Synthesis and structure-activity relationship of coumarins as potent Mcl-1 inhibitors for cancer treatment

Article abstract of DOI:10.1016/j.bmc.2020.115851

Myeloid cell leukemia-1 (Mcl-1) is a validated and attractive target for cancer therapy. Over-expression of Mcl-1 in many cancers allows cancer cells to evade apoptosis and contributes to their resistance to current chemotherapeutics. In this study, more than thirty coumarin derivatives with different substituents were designed and synthesized, and their Mcl-1 inhibitory activities evaluated using a fluorescence polarization-based binding assay. The results showed that the catechol group was a key constituent for Mcl-1 inhibitory activity of the coumarins, and methylation of the catechol group led to decreased inhibitory activity. The introduction of a hydrophobic electron-withdrawing group at the C-4 position of 6,7-dihydroxycoumarin, enhanced Mcl-1 inhibitory capacity, and a hydrophilic group in this position was unbeneficial to the inhibitory potency. In addition, the introduction of a nitrogen-containing group to the C-5 or C-8 position, which allowed an intramolecular hydrogen bond, was also unfavorable for Mcl-1 inhibition. Among all coumarins tested, 4-trifluoromethyl-6,7-dihydroxycoumarin (Cpd 4) displayed the most potent inhibitory activity towards Mcl-1 (K_i = 0.21 ± 0.02 μM, IC₅₀ = 1.21 ± 0.56 μM, respectively), for which the beneficial effect on taxol resistance was also validated in A549 cells. A strong interaction between Cpd 4 and Mcl-1 in docking simulations further supported the observed potent Mcl-1 inhibition ability of Cpd 4. 3D-QSAR analysis of all tested coumarin derivatives further provides new insights into the relationships linking the inhibitory effects on Mcl-1 and the steric-electrostatic properties of coumarins. These findings could be of great value for medicinal chemists for the design and development of more potent Mcl-1 inhibitors for biomedical applications.

Full text of DOI:10.1016/j.bmc.2020.115851

Journal Pre-proofs
Synthesis and structure-activity relationship of coumarins as potent Mcl-1 in-
hibitors for cancer treatment
Yang-Liu Xia, Jing-Jing Wang, Shi-Yang Li, Yong Liu, Frank J. Gonzalez,
Ping Wang, Guang-Bo Ge
PII:
S0968-0896(20)30681-7
https://doi.org/10.1016/j.bmc.2020.115851
BMC 115851
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
15 July 2020
16 October 2020
1 November 2020
Please cite this article as: Y-L. Xia, J-J. Wang, S-Y. Li, Y. Liu, F.J. Gonzalez, P. Wang, G-B. Ge, Synthesis and
structure-activity relationship of coumarins as potent Mcl-1 inhibitors for cancer treatment, Bioorganic &
Medicinal Chemistry (2020), doi: https://doi.org/10.1016/j.bmc.2020.115851
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. 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.
©
2020 Published by Elsevier Ltd.

Synthesis and structure-activity relationship of coumarins as potent
Mcl-1 inhibitors for cancer treatment
a,d
a
c
a
d
Yang-Liu Xia , Jing-Jing Wang , Shi-Yang Li , Yong Liu , Frank J. Gonzalez , Ping
b.d
b
Wang *, Guang-Bo Ge
aSchool of Life Science and Medicine, Dalian University of Technology, Panjin,
24221, China.
1
bInstitute of Interdisciplinary Integrative Medicine Research, Shanghai University of
Traditional Chinese Medicine, Shanghai, 201203, China.
cAnalytical central laboratory, Shengyang Harmony Health Medical Laboratory Co
Ltd, Shenyang, 210112, China.
dLaboratory of Metabolism, Center for Cancer Research, National Cancer Institute,
National Institutes of Health, Bethesda, MD, USA.
∗
Corresponding author.
E-mail: wangping12@dicp.ac.cn
1

Abstract: Myeloid cell leukemia-1 (Mcl-1) is a validated and attractive target for
cancer therapy. Over-expression of Mcl-1 in many cancers allows cancer cells to
evade apoptosis and contributes to their resistance to current chemotherapeutics. In
this study, more than thirty coumarin derivatives with different substituents were
designed and synthesized, and their Mcl-1 inhibitory activities evaluated using a
fluorescence polarization-based binding assay. The results showed that the catechol
group was a key constituent for Mcl-1 inhibitory activity of the coumarins, and
methylation of the catechol group led to decreased inhibitory activity. The
introduction of a hydrophobic electron-withdrawing group at the C-4 position of
6
,7-dihydroxycoumarin, enhanced Mcl-1 inhibitory capacity, and a hydrophilic group
in this position was unbeneficial to the inhibitory potency. In addition, the
introduction of a nitrogen-containing group to the C-5 or C-8 position, which allowed
an intramolecular hydrogen bond, was also unfavorable for Mcl-1 inhibition. Among
all coumarins tested, 4-trifluoromethyl-6,7-dihydroxycoumarin (Cpd 4) displayed the
most potent inhibitory activity towards Mcl-1 (K = 0.21 ± 0.02 μM, IC = 1.21 ± 0.56
i
50
μM, respectively), for which the beneficial effect on taxol resistance was also
validated in A549 cells. A strong interaction between Cpd 4 and Mcl-1 in docking
simulations further supported the observed potent Mcl-1 inhibition ability of Cpd 4.
3
D-QSAR analysis of all tested coumarin derivatives further provides new insights
into the relationships linking the inhibitory effects on Mcl-1 and the
steric-electrostatic properties of coumarins. These findings could be of great value for
medicinal chemists for the design and development of more potent Mcl-1 inhibitors
for biomedical applications.
Keywords: Myeloid cell leukemia-1 (Mcl-1); coumarins; structure-activity
relationship (SAR)
2

1
. Introduction
Mcl-1 (myeloid cell leukemia 1) is a member of the Bcl-2 family of proteins and
plays a central role in apoptosis regulation via interacting with other Bcl-2 family
1
members. Mcl-1 was established as a major regulator of carcinogenesis and its
dysregulation prevents cancer cells from undergoing programmed cell death. Mcl-1
overexpression was observed in a variety of human cancers, including lung, breast,
prostate, pancreatic, ovarian, and cervical cancers, and also in melanoma and
leukemia.2-6 Moreover, Mcl-1 overexpression has emerged as a resistance mechanism
against a number of anticancer therapies including paclitaxel and vincristine, which
are widely prescribed microtubule-targeted chemotherapeutic agents and also
gemcitabine,7,8 a first-line drug for pancreatic cancer. Mcl-1 overexpression also
confers resistance to ABT-263, which is an inhibitor of other Bcl-2 family members
Bcl-2/Bcl-xL, and is currently in clinical trials.9-11 Inhibited cell growth and induced
apoptosis of cancer cells in vitro, and markedly decreased tumorigenicity in vivo were
1
2
observed after specific inhibition of Mcl-1 or by Mcl-1 silencing. These data suggest
that direct inhibition of Mcl-1 could be an effective therapeutic option for a wide
variety of cancers. In recent years, several groups reported on small molecules and
stapled peptides that bind to Mcl-1 in an attempt to target Mcl-1 for cancer
treatment,13-17 although few of these identified Mcl-1 inhibitors showed significant
downregulating effects, and none have entered into the clinic.
Coumarins are bioactive compounds of both natural and synthetic origin. There has
been a growing interest in their discovery and development into drugs due to their
versatile synthetic scaffold and immeasurable pharmaceutical and biological potential
with minimum side effects. A number of natural coumarins, as well as coumarin
derivatives that have been modified by medicinal chemists over the years, have been
used as antioxidants, antivirals, monoamine oxidases, cholinesterases and aromatase
inhibitors to treat various diseases and metabolic pathway disorders.18-23 Coumarins
also exhibit antitumor activities at different stages of cancer formation through
various mechanisms, for example blocking cell cycle, inducing cell apoptosis,
modulating estrogen receptor (ER), or inhibiting the DNA-associated enzymes, such
3

as topoisomerase.24-29 Recently, natural coumarins were demonstrated to have
anticancer activities through a mechanism involving Mcl-1 inhibition,30 but the
corresponding systematic study on their structure-Mcl-1 inhibitory activity
relationship is lacking.
In the present study, a series of natural or synthetic coumarins with varied
substitutions on their scaffolds were used to evaluate the inhibitory activity towards
Mcl-1 as assessed by the FAP assay. The structure-Mcl-1 inhibitory activity
relationship of these coumarins are summarized and discussed in detail. To get a
deeper understanding of the interaction between these analogs and Mcl-1, the
compound with the most potent inhibitory activity was docked to the binding site of
Mcl-1. A 3D-QSAR model was further developed to explore the structural
requirement of coumarins for potent Mcl-1 inhibition. This research could provide
useful guidance to the design and optimization of potent coumarin-type Mcl-1
inhibitors for anticancer therapy.
2
2
. Results and discussion
.1. Chemistry
To explore the effects of different property groups and their position on
hydroxycoumarin on Mcl-1 inhibition activities, a series of 4-substituted
,7-dihydroxycoumarin derivatives were designed and prepared from 1,2,4-phenenyl
6
triacetate via acid catalytic pechmann condensation with good yields (Scheme 1). After
that, compounds 6 and 7 were obtained from chloride 5 by reflux in a mixture of DMF
and water (3:10, v:v) and treated with sodium azide, respectively. 4-Carboxylic 9 was
synthesized by treating 1,2,3-phenenyl triacetate with 1,3-acetonedicarboxylic acid in
3
1
the present of perchloric acid. The substituents were selected mainly based on the
3
2
electronic and lipophilic considerations defined by Craig′s plot. In the present study,
trifluoromethyl and phenyl were selected for their electron-withdrawing and
hydrophobic property; carboxyl was selected to represent electron-withdrawing and
hydrophilic substituents, while groups like hydroxymethyl, azidomethyl, and
chloromethyl were representative of electron-donating and hydrophilic substitution,
methyl represented electron-donating and hydrophobic substituents.
4

