Design, synthesis and characterization of potent microtubule inhibitors with dual anti-proliferative and anti-angiogenic activities

Article abstract of DOI:10.1016/j.ejmech.2018.07.043

Microtubule has been an important target for anticancer drug development. Here we report the discovery and characterization of a series of fused 4-aryl-4H-chromene-based derivatives as highly potent microtubule inhibitors. Among a total of 37 derivatives synthesized, 23 exhibited strong in vitro anti-proliferative activities against A375 human melanoma cells. The relationship between the biological activities of these microtubule inhibitors and their chemical structure variations was analyzed. Studies of compounds 27a, 19a and 9a in parallel with colchicine as the positive control compound in a panel of biological assays revealed that these compounds blocked cell cycle progression, increased apoptosis, and inhibited HUVEC capillary tube formation at low nanomolar concentrations. The most potent compound 27a was also tested in eight additional cancer cell lines besides A375 cells and two non-cancer cells and showed potent and selective activity on these cancer cells. To understand the molecular and structure mechanism of action of these compounds, tubulin polymerization and molecular docking studies were carried out for 27a as the representative. The results were consistent with the mechanism by which 27a interacts with the colchicine binding site on tubulin and disrupts tubulin polymerization. With potent dual actions of microtubule destabilization and vascular disruption described above, this small molecule can serve as a valuable research probe of the function and role of microtubules in human diseases and promising lead for developing new therapeutic agents.

Full text of DOI:10.1016/j.ejmech.2018.07.043

Accepted Manuscript
Design, synthesis and characterization of potent microtubule inhibitors with dual anti-
proliferative and anti-angiogenic activities
Huijun Zhang, Xiong Fang, Qian Meng, Yujia Mao, Yan Xu, Tingting Fan, Jing An,
Ziwei Huang
PII:
S0223-5234(18)30599-3
10.1016/j.ejmech.2018.07.043
EJMECH 10578
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 16 April 2018
Revised Date: 30 June 2018
Accepted Date: 16 July 2018
Please cite this article as: H. Zhang, X. Fang, Q. Meng, Y. Mao, Y. Xu, T. Fan, J. An, Z. Huang,
Design, synthesis and characterization of potent microtubule inhibitors with dual anti-proliferative
and anti-angiogenic activities, European Journal of Medicinal Chemistry (2018), doi: 10.1016/
j.ejmech.2018.07.043.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Design, synthesis and characterization of potent
microtubule inhibitors with dual
anti-proliferative and anti-angiogenic activities
1
1,2
1
1
1
1
Huijun Zhang , Xiong Fang , Qian Meng , Yujia Mao , Yan Xu , Tingting Fan ,
3
,4,*
1,3,*
Jing An , Ziwei Huang
1
School of Life Sciences, Tsinghua University, Beijing, 100084, P. R. China
2
Tsinghua-Peking Joint Center for Life Sciences, Tsinghua University, Beijing, 100084, P.
R. China
3
Department of Medicine, University of California San Diego, La Jolla, CA 92093
4
Nobel Institute of Biomedicine, Zhuhai, 519080, P. R. China
*
Corresponding authors:
Dr. Ziwei Huang. Email: zwhny@yahoo.com. Phone: 0086-10-6279-6890.
Dr. Jing An. Email: jan@ucsd.edu. Phone: 858-822-0333.
Keywords: Fused 4-aryl-4H-chromenes; microtubule inhibitors; anti-proliferation;
anti-angiogenesis; drug development
1

ACCEPTED MANUSCRIPT
ABSTRACT
Microtubule has been an important target for anticancer drug development. Here we
report the discovery and characterization of a series of fused 4-aryl-4H-chromene-based
derivatives as highly potent microtubule inhibitors. Among a total of 37 derivatives
synthesized, 23 exhibited strong in vitro anti-proliferative activities against A375 human
melanoma cells. The relationship between the biological activities of these microtubule
inhibitors and their chemical structure variations was analyzed. Studies of compounds
2
7a, 19a and 9a in parallel with colchicine as the positive control compound in a panel of
biological assays revealed that these compounds blocked cell cycle progression,
increased apoptosis, and inhibited HUVEC capillary tube formation at low nanomolar
concentrations. The most potent compound 27a was also tested in eight additional cancer
cell lines besides A375 cells and two non-cancer cells and showed potent and selective
activity on these cancer cells. To understand the molecular and structure mechanism of
action of these compounds, tubulin polymerization and molecular docking studies were
carried out for 27a as the representative. The results were consistent with the mechanism
by which 27a interacts with the colchicine binding site on tubulin and disrupts tubulin
polymerization. With potent dual actions of microtubule destabilization and vascular
disruption described above, this small molecule can serve as a valuable research probe of
the function and role of microtubules in human diseases and promising lead for
developing new therapeutic agents.
1
. INTRODUCTION
Cancer, one of the major causes of death worldwide [1], is characterized by abnormal
2

