a has the most negative binding free energy, implying that
MD simulations analysis
To get a more detailed insight into the stability of the tubulin-inhibitor
complex, 20 ns MD simulations and binding free energy calculations were
5

MedChemComm
Page 6 of 14
View Article Online
DOI: 10.1039/C7MD00416H
it perhaps possesses the best inhibitory activity.
Table 3. Binding free energy (ΔGbind, kcal·mol-1) (last 2ns) for ligand-tubulin complexes along with the different energy components
ΔGsolv
ΔGbind
(GB)
ΔGsolv
(PB)
ΔGbind
(PB)
a
b
c
No.
VDW
EEL
ΔGgas
ΔGGB
ΔGSA
ΔGPB
ΔGSA
(GB)
CA4
-46.64
-69.28
-72.63
-66.73
-68.71
-64.86
-72.79
-18.64
-22.54
-20.17
-17.49
-19.20
-17.65
-16.77
-11.73
-65.28
-92.41
-92.80
-84.22
-87.91
-82.51
-89.56
36.96
43.02
44.16
37.87
40.06
39.97
42.44
35.09
-6.01
-8.20
-8.88
-8.07
-7.93
-8.22
-8.44
-7.02
30.95
34.82
35.28
29.81
32.13
31.74
34.00
28.07
-34.32
-57.59
-57.52
-54.41
-55.78
-50.77
-55.56
48.02
53.83
55.18
47.47
52.75
48.49
58.79
46.23
-4.29
-5.43
-5.82
-5.28
-5.83
-5.77
-5.46
-5.46
43.74
48.40
49.35
42.19
46.92
42.72
53.33
40.77
-21.54
-44.02
-43.45
-42.03
-40.99
-39.79
-36.23
-27.82
1
3
1
3a
3b
1
1
3c
1
3d
3e
1
1
3f
-56.86
-68.59
-40.52
a
b
c
-1
Vander Waals energy; EEL: electrostatic energy; The units of all interaction energy are kcal mol
The benzo[b]furan derivatives (13a, 13b) and indole derivative (13f) were
evaluated for their antiproliferative activity in vitro against six human
cancer cells (A549, lung carcinoma; HCT-116, colon carcinoma; HeLa,
cervical carcinoma; HepG2, liver carcinoma; MGC-803, gastric carcinoma;
MCF-7, mammary carcinoma) using CCK-8 assays. CA-4 and CA-4P served
Chemistry
According to the results of binding free energy calculations and the known
synthesized CA-4 analogues, three designed compounds 13a, 13b, 13f
were firstly selected to be synthesized. The benzo[b]furan derivatives 13a,
1
3b described in this paper were prepared according to the synthetic route
3
0
(
Scheme-1), and the route of indole derivative 13f was shown in
3
1, 32
Scheme-2.
The structures of the synthesized intermediates and new
1
13
benzo[b]furan or indole derivatives were confirmed by H NMR, H NMR,
19
F
NMR and HRMS. Iodination of 3-methoxy phenol (13a-1) or
4
-fluorine-5-methoxy phenol (13b-1) was carried out using iodine in
presence of catalytic quantity of silver trifluoroacetate (AgOTFA) in
chloroform at room temperature for 24 h to obtain the iodide compounds
3
3
1
3a-2 or 13b-2. Acetylation of iodo-phenol was performed in presence of
acetic anhydride in pyridine to afford the corresponding acetate
3
4
derivatives 13a-3 , 13b-3. Sonogashira coupling of 13a-3 or 13b-3 with
-ethynyl-4-ethoxybenzene or 1-ethynyl-4-methoxybenzene in presence of
Pd(PPh Cl /CuI catalyst load followed by the intramolecular cyclisation in
Scheme-1 Synthesis of benzo[b]furan derivatives 13a, 13b
1
3
)
2
2
presence of potassium carbonate resulted in the formation of
benzo[b]furan. Friedel Crafts procedure was used in presence of 3, 4,
5
-trimethoxybenzoyl chloride followed by the addition of 13a-5 or 13b-5
and tin (IV) chloride to afford 13a or 13b.
Biological assay
In vitro antiproliferative assay
Scheme-2 Synthesis of indole derivative 13f
6

Page 7 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
as positive control compounds. The IC50 values of the compounds are
more active against HepG2 cells compared with CA-4, while its activity
against HeLa, HCT-116, and MCF-7 cells is comparable to that of CA-4. For
compound 13a, the antiproliferative activities against the six human
cancer cells are all superior to those of compound 13b. It shows an
enhancement in antiproliferative activity by more than 5-fold as compared
with 13b except for HeLa and HepG2 cells, while 13b is more potent than
CA-4 against the HepG2 cells. Compound 13f exhibits no antiproliferative
activities against the cell lines at all. Except for the activity against MCF-7
cells of 13a and CA-4, all the other activities are lower than CA-4P.
summarized in Table 4.
As shown in Table 4, among the evaluated compounds, compound
1
3a exhibits the most potent antiproliferative activity, with IC50 values of
1
.37, 12.55, 8.99, 1.31, 9.97, and 0.91 µM against A549, HCT-116, HeLa,
HepG2, MGC-803, and MCF-7 cells, respectively. It is thus more potent
than the reference CA-4 against the A549, HeLa, and HepG2 cells.
Specifically, 13a is 4.4-fold more active against A549 cells and 12.2-fold
Table 4. In vitro antiproliferative activities of synthetic compounds
IC50 ± SD (µM)
Com.
A549
HCT-116
12.55 ± 0.60
78.23 ± 3.83
-
HeLa
HepG2
1.31 ± 0.17
4.75 ± 0.20
-
MGC-803
9.97 ± 0.81
47.26 ± 1.17
-
MCF-7
0.91 ± 0.16
6.49 ± 0.55
-
1
3a
1.37 ± 0.14
6.87 ± 0.34
-
8.99 ± 1.11
19.1 ± 1.59
-
1
3b
1
3f
CA-4
5.99 ± 0.46
0.40 ± 0.17
9.00 ± 0.85
0.24 ± 0.05
11.33 ± 1.03
0.56 ± 0.12
16.04 ± 0.50
0.27 ± 0.04
1.76 ± 0.85
0.47 ± 0.14
0.67 ± 0.13
0.96 ± 0.22
CA-4P
In vitro tubulin polymerization assay and docking analysis
assays. 13f has no obvious inhibitory activity for tubulin polymerization
either.