Introduction of a nitrogen containing group as hydrogen bond acceptor on the
,7-dihydroxycoumarins by Mannich reaction (Scheme 2) was also carried out. After
6
formaldehyde and the corresponding secondary amine were stirred in ethanol for 0.5 h,
the reaction was added to a solution of 6,7-dihydroxycoumarin/4-substituted
o
6
,7-dihydroxycoumarin in ethanol or methanol, and stirring for 8 ~ 12 h at 25 ~ 60 C,
which obtained compounds 9 ~ 16 after being purified through flash column
chromatography on a silica gel. It is worth mentioning that esculetin condenses with
formaldehyde and amines with the low yield and poor regioselectivity. For example,
5
-substituted compounds 10 and 12 were formed simultaneously, when 8-substituted
compounds 9 and 11 were prepared, respectively. While the C4-H was substituted with
methyl, 4-methyl-6,7-dihydroxycoumarin condensed with formaldehyde and amines to
compound 14 ~ 16 with the good yield and the excellent regioselectivity. Furthermore,
monomethylation 17 and 18 were obtained from 6,7-dihydroxycoumarin via selectively
methylated at 7-OH and 6-OH in a good yield with CH I in the presence of Na CO at
3
2
3
o
o
33
2
5 C and NaH at 0 C, respectively (Scheme 3). The nitro-esculetin 20 was obtained
from dimethylation 19 through nitration reaction.
To further explore the effect of the existence of phenolic groups on the coumarin
skeleton for Mcl-1 inhibition activities, 4-substituted 7,8-dihydroxycoumarins and their
methylation were prepared by acid catalytic pechmann condensation mentioned above
(
Scheme 4).34 Furthermore, a few 7-hydroxycoumarin derivatives including
7
6
3
-hydroxyl-4-methyl coumarin 29, 7-hydroxyl-4-methyl-8-nitro coumarin 30,
-chloro-7-hydroxy-4-methyl coumarin 31 and 6-ethyl-7-hydroxy-4-methy coumarin
2 were also prepared as depicted in Scheme 5. In addition, some hydroxycoumarin
derivatives including daphnetin 21, 7-methoxy daphnetin 25, fraxetin 33, coumarin 34,
-hydroxycoumarin 35 and 6-hydroxycoumarin 36, which were purchased from
7
Chengdu Pufei De Biotech Co. Ltd. (Sichuan, China), and were employed in our study
in order to explore the potential structure-Mcl-1 inhibition activity relationships of
hydroxyl coumarin derivatives.
5

R
O
AcO
HO
HO
O
O
a
+
R
OEt
OAc
O
OAc
R = CH3
R = Ph
2
3
4
5
R = CH3
R = Ph
R = CF3
R = CF3
R= CH Cl
R= CH Cl
2
2
N3
OH
HO
HO
HO
HO
b
c
5
O
O
O
O
6
7
O
AcO
OH
O
d
HO
HO
O
O
O
+
OAc
HO
OH
O
OAc
8
Scheme 1. Synthesis of 4-substituted 6,7-dihydroxycoumarin derivatives. Reagents
and conditions: (a) HClO , rt, 6 ~ 8 h; (b) DMF/H O (1:3), reflux, 20 h; (c) DMF, NaN ,
4
2
3
o
rt, 10 h; (d) HClO , 50 C, 4 h.
4
6

R
O
R
O
HO
HO
HO
HCOH + R R NH
1
2
HO
R1
O
O
N
1
2
R = H
R2
R = CH3
R R N = Me N
9
R = H
1
2
2
R R N = HO
N
N
1
1
R = H
1 2
R R N =
1 2
N
13
R = H
14
R R N = Me N
1 2 2
R = CH3
R = CH3
R = CH3
R R N = HO
1
5
6
1 2
R R N =
N
1
1
2
HO
N
N
HO
HO
HO
HO
O
O
O
O
10
12
Scheme 2. Synthesis of 6,7-dihydroxycoumarins bearing Mannich bases.
HO
HO
H3CO
HO
HO
a
b
H CO
3
O
O
O
O
O
O
18
17
1
c
H CO
3
NO2
H CO
3
d
H CO
3
O
O
H CO
3
O
O
NO2
19
20
Scheme 3. Synthesis of methylated 6,7-dihydroxycoumarin derivatives. Reagents and
o
o
conditions: (a) CH I, Na CO , DMF, 25 C, 4 h. (b) CH I, NaH, DMF, 0 C, 1 h; (c)
3
2
3
3
o
CH I, NaH, DMF, rt, 5 h; (d) DCM, H SO , HNO , 0 C, 3h.
3
2
4
3
7

R2
O
O
O
a
+
R1
OH
R2
OEt
R1
O
OH
OH
22 R = OH, R = CH
3
R = OH
1
R = CH
2
3
1
2
R = OCH
1
R = CH Cl
23 R = OH, R = CH Cl
1 2 2
3
2
2
b
2
2
2
2
4 R = OH, R = CH OH
1 2 2
6 R = OCH , R = CH
3
1
3
2
7 R = OCH , R = CH Cl
1
3
2
2
b
8 R = OCH , R = CH OH
1
3
2
2
Scheme 4. Synthesis of 4-substituted 7,8-dihydroxycoumarin and methylated
derivatives. Reagents and conditions: (a) HClO , rt, 6 ~ 8 h; (b) DMF/H O (1:3), reflux,
4
2
2
0 h.
a
HO
O
O
HO
O
O
NO2
29
30
OH
R
O
O
b
+
R
OEt
HO
O
O
OH
R = Cl
31
R = Cl
R = CH CH
32 R = CH CH3
2
3
2
Scheme 5. Synthesis of 7-hydroxycoumarin derivatives. Reagents and conditions: (a)
o
H SO , HNO , 0 C, 3h; (b) HClO , rt, 6 h;
2
4
3
4
2
.2. Mcl-1 inhibitory activity assays
Coumarins, a class of widespread natural compounds containing fused benzene and
3
5
pyrone ring systems, are an excellent reservoir of biologically active compounds. To
investigate the potency of Mcl-1 inhibition, the binding affinities of 36 natural and
synthesized coumarins to Mcl-1 protein were determined using Fluorescence
polarization-based binding assay, in which the ability of inhibitors to disrupt the
interaction between Mcl-1 and two different BH3 peptides, fluorescently labeled Bid
and biotin-labeled Bim, was tested.
8

The effect of the position and number of hydroxyl groups on the coumarin scaffold
toward inhibitory activity of Mcl-1 were investigated firstly. As shown in Table 1,
coumarin 33 exhibited Mcl-1inhibitory activity with K value of 4.06 ± 0.05 μM. A
i
6
-fold decrease in Mcl-1 inhibitory activity was observed when hydroxyl was
introduced to the C-6 position of coumarin (34), while the C-7 hydroxyl substituent of
coumarin (35) displayed little effect on its inhibition potency of Mcl-1. When
catechol group was introduced on coumarin to form 6,7-dihydroxycoumarin 1 and
7
,8-dihydroxycoumarin 21, the Mcl-1inhibitory activities were improved significantly
with K values of 1.49 ± 0.04 μM and 1.75 ± 0.032 μM, respectively. Furthermore,
i
when the catechol groups of compounds 1 and 21 were blocked by methylation, the
inhibitory activities towards Mcl-1 were significantly reduced, with K values of 15.9
i
±
0.03 μM, 13.4 ± 0.08 μM and 9.54±0.08 μM for the methylated products 17, 18 and
1
9 of compound 1, respectively. The competitive binding curves of these compounds
to Mcl-1 were outlined in Fig.1A and 1B. These results indicated that catechol
moieties are key groups for the Mcl-1inhibitory activities of these coumarin
derivatives.
For 6,7-dihydroxycoumarin derivatives, the order of Mcl-1 inhibitory activity
among compounds with varying substituents in the 4-position was as follows:
trifluoromethyl 4 (K = 0.21 ± 0.02 μM) > phenyl 3 (K = 0.65±0.03 μM) >
i
i
hydroxymethyl 6 (K = 0.76 ±0.03 μM) > chloromethyl 5 (K = 1.04 ± 0.04 μM) >
i
i
methyl 2 (K = 1.19 ± 0.04 μM) > hydrogen 1 (K = 1.49 ±0.04 μM) > azidomethyl 7
i
i
(
K = 9.17 ±0.06 μM) ≈ carboxyl 8 (K = 9.36 ±0.05 μM). This trend indicated that
i
i
hydrophobic
and
electron–withdrawing
groups
in
4-position
of
6
,7-dihydroxycoumarin 1 were beneficial for Mcl-1 inhibitory activity, while
hydrophilic substitution in 4-position was unfavorable for the potency of
Mcl-1inhibition. As shown in Fig. 1C, when 4-hydrogen of parent 1 was replaced by
trifluoromethyl group or carboxyl group, the IC of parent 1 (8.77 ± 1.16 μM) was
5
0
decreased to 1.21 ± 0.56 μM or increased to 33.7 ± 0.86 μM, respectively. In order to
increase the metabolic stability of coumarin analogues, nitrogen-containing groups
including dimethylamino, pyrrolidinyl and piperidinol as the hydrogen bond acceptor
9