ACCEPTED MANUSCRIPT
cell proliferation and inappropriate cell survival. One approach to prevent this
proliferation to promote cancer cell death is to use chemotherapeutic drugs that target
microtubules [2, 3]. The well-organized function and regulated dynamics of microtubules
contribute to the structure of the cytoskeleton, mediation of intracellular transport [4, 5],
cell division, signaling networks, and many other biological processes [6, 7], making
microtubules successful and efficacious cellular targets for the development of traditional
chemotherapeutic agents [8-11]. For example, taxanes [12, 13] and vinca alkaloids [14,
1
5] have been used for the treatment of leukemia and solid tumors, and additional
microtubule-targeting agents are still being pursued intensively [16-19].
Colchicine, extracted from the meadow saffron colchicium autumnale, is one of the
best known microtubule-destabilizing agents [20]. Colchicine binding-site inhibitors
(CBSI) are also potent destroyers of tumor vasculature [6, 20-27]. They bind to tubulins
in endothelial cells of tumor vessels, disrupt cytoskeleton, cell shape, and cellular
connections and eventually interrupt of the blood flow to the tumor [28]. Also they can
selectively affect tumor neovasculature, rather than normal tissues [28-32]. Several small
molecule vascular disrupting agents (VDAs), including CA-4P, Oxi4503, AVE8062, and
EPC2407, are currently undergoing clinical trials [32, 33]. Additional extensive efforts on
the design and discovery of novel VDAs would further expedite the development of these
drugs that target microtubule and show special advantages over other cancer therapies
[34-39].
One of the drug development approaches has been based on the use of
4
-aryl-4H-chromenes, which was emerged from the studies of natural products and
synthetic small molecules. Their broad pharmacological properties have made them
3

ACCEPTED MANUSCRIPT
medically privileged scaffolds [40-43]. Extensive research indicates that they exhibit
strong antitumor activities through multiple pathways, including the inhibition of
microtubule polymerization, disruption of tumor vasculature, and antagonism of the Wnt
pathway. One representative compound, EPC2407, which has entered into Phase II
clinical trials as a vascular-disrupting agent, has also been identified as a tubulin inhibitor
binding to the colchicine-binding site [44-46]. Similarly, the 4H-chromene analog 1,
featuring a sesamol-derived ring and resembling closely the structure of podophyllotoxin
(PT), exhibited strong antimitotic microtubule destabilizing activity (Figure 1) [47].
Subsequent studies reported that substitution of 7-methoxy, 2-NH , and 3-CN groups for
2
the methylenedioxy ring A and lactone D, respectively, in oxa-PTs yielded
7
6
-methoxy-4-aryl-4H-chromene 2 that showed markedly improved cytotoxicity against
0 human cancer cell lines [48]. Similarly, 5-pyrido 4H-chromene analog 3 displayed
Wnt pathway inhibitory activity (IC = 6 nM) and potent anti-proliferative activity
5
0
against HCT116 human colon carcinoma cells (IC50 = 15 nM) [49, 50]. Furthermore,
virtual screening and SAR-guided modification has identified 4-heteroaryl-4H-chromene
(HFI437) as a first generation small molecule inhibitor of insulin-regulated
amino-peptidase (IRAP) [51]. Overall, these highly functionalized 4H-chromenes have
shown diverse bioactivities. Most of the 4-aryl-4H-chromene-based derivatives have
great advantages for clinical applications as they possess dual biological activities (i.e.,
tubulin inhibition and vascular disruption). Furthermore, these molecules can overcome
4