To investigate whether the designed CA-4 analogues interact with tubulin,
compounds 13a, 13b, 13f were selected to evaluate their inhibitory effects
To better understand the discrepancy of activity between 13a, 13b,
and 13f for tubulin polymerization, a series of molecular docking
simulations were carried out. Fig. 9 shows the 3D binding modes of the
representative 13a, 13b, 13f, and CA-4 at the colchicine binding site of
tubulin (PDB ID: 3UT5). As shown in Fig. 9, compound 13a and 13b could
be perfectly docked into the active site of tubulin. For example, they
bound with the active site in a very similar pattern to the binding manner
of CA-4. The trimethoxyphenyl group takes up a same position as the
corresponding moiety of CA-4, which is in proximity to the Cys241. The
carbonyl groups of 13a and 13b keep a cis-configuration for activity like
the cis-double bonds of CA-4. The 6-methoxy-benzo[b]furan moiety seems
to be deeply placed in a tight hydrophobic pocket defined by Met259,
Asn258, Ala180, Lys352, and Asn350 and the ethoxyphenyl moiety is
located in the hydrophilic region (Fig. 10). The benzo[b]furan ring of 13a
and 13b establishes CH-π interactions with Lys352 in Fig. 10, respectively.
In Fig. 10A, one H atom of 6-methoxy of 13a forms CH…O non classical H
bond with carbonyl of Asn350, and the other two H atoms form CH…O
non classical H bonds with carbonyl of Asn258, respectively. Besides, the
3
5
on tubulin polymerization. The corresponding results are shown in Table
5
. For comparison, CA-4 and CA-4P were evaluated as the reference
compounds. The result shows that compound 13a can strongly inhibit
tubulin polymerization with IC50 value of 0.86 µM, which is comparable
with that of CA-4 (IC50, 0.88 µM) and superior to CA-4P (IC50, 4.79 µM). It is
a pity that the compound 13b has no inhibition activity for tubulin
polymerization, which exhibits appreciable activity in vitro antiproliferative
Table 5. Inhibition of tubulin polymerization
Compounds
IC50 ± SD (µM)
0.86 ± 0.07
> 200
1
3a
1
3b
1
3f
> 200
CA-4
0.88 ± 0.22
4.79 ± 0.40
-
CH
bonds with carbonyl of Asn249. However, in Fig. 10B, two H atoms of
-methoxy of 13b form CH…O non classical H bonds with carbonyl of
2
- of 2-(4-ethoxyphenyl) of 13a have formed two CH…O non classical H
CA-4P
6
7

MedChemComm
Page 8 of 14
View Article Online
DOI: 10.1039/C7MD00416H
Asn350 and Asn258, respectively. The rest of H atom of 6-methoxy of 13b
forms a intramolecular non classical H bond with 5-F atom of 13b, and 13b
does not generate interaction with Asn249 like that of 13a. These may
cause 13b to lose the inhibition of tubulin polymerization activity. In
addition, Fig. 9 and Fig. 11 show that the conformation of 13f seems to be
out of the colchicine binding site. Furthermore, the docking poses of the
indole moiety and the ethoxyphenyl moiety of 13f have an opposite
orientation compared with that of 13 (Fig. 11), 13a and 13b (Fig. 9), which
may cause the loss of bioactivity. The conformation of 13 is similar to that
of 13a and 13b.
In the present work, a series of five-membered heterocycle-based CA-4
analogues were studied by a combination of computer-aided drug design
techniques, i.e. 3D-QSAR study, molecular docking, MD simulations, and
binding free energy calculation. The established 3D-QSAR models show
significant statistical quality and excellent predictive ability. According to
the information derived from contour maps generated by 3D-QSAR
models, some information about the template modification were
presented for region A, B, C and D on inhibitory activities, six new
inhibitors (13a-13f) were designed and predicted. Molecular docking for
synthetic compounds was carried out. 20 ns MD simulations had been
successfully run for representative compounds and their binding free
energies were calculated. The total binding free energies of all designed
compounds are much more negative than that of CA-4. Based on the
theoretical results and the known synthesized scheme, 13a, 13b, and 13f
are synthesized, and biologically evaluated in cellular and tubulin inhibition
assays. 13a displays potent antiproliferative activity in cellular assays (IC50
value of 1.31 µM against HepG2 cells, IC50 value of 1.37 µM against A549
cells) and inhibition of tubulin polymerization activity (IC50 value of 0.86
µM). 13b also exhibits a decent antiproliferative activity with IC50 value
4
.75 µM against HepG2 cells.
Figure 9. Proposed binding modes of 13a (in green), 13b (in magenta), and 13f
(
in orange) in the colchicine site of tubulin. CA-4 as reference compound is
Experiment Section
shown in white.
Data set and alignment
A total of 64 five-membered heterocycle-based combretastatin A-4
analogues were collected from different research groups for constructing
3
D-QSAR models. Three dimensional structures of the molecules and
subsequent 3D-QSAR studies were performed in SYBYL-X 2.0 (Tripos, Inc.,
USA). All compounds were constructed using the Sketch Molecule module
and then minimized by the Powell gradient algorithm. The energy gradient
termination criterion and the maximum iterations were set to 0.005
kcal/(mol·Å) and 100000, respectively. The energy minimization was
performed using Tripos force field with Gasteiger-Huckel charges. The
treated molecules were divided into a training set (51 compounds) and a
validated test set (13 compounds). To obtain a good alignment, the
database alignment method was performed using the most active
compound as a template.
Figure 10. The detailed binding poses of 13a (A) and 13b (B). Yellow dashed
lines represent non classical hydrogen bonds. Blue dashed lines represent CH-π
interactions.