were introduced at the 5-position or 8-position of 6,7-dihydroxycoumarin
3
1
derivatives. All the nitrogen-containing derivatives tested exhibited lower inhibitory
activities than parent 1 except Cpd 14, among which more than 7-fold decrease of
inhibitory activity was observed for 8-piperidinol-6,7-dihydroxycoumarin 15.
Similar trend in potency was also observed for 7,8-dihydroxycoumarin derivatives,
as blockage of the catechol group by methylation was unbeneficial for
Mcl-1inhibitory activity regardless of the types of substitution on 4-postion (21 (1.75
±
0.03 μM) > 25 (7.02 ± 0.06 μM), 22 (2.11 ± 0.03 μM) > 26 (2.39 ± 0.10 μM), 23
(3.19 ± 0.08 μM) > 27 (6.59 ± 0.10 μM), 24 (5.67 ± 0.11 μM) > 28 (7.15 ±0.11 μM)).
It’s worth noting that the substitutes at 4-position showed different effects on Mcl-1
inhibitory activity of 7,8-dihydroxycoumarins compared to that of
6
,7-dihydroxycoumarins, with the following order of inhibitory potency: methyl 22,
6 > chloromethyl 23, 27 > hydroxymethyl 24, 28. Furthermore, methyl group
2
introduced at C-4 position of 7-hydroxycoumarin 35 led to a 5-fold decrease in
inhibitory activity of Mcl-1, while this decrease in inhibitory activity could be
alleviated by introduction of a chlorine, ethyl or nitro group to the C-6 or C-8 position
of compound 35. In addition, a known inhibitor gossypol was tested under identical
3
6
conditions as a positive control. The result indicated that the most potent Mcl-1
inhibitor obtained, Cpd 4, possessed comparable inhibitory activity against Mcl-1 to
that of the positive control (Table 1). The structure-activity relationships (SAR) of
these coumarin derivatives as Mcl-1 inhibitors are summarized in Fig. 2, which could
be helpful for medical chemists to design and develop more potent Mcl-1 inhibitors
for biomedical applications.
2
.3. Cytotoxicity assay
Most solid tumor malignancies show high expression levels of Mcl-1and Mcl-1
levels are correlated with resistance to anti-cancer therapeutics. A549 cells, which
also express relatively high levels of Mcl-1, demonstrated resistance to taxol, a widely
3
7
prescribed microtubule-targeted agent. To validate whether the coumarin type Mcl-1
inhibitors has the ability to enhance the apoptotic response induced by other
anti-cancer drugs known to prime cells for apoptosis, here we investigated the effect
1
0

of taxol and Cpd 4, the most potent Mcl-1 inhibitor, alone or in combination on the
viability of A549 cell lines. As shown in Fig. 3, cell killing activity of taxol toward
A549 cell lines was significantly enhanced by combination with Cpd 4, with a more
than 10 fold decreased IC50 value (0.0025 μM) observed for the combination
compared with that for taxol alone (0.034 μM). These results demonstrated that Cpd 4
is potentially useful as a reversal agent in lung cancer therapy.
2
.4. Molecular Docking Analysis
To gain a deeper understanding of the interaction mode between Mcl-1 and the
coumarins, the most potent Mcl-1 inhibitor-Cpd 4 with a K value of 0.204 μM, was
i
selected and subjected to docking at the hydrophobic binding pocket of Mcl-1. 79
low-energy conformational poses were generated from docking runs and the best
conformation was selected based on the docking score after investigation of ensemble
of ligand poses. Good alignment was observed between the crystal ligand Mcl-1 and
the low-energy conformational pose of Cpd 4. The docked ligand Cpd 4 was
surrounded by Met231, Leu235, Leu246, Val249, Met250, Val253, Phe254, Thr266,
Leu267, Phe270, Gly271, Val274, Leu290, and Ile294 residues at the Mcl-1 binding
pocket which cover a larger area of interaction (Fig. 4A). A similar binding region of
Mcl-1
-hydroxy-4-sulfamoyl-2-naphthoate compounds were docked into the crystal
structure of Mcl-1.38
was
also
reported
previously
when
N-substituted
1
Further graphical inspection of bonded and non-bonded contacts shows that the π–
π interactions stabilizes the predicted low-energy conformational pose of Cpd 4 inside
the binding pocket of Mcl-1. As shown in Fig. 4B, the π– π interactions take place
between A ring and B ring of Cpd 4 that insert deep into the binding pocket with the
side-chain benzene ring of Phe270, enhancing the complex’s stability by helping
anchor Cpd 4 in place. The distances between residues involved in the π–π
interactions are also depicted in Fig. 4B. Despite the strong π– π interactions, the
predicted low-energy conformational pose of Cpd 4 was further stabilized by a series
of interactions between A and B rings of Cpd 4 and the alkyl groups of Met231 and
Met250. Furthermore, the rest of the surrounding hydrophobic residues provide
1
1

additional strength for complex stability via non-bonded interactions. All these
interactions may contribute to the relatively high binding affinity of Cpd 4 to Mcl-1
protein observed in FP assay. These results of docking simulation are also in good
agreement with the SAR analysis, given that both the hydroxyl in B ring and the
hydrophobic groups in A ring could be beneficial for Mcl-1 inhibition as these
substituents could interact favorably with specific residues of Mcl-1 protein. In
addition, to explore why the introduction of 6-hydroxyl or 7-hydroxyl alone showed
little impact on the Mcl-1inhibitory activity of coumarin while the introduction of
6
,7-dihydroxyls led to a significant improvement of inhibitory activity of coumarin
towards Mcl-1, Cpd 33, Cpd 34, Cpd 35 and Cpd 1 were docked to the binding pocket
of Mcl-1. As shown in Fig. S1 and Table S1, the more favored conformation formed
between Cpd 1 and Mcl-1indicated by Libdock score and binding energy, as well as
the relatively shorter interaction distances between residues involved in the bonded
interactions for Cpd 1 compared to those for the other three coumarins, may
contribute to the significantly improved Mcl-1 inhibitory activity for Cpd 1.
2
.5. Establishment of 3D QSAR model
Quantitative Structure-Activity Relationship (QSAR) studies have been extensively
applied to explore the correlations between biological activities and molecular
descriptors for different classes of compounds.39,40 In this investigation, 3D QSAR
models were built using Drug Discovery Studio to acquire a systematic SAR profile
on the coumarin derivates as Mcl-1 inhibitors. Among 36 coumarin-type Mcl-1
inhibitors, the training and test sets were chosen to develop the 3D QSAR model (Fig.
2
2
5
). Coefficient determination (r ), external validation coefficient (q ) and root mean
squared error (RMSE) of 0.977, 0.968 and 0.392 were achieved for training set, and r2
2
of 0.672, q of 0.633 and RMSE of 0.501 were obtained for test set, indicating that the
built 3D QSAR model was acceptable.
Also, the compounds aligned with the iso-surfaces of the 3D QSAR model
coefficients on van der Waals grids and electrostatic potential grids are shown in Fig.
6
. In the electrostatic map, red contours around regions where high electron density
(negative charge) is favorable for Mcl-1 inhibition, and blue contours represent areas
1
2

where low electron density (partial positive charge) is expected to increase activity
Fig. 6A). Similarly, the green-yellow steric contours depicted in Fig. 6B illustrate
(
areas where steric bulk is predicted to increase (green) or decrease (yellow) activity.
The resulting mappings suggested that both 3-D steric and electronic interactions
could strongly affect the inhibitory effects of coumarins on Mcl-1. Notably, these
findings agreed well with the experimental data, such as the substitutional groups on
the aromatic B ring or heterocyclic A ring ask for high negative charged groups, and
the small substitutional groups on the two rings are suitable. These findings provide
new insights into the fine correlations between the inhibitory effects against Mcl-1
and the steric-electrostatic properties of coumarins, which could be helpful for the
rational design of novel coumarins as potent Mcl-1 inhibitors.
3
. Conclusion
In summary, more than thirty coumarin derivatives with different substituents were
designed and synthesized. The Mcl-1 inhibitory activities of these compounds were
assayed and the potential structure–inhibition relationships of coumarin derivatives
were summarized and discussed. The results demonstrated that the catechol group in
the coumarin scaffold was a key group for Mcl-1 inhibitory activity, while
introduction of a hydrophobic and electron-withdrawing group at the C-4 position of
coumarins was beneficial for the Mcl-1 inhibition effect. Among all the tested
compounds, 4-trifluoromethyl-6,7-dihydroxycoumarin (Cpd 4) exhibited the strongest
inhibitory activity against Mcl-1, and its beneficial effect on taxol resistance was
validated in A549 cell lines. The strong interaction between Cpd 4 and Mcl-1 protein
in docking simulation further supported the observed potent Mcl-1 inhibitory ability
of Cpd 4. Meanwhile, a 3D-QSAR model was well established for these coumarin
derivatives, and the results demonstrated that the steric bulk and the electrostatic
potential of substituted groups in the coumarin scaffold were key determinants for
Mcl-1 inhibitory activity. Among all the tested compounds, Cpd 4 exhibited the
strongest Mcl-1 inhibitory activity, which meet all the preferred structural
requirements according to the 3D-QSAR model. All these findings could be helpful
for medicinal chemists to design and develop potent anticancer agents by targeting
1
3