ACCEPTED MANUSCRIPT
drug resistance by binding directly to the colchicine binding site on tubulin and
selectively target established tumor vasculature [35, 37, 41].
Our laboratory has previously developed mHA11, a 4-aryl-4H-chromene derivative
that inhibits tubulin by binding to the colchicine-binding site [52]. mHA11 was derived
by modification of HA14-1, a small molecule originally reported by us as an inhibitor of
both tubulin and Bcl-2 [52, 53]. The reported potent bioactivities of 4-aryl-4H-chromenes
with dual actions as tubulin-binding VDAs prompted us to investigate whether novel and
potent chromene-based VDAs could be developed by modifying the structure of mHA11.
To our knowledge, a number of structure-activity relationship (SAR) studies have
established that the polyalkoxybenzene moiety of the 4H-chromenes is a crucial
pharmacophore for mimicking colchicine and podophyllotoxin, and that small
3 2 3 2 2
hydrophobic groups, such as N(CH ) , -OCH , and -NH at the 7- or 8-positions or 2-NH
and 3-CN, generate compounds that inhibit cancer cell growth. In addition, fusion of the
pyrrole ring at the 7,8-position also increases the potency [44]. Fusion of rings at the
6
4
,7-position has not yet been explored in the design of new classes of fused
-aryl-4H-chromene-based derivatives. Thus, we designed different five-member or
six-member rings fused at the 6,7- or 7,8-positions of mHA11 to investigate their
activities (Figure 2). We fused a 1,3-dioxolane at the 6,7-position, following the
introduction of an -OH group at the C-8 position. We also prolonged the fused
5

ACCEPTED MANUSCRIPT
4
H-chromenes by designing two classes of linear and angular 4-aryl-4H-chromene
derivatives by incorporating some different groups into the fused imidazole ring.
Figure 1. 4-aryl-4H-chromene-based derivatives and colchicine binding site inhibitors
Figure 2. Structural modifications of the lead compound mHA11
2
. RESULTS
6

ACCEPTED MANUSCRIPT
2
.1 Chemistry
Fused 4-aryl-4H-chromene derivatives were prepared by a one-pot three-component
cascade reaction of starting materials 4 (different substituted benzaldehydes), ethyl
cyanoacetate or malononitrile, and synthesized intermediates 5 (different fused phenol
analogs). The synthetic route was described in Scheme 1. The three-component reaction
was carried out by three steps in the presence of piperidine at room temperature (rt) or 80
℃. The Knoevenagel condensation of aldehydes and active methylene compounds (ethyl
cyanoacetate or malononitrile) gave the first intermediate (benzylidenemalononitrile).
The ortho-position of the phenols with electron-rich density then underwent Michael
addition to form the next intermediate. Subsequently, the phenolic hydroxyl group
participated in an intramolecular attack at the nitrile group and underwent
tautomerization (namely, hetero-Thorpe-Ziegler reaction) [47].
Scheme 1. Synthesis of 6,7-fused or 7,8-fused-4-aryl-4H-chromene derivatives via a
three-component cascade reaction.
R2
R2
R2
1. Knoevenagel reaction
2. Michael reaction
3. hetero-Thorpe-Ziegler reaction
R3
R1
R3
R1
R1
R3
B
O
+
NC
+
n
A
O
piperidine, EtOH, N2, rt or 80 o
C
OH
OH
B
6
7
X
X
or
R4
or
CHO
4
n
A
7
CN
CN
O
NH2
O
NH2
5
8
R4
C
8
D
n
C
2 3
X = -COOCH CH , -CN
D
6,7-fused or 7,8-fused-4-aryl-4H-chromene
n
derivatives
The synthetic approaches of different fused phenol analogs were summarized in Scheme
2
. Indolin-6-ol 9 was prepared according to a modified Leimgruber-Batcho indole
7