CoMFA and CoMSIA models
The CoMFA and CoMFA models were developed for the same training set
and test set. The CoMFA model was carried on steric and electrostatic
fields. The CoMSIA model was carried out using steric, electrostatic,
hydrophobic, hydrogen bond donor, and hydrogen bond accepter fields.
Figure 11. Proposed binding modes of 13 (in white) and 13f (in orange).
3
D contour maps of CoMFA and CoMSIA models were graphed using the
Conclusion
field type “Stdev*Coeff”. Some statistical parameters were used to
evaluate the built of models. Partial least squares analysis and leave one
8

Page 9 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
out validation were used to obtain the cross-validation correlation
distance of 10 Å at 300 K. Periodic boundary condition was used for the
whole simulation run in order to encounter bulk effect. The integration
time step was set to 2 fs and coordinate trajectories were recorded in
every 2 ps in the whole simulation. Visualization and analysis were carried
out by VMD. The figure of root mean square deviation (RMSD) and cluster
analysis were performed by Xmgrace program.
2
coefficient (q ) and an optimal number of components (NOC), as well the
2
square correlation coefficient (r ), F-statistic values (F), and standard error
of estimate (SEE). These statistical parameters described the predictive
ability of the models. To further check the predictive ability of the models,
2
the predicted correlation coefficient (r pred) was also calculated. For a
2
36, 37
predictive QSAR model r pred value should be higher than 0.5.
Binding free energy calculations
Molecular docking
4
3
The Molecular Mechanics/Generalized Born Surface Area (MM/GBSA)
Molecular docking was carried out using the Surflex-Dock module in
Sybyl-X 2.0. The crystal structure of tubulin (PDB ID: 3UT5) was
and the Molecular Mechanics/Poisson Boltzmann Surface Area
44
(MM/PBSA) methodologies were carried out for binding free energy
calculations of systems based on the minimized structures or the MD
trajectories of protein-ligand complexes. The calculations were performed
on 50 snapshots (frequency 20) from the last 2 ns trajectories (1000
frames). All other parameters were kept at their default values. Here the
change in the conformational entropy upon ligands binding was not
considered because of the expensive computational cost and low
prediction accuracy. The corresponding calculation equations are listed as
follows:
3
8
downloaded from RSCB Protein Data Bank. The complex with colchicine
forming by Chains A and B of the tubulin heterodimer were considered for
docking. The ligands were docked into the colchicine binding site by an
empirical scoring fuction and a patented search engine in Surflex-Dock.
Before the docking process, the colchicine ligand containing in the tubulin
complex was extracted and water molecules were removed. Subsequently,
protein was prepared by using Biopolymer module implemented in Sybyl.
The polar hydrogen atoms were added. AMBER7 FF99 charges were
3
9
assigned to protein atoms. Other parameters were taken by default.
Then a protomol was generated based on original ligand. The molecule
was flexible docked into the active pocket. The Suflex-Dock total scores
and corresponding conformations were employed for further studies.
ꢀ
G
G
bind = ꢀH - TꢀS ≈ ꢀGgas + ꢀGsol - TꢀS;
ꢀ
gas = ꢀEele + ꢀEvdw; ꢀGsol=ꢀGPB/GB+ꢀGSA
Where ꢀGgas is the gas-phase interaction energy between protein
Molecular Dynamics (MD) simulations
and ligand, it includes ꢀEele (electrostatic), and ꢀEvdw (van der Waals)
energies. ꢀGsol is the sum of electrostatic solvation energy (polar
contribution), ꢀGPB/GB, and the nonelectrostatic solvation component
MD simulations were carried out by Amber 12.0 package. The general
amber force field (gaff) and the ff99SB force field were used for the ligands
(nonpolar contribution), ꢀGSA.
+
−
and the protein, respectively. The counter-ions, Na or Cl , were added to
4
0
neutralize the the unbalanced charges in the complexes. Each system
was added 10 Å out of the solute with an octahedral TIP3P water box to
reduce computational demand.
Chemistry
Commercially available chemicals were used without further purification
unless otherwise specified. Thin layer chromatography (TLC) analysis were
carried out on pre-coated plates silica gel G F254 and spots were visualized
under UV light (254 nm). Chromatography separations were performed on
silica gel (300-400 mesh) flash columns by GENERAL-REAGENT® and the
All the solvated simulations follow the procedure of minimization,
heating, density, equilibration and production. Initially to begin with, the
minimization of the entire system was performed by two steps. In the first
step, the atom position of all solute species was minimized with 500 cycles
of each steepest descent and conjugate gradient methods by restraining
1
13
19
F
eluting solvents are indicated in the procedures. H NMR, C NMR and
NMR spectra were recorded on a Bruker AVANCE III at 500 MHz, 125 MHz
-
1
-2
the protein-ligand complex with a force constant of 10 kcal·mol ·Å . In
or 470 MHz in chloroform-d using TMS (δ = 0.0 ppm) as an internal
the second step, the whole system was minimized without any restraints
6
standard or (Methyl Sulfoxide)-d . Chemical shifts are reported in δ parts
4
1
by 1000 cycles of steeps descent and 5000 cycles of conjugate gradient.
per million (ppm) with TMS and coupling constants (J) in Hz. HRMS were
recorded on a solariX 70 FT-MS spectrometer (Bruker) using anhydrous
dichloromethane as solvent. Melting points were measured using a
WRS-1C melting point apparatus (Shanghai ShenGuang Instrument Co.,
Ltd., Shanghai China). The purity of the targeted compounds were
Then the minimized system was gradually heated under NVT ensemble
from 0 to 300 K over a period of 20 ps, followed by density equilibration
under NPT ensemble at 300 K over 100 ps with weak restraint of 2
-1
-2
kcal·mol ·Å of constant pressure (1 atm). In this progress the SHAKE
algorithm was applied to all bonds involving hydrogen atoms, and the
measured by Agilent 1260 infinity HPLC with 75% CH
3 2
CN + 25% H O as
4
2
langevin dynamics was used for temperature control. Finally, a 20 ns
mobile phase.
production run was performed under NPT (1 atm) ensemble with cut-off
9

MedChemComm
Page 10 of 14
View Article Online
DOI: 10.1039/C7MD00416H
To a solution of compound 13a-4/13b-4 (1.6 mmol) in methanol (10 mL)
was added potassium carbonate (553 mg, 4.0 mmol) and heated to 60 °C
for 16 h, reflux. The excess methanol was evaporated under reduced
pressure, diluted with water, and extracted with EtOAc, dried over
General procedure for the preparation of 13a, 13b
General procedure for the synthesis of compounds 13a-2,
13b-2
anhydrous Na
purification of 13a-5 was performed by flash chromatography over
00-400 silica gel (pet ether : EtOAc = 30 : 1, v/v) to obtain a white solid.