Mcl-1 based on the structural modifications of coumarin derivatives.
4
4
. Materials and methods
.1 Reagents and materials
A Bid BH3 peptide of 21-residue (residues 79-99) with a 6-carboxy-fluorescein
succinimidyl ester fluorescence tag (FAM-Bid) was synthesized by HD Biosciences
Shanghai, China). Recombinant Mcl-1 protein Escherichia coli BL21 was
synthesized and purified following methods described earlier.41,42
.2. Chemicals and Measurements
(
4
1
13
The H NMR and C NMR spectra were recorded on a Bruker ARX (Bruker,
Rheinstetten, Germany) 400 MHz or Bruker AMX 400 MHz or 600 MHz
spectrometer in dimethyl sulfoxide (DMSO-d ) if not noted otherwise, and the
6
chemical shifts were expressed as ppm using trimethylsilane (TMS) as an internal
reference. High-resolution mass spectral (HRMS) analyses were measured on a
TripleTOF™ Mass Spectrometer (TripleTOF 5600, SCIEX, Foster City, USA).
Coumarin derivatives were analyzed using an ultra-fast liquid chromatography
spectrometry system (Shimadzu, Kyoto, Japan) equipped with two LC-20AD pumps,
a DGU-20A3 vacuum degasser, a SIL-20ACHT auto-sampler, a CTO-20AC column
oven, and an SPD-M 20A diode-array detector (DAD). All reagents used in the
synthesis were obtained commercially and used without further purification. The
reactions were monitored using thin layer chromatography (TLC) on glass packed
precoated silica gel GF254 plates. Flash column chromatography was performed
using silica gel (200–300 mesh), which was purchased from the Qingdao Ocean
Chemical Co. Ltd (Qingdao, China). Compounds 1, 21, 25, 33, 34, 35 and 36 were
purchased from Chengdu Pufei De Biotech Co. Ltd. (Sichuan, China).
General procedure for the synthesis of 4-substituted esculetins (2 ~ 5). To a
mixture of 1,2,4-phenenyl triacetate (5.0 mmol) and properly substituted β-ketoesters
(10 mmol) was added drop-wise perchloric acid (5.0 mL) at room temperature and
stirred for 6 ~ 8 h. After completion of the reaction, as indicated by TLC monitoring,
the reaction mixture was poured slowly into 100 mL ice-water with stirring. The
resultant suspension was filtered and the collected solid washed with water and dried,
1
4

then the crude compound was recrystallized from methanol.
6
,7-dihydroxy-4-methyl-2H-chromen-2-one (2). According to the general procedure,
1
,2,4-phenenyl triacetate was reacted with ethyl acetoacetate in the presence of
perchloric acid, and crude compound was recrystallized from methanol to produce
1
compound 2 as light white solid, 87% yield. H NMR (400 MHz, DMSO-d ) δ: 2.31
6
(
s, 3H, CH ), 6.09 (s, 1H, COCH=C), 6.73 (s, 1H, ArH), 7.00 (s, 1H, ArH), 9.34 (s, br,
3
1
3
1
1
H, ArOH), 10.19 (s, br, 1H, ArOH); C NMR (100 MHz, DMSO-d ) δ: 18.68,
6
03.13, 109.89, 110.86, 111.96, 143.24, 148.17, 150.58, 153.68, 161.08; ESI-MS: M
=
192, found 191.0 [M-H]-.
6
,7-dihydroxy-4-phenyl-2H-chromen-2-one (3). According to the general procedure,
1
,2,4-phenenyl triacetate was reacted with ethyl benzoylacetate in the presence of
perchloric acid, and crude compound was recrystallized from methanol to generate
1
compound 3 as white solid, 91% yield. H NMR (400 MHz, DMSO-d ) δ: 6.12 (s, 1H,
6
COCH=C), 6.80 (s, 1H, ArH), 6.84 (s, 1H, ArH), 7.50-7.58 (m, 5H, ArH) 9.43 (s, br,
1
H, ArOH), 10.26 (s, br, 1H, ArOH); ESI-MS: M = 254, found 253.0 [M-H]-.
6
,7-dihydroxy-4-trifluoromethyl-2H-chromen-2-one (4). According to the general
procedure, 1,2,4-phenenyl triacetate was reacted with ethyl 4,4,4-trifluoroacetoacetate
in the presence of perchloric acid, and crude compound was recrystallized from
1
methanol to produce compound 4 as light yellow solid, 63% yield. H NMR (400 MHz,
DMSO-d ) δ: 6.71 (s, 1H, COCH=C), 6.86 (s, 1H, ArH), 7.03 (s, 1H, ArH), 9.78 (s, br,
6
1
3
1
1
H, ArOH), 10.61 (s, br, 1H, ArOH); C NMR (100 MHz, DMSO-d ) δ: 104.01,
6
05.05, 108.89, 112.12, 120.94, 123.68, 144.03, 149.59, 152.08, 159.62; ESI-MS: M
=
246, found 245.1 [M-H]-.
,7-dihydroxy-4-chloromethyl-2H-chromen-2-one (5). According to the general
6
procedure, 1,2,4-phenenyl triacetate was reacted with ethyl 4-chloroacetoacetate in
the presence of perchloric acid, and crude compound was recrystallized from
1
methanol to generate compound 5 as white solid, 78% yield. H NMR (400 MHz,
DMSO-d ) δ: 4.90 (s, 2H, CH Cl), 6.40 (s, 1H, COCH=C), 6.78 (s, 1H, ArH), 7.12 (s,
6
2
1
H, ArH), 9.48 (s, br, 1H, ArOH), 10.39 (s, br, 1H, ArOH). ESI-MS: M = 225.5,
-
-
found 224.9 [M-H] and 226.9 [M-H] .
1
5

6
,7-dihydroxy-4-hydroxymethyl-2H-chromen-2-one (6). Chloride 5 (0.5 g, 2.2 mmol)
was dissolved in a mixture of DMF (3 ml) and H O (10 ml) with stirring and then
2
refluxed for 20 h. After completion of the reaction as indicated by TLC, the mixture
was diluted with water. The resultant suspension was filtered and the collected solid
was washed with water and dried, then crude compound was recrystallized from
1
methanol to produce hydroxymethyl 6 (380 mg, 83%) as light brown solid. H NMR
(
400 MHz, DMSO-d ) δ: 4.63 (s, 2H, CH OH), 6.21 (s, 1H, COCH=C), 6.75 (s, 1H,
6 2
ArH), 6.94 (s, 1H, ArH), 9.31 (s, br, 1H, ArOH), 10.19 (s, br, 1H, ArOH); ESI-MS: M
208, found 208.9 [M+H]+.
-(azidomethyl)-6,7-dihydroxy-2H-chromen-2-one (7). Chloride 5 (0.5 g, 2.2 mmol)
=
4
was dissolved in a mixture of DMF (6.0 ml), then NaN (0.3 g 4.5 mmol) was added
3
and stirred overnight. After completion of the reaction as indicated by LC-MS, the
mixture was diluted with water. The resultant suspension was filtered and the
collected solid was washed with water and dried, then crude compound was
recrystallized from methanol to generate azidomethyl 7 in 90% yield as light brown
1
solid. H NMR (400 MHz, d-DMSO) δ: 4.73 (s, 2H, CH N ), 6.26 (s, 1H, COCH=C),
2
3
6
.78 (s, 1H, ArH), 6.98 (s, 1H, ArH), 9.49 (s, br, 1H, ArOH), 10.38 (s, br, 1H, ArOH).
ESI-MS: M = 233, found 234.2 [M+H]+.
-(7,8-dihydroxy-2-oxo-2H-chromen-4-yl)acetic acid (8). According to the general
procedure mentioned above, 1,2,4-phenenyl triacetate was reacted with
,3-acetonedicarboxylic acid in the presence of perchloric acid, and crude compound
2
1
was recrystallized from methanol to produce compound 8 in 70 % yield as white solid.
1H NMR (400 MHz, DMSO) δ 12.75 (s, 1H), 10.11 (s, 1H), 9.35 (s, 1H), 7.03 (d, J =
1
3
8
.4 Hz, 1H), 6.81 (d, J = 8.4 Hz, 1H), 6.22 (s, 1H), 3.80 (s, 2H). C NMR (101MHz,
DMSO) δ: 37.88, 112.43,112.58,112.71, 116.03, 132.74, 143.95, 149.93, 150.99,
60.63,171.23; ESI-MS: M = 236, found 235.1 [M-H]-.
General procedure for the synthesis of 5/8-substituted esculetin Mannich bases
9 ~ 16). To a solution of the appropriate amines (0.013 mol) in 30 mL of absolute
1
(
methanol was added aqueous formaldehyde (37%) (0.015 mol). After 0.5 h of gentle
refluxing at 50°C, esculetin/4-substituted esculetin was added (0.01 mol), dissolved in
1
6