ACCEPTED MANUSCRIPT
synthesis [54]. Commercially available 4-methyl-3-nitrophenol was protected with a
benzyl group to provide the intermediate 7, followed by condensation reaction of 7 with
DMF-DMA and pyrrolidine, reductive cyclization and deprotection to obtain
6
2
-hydroxyindole 8 in one step using 10% Pd-C and H , and then reduction of indole ring
in the presence of NaBH(CN) and AcOH to yield the desired 9 (Scheme 2a).
3
Simultaneously, following the same reductive condition, the indolin-4-ol 11 was obtained
by using the purchased 10 as the starting material (Scheme 2b). Amide reduction of 12 in
the presence of LiAlH gave 1,2,3,4-tetrahydroquinolin-7-ol 13 (Scheme 2c), and
4
5
1
-hydroxylquinoline
14
was
hydrogenated
with
PtO2
to
yield
,2,3,4-tetrahydroquinolin-5-ol 15 under acidic conditions (Scheme 2d).
Next, sesamol (16) was treated with MOMCl to give the MOM-protected intermediate
7, followed by the regioselective lithiation of 17 with n-BuLi, boration with B(OMe)
in a one-pot synthesis to afford the intermediate 18 according to
the reported method [55]. The protection of the MOM group afforded the sesamol analog
9 under acidic condition (Scheme 2e). Another sesamol analog 22 was prepared by
1
3
,
2 2
and oxidation with H O
1
treating the 3,4-dihydroxybenzaldehyde 20 with 1,2-dibromoethane in the presence of
K CO to yield the intermediate 21 which was further oxidized with mCPBA (Scheme
2
3
2
f). Replacement of one oxygen from the dioxane ring of 22 with nitrogen gave the fused
phenol analog 27, which was synthesized with the route described in Scheme 2g.
According to the reference method, the starting material was 4-methoxy-2-nitrophenol
8

ACCEPTED MANUSCRIPT
2
3, which was reduced to 2-amino-4-methoxyphenol 24 using Fe powder as the
reductant, followed by ring closure of 24 with 2-chloroacetyl chloride to yield 25 under
alkaline conditions [56]. Reduction of 25 then afforded the intermediate 26, which was
further treated with BBr3 to remove the methyl group to obtain 27 in the final step.
Eventually, as shown in Scheme 1, these fused phenol analogs (9, 11, 13, 15, 19, 22,
2
1
7) were converted to the corresponding desired compounds 9a-i, 11a-c, 13a-c, 15a-b,
9a-b, 22a, and 27a-b, respectively, following the general procedure for the one-pot
three-component cascade reaction. These synthesized compounds are summarized in
Table 1.
9

ACCEPTED MANUSCRIPT
Scheme
2.
Synthetic
route
for
fused
phenol
analogs.
Reagents and Conditions: (a) BnBr, K CO , DMF, rt, 16 h; (b) (i) DMF-DMA,
2
3
pyrrolidine, 120℃, 3 h; (ii) 10% Pd-C, H , THF; (c) NaBH(CN) , AcOH, 0℃-rt, 3 h; (d)
2
3
LiAlH , THF, 65℃, 16 h; (e) PtO , H , conc. HCl, EtOH/MeOH, rt, 26 h; (f) MOMCl,
4
2
2
NaH, THF, 0℃-rt, 16 h; (g) n-BuLi, B(OMe) , THF, 0℃-rt, 0.5 h-1 h, and then AcOH,
3
H
2
O
2
, rt, 5 h; (h) 2 N HCl, MeOH, rt, 12 h; (i) BrCH
j) (i) mCPBA, CH Cl , 40℃, 12 h; (ii) NaOH, MeOH, rt, 15 min; (k) Fe powder, NH
MeOH/THF/H O(2:1:1), 80℃, 4 h; (l) 2-chloroacetyl chloride, K CO , ACN, 80℃, 2 h;
m) LiAlH , THF, rt, 16 h; (n) BBr , CH Cl , -78℃-rt, 2 h.
2
CH
2
Br, K
2
CO
3
, acetone, 60℃, 36 h;
(
2
2
4
Cl,
2
2
3
(
4
3
2
2
1
0

ACCEPTED MANUSCRIPT
The preparation of linearly and angularly fused 4-aryl-4H-chromene derivatives
presented in Scheme
3
began with 4-amino-3-nitrophenol (28) and
2
-amino-3-nitrophenol (32) as the starting materials, respectively. These were reduced by
Raney-Ni to yield the intermediates 29 (or 33), which were further transformed to the
compounds 30a-30b (or 34a-34b) through the above mentioned three-component cascade
reaction. 30a was treated with triethyl orthoformate or triethyl orthoacetate to afford the
desired compounds 31a-31b. When 30a (or 30b, 34a) reacted with BrCN to obtain
compounds 31c (or 31d, 35a), subsequent acylation of 31c (or 35a) provided compounds
3
1e (or 35b) in the presence of cyclopropanecarbonyl chloride. 30a (or 34b) was then
reacted with different substituted benzaldehydes to afford compounds 31f-31h (or 35c)
utilizing Na S O as catalytic reagent.
2
2
5
1
1