2 4
SO and filtered, evaporated under reduced pressure. The
I
2
3
(5.1 g, 20.0 mmol) was dissolved in CHCl (150 mL) with stirring over 1.5
3
h. A 250 mL tinfoil wrapped round-buttomed flask was charged with
3a-1/13b-1 (20.0 mmol) and AgOTFA (4.5 g, 20.0 mmol). The solution was
The purification of 13b-5 was eluted with n-hexane, EtOAc to afford a
1
white solid.
added slowly into the tinfoil wrapped round-bottomed flask over 3 h, and
the reaction was stirred at room temperature for 24 h. The mixture was
General procedure for the synthesis of compounds 13a,
13b
filtered saturated Na
over anhydrous Na SO
The residue was purified via flash chromatography using CH
compound 13a-2, 13b-2 as a white solid.
2 2 3 3
S O , saturated NaHCO , washed with brine, dried
2
4
, filtered and concentrated under reduced pressure.
2
2
Cl to give
To a solution of compound 13a-5/13b-5 (50 mg) in dichloromethane (2 mL)
was added 3, 4, 5-trimethoxybenzoyl chloride (1.2 eq) followed by
General procedure for the synthesis of compounds 13a-3,
3b-3
4
anhydrous SnCl (1.0 eq) drop wise at 0 °C. The reaction mixture was
stirred at room temperature for 3 h and then quenched with ice and
stirred for 1 h. The organic layer was separated and the aqueous layer was
1
extracted with CH
dried over anhydrous Na
pressure. The purification was carried out by flash chromatography over
2 2
Cl several times. The combined organic extracts was
To a solution of compound 13a-2/13b-2 (7.46 mmol) in dichloromethane
40 mL) was added pyridine (1.2 g, 14.92 mmol) followed by drop wise
addition of acetic anhydride (1.54 g, 14.92 mmol) at 0 °C. The reaction
mixture was stirred for 4 h at ambient temperature and poured into CH Cl
washed with saturated NaHCO and brine solution. The organic layer was
dried over anhydrous Na SO , filtered and evaporated under reduced
pressure. The purification was performed by flash chromatography over
4
SO , filtered and evaporated under reduced
(
2
3
00-400 silica gel and eluted with pet-ether/EtOAc to afford compound
2
2
1
3a, 13b as a light yellow-green solid.
3
2
4
2
-iodo-5-methoxyphenol (13a-2)
3
00-400 silica gel and eluted with pet-ether/EtOAc to obtain compounds
1
White solid, yield: 52.87%. H NMR (500 MHz, CDCl
3
) δ = 7.49 (d, J = 9.0 Hz,
1
3a-3 as a colorless oily liquid, 13b-3 as a white crystal.
1
H), 6.60 (d, J = 2.5 Hz, 1H), 6.34 (dd, J = 6.0, 2.5 Hz, 1H), 5.31 (s, 1H), 3.77
(s, 3H).
General procedure for the synthesis of compounds 13a-4,
3b-4
2
-iodo-5-methoxyphenyl acetate (13a-3)
1
1
Colorless oily liquid, yield: 78.55%. H NMR (500 MHz, CDCl
3
) δ = 7.66 (d, J
To a solution of compound 13a-3/13b-3 (5.0 mmol) in DMF (10 mL) was
=
3
9.0 Hz, 1H), 6.69 (d, J = 3.0 Hz, 1H), 6.60 (dd, J = 6.0, 2.5 Hz, 1H), 3.77 (s,
added 1-ethynyl-4-ethoxybenzene (950 mg, 6.5 mmol) or
H), 2.36 (s, 3H).
1
1
0
-ethynyl-4-methxoybenzene (860 mg, 6.5 mmol), triethylamine (1.02 g,
0.0 mmol) followed by CuI (50 mg, 0.25mmol) and Pd(PPh Cl (175 mg,
.25 mmol) under N . The reaction mixture was heated at 100 °C for 6 h.
3
)
2
2
2
-((4-ethoxyphenyl)ethynyl)-5-methoxyphenyl acetate (13a-4)
2
1
Yellow solid, yield: 46.08%, mp 121.4-121.7 °C. H NMR (500 MHz, CDCl
3
) δ
The reaction was monitored by TLC and then slowly warmed to an
ambient temperature, extracted with EtOAc and water several times. The
combined organic layer was dried over anhydrous Na SO , filtered and
=
6
3
=
1
7.46 (d, J = 8.5 Hz, 1H), 7.40 (d, J = 8.5 Hz, 2H), 6.85 (d, J = 8.5 Hz, 2H),
.78 (dd, J = 6.0, 2.5 Hz, 1H), 6.67 (d, J = 2.5 Hz, 1H), 4.05 (q, J = 7.0 Hz, 2H),
2
4
1
3
3
.82 (s, 3H), 2.36 (s, 3H), 1.42 (t, J =7.0 Hz, 3H). C NMR (125 MHz, CDCl ) δ
concentrated under reduced pressure. The purification was carried out by
flash chromatography over 300-400 silica gel and eluted with
pet-ether/EtOAc to afford compound 13a-4, 13b-4 as a yellow solid.
168.77, 160.28, 159.02, 152.58, 133.40, 132.84, 115.23, 114.54, 112.01,
09.89, 108.26, 92.87, 82.84, 63.53, 55.56, 20.89, 14.76. ESI-HRMS (m/z):
+
calculated for C19
H
18
O
4
(M+H) , 311.12386, found: 311.12381.