5
0 mL of absolute methanol. The duration of the reaction was 8 ~ 12 h and the
reaction was monitored by TLC.
6
6
,7-dihydroxy-8-((dimethylamino)methyl)-2H-chromen-2-one
(9)
and
,7-dihydroxy-5-((dimethylamino)methyl)-2H-chromen-2-one (10). According to the
general procedure, esculetin was reacted with aqueous formaldehyde and aqueous
dimethylamine. After completion of the reaction as indicated by TLC, the solvent was
removed under reduced pressure. The residue was applied to flash column
chromatography (silicagel) to generate compound 9 and compound 10 with the yield
of 17% and 6%, respectively.
1
Compound 9: H NMR (400 MHz, DMSO-d ) δ: 2.49 (s, 6H, 2CH ), 4.03 (s, 2H,
6
3
CH ), 6.03 (d, J=8.0 Hz, 1H, COCH=C), 6.86 (s, 1H, ArH), 7.79 (d, J=12.0 Hz, 1H,
2
PhCH=C), 8.16 (s, 1H, ArOH); ESI-MS: M = 235, found 236.1 [M+H]+.
1
Compound 10: H NMR (400 MHz, d-DMSO) δ: 2.25 (s, 6H, 2CH ), 3.81 (s, 2H,
3
CH ), 6.17 (d, J=12 Hz, 1H, COCH=C), 6.70 (s, 1H, ArH), 8.10 (d, J=12 Hz, 1H,
2
PhCH=C), 8.15 (s, 1H, ArOH), ESI-MS: M = 235, found 236.1 [M+H]+.
6
5
,7-dihydroxy-8-((4-hydroxypiperidin-1-yl)methyl)-2H-chromen-2-one (11) and
-((4-hydroxypiperidin-1-yl)methyl)-6,7-dihydroxy-2H-chromen-2-one (12).
According to the general procedure, esculetin was reacted with aqueous formaldehyde
and 4-hydroxy piperidine. After completion of the reaction as indicated by TLC, the
solvent was removed under reduced pressure. The residue was applied to flash
column chromatography (silicagel) to produce compound 11 and compound 12 with
the yield of 12% and 3%, respectively.
1
Compound 11: H NMR (400 MHz, DMSO-d ) δ: 1.48 (m, 2H, CH ), 1.82 (m, 2H,
6
2
CH ), 2.49 (m, 2H, NCH ), 2.91 (m, 2H, NCH ), 3.62 (m, 1H, CHOH), 4.00 (s, 2H,
2
2
2
CH ), 6.08 (d, J=8 Hz, 1H, COCH=C), 6.89 (s, 1H, ArH),7.81 (d, J=8 Hz, 1H,
2
1
3
PhCH=C); C NMR (100 MHz, DMSO-d ) δ: 33.50, 49.72, 52.56, 106.77, 108.87,
6
1
2
2
09.71, 110.38, 142.79, 144.87, 146.79, 153.60, 160.62; ESI-MS: M = 291, found
+
+
92.1 [M+H] ; HRMS (ESI): C H NO calcd for [M + H] : 292.1185, found:
1
5
17
5
92.1181.
1
Compound 12: H NMR (400 MHz, DMSO-d ) δ: 1.68 (m, 2H, CH ), 1.92 (m, 2H,
6
2
1
7

CH ), 3.09 (m, 2H, NCH ), 3.29 (m, 2H, NCH ), 3.76 (m, 1H, CHOH), 4.44 (s, 2H,
2
2
2
CH ), 6.27 (d, J=8 Hz, 1H, COCH=C), 6.97 (s, 1H, ArH), 8.33 (d, J=8 Hz, 1H,
2
1
3
PhCH=C); C NMR (100 MHz, DMSO-d ) δ: 30.38, 48.53, 49.61, 103.75, 110.68,
6
1
2
2
6
11.56, 112.92, 141.95, 143.54, 148.89, 150.28, 160.18; ESI-MS: M = 291, found
+
+
92.1 [M+H] ; HRMS (ESI): C H NO calcd for [M + H] : 292.1185, found:
1
5
17
5
92.1183.
,7-dihydroxy-8-(pyrrolidin-1-ylmethyl)-2H-chromen-2-one (13). According to the
general procedure, esculetin was reacted with aqueous formaldehyde and pyrrolidine.
After completion of the reaction as indicated by TLC, the solvent was removed under
reduced pressure. The residue was applied to flash column chromatography (silicagel)
1
to afford compound 13 with the yield of 18 % as light brown solid. H NMR (400
MHz, DMSO-d ) δ: 1.85 (m, 4H, CH CH ), 2.88 (t, J = 4.8 Hz 4H, NCH CH ), 4.17
6
2
2
2
2
(
s, 2H, CH ), 5.97 (d, J=8.0 Hz, 1H, COCH=C), 6.82 (s, 1H, ArH), 7.76 (d, J=8.0 Hz,
2
1
3
1
H, PhCH=C); C NMR (100 MHz, DMSO-d ) δ: 23.05 (2C), 49.88, 52.84(2C),
6
1
05.98, 107.16, 107.75, 109.05, 143.49, 144.90, 147.58, 156.57, 160.86; ESI-MS: M
+
+
=
261, found 262.1 [M+H] ; HRMS (ESI): C H NO calcd for [M + H] : 262.1079,
14 15 4
found: 262.1072.
8
-((dimethylamino)methyl)-4-methyl-2H-chromen-2-one (14). According to the
general procedure, 4-methyl esculetin was reacted with aqueous formaldehyde and
aqueous dimethylamine. After completion of the reaction, as indicated by TLC, the
mixture was diluted with water. The resultant suspension was filtered and the
collected solid washed with water and dried, then crude compound was recrystallized
1
from methanol to generate compound 14 as yellow solid with 78% yield. H NMR
(
400 MHz, DMSO-d ) δ: 2.30 (s, 3H, CH ), 2.42 (s, 6H, 2CH ), 4.00 (s, 2H, CH ),
6 3 3 2
1
3
5
4
.99 (s, 1H, COCH=C), 6.92 (s, 1H, ArH); C NMR (100 MHz, DMSO-d ) δ: 18.87,
6
3.83, 54.51, 106.93, 107.48, 108.82, 109.27, 143.39, 146.82, 154.21, 155.03, 160.98 ;
+
ESI-MS: M = 249, found 250.0 [M+H] ; HRMS (ESI): C H NO calcd for [M +
1
3
15
4
+
H] : 250.1079, found: 250.1074.
,7-dihydroxy-8-((4-hydroxypiperidin-1-yl)methyl)-4-methyl-2H-chromen-2-one
15). According to the general procedure, 4-methyl esculetin was reacted with
6
(
1
8

aqueous formaldehyde and 4-hydroxy piperidine. After completion of the reaction as
indicated by TLC, the mixture was diluted with water. The resultant suspension was
filtered and the collected solid washed with water and dried, then crude compound
was recrystallized from methanol to produce compound 15 as light yellow solid with
1
8
2
3
0% yield. H NMR (400 MHz, DMSO-d ) δ: 1.47 (m, 2H, CH ), 1.81 (m, 2H, CH ),
6
2
2
.31 (s, 1H, CH ), 2.45 (m, 2H, NCH ), 2.87 (m, 2H, NCH ), 3.61 (m, 1H, CHOH),
3
2
2
1
3
.99 (s, 2H, CH ), 6.04 (s, 1H, COCH=C), 6.94 (s, 1H, ArH); C NMR (100 MHz,
2
DMSO-d ) δ: 18.30, 33.61, 49.77, 52.88, 107.02, 107.58, 109.18, 109.77, 142.49,
6
+
1
45.82, 152.80, 153.61, 160.32; ESI-MS: M = 305, found 306.1 [M+H] ; HRMS
+
(
ESI): C H NO calcd for [M + H] : 306.1341, found: 306.1337.
1
6
19
5
6
,7-dihydroxy-8-(pyrrolidin-1-ylmethyl)-4-methyl-2H-chromen-2-one
(16).
According to the general procedure, 4-methyl esculetin was reacted with aqueous
formaldehyde and pyrrolidine. After completion of the reaction as indicated by TLC,
the mixture was diluted with water. The resultant suspension was filtered and the
collected solid washed with water and dried, then crude compound was recrystallized
1
from methanol to afford compound 16 as yellow solid with 88% yield. H NMR (400
MHz, DMSO-d ) δ: 1.84 (m, 4H, 2CH ), 2.29 (s, 3H, CH ), 2.83 (m, 4H, 2NCH ),
6
2
3
2
1
3
4
.15 (s, 2H, CH ), 5.96 (s, 1H, COCH=C), 6.89 (s, 1H, ArH); C NMR (100 MHz,
2
DMSO-d ) δ: 18.89, 23.60, 50.17, 53.33, 106.97, 108.32, 108.70, 143.65, 146.90,
6
+
1
54.23, 155.87, 161.07; ESI-MS: M = 275, found 276.1 [M+H] ; HRMS (ESI):
+
C H NO calcd for [M + H] : 276.1236, found: 276.1230.
1
5
17
4
6
-hydroxy-7-methoxy coumarin (17). To the solution of 6,7-dihydroxycoumarins
(
500 mg, 2.81 mmol) in DMF, Na CO (3 eq.) was added at 15°C and stirred for 0.5
2
3
h. iodomethane (525 L) was added dropwise to the reaction mixture at 15°C,
maintaining the temperature at 20°C for 4 h. The reaction mixture was poured into
water-ice and acidified with 2N hydrochloric acid. The reaction mixture was extracted
with ethyl acetate (50 mL x 3). The combined organic layer was washed with water
and brine, dried over anhydrous sodium sulfate, and evaporated in vacuo. The residue
was purified by column chromatography using silica gel with petroleum ether-
dichloromethane- acetone as mobile phase to obtain compound 17 in 75% yield as
1
9