ACCEPTED MANUSCRIPT
Scheme 3. Synthetic route for linearly and angularly fused 4-aryl-4H-chromene
derivatives.
Reagents and Conditions: (a) Raney-Ni, H , THF, rt, 5 h; (b) Piperidine, EtOH, N , 80℃,
2
2
1
3
5
6 h; (c) 31a or 31b: triethyl orthoformate or triethyl orthoacetate, ZrCl , MeOH, rt, 3 h;
4
1c and 31d: BrCN, MeOH/THF, rt, 16 h; 31f-31h and 35c:
-methylthiophene-2-carboxaldehyde/3,4,5-trimethoxybenzaldehyde/3-bromo-4,5-dimeth
oxybenzaldehyde/3-fluorobenzaldehyde, Na S O , DMF, 100℃, 6 h; (e) 31e and 35b:
2
2
5
Cyclopropanecarbonyl chloride, DMAP, pyridine, rt, 16 h.
1
2

ACCEPTED MANUSCRIPT
Table
1.
Chemical
structures
of
synthesized
compounds.
1
2
3
4
Compd R
R
R
R
X
B-n-A
C-n-D
(
6,7-)
(7,8-)
9
9
9
9
9
9
9
9
a
b
c
d
e
f
H
H
Cl
Cl
Br
F
H
H
H
H
CO
CN
CN
CN
CN
CN
CO
CN
2
C
2
H
5
C-1-NH
C-1-NH
C-1-NH
C-1-NH
C-1-NH
C-1-NH
C-1-NH
C-1-NH
-
-
-
-
-
-
-
-
H
H
H
H
H
H
H
H
OCH
Br
H
H
H
H
3
OCH
OCH
OCH
OCH
OCH
OCH
3
3
3
3
3
3
g
h
Br
2 2 5
C H
OCH
3
1
3

ACCEPTED MANUSCRIPT
9
1
1
1
1
1
1
1
1
1
1
1
1
1
i
F
F
F
H
-
CN
CO
CO
CN
CO
CO
CN
CO
CN
CO
CO
CN
CN
CN
C-1-NH
-
-
1a
1b
1c
3a
3b
3c
5a
5b
6a
6b
6c
6d
9a
H
H
Cl
2
2
C
2
2
H
5
5
NH-1-C
OCH
OCH
H
OCH
OCH
H
Br
-
C
H
-
NH-1-C
3
3
3
3
Br
-
-
NH-1-C
Cl
H
H
H
-
C
C
H
H
C-2-NH
C-2-NH
C-2-NH
-
-
2
2
2
5
OCH
OCH
H
3
OCH
OCH
H
3
Br
2
5
-
Br
-
3
3
Cl
C
2
H
5
NH-2-C
2
OCH
3
OCH
3
Br
-
-
NH-2-C
H
H
Cl
H
H
H
H
OH
C
C
H
H
O-1-O
O-1-O
O-1-O
O-1-O
O-1-O
-
-
-
-
-
2
2
2
5
OCH
OCH
OCH
OCH
3
3
3
3
OCH
OCH
OCH
OCH
3
3
3
3
OCH
OCH
Br
3
2
5
3
Br
1
4

ACCEPTED MANUSCRIPT
1
2
2
2
9b
2a
7a
7b
OCH
OCH
OCH
OCH
3
3
3
3
OCH
OCH
OCH
OCH
3
3
3
3
OCH
3
OH
H
CN
CN
CN
CN
O-1-O
-
-
-
-
Br
O-2-O
Br
H
O-2-NH
O-2-NH
OCH
3
H
2
2
.2 Biological evaluations
.2.1 In vitro anti-proliferative activities and SAR analysis
All synthesized compounds, including the positive control compounds mHA11,
EPC2407, colchicine, and paclitaxel, were evaluated for their anti-proliferative activities
against human melanoma A375 cells, after the incubation with gradient concentrations of
compounds for 72 h. Cell viability was measured with the CellTiter 96® AQueous One
Solution Cell Proliferation Assay（MTS）kit (Promega) according to the manufacturer’s
instructions. The results presented in Table 2 indicated that 27a, 27b and 19a displayed
stronger anti-proliferative activities in A375 cells than colchicine (or paclitaxel), with
IC50 values of 16 pM and 369 pM for 27a and 27b respectively. Comparison with
mHA11 (IC50 value of 902.3 nM) revealed that the 6,7-fused 4-aryl-4H-chromene
derivatives (9a, 9g, 13a, 16a) had approximately 2-3-fold higher anti-proliferative
activities except for 13b. On the contrary, 7,8-fused 4-aryl-4H-chromene derivatives
1
5