General procedure for the synthesis of compounds 13a-5,
2
-(4-ethoxyphenyl)-6-methoxybenzofuran (13a-5)
13b-5
1
White solid, yield: 40.25%, mp 123.9-125.2 °C. H NMR (500 MHz, CDCl
3
) δ
=
7.73 (d, J = 9.0 Hz, 2H), 7.41 (d, J = 8.0 Hz, 1H), 7.06 (s, 1H), 6.96 (d, J =
1
0

Page 11 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
1
9
9
2
1
1
C
.0 Hz, 2H), 6.86 (dd, J = 6.0, 2.5 Hz, 1H), 6.80 (s, 1H), 4.09 (q, J = 7.0 Hz,
7.17 (s, 1H), 7.06 (d, J = 8.5 Hz, 2H), 3.90 (s, 3H), 3.82 (s, 3H). F NMR (470
1
3
13
H), 3.87 (s, 3H), 1.45 (t, J = 7.0 Hz, 3H). C NMR (125 MHz, CDCl
3
) δ =
MHz, DMSO) δ = -140.21. C NMR (125 MHz, DMSO) δ = 160.05, 156.03,
59.03, 157.70, 155.68, 155.39, 125.93, 123.47, 122.88, 120.64, 114.77,
150.84, 149.80 (d, J = 236.3 Hz), 145.83 (d, J = 13.8Hz), 126.29, 122.96,
11.67, 99.42, 95.94, 63.55, 55.75, 14.85. ESI-HRMS (m/z): calculated for
121.40 (d, J = 10.0 Hz), 115.00, 107.03 (d, J = 22.5 Hz), 100.68, 97.60,
+
+
17
H
16
O
3
(M+H) , 268.10994, found: 268.10936.
56.93, 55.74. ESI-HRMS (m/z): calculated for C16
found: 273.09217.
H13FO
3
(M+H) , 273.08823,
(2-(4-ethoxyphenyl)-6-methoxybenzofuran-3-yl)(3,4,5-trimethoxyphenyl)
methanone (13a)
(5-fluoro-6-methoxy-2-(4-methoxyphenyl)benzofuran-3-yl)(3,4,5-trimeth
oxy phenyl)methanone (13b)
1
Purity: 97%, yellow-green solid, yield: 70.76%, mp 135.9-137.4 °C. H NMR
500 MHz, CDCl ) δ = 7.55 – 7.53 (m, 3H), 7.12 (s, 2H), 7.08 (d, J = 2.0 Hz,
H), 6.91 (dd, J = 6.5, 2.0 Hz, 1H), 6.79 (d, J = 9.0 Hz, 2H), 4.00 (q, J = 7.0 Hz,
1
(
3
Purity: 98%, yellow-green solid, yield: 70.07%, mp 154.5-154.7 °C. H NMR
1
2
(500 MHz, CDCl
3
) δ = 7.50 (d, J = 8.5 Hz, 2H), 7.38 (d, J = 10.5 Hz, 1H), 7.14
1
3
H), 3.88 (s, 3H), 3.86 (s, 3H), 3.69 (s, 6H), 1.39 (t, J = 7.0 Hz, 3H). C NMR
(d, J = 6.5 Hz, 1H), 7.08 (s, 2H), 6.79 (d, J = 8.5 Hz, 2H), 3.94 (s, 3H), 3.85 (s,
1
9
13
(
125 MHz, CDCl
3
) δ = 190.90, 159.99, 158.50, 157.34, 154.69, 152.86,
3H), 3.76 (s, 3H), 3.68 (s, 6H). F NMR (470 MHz, CDCl
3
) δ = -138.74. C
1
1
42.39, 132.68, 129.84, 122.21, 121.99, 121.70, 114.54, 114.35, 112.60,
NMR (125 MHz, CDCl ) δ = 190.41, 160.76, 158.10, 152.91, 150.58 (d, J =
3
07.42, 95.65, 63.55, 60.93, 56.10, 55.76, 14.66. ESI-HRMS (m/z):
240 Hz), 149.85, 146.81 (d, J = 13.8 Hz), 142.60, 132.39, 129.92, 122.14,
120.79 (d, J = 8.8 Hz), 114.74 (d, J = 3.8 Hz), 113.90, 107.60, 107.42, 96.29,
+
calculated for C27
H
26
O
7
(M+H) , 463.17121, found: 463.17670.
6
0.93, 56.61, 56.12, 55.32. ESI-HRMS (m/z): calculated for C26H23FO
7
4
-fluoro-2-iodo-5-methoxyphenol (13b-2)
+
(
M+H) , 467.14614, found: 467.15066.
1
White solid, yield: 39.92%, mp 82.4-83.4 °C. H NMR (500 MHz, CDCl
3
) δ =
) δ =
Procedure for the preparation of 13f
7
3
1
1
.31 (d, J = 10 Hz, 1H), 6.67 (d, J = 7.5 Hz, 1H), 5.13 - 5.12 (m, 1H), 3.85 (s,
1
9
13
3
H). F NMR (470 MHz, CDCl ) δ = -143.08. C NMR (125 MHz, CDCl
3
Synthesis of 2-Bromo-1-(4-ethoxyphenyl)ethanone (13f-2)
51.73 (d, J = 2.5 Hz), 149.2, 146.82 (d, J = 240.6Hz), 123.70 (d, J = 22.5 Hz),
00.30, 70.97 (d, J = 7.5 Hz), 56.27. ESI-HRMS (m/z): calculated for C
7
H
6
FIO
2
To a solution of the 1-(4-ethoxyphenyl)ethan-1-one (3.284 g, 20.0 mmol) in
Et O (30 mL) was added Bromine (3.2 g, 20.0 mmol). The solution was
+
(
M+H) , 268.94301, found: 268.94769.