1
white solid. H NMR (400 MHz, DMSO-d ) δ 3.82 (s, 3H, OCH ), 6.23 (d, J = 9.5
6
3
Hz, 1H, COCH=C), 7.00 (s, 2H, Ar-H), 7.89 (d, J = 9.5 Hz, 1H, C=CH). M = 192,
found 193 [M+H]+.
6
-methoxy-7-hydroxycoumarin (18). To the solution of 6,7-dihydroxycoumarins
500 mg, 2.81 mmol) in DMF, NaH (60% suspension in oil) (7.0 mmol) was added at
°C under dry argon and allowed to stir for 0.5 h for the anion generation.
iodomethane (210 L) was added dropwise to the reaction mixture, maintaining the
temperature at 0-5°C until the reaction was completed, as indicated by TLC (VDCM
(
0
:
VMeOH = 50: 1). The reaction mixture was poured into water-ice and acidified with 2N
hydrochloric acid. The resultant suspension was filtered and the collected solid was
washed with water and dried, then crude compound was recrystallized from methanol
1
to afford compound 18 in about 90% yield as white solid. H NMR (400 MHz,
DMSO-d ) δ 3.82 (s, 3H, OCH ), 6.22 (d, J = 9.5 Hz, 1H, COCH=C),
6.78 (s, 1H,
6
3
Ar-H), 7.22 (s, 1H, Ar-H), 7.91 (d, J = 9.5 Hz, 1H, C=CH), 10.29 (d, br, 1H, Ar-OH).
-
+
M = 192, found 191.1 [M-H] , found 193 [M+H] .
6
,7-dimethoxylcoumarin (19). To the solution of 6,7-dihydroxycoumarin (500 mg,
2
.81 mmol) in DMF, NaH (60% suspension in oil) (7.0 mmol) was added at 0°C
under dry argon and allowed to stir for 0.5 h for the anion generation. iodomethane
300 L) was added dropwise to the reaction mixture and stirred for 5 h at room
(
temperature. The reaction mixture was poured into water-ice and acidified with 2N
hydrochloric acid. The resultant suspension was filtered and the collected solid was
washed with water and dried, then crude compound was recrystallized from methanol
1
to produce compound 19 in about 95% yield as white solid. H NMR (400 MHz,
d-DMSO) δ: 3.80 (s, 3H, CH O), 3.86 (s, 3H, CH O), 6.29 (d, J = 8.0 Hz, 1H,
3
3
COCH=C), 7.06 (s, 1H, ArH), 7.25 (s, 1H, ArH), 7.95 (d, J = 8.0 Hz, 1H,
1
3
PhCH=C). C NMR (100 MHz, d-DMSO) δ: 56.34, 56.63, 100.50, 109.38, 111.64,
13.11, 144.79, 146.31, 149.88, 152.99, 161.01. ESI-MS: M = 206, found 207.1
M+H]+.
1
[
6
,7-dimethoxy-3,8-dinitro-2H-chromen-2-one
(20).
To
a
solution
of
6
,7-dimethoxylcoumarin 19 (0.5 g, 2.4 mmol) in DCM (10.0 ml) was added
2
0

o
drop-wise a mixture of concentrated sulfuric acid and fuming nitric acid (4 ml) at 0 C,
After completion of the reaction as indicated by TLC, the reaction mixture was
poured slowly into a mixture of ice-water (50 ml) with stirring. The resultant
suspension was filtered and the collected solid was washed with water and dried, then
1
the crude compound was recrystallized from methanol to afford 20 as yellow solid. H
NMR (400 MHz, d-DMSO) δ: 3.91 (s, 3H, CH O), 4.01 (s, 3H, CH O), 7.62 (s, 1H,
3
3
1
3
ArH), 8.77 (s, 1H, PhCH=C). C NMR (100 MHz, d-DMSO) δ: 58.60, 62.99, 102.69,
04.28, 134.56, 137.36, 139.70, 142.02, 151.78, 152.65, 159.79. HRMS: M=296,
found 327.0461 [M+MeOH-H]－.
,8-dihydroxy-4-methyl-2H-chromen-2-one (22). To a mixture of 1,2,3-phenenyl
1
7
triacetate (5.0 mmol) and ethyl acetoacetate (7.5 mmol) was added drop-wise
perchloric acid (6.0 ml) at room temperature and stirred for 6 h. After completion of
the reaction as indicated by TLC, the reaction mixture was poured slowly into a
mixture of ice-water (50 ml) with stirring. The resultant suspension was filtered and
the collected solid was washed with water and dried, then the crude compound was
recrystallized from methanol to produce compound 22 as a light white solid with a
1
yield of 78%. H NMR (600 MHz, DMSO) δ 10.05 (s, 1H), 9.28 (s, 1H), 7.09 (d, J = 8.6
1
3
Hz, 1H), 6.82 (d, J = 8.6 Hz, 1H), 6.13 (d, J = 1.1 Hz, 1H), 2.36 (s, 3H); C NMR (151
MHz, DMSO) δ 160.67, 154.40, 149.87, 143.79, 132.62, 115.95, 113.23, 112.58,
-
1
7
1
10.65, 18.72; HRMS (ESI) for C H O , Calcd 192.0423, found 192.0355 [M-H] .
1
0
8
4
,8-dihydroxy-4-chloromethyl-2H-chromen-2-one (23). To
a
mixture of
,2,3-phenenyl triacetate (5.0 mmol) and ethyl 4-chloroacetoacetate (7.5 mmol) was
added drop-wise perchloric acid (6.0 ml) at room temperature and stirred for 6 h.
After completion of the reaction as indicated by TLC, the reaction mixture was
poured slowly into a mixture of ice-water (50 ml) with stirring. The resultant
suspension was filtered and the collected solid was washed with water and dried, then
the crude compound was recrystallized from methanol to produce compound 23 as a
1
white solid with a yield of 69%. H NMR (600 MHz, DMSO) δ 10.19 (s, 1H), 9.40 (s,
1
3
1
H), 7.18 (d, J = 8.7 Hz, 1H), 6.85 (d, J = 8.7 Hz, 1H), 6.42 (s, 1H), 4.94 (s, 2H); C
NMR (151 MHz, DMSO) δ 160.55, 151.86, 150.23, 144.15, 132.92, 115.95, 112.79,
2
1