ACCEPTED MANUSCRIPT
(11a-b, 15a) lost their anti-proliferative activities (IC50＞10 µM). Likewise, replacement
of the 4-(3-chlorophenyl) group of 16a with a 3,4,5-trimethoxyphenyl group led to the
loss of activity of derivative 16b. The most potent compound 27a as described above was
further evaluated on eight additional cancer cell lines (CHL-1, B16-F10, HeLa，A549,
HT-29, HCT116, SW480, Jurkat) and two non-cancer cells (HUVECs and PBMCs). As a
positive control, colchicine was included in these assays. As shown in Table 3, compound
2
7a exhibited potent biological activities on these cancer cell lines. In contrast,
compound 27a displayed no cytotoxic effects on normal, non-cancer HUVECs and
PBMCs, suggesting the specificity of the effects of compound 27a for cancer cells.
Fusion of a five-member dihydropyrrole at the C6,7-positions and replacement of
C3-carboxylate with carbonitrile group, or transformation of chlorine on the 4-aryl to
bromine yielded three derivatives 9b-c that showed anti-proliferative activities similar to
9
a. The 9d with 3-fluorophenyl was about 10-fold better than 9b, and subsequent
introduction of polymethoxy substituents onto the 4-aryl ring (compounds 9f and 9h)
gave a further slight improvement. Fusion of a five-member dihydropyrrole fused at the
C7,8-positions, resulting in 11c, reduced the activities by more than 3-fold when
compared to 9f. Ring expansion of dihydropyrrole to form tetrahydropyridine (15b) also
decreased the activity by 2-fold compared with 11c. Taking together, the data indicated
that the scaffolds containing the 6,7-fused 4-aryl-4H-chromene were superior to the
7
,8-fused 4-aryl-4H-chromene scaffolds.
1
6

ACCEPTED MANUSCRIPT
We then explored the effect of different 6,7-fused rings on anti-proliferative activity.
Given that 1,3-dioxolane is a commonly fused ring found in natural products, such as
podophyllotoxin derivatives, we also synthesized the two compounds 16c-d with fused
1
,3-dioxolane and found their activities were similar to that of 9f or 9h. Nevertheless,
introduction of the –OH group at the C-8 position gave compound 19a with a
-bromo-4,5-dimethoxyphenyl group that was more active than 19b and exhibited a
3
nearly 8-fold increase compared with 16d. Therefore, the introduction of small polar
groups (-OH, -NH ) was favorable for increasing the potency of 6,7-fused
2
4
-aryl-4H-chromenes.
We next made a comparison between the fusion of five-member versus six-member
rings at the 6,7-position. 16d with a fused 1,3-dioxolane was more potent than 22a with
,4-dioxane. Conversely, 9f with fused tetrahydropyridine was 2-fold less potent than 13c
with fused dihydropyrrole. Most interestingly, when the morpholine ring fused at the
,7-position, compound 27a was 750-fold more active than 13c and 18-fold more active
1
6
than 27b. Introduction of O,N-containing hetero ring fused at the 6,7-position therefore
greatly increased the potency.
Finally, we tested the linearly and angularly fused 4-aryl-4H-chromene derivatives
(31a-h and 35a-c). Fusion of a 1H-imidazole ring at the 6,7-position of compound 31a
resulted in a 18-fold reduction in anti-proliferative activity compared with 9h.
Introduction of different sized groups at the 2-position of the 1H-imidazole ring, such as
1
7

ACCEPTED MANUSCRIPT
methyl (31b), amino (31c-d and 35a), cyclopropanecarboxamide (31e and 35b),
5
-methylthiophen-2-yl (31f), or substituted phenyl groups (31g-h and 35c), revealed that
three compounds (31b, 31d and 35a-b) with small substituents retained a certain degree
of activity (IC <1 µM). The rest of the compounds with large substituents completely
5
0
lost their anti-proliferative activities. These results suggested that fusion of unsaturated
heterocycle at the 6,7-position or 7,8-position might be detrimental for activity, and an
increased steric bulkiness at these two positions results in a great decrease in potency.
a
Table 2. In vitro anti-proliferative activity of the synthesized compounds against human
A375 melanoma cells.
b
b
Compd
IC50±ꢀꢁ (nM)
Compd
16c
IC50±ꢀꢁ (nM)
9
9
9
9
9
9
a
b
c
d
e
f
268.3±18.01
332.0±91.95
233.7±178.27
34.5±7.48
ND
32.7±13.20
23.3±11.55
3.0
16d
19a
19b
16.7±5.77
38.6±19.63
0.016±0.005
22a
28.0±22.91
27a
1
8