2
stirred overnight at room temperature (until the color of the solution
changed from red to light-yellow). The reaction progress was monitored by
TLC. When the starting material disappeared , the reaction was quenched
with water, and extracted with EtOAc. The combined organic phases were
4
-fluoro-2-iodo-5-methoxyphenyl acetate (13b-3)
1
White crystal, yield: 81.28%, mp 84.6-84.9 °C. H NMR (500 MHz, CDCl
3
) δ =
1
9
7
.47 (d, J = 10 Hz, 1H), 6.74 (d, J = 7.5 Hz, 1H), 3.85 (s, 3H), 2.35 (s, 3H). F
washed with saturated aqueous NaHCO
3
solution, saturated aqueous
1
3
3 3
NMR (470 MHz, CDCl ) δ = -135.80. C NMR (125 MHz, CDCl ) δ = 168.57,
Na solution, brine, and dried over anhydrous Na SO . After being
2
S
2
O
3
2 4
1
1
50.05 (d, J = 247.5 Hz), 148.50 (d, J = 11.3 Hz), 147.60 (d, J = 2.5 Hz),
filtered and concentrated, the residue was purified by flash
chromatography over 300-400 silica gel (pet ether : EtOAc = 50 : 1, v/v) to
get compound 13f-2 1.54 g, white crystal, yield 31.57%.
25.22 (d, J = 21.3 Hz), 108.28, 99.99, 56.43, 21.15. ESI-HRMS (m/z):
+
calculated for C
9
8
H FIO
3
(M+H) , 310.95357, found: 310.95754.
4
-fluoro-5-methoxy-2-((4-methoxyphenyl)ethynyl)phenyl acetate (13b-4)
Synthesis of 2-(4-ethoxyphenyl)-6-methoxy-1H-indole (13f-3)
1
Yellow solid, yield: 32.80%, mp 112.9-114.2 °C. H NMR (500 MHz, CDCl
3
) δ
To
a boiling mixture of m-anisidine (740 mg, 6.0 mmol) and
=
7.43 (d, J = 9.0 Hz, 2H), 7.27 (d, J = 11.5 Hz, 1H), 6.89 (d, J = 8.5 Hz, 2H),
1
9
N,N-dimethylaniline (3.5 mL) was added compound 2 (730 mg in EtOAc,
3.0 mmol) slowly by syringe. After addition, the mixture was kept at 170 °C
for 2 h. The reaction mixture was cooled to room temperature and a
dark-colored solid was formed. EtOAc was added along with HCl (2N). The
aqueous layer was extracted with EtOAc several times. The combined
6
.75 (d, J = 7.5 Hz, 1H), 3.90 (s, 3H), 3.84 (s, 3H), 2.37 (s, 3H). F NMR (470
1
3
3 3
MHz, CDCl ) δ = -137.90. C NMR (125 Hz, CDCl ) δ = 168.81, 159.88, 149.6
(
d, J = 243.8 Hz), 148.24 (d, J = 12.5 Hz), 147.99 (d, J = 3.8 Hz), 132.94,
1
8
19.00 (d, J = 21.3 Hz), 114.91, 114.09, 109.59 (d, J = 10Hz), 107.86, 93.63,
1.94, 56.38, 55.31, 20.77. ESI-HRMS (m/z): calculated for C18H15FO
4
+
2 4
organic layers were washed with brine, and dried over anhydrous Na SO .
(
M+Na) , 337.09544, found: 337.08508.
After being filtered and concentrated, the crude compound was purified
by flash chromatography over 300-400 silica gel (pet ether : EtOAc = 30 : 1,
v/v) to afford compound 13f-3 95 mg, white solid, yield 11.85%.
5
-fluoro-6-methoxy-2-(4-methoxyphenyl)benzofuran (13b-5)
1
White solid, yield: 25.25%, mp 118.6-120.4 °C. H NMR (500 MHz, DMSO) δ
=
7.79 (d, J = 9.0 Hz, 2H), 7.49 (d, J = 7.0 Hz, 1H), 7.45 (d, J = 11.0 Hz, 1H),
1
1

MedChemComm
Page 12 of 14
View Article Online
DOI: 10.1039/C7MD00416H
Synthesis of (2-(4-ethoxyphenyl)-6-methoxy-1H-indol-3-yl)
3,4,5-trimeth oxyphenyl) methanone (13f)
at 37 °C for 24 h. 100 µL/well containing corresponding drug mediums was
added. The negative control groups, menstruum control groups and
positive control groups were set up, 5 wells of each group. The DMSO
concentration was kept below 0.01% in cell culture, so it did not effect on
cell growth. After being cultured in 5% CO2 incubator at 37 °C for 72 h,
CCK-8 solution was added (10 µL/well) to each well of 96-well plates and
incubated for another 4 h. The OD values were measured by microplate
reader (Thermo MK3, US) at 450 nm. the IC50 values and cancer cells
inhibition ratio were obtained. Experiments were repeated two times. All
the cell lines used for experiment were obtained from Nanjing OGpharma
Co., Ltd., China.
(
To a well stirred solution of compound 13f-3 (60 mg, 0.225 mmol), in
,2-dichlorobenzene (5 mL) was added 3,4,5-trimethoxybenzoyl chloride
104 mg, 0.45 mmol), anhydrous potassium carbonate (83 mg, 0.6 mmol).
1
(
The reaction mixture was heated to reflux for 12 h. After cooling down to
room temperature, solvent was removed by rotary evaporators in vacuo. A
dark solid formed which was dissolved in dichloromethane (1 mL),
purification through thin layer chromatography (silica gel HSGF254)
afforded compound 13f 10 mg, dark yellow solid, yield 7.23%.
2-Bromo-1-(4-ethoxyphenyl)ethanone (13f-2)
In vitro Tubulin Polymerization Assay
1
White crystal, yield: 31.57%. H NMR (500 MHz, CDCl
3
) δ = 7.95 (d, J = 9.0
Pig brain microtubule protein was isolated by three cycles of
temperature-dependent assembly/disassembly according to Shelanski et
Hz, 2H), 6.93 (d, J = 8.5 Hz, 2H), 4.39 (s, 2H), 4.10 (q, J = 7.0 Hz, 2H), 1.44 (t,
J = 7.0 Hz, 3H).