1
11.42, 110.57, 41.95; HRMS (ESI) for C H ClO , Calcd 226.0033, found 224.9967
10 7 4
[
M-H]-.
7
,8-dihydroxy-4-hydroxymethyl-2H-chromen-2-one (24). Chloride C23 (0.5 g, 2.2
mmol) was dissolved in a mixture of DMF (3 mL) and H O (10 mL) with stirring and
2
refluxed for 20 h. After TLC indicated that the reaction was complete, the mixture was
diluted with water. The resultant suspension was filtered, and the collected solid was
washed with water and dried. The crude compound was recrystallized from methanol to
1
yield hydroxymethyl 24 as a light brown solid with a yield of 80%. H NMR (600 MHz,
DMSO) δ 10.03 (s, 1H), 9.30 (s, 1H), 7.01 (d, J = 8.7 Hz, 1H), 6.78 (d, J = 8.6 Hz, 1H),
1
3
6
.24 (t, J = 1.4 Hz, 1H), 5.55 (t, J = 5.5 Hz, 1H), 4.69 (d, J = 4.2 Hz, 2H); C NMR (151
MHz, DMSO) δ 161.03, 157.70, 149.75, 143.82, 132.75, 114.75, 112.65, 110.73,
-
1
8
3
06.85, 59.56; HRMS (ESI) for C H O , Calcd 208.0372, found 207.0304 [M-H] .
1
0
8
5
-hydroxy-7-methoxy-4-methylcoumarin
(26).
To
a
mixture
of
-methoxybenzene-1,2-diol (0.5 g, 3.5 mmol) and ethyl acetoacetate (0.95 ml, 7.0
mmol) was added drop-wise perchloric acid (3.0 ml) at room temperature and stirred
for 6 h. After completion of the reaction as indicated by TLC, the reaction mixture
was poured slowly into a mixture of ice-water (50 ml) with stirring. The resultant
suspension was filtered and the collected solid was washed with water and dried, then
the crude compound was recrystallized from methanol to produce 26 as light white
1
solid. H NMR, δ: 2.38 (s, 3H, CH ), 3.89 (s, 3H, OCH ), 6.20 (s, 1H, COCH=C),
3
3
1
3
7
.03 (d, J = 8 Hz, 1H, Ar-H), 7.21 (d, J = 8 Hz, 1H, Ar-H), 9.38 (s, 1H, OH). C
NMR, δ: 160.4, 154.3, 154.1, 148.1, 134.7, 120.9, 113.5, 113.4, 110.8, 61.1, 18.7. M
-
+
–1
=
206, found 205.1 [M-H] , found 206.9 [M+H] . IR (KBr), ṽ/cm : 3552, 3447, 3362,
2
4
3
7
968, 2840, 1663, 1613, 1574, 1511, 1457, 1444, 1385.
-(chloromethyl)-8-hydroxy-7-methoxy-2H-chromen-2-one (27). To a mixture of
-methoxybenzene-1,2-diol (0.5 g, 3.5 mmol) and ethyl 4-chloroacetoacetate (1.0 ml,
.0 mmol) was added drop-wise perchloric acid (5.0 ml) at room temperature and
stirred for 6 h. After completion of the reaction as indicated by TLC, the reaction
mixture was poured slowly into a mixture of ice-water (50 ml) with stirring. The
resultant suspension was filtered and the collected solid was washed with water and
2
2

dried, then the crude compound was recrystallized from methanol to generate 27 as
1
light white solid. H NMR (400 MHz, d-DMSO) δ: 3.90 (s, 3H, OCH ), 4.97 (s, 2H,
3
CH Cl), 6.48 (s, 1H, COCH=C), 7.08 (d, J = 8 Hz, 1H, ArH), 7.30 (d, J = 8 Hz, 1H,
2
1
3
ArH), 9.51 (s, br, 1H, ArOH); C NMR (100 MHz, d-DMSO) δ: 41.90, 56.79, 108.87,
11.88, 112.67, 115.63, 134.13, 143.20, 151.34, 151.60, 160.36. ESI-MS: m/z 241
and 243 [M+H]＋
1
8
-hydroxy-7-methoxyl-4-hydroxymethyl-2H-chromen-2-one (28). Chloride 27 (0.4 g,
1
.7 mmol) was dissolved in a mixture of DMF (3 mL) and H O (10 mL) with stirring
2
and refluxed for 20 h. After TLC indicated that the reaction was complete, the mixture
was diluted with water. The resultant suspension was filtered, and the collected solid
was washed with water and dried. The crude compound was recrystallized from
1
methanol to yield hydroxymethyl 28 as a light brown solid with a yield of 83%. H
NMR (400 MHz, d-DMSO) δ: 3.87 (s, 3H, OCH ), 4.70 (d, J = 4 Hz, 2H, CH ), 5.59
3
2
(s, 1H, OH),6.30 (s, 1H, COCH=C), 6.96 (d, J = 8 Hz, 1H, ArH), 7.09 (d, J = 8 Hz,
1
3
1
1
2
4
4
H, ArH), 9.39 (s, br, 1H, ArOH); C NMR (100 MHz, d-DMSO) δ: 56.68, 59.56,
08.07, 108.67, 112.10, 114.39, 133.98, 142.88, 150.93, 157.35, 160.87. ESI-MS: m/z
23 [M+H]＋
-methyl-7-hydroxy-8-nitrocoumarin
(30).
To
a
stirred
solution
of
-methyl-7-hydroxy coumarin (1.5 g, 8.5 mmol) in concentrated sulphuric acid (30
mL), a solution of concentrated nitric acid (6.0 mL) in concentrated sulphuric acid
o
(
6.0 mL) was added at such a rate as to keep the temperature below 5 C. After
o
warming to 20 C, the reaction mixture was poured into a stirred mixture of ice cold
water. The yellow solid that separated was filtered and thoroughly washed with water.
The product was recrystalized first from glacial acetic acid and then methanol to
1
obtain compound 30 in 78% yield. H NMR (400 MHz, DMSO) δ 12.18 (s, 1H), 7.81
(d, J = 9.0 Hz, 1H), 7.05 (d, J = 8.9 Hz, 1H), 6.32 (d, J = 1.1 Hz, 1H), 2.41 (d, J = 1.1
1
3
Hz, 3H). C NMR (101 MHz, DMSO) δ 158.57, 154.05, 152.86, 146.09, 128.62,
1
6
4
13.52, 112.46, 111.69, 18.74.
-chloro-7-hydroxy-4-methyl-2H-chromen-2-one (31). To
a
mixture of
-chlorobenzene-1,3-diol (5.0 mmol) and ethyl acetoacetate (7.5 mmol) was added
2
3

drop-wise perchloric acid (6.0 ml) at room temperature and stirred for 6 h. After
completion of the reaction as indicated by TLC, the reaction mixture was poured
slowly into a mixture of ice-water (50 ml) with stirring. The resultant suspension was
filtered and the collected solid was washed with water and dried, then the crude
compound was recrystallized from methanol to produce compound 31 with a yield of
1
6
5%. H NMR (400 MHz, DMSO) δ 11.37 (s, 1H), 7.76 (s, 1H), 6.90 (s, 1H), 6.21 (d,
1
3
J = 1.1 Hz, 1H), 2.38 (d, J = 1.1 Hz, 3H). C NMR (101 MHz, DMSO) δ 160.30,
1
6
4
56.71, 153.46, 153.37, 126.59, 117.30, 113.30, 111.78, 103.77, 18.56.
-ethyl-7-hydroxy-4-methyl-2H-chromen-2-one (32). To
-ethylbenzene-1,3-diol (5.0 mmol) and ethyl acetoacetate (7.5 mmol) was added
a
mixture
of
drop-wise perchloric acid (6.0 ml) at room temperature and stirred for 6 h. After
completion of the reaction as indicated by TLC, the reaction mixture was poured
slowly into a mixture of ice-water (50 ml) with stirring. The resultant suspension was
filtered and the collected solid was washed with water and dried, then the crude
compound was recrystallized from methanol to produce compound 32 with a yield of
1
8
5%. H NMR (400 MHz, DMSO) δ 10.51 (s, 1H), 7.45 (s, 1H), 6.73 (s, 1H), 6.11 (d,
J = 1.1 Hz, 1H), 2.60 (q, J = 7.5 Hz, 2H), 2.38 (d, J = 1.1 Hz, 3H), 1.16 (t, J = 7.5 Hz,
1
3
3
1
4
H). C NMR (101 MHz, DMSO) δ 160.90, 159.39, 154.11, 153.41, 128.18, 125.60,
12.17, 110.57, 102.03, 22.98, 18.63, 14.60.
.3. Fluorescence polarization-based binding (FP) assay
For the competitive binding assay for Mcl-1 protein, FAM-Bid peptide (10 nM)
and Mcl-1 protein (50 nM) were preincubated in the assay buffer (25 mM Tris, pH 8.0;
50 mM NaCl). Each inhibitor was dissolved in DMSO to obtain a stock solution (10
mM). The stock solution was then diluted successively to get different concentrations
1000 μM, 400 μM, 100 μM, 100 μM, 40 μM, 10 μM, 4 μM , 1 μM, 0.4 μM, 0.1 μM
.04 μM and 0.01 μM). The serial dilutions of each compound were added to the
1
(
0
incubation mixture. After 30 min incubation, the polarization values were measured
using a black 96-well plate in Spectra Max M5 Detection System. A K value of 13
d
nM for FAM-Bid binding to Mcl-1 protein was determined by saturation experiments.
The K value for each inhibitor was calculated using the equation developed for
i
2
4