ACCEPTED MANUSCRIPT
9
9
9
g
h
i
546.7±213.62
27b
31a
31b
31c
0.369±0.301
382.0±92.26
1345.3±1104.92
＞10000
22.0±12.29
ND
11a
＞10000
1
1b
＞10000
31d
31e
676.7±265.02
＞10000
11c
70.0±40.00
441.0±431.57
1
1
1
1
1
1
1
3a
3b
3c
5a
5b
6a
6b
31f
＞10000
787.7
±349.54
31g
31h
35a
35b
35c
＞10000
12.3±14.57
＞10000
＞10000
103.3±47.26
110.0±75.50
5113.3±2172.15
902.3±230.73
150.0±77.95
330.0±50.21
＞10000
mHA11
colchicine
9.7±2.89
EPC2407(racemate) 27.7
1
9

ACCEPTED MANUSCRIPT
paclitaxel
47.8±0.06
a
b
Anti-proliferative activity was determined using a MTS assay. IC values are presented
5
0
as compound concentration required to inhibit A375 cells proliferation by 50%. Data
were analyzed in Microsoft Excel and plotted in GraphPad Prism 5, and expressed as
means with standard deviations from at least three independent experiments. ND, not
detected.
Table 3. In vitro anti-proliferative activity of compound 27a and colchicine against eight
cancer cell lines (CHL-1, B16-F10, HeLa, A549, HT-29, HCT116, SW480, Jurkat) and
two non-cancer cells (HUVECs and PBMCs).
Cell lines
IC ±SD (nM)
5
0
2
7a
Colchicine
3.60±0.95
around 32
9.5±1.63
CHL-1
B16-F10
HeLa
3.98±0.60
around 32
8.6±1.23
A549
31.5±13.80
11.6±0.44
11.7±0.65
15.3±1.43
10.7±3.73
62.7±14.55
11.5±1.52
13.9±0.83
14.9±2.37
19.8±13.60
HT-29
HCT116
SW480
Jurkat
2
0

ACCEPTED MANUSCRIPT
HUVEC
PBMCs
>10000
>10000
>10000
>10000
2
.2.2 Induction of G2-M arrest and cell apoptosis of A375 cells
Cell cycle assays were conducted to verify whether the newly synthesized compounds
could arrest mitosis of A375 cells. Cell cycle progression was followed by PI staining and
flow cytometry analysis. A375 cells were incubated with different concentrations of the
selected compounds for 24 h. The untreated cells showed a normal cell cycle pattern
while all the compound treated groups showed a dose-dependent inhibition of cell cycle
similar to that of colchicine control (Figure 3). Compounds 27a, 19a, and 9f all caused
obvious G2-M arrest when supplied at 50 nM.
2
1

ACCEPTED MANUSCRIPT
Figure 3. Analysis of cell cycle. (A) Representative results of effects of compounds 27a,
1
9a and 9f on cell cycle progression of A375 cells, measured by flow cytometry analysis.
A375 cells were treated with different concentrations of compounds 27a, 19a and 9f for
4 h. The percentage of cells in each cycle phase is indicated. (B) Graph summarizing the
2
cell cycle distribution affected by compounds 27a, 19a, and 9f of at least three
independent experiments.
2
2

ACCEPTED MANUSCRIPT
We then tested whether these active compounds could induce apoptosis in cancer cells
by exposing A375 cells to 10, 100 and 1000 nM of 27a, 19a, and 9f for 24 h and 48 h. All
samples were stained with annexin V-FITC and PI for analysis. This method divided cells
into four groups: live populations (annexin V-/PI-), early apoptotic populations (annexin
V+/PI-), late apoptotic populations (annexin V+/PI+), and necrotic populations (annexin
V+/PI+). The early apoptotic and late apoptotic cells were regarded as the apoptotic
population. Results in Figure 4 indicated that apoptosis was increased after treatment
with 27a, 19a, and 9f but not with the vehicle control, and the extent of apoptosis was
concentration dependent. At 24 h, a concentration of 1000 nM of 27a induced apoptosis
in 19.8% of the cells, whereas at 48 h it induced apoptosis in 32.8% of the cells.
Treatment with 19a and 9f also gave similar results in A375 cells.
2
3