4
7
al. in 100 mM PIPES (pH 6.5), 1 mM MgSO , 2 mM EGTA, 1 mM GTP and 1
4
mM 2-mercaptoethanol. In the first cycle of polymerization, glycerol and
2-(4-ethoxyphenyl)-6-methoxy-1H-indole (13f-3)
phenylmethylsulfonyl fluoride were added to
4 M and 0.2 mM,
respectively. Homogeneous tubulin was prepared from microtubule
protein by phosphocellulose (P11) chromatography as described 28. The
purified proteins were stored in aliquots at -70 °C.
1
White solid, yield: 11.85%, mp 129.7-130.5 °C. H NMR (500 MHz, CDCl
3
) δ
=
8.30 (s, 1H), 7.55 (d, J = 8.5Hz, 2H), 7.11 (t, J = 8.0Hz, 1H), 7.01 (d, J = 8.0
Hz, 1H), 6.95 (d, J = 8.5 Hz, 2H), 6.85 (s, 1H), 6.57 (d, J = 7.5Hz, 1H), 4.06 (q,
1
3
J = 7.0 Hz, 2H), 3.99 (s, 3H), 1.45 (t, J = 7.0 Hz, 3H). C NMR (125 MHz,
Tubulin protein (4.8 mg/mL) was mixed with different
CDCl ) δ = 158.59, 153.18, 138.01, 136.66, 126.31, 125.12, 122.66, 120.14,
3
2
concentrations of compound in PEM buffer (100 mM PIPES, 1 mM MgCl ,
1
15.05, 104.30, 100.08, 95.99, 63.60, 55.40, 14.84. ESI-HRMS (m/z):
and 1 mM EGTA) containing 1 mM GTP and 5 % glycerol. Microtubule
polymerization was monitored at 37 °C by light scattering at 340 nm using
a SPECTRA MAX 190 (Molecular Device) spectrophotometer. The plateau
absorbance values were used for calculations.
+
calculated for C17
H17NO
2
(M+H) , 268.12928, found: 268.13335.
(2-(4-ethoxyphenyl)-6-methoxy-1H-indol-3-yl)(3,4,5-trimethoxyp
henyl) methanone (13f)
Acknowledgments
1
Purity: 95%, yellow solid, yield: 7.23%, mp n.d. H NMR (500 MHz, CDCl
3
) δ
10.64 (s, 1H), 7.67 (d, J = 8.5 Hz, 2H), 7.63 (d, J = 8.5 Hz, 1H), 7.03 (s, 2H),
.99 (d, J = 8.0 Hz, 2H), 6.87 (s, 1H), 6.57 (d, J = 8.5 Hz, 1H), 4.13 - 4.09 (m,
Thanks for financial support given by Natural Science Foundation of
Shanghai (No. 15ZR1440400), National Natural Science Foundation of
China (No. 21472126 and 21602135), Collaborative Innovation fund
=
6
2
1
3
H), 4.07 (s, 3H), 3.96 (s, 3H), 3.92 (s, 6H), 1.45 (t, J = 7.0 Hz, 3H). C NMR
(
XTCX2016-14), and the Shanghai Municipal Education Commission
Plateau Discipline Construction Program).
(
125 MHz, CDCl
3
) δ = 195.72, 158.91, 157.93, 152.90, 148.98, 145.59,
(
1
9
38.50, 137.95, 134.78, 130.48, 126.55, 120.43, 115.08, 113.88, 106.87,
9.84, 95.57, 63.62, 60.99, 56.34, 55.70, 14.83. ESI-HRMS (m/z): calculated
+
Conflict of Interest
for C27H27NO
6
(M+H) , 462.19111, found: 462.19010.
The authors declare no competing interest.
Biological assays
References
Cell proliferation inhibition assays (CCK-8 assay)
1
K. H. Downing, Annu. Rev. Cell Dev. Biol. , 2000, 16, 89-111.
The cell viability detection kit, EnoGeneCellTM Counting Kit-8 (CCK-8) was
2 L. A. Amos, Org. Biomol. Chem. , 2004, 2, 2153-2160.
4
5, 46
3 S. Honore, E. Pasquier, D. Braguer, Cell. Mol. Life Sci. , 2005, 62,
used for cell proliferation inhibition assays.
The experiments were
3
4
039-3056.
carried out on the cells with the ratio of living cells up to 90%. The cells
were digested, counted and made into cell suspensions (1*105 cells/mL).
A. L. Risinger, F. J. Giles, S. L. Mooberry, Cancer Treat. Rev. ,
2008, 35, 255-261.
D. G. Kingston, J. Nat. Prod. , 2009, 72, 507-515.
5
1
00 µL cell suspensions/well (1*104 cells/well) were seeded in 96-well
plates (Costar), cultured in 5% CO incubator (MCO-15AC, SANYO, Japan)
2
1
2

Page 13 of 14
MedChemComm
View Article Online
DOI: 10.1039/C7MD00416H
6
G. Pettit, S. B. Singh, E. Hamel, C. M. Lin, D. S. Alberts, D.
34 K. Hiroya, N. Suzuki, A. Yasuhara, J. Chem. Soc. , Perkin Trans.
2000, 1, 4339-4346.
Garcia-Kendal, Experientia. , 1989, 45, 209-211.
7
8
N. H. Nam, Curr. Med. Chem. , 2003, 10, 1697-1722.
K. Ohsumi, R. Nakagawa, Y. Fukuda, T. Hatanaka, Y. Morinaga, Y.
35 R. Remagnoli, P. G. Baraldi, M. K. Salvador, J. Med. Chem. ,
2012, 55, 5433-5445.
Nihei, J. Med. Chem. , 1998, 41, 3022-3032.
36 P. K. Ojha, K. Roy, Chemometr Intell Lab. , 2011, 109, 146-161.
37 A. Tropsha, Mol Inf. , 2010, 29, 476-488.
9
1
1
4
1
1
1
G. R. Pettit, M. R. Rhodes, Anti-cancer Drug Des. , 1998, 13,
83-191.
0 G. C. Tron, T. Pirali, G. Sorba, F. Pagliai, J. Med. Chem. , 2006,
9, 3033-3044.