4
3
FP-based assays by Kuntz et al. using a computer program for K value calculation,
i
which is available at: http://sw16.im.med.umich.edu/software/calc_ki/.
4
.4. Cell culture
Human lung adenocarcinoma cells (A549 cell line) was cultured in RPMI1640
medium supplemented with 10% fetal bovine serum (FBS), 60 mg/mL streptomycin
and 100 mg/mL penicillin in a humidified atmosphere of a 5% CO at 37°C.
2
4
.5. Cytotoxicity assay by SRB method
A549 cells were seeded in 96-well plates (6.5 × 103 cells/well) and incubated for
2
4 hours at 37°C in a humid atmosphere with 5% CO for adhesion. After this period,
2
the medium was withdrawn and the wells were treated with different concentrations
0.0001, 0.001, 0.01, 0.1, 1, 10 and 100 μM) of Cpd 4 and taxol, either alone or in
(
combination in 10% FBS. After 48 h, 50 μl of 50% pre-cold trichloroacetic acid was
added to the wells gently. After incubation at 4°C for one h, the plates were flicked
and washed five times with cold water and were then air dried. The air-dried plates
were stained with 100 μl SRB and kept for 30 min at 37°C. The unbound dye was
removed by four times wash with 1% acetic acid. After being air-dried, 100 μl of 10
mM Tris base (pH 10.5) was then added to the wells to solubilize the dye and the
plates were shaken vigorously for 5 min. The absorbance reading was performed in a
microplate reader (Powerwave HT; Biotek) at 570 nm. The percentage growth
inhibition was calculated using following formula:
All experiments were performed in triplicate. The results were expressed as the means
of the IC (drug concentration that reduced cell viability to 50% of the control).
5
0
4
.6. Molecular docking
The crystal structure of Mcl-1 was obtained from the Protein Data Bank (PDB ID:
4
4
4
WMR). The molecular docking process was performed using Discovery Studio
(BIOVIA Discovery Studio 2016, Dassault Systèmes, San Diego, USA). The
CHARMM 40.1 force field was used to represent the protein and ligand structures.
Docking simulations were performed by a standard LibDock protocol, where protein
2
5

features are referred to as hotspots. After a final energy-minimization step (allowing
the ligand poses to be flexible), the top scoring ligand poses are saved. The rigid
poses were placed into the active site of Mcl-1, and the hotspots are matched as
triplets. The protein-ligand complexes with the highest LibDock score were taken
from the docking results and depicted in full text.
4
.7. 3D-QSAR Model Building
Three 3D-QSAR models were built using the corresponding package of Drug
Discovery Studio. Primarily, compounds were aligned by consensus on both steric
and electrostatic fields, with relative weight of 50-50%. The aligned molecules were
placed in a 3D grid space, with grid spacing of 1.5 Å. The extent of the grid was set to
the bounding box of all the ligands plus 6.0 Å of extension. The CHARMm force
field was used. The electrostatic potential and the van derWaals potential were treated
as separate terms. A +1e point charge was used as the electrostatic potential probe.
Distance-dependent dielectric constant was used to mimic the solvation effect. For the
van der Waals potential, a carbon atom with a 1.73 Å radius was used as a probe. The
energy grid potentials were filtered to remove highly correlated descriptors
(maximum descriptor correlation was set as 0.9). The energy grid potentials were
filtered to remove highly correlated descriptors. Partial least squares (PLS) models
were then built using these remaining descriptors, and Log (Ki for Mcl-1) was used as
activity properties.
Declaration of interest
The authors report no conflicts of interests.
Acknowledgement
This work was supported by National Key Research and Development Program of
China (2017YFC1700200, 2017YFC1702000), National Natural Science Foundation
of China (81603187), Natural Science Foundation of Shanghai (18ZR1436500),
Fundamental Research Funds for the Central Universities (DUT18LK39), and
National Cancer Institute Intramural Research Program (ZIABC005708).
2
6

2
7

References
. G. Wei, A. A. Margolin, L. Haery, E. Brown, L. Cucolo, B. Julian, S. Shehata, A.L.
Kung, R. Beroukhim, T.R. Golub, Cancer Cell 21 (2012) 547-562.
. L. Song, D. Coppola, S. Livingston, D. Cress, E.B. Haura, Mcl-1 regulates survival
1
2
and sensitivity to diverse apoptotic stimuli in human non-small cell lung cancer cells,
Cancer Biol. Ther. 4 (2005) 267-276.
3
. Q. Ding, X. He, W. Xia, J.M. Hsu, C.T. Chen, L.Y. Li, D.F. Lee, J.Y. Yang, X. Xie,
J.C. Liu, M.C. Hung, Myeloid cell leukemia-1 inversely correlates with glycogen
synthase kinase-3beta activity and associates with poor prognosis in human breast
cancer, Cancer Res. 67 (2007) 4564-4571.
4
. S. Derenne, B. Monia, N.M. Dean, J.K. Taylor, M.J. Rapp, J.L. Harousseau, R.
Bataille, M. Amiot, Antisense strategy shows that Mcl-1 rather than Bcl-2 or Bcl-x(L)
is an essential survival protein of human myeloma cells, Blood 100 (2002) 194-199.
5
. M.H. Andersen, J.C. Becker, P. Thor Straten, The antiapoptotic member of the
Bcl-2 family Mcl-1 is a CTL target in cancer patients, Leukemia 19 (2005) 484-485.
. M.H. Kang, Z. Wan, Y.H. Kang, R. Sposto, C.P. Reynolds, Mechanism of synergy
6
of N-(4-hydroxyphenyl)retinamide and ABT737 in acute lymphoblastic leukemia cell
lines: Mcl-1 inactivation, J. Natl. Cancer Inst. (2008) 580-595.
7
. I.E. Wertz, S. Kusam, C. Lam, T. Okamoto, W. Sandoval, D.J. Anderson, E.
Helgason, J.A. Ernst, M. Eby, J. Liu, L.D. Belmont, J.S. Kaminker, K.M. O’Rourke,
K. Pujara, P.B. Kohli, A.R. Johnson, M.L. Chiu, J.R. Lill, P.K. Jackson, W.J.
Fairbrother, S. Seshagiri, M.J.C. Ludlam, K.G. Leong, E.C. Dueber, H. Maecker,
D.C.S. Huang, V.M. Dixit, Sensitivity to antitubulin chemotherapeutics is regulated
by MCL1 and FBW7, Nature 471 (2011) 110-114.
8
. S.H. Wei, K. Dong, F. Lin, X. Wang, B. Li, J.J. Shen, Q. Zhang, R. Wang, H.Z.
Zhang, Inducing apoptosis and enhancing chemosensitivity to gemcitabine via RNA
interference targeting Mcl-1 gene in pancreatic carcinoma cell, Cancer Chemother.
Pharmacol. 62 (2008) 1055-1064.
9
. M.F. van Delft, A.H. Wei, K.D. Mason, C.J. Vandenberg, L. Chen, P.E. Czabotar,
S.N. Willis, C.L. Scott, C.L. Day, S. Cory, J.M. Adams, A.W. Roberts, D.C.S. Huang,
2
8

The BH3 mimetic ABT737 targets selective Bcl-2 proteins and efficiently induces
apoptosis via Bak/Bax if Mcl-1 is neutralized, Cancer Cell 10 (2006) 389-399.
1
0. E. Varin, C. Denoyelle, E. Brotin, M. Meryet-Figuiere, F. Giffard, E. Abeilard, D.
Goux, P. Gauduchon, P. Icard, L. Poulain, Downregulation of Bcl-xL and Mcl-1 is
sufficient to induce cell death in mesothelioma cells highly refractory to conventional
chemotherapy, Carcinogenesis 31 (2010) 984-993.
1
1. K.E. Olberding, X. Wang, Y. Zhu, J. Pan, S.N. Rai, C. Li, Actinomycin D
synergistically enhances the efficacy of the BH3 mimetic ABT-737 by
downregulating Mcl-1 expression, Cancer Biol. Ther. 10 (2010) 918-929.
1
2. H. Zhang, S. Guttikonda, L. Roberts, T. Uziel, D. Semizarov, S.W. Elmore, J.D.
Leverson, L.T. Lam, Mcl-1 is critical for survival in a subgroup of non-small-cell
lung cancer cell lines, Oncogene 30 (2011) 1963-1968.
1
3. M.F. Rega, B. Wu, J. Wei, Z. Zhang, J.F. Cellitti, M. Pellecchia, SAR by
interligand nuclear Overhauser effects (ILOEs) based discovery of acylsulfonamide
compounds active against Bcl-x(L) and Mcl-1, J. Med. Chem. 54 (2011) 6000-6013.
1
4. Y.B. Kim, M.E. Balasis, K. Doi, N. Berndt, C. Duboulay, C.C.A. Hu, W. Guida,
H.G. Wang, S.M. Sebti, J.R. Del Valle, Synthesis and evaluation of substituted
hexahydronaphthalenes as novel inhibitors of the Mcl-1/BimBH3 interaction,
Bioorg. Med. Chem. Lett. 22 (2012) 5961-5965.
1
5. N.A. Cohen, M.L. Stewart, E. Gavathiotis, J.L. Tepper, S.R. Bruekner, B. Koss,
J.T. Opferman, L.D. Walensky, A competitive stapled peptide screen identifies a
selective small molecule that overcomes MCL-1-dependent leukemia cell survival,
Chem. Biol. 19 (2012) 1175-1186.
1
6. M.L. Stewart, E. Fire, A.E. Keating, L.D. Walensky, The MCL-1 BH3 helix is an
exclusive MCL-1 inhibitor and apoptosis sensitizer, Nat. Chem. Biol. 6 (2010)
5
1
95-601.
7. A. Muppidi, K. Doi, S. Edwardraja, E.J. Drake, A.M. Gulick, H.G. Wang, Q. Lin,
Rational design of proteolytically stable, cellpermeable peptide-based selective Mcl-1
inhibitors, J. Am. Chem. Soc. 134 (2012) 14734-14737.
1
8. I. Kostova, S. Bhatia, P. Grigorov, S. Balkansky, V.S. Parmar, A.K. Prasad, L.
2
9