ACCEPTED MANUSCRIPT
6
Figure 4. Analysis of cell apoptosis. A375 cells (10 /well) were seeded in a six-well plate
and incubated overnight. Different concentrations (10, 100, 1000 nM) of 27a, 19a, and 9f
and colchicine were added and the cells were incubated for 24 h and 48 h. Samples of
6
1
×10 cells were then collected by centrifugation, resuspended in 500 µL of 1×binding
2
4

ACCEPTED MANUSCRIPT
buffer, stained with annexin V-FITC and PI for 5 min protecting from light, and analyzed
by flow cytometry. (A) 24 h. (B) 48 h. The apoptotic group was indicated in the A+/PI-
and A+/PI+ regions. The percentage of cell populations was calculated from at least two
independent experiments.
2
.2.3 Immunofluorescence staining of interfered microtubule
We examined whether our compounds could directly disrupt the distribution of
microtubules by immunofluorescence staining of tubulin in A375 cells. After treatment
with different concentrations of 27a, 19a, and 9f for 24 hours, the cells changed from a
long-spindle shape to a shrunken round or multi angular shape. Staining results (Figure 5)
showed that untreated cells had a normal microtubule distribution whereas the candidates
treated cells displayed a disorganized distribution of tubulin and had a coenocytic
appearance, which may be arisen from incomplete cell division. These observations
confirmed that 27a, 19a, and 9f could interfere with microtubule organization and retard
mitosis in A375 cells.
2
5

ACCEPTED MANUSCRIPT
Figure 5. Effects of 27a, 19a and 9f on cellular microtubule distribution. Green
represents tubulins and blue represents nuclei. Images (magnification: 630X) were
obtained using Olympus cellSens microscope.
2
.2.4 Anti-colonogenic activities
We further conducted colony formation assays to determine if our candidate
compounds could inhibit the clone-forming ability of tumor cells. Figure 6 indicated that
7a, 19a, and 9f at a concentration of 10 nM could inhibit colony formation by A375
cells. Three compounds displayed similar inhibitory effects.
2
2
6

ACCEPTED MANUSCRIPT
Figure 6. Inhibition effects on A375 cells colony formation. A375 cells were seeded in a
6
0 mm culture dish and left for several days to form colonies. Cultures were
photographed after staining with 0.1% crystal violet. (A) Untreated control. (B)
Representative results of colony formation. (C) Statistic summary of three independent
experiments. Colony numbers were counted by ImageJ.
2
.2.5 Anti-vascular activities
Numerous studies indicate that many microtubule-binding agents can disrupt vascular
activities [21]. These include flavonoids, chalcones, and many other tubulin inhibitors
that inhibit angiogenesis through different mechanisms [22]. We tested whether our
synthesized compounds could also inhibit vascular formation by seeding human
umbilical vein endothelial cells (HUVECs) on solidified matrigel and allowing the cells
form tube structures for 6-8 h. The control groups showed an integral net of tubes
whereas the compounds treated cells failed to develop whole tubes when the compounds
were given at low nanomolar concentration (Figure 7). These responses were also
2
7

ACCEPTED MANUSCRIPT
dose-dependent. These results suggest that mHA series compounds have potent
anti-angiogenic effects.
To exclude cytotoxic effects of the compounds on HUVECs, cell viability at 24 h was
determined with the CellTiter 96® AQueous One Solution Cell Proliferation Assa y（ MTS）
kit (Promega). At concentration below 100 nM, none of them showed noticeable toxicity,
2
7a had the weakest toxic among our three compounds (Figure 7D).
4
Figure 7. Inhibition of in vitro vascular tube formation. HUVECs (10 /well) were seeded
on matrigel with 1, 10 or 100 nM 27a, 19a and 9f (B) or without drugs (vehicle control)
(
A) in a 96-well plate. Images were captured after 6-8 h at 37°C (magnification: 100X).
C) Measurements of total lengths of tubular structures using ImageJ from three
(
2
8