1 T. Hatanaka, K. Fujita, K. Ohsumi, Bioorg. Med. Chem. Lett. ,
998, 8, 3371-3374.
38 F. M. Ranaivoson, B. Gigant, S. Berritt, M. Joullie, M. Knossow,
Acta Crystallogr Sect D: Biol Crystallogr D. , 2012, 68, 927-934.
39 A. Jain, N. Jay, J. Med. Chem. , 2003, 46, 499-511.
40 H. Y. Sun, Y. Y. Li, M. Y. Shen, S. Tian, T. J. Hou, Phys. Chem.
Chem. Phys. , 2014, 16, 22035-22045.
2 L. Wang, K. W. Woods, Q. Li, K. J. Barr, R. W. McCroskey, J.
41 N. Sanghai, V. Jain, R. Preet, S. Kandekar, Med. Chem. Comm. ,
2014, 5, 766-782.
Med. Chem. , 2002, 45, 1697-1711.
1
2
1
3 R. Romagnoli, P. G. Baraldi, A. Brancale, A. Ricci, J. Med. Chem. ,
011, 54, 5144-5153.
4 R. Romagnoli, P. G. Baraldi, M. K. Salvador, D. Preti, M. A.
42 S. Miyamoto, P. A. Kollman, J Comput Chem. , 1992, 13,
952-962.
43 P. A. Kollman, I. Massova, C. Reyes, B. Kuhn, S. Huo, L. Chong,
Acc Chem Res. , 2000, 33, 889-897.
44 N. Homeyer, H. Gohlke, Mol Inf. , 2012, 31, 114-122.
45 M. Linsalata, M. Notarnicola, V. Tutino, M. BifꢁLco, Anticancer
Res. , 2010, 30 (7), 2583-2589.
46 M. Sꢁgiyama, H. Takahashi, K. Hosono, H. Endo, S. Kato, Int J
Oncol. , 2009, 34 (2), 339-344.
Tabrizi, J. Med. Chem. , 2012, 55, 5433-5445.
5 R. Romagnoli, P. G. Baraldi, O. Cruz-Lopez, C. L. Cara, M. D.
Carrion, J. Med. Chem. , 2010, 53, 4248-4258.
6 R. Romagnoli, P. G. Baraldi, M. K. Salvador, D. Preti, M. A.
Tabrizi, A. Brancale, J. Med. Chem. , 2012, 55, 475-488.
7 R. Romagnoli, P. G. Baraldi, M. K. Salvador, D. Preti, J. Med.
Chem. , 2013, 56, 2606-2618.
8 R. Romagnoli, P. G. Baraldi, C. Lopez-Cara, D. Preti, M. A.
Tabrizi, J. Balzarini, M.Bassetto, J. Med. Chem. , 2013, 56,
1
1
1
47 M. L. Shelanski, F. Gaskin, C. R. Cantor, Proc Natl Acad Sci U S
A. , 1973, 70, 765-768.
1
9296-9309.
19 R. Romagnoli, P. G. Baraldi, M. K. Salvador, F. Prencipe, C.
Lopez-Cara, S. S. Ortega, A. Brancale, E. Hamel, J. Med. Chem. ,
2
2
8
2
015, 58, 3209-3222.
0 M. L. Brown, J. M. Rieger, T. L. Macdonald, Med. Chem. , 2000,
, 1433-1441.
1 S. Y. Liao, J. C. Chen, T. F. Miao, Y. Shen, K. G. Zheng, J. Enzyme
Inhib. Med. Chem. , 2010, 25, 421-429.
2
6
2
2 B. L. Flynn, P. Verdier-Pinard, E. Hame, Org. Lett. , 2001, 3,
51-654.
3 R. Romagnoli, P. G. Baraldi, M. G. Pavani, M. A. Tabrizi, D. Preti,
F. Fruttarolo, L. Piccagli, K. Jung, E. Hamel, M. Borgatti, R. Gambari,
J. Med. Chem. , 2006, 49, 3906-3915.
24 B. L. Flynn, G. S. Gill, D. W. Grobelny, J. H. Chaplin, D. Paul, A. F.
Leske, J. Med. Chem. , 2011, 54, 6014-6027.
2
2
2
5 D. J. Kerr, E. Hamel, M. K. Jung, B. L. Flynn, Bioorg. Med. Chem. ,
007, 15, 3290-3298.
6 R. Romagnoli, P. G. Baraldi, T. Sarkar, M. D. Carrion, C. L. Cara,
O. Cruz-Lopez, D. Preti, J. Med. Chem. , 2008, 51, 1464-1468.
7 R. Romagnoli, P. G. Baraldi, M. D. Carrion, C. L. Cara, D. Preti, F.
2
Fruttarolo, M. G. Pavani, M. A. Tabrizi, M. Tolomeo, S. Grimaudo,
J. Med. Chem. , 2007, 50, 2273–2277.
28 R. Romagnoli, P. G. Baraldi, T. Sarkar, M. D. Carrion, C. L. Cara,
O. Cruz-Lopez, C. L. Cara, M. Tolomeo, S. Grimaudo, A. D. Cristina,
Bioorg. Med. Chem. , 2008, 16, 8419-8426.
2
3
2
3
9 L. Ståhle, S. Wold, J. Chemometr. , 1987, 1, 185-196.
0 V. K. Reddy, J. V. Rao, L. B. Reddy, Asian J. Chem. , 2015, 27 (6),
245-2248.
1 Patent: Indole-Containing Compounds with Anti-tubulin and
Vascular Targeting Activity. , WO2004099139A1.
2 J. Z. Chen, D. L. Liu, N. Butt, Angewandte Chemie. , 2013, 52
44), 11632-11636.
3
(
33 J. Liu, X. Zhou, C. Wang, Chem. Commun. , 2016, 52 (29),
5152-5155.
1
3

MedChemComm
Page 14 of 14
View Article Online
DOI: 10.1039/C7MD00416H
Graphical abstract
Compound 13a, more effective than CA-4 against the HepG2 cells and tubulin, and
the proposed binding mode for 13a.