Synthesis and investigation of novel 6-(1,2,3-triazol-4-yl)-4-aminoquinazolin derivatives possessing hydroxamic acid moiety for cancer therapy

Article abstract of DOI:10.1016/j.bmc.2016.10.006

By merging the critical pharmacophore of EGFR/HER2 and HDAC inhibitors into one compound, a novel series of EGFR, HER-2, and HDAC multitarget inhibitors were synthesized. Compounds 9a–l contained 4-anilinoquinazolines with C-6 triazole-linked long alkyl chains of hydroxamic acid and displayed excellent inhibition against these enzymes (compound 9d exhibited the best inhibitory potency on wild-type EGFR, HDAC1, and HDAC6 with IC₅₀values 0.12 nM, 0.72 nM and 3.2 nM individually). Furthermore, compounds 9b and 9d potently inhibited proliferation of five human cancer cell lines (with IC₅₀values between 0.49 and 8.76 μM). Further mechanistic study revealed that compound 9d also regulated the phosphorylation of EGFR and HER2 and histone H3 hyperacetylation on the cellular level and induced remarkable apoptosis in BT-474 cells. Therefore, our study suggested that a system network-based multi-target drug design strategy might provided an alternate drug design method, by taking into account the synergy effect of EGFR, HER-2 and HDAC.

Full text of DOI:10.1016/j.bmc.2016.10.006

Accepted Manuscript
Synthesis and investigation of novel 6-(1,2,3-triazol-4-yl)-4-aminoquinazolin
derivatives possessing hydroxamic acid moiety for cancer therapy
Chao Ding, Shaopeng Chen, Cunlong Zhang, Guangnan Hu, Wei Zhang, Lulu
Li, Yu Zong Chen, Chunyan Tan, Yuyang Jiang
PII:
S0968-0896(16)30915-4
DOI:
http://dx.doi.org/10.1016/j.bmc.2016.10.006
BMC 13333
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
21 July 2016
6 October 2016
7 October 2016
Please cite this article as: Ding, C., Chen, S., Zhang, C., Hu, G., Zhang, W., Li, L., Chen, Y.Z., Tan, C., Jiang, Y.,
Synthesis and investigation of novel 6-(1,2,3-triazol-4-yl)-4-aminoquinazolin derivatives possessing hydroxamic
acid moiety for cancer therapy, Bioorganic & Medicinal Chemistry (2016), doi: http://dx.doi.org/10.1016/j.bmc.
2
016.10.006
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.

Synthesis and investigation of novel 6-(1,2,3-triazol-4-yl)-4-aminoquinazolin
derivatives possessing hydroxamic acid moiety for cancer therapy
a,b
b
c
d
a,b
Chao Ding , Shaopeng Chen , Cunlong Zhang , Guangnan Hu , Wei Zhang , Lulu
b
c
a,b,*
b,e,*
Li , Yu Zong Chen , Chunyan Tan , Yuyang Jiang
a
Department of Chemistry, Tsinghua University, Beijing 100084, PR China
b
The State Key Laboratory Breeding Base-Shenzhen Key Laboratory of Chemical
Biology, the Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, P.
R. China
c
Shenzhen Technology and Engineering Laboratory for Personalized Cancer
Diagnostics and Therapeutics, Shenzhen Kivita Innovative Drug Discovery Institute,
Shenzhen 518055, P. R. China
d
Department of Medicine, University of Massachusetts Medical School, Worcester,
Massachusetts 01605
e
School of Medicine, Tsinghua University, Beijing 100084, PR China
*
Corresponding author. Tel.: +86 755 2603 2094; fax: +86 755 2603 2094;
Email address: jiangyy@sz.tsinghua.edu.cn; tancy@sz.tsinghua.edu.cn

Abstract
By merging the critical pharmacophore of EGFR/HER2 and HDAC inhibitors into
one compound, a novel series of EGFR, HER-2, and HDAC multitarget inhibitors
were synthesized. Compounds 9a-l contained 4-anilinoquinazolines with C-6
triazole-linked long alkyl chains of hydroxamic acid and displayed excellent
inhibition against these enzymes (compound 9d exhibited the best inhibitory potency
on wild-type EGFR, HDAC1, and HDAC6 with IC50 values 0.12 nM, 0.72 nM and
3
.2 nM individually). Furthermore, compounds 9b and 9d potently inhibited
proliferation of five human cancer cell lines (with IC50 values between 0.49-8.76 μM).
Further mechanistic study revealed that compound 9d also regulated the
phosphorylation of EGFR and HER2 and histone H3 hyperacetylation on the cellular
level and induced remarkable apoptosis in BT-474 cells. Therefore, our study
suggested that a system network-based multi-target drug design strategy might
provided an alternate drug design method, by taking into account the synergy effect of
EGFR, HER-2 and HDAC.
Keywords: EGFR; HER2; HDAC; multitarget; synergy; resistance.

1
. Introduction
The ErbB tyrosine kinase family consists of four closely related members:
epidermal growth factor receptor (EGFR, or ErbB1/HER1), human epidermal growth
1
factor receptor 2 (HER2, or ErbB2/Neu), HER3 (ErbB3), and HER4 (ErbB4).
Binding ligands to the receptors stimulates ErbB dimer formation through
homodimerization or heterodimerization, which triggers autophosphorylation of the
2
receptor on a number of tyrosine residues. Subsequently, this leads to the recruitment
3
of appropriate adaptor or signaling proteins with SRC homology 2 (SH2) domains,
4
resulting in activation of several downstream pathways, such as the Ras/Raf/Erk and
5
PI3K/Akt/mTOR pathways involved in the proliferation and survival response.
Selective disruption of EGFR and HER2 by monoclonal antibodies and
small-molecule inhibitors has been clinically validated as a rational strategy for
6
cancer therapy. As the first generation small-molecule EGFR inhibitors, gefitinib
(
ZD1839) and erlotinib (OSI-774) have markedly improved the overall survival (OS)
of EGFR activating mutation (L858R and deletions in exon 19) in non-small cell lung
7
,8
cancer (NSCLC) patients.
Unfortunately, these inhibitors lack prolonged
effectiveness due to acquired resistance mainly caused by T790M mutation at the
9
-11
gatekeeper and activation of alternative compensatory signaling pathways.
Second
12
generation EGFR inhibitor, afatinib (BIBW-2992), and third generation EGFR
1
3
inhibitor, osimertinib (AZD9291), have been approved by the U.S. FDA for the
treatment of advanced NSCLC in patients with T790M mutation. However, the
resistance caused by upregulation of bypass signal, which consists of numerous

genetic and epigenetic signaling aberrations including MET amplification, PIK3CA
mutation, BRAF mutation, etc., is still a major problem facing current cancer
9
,14,15
research.
Several strategies have been proposed to overcome the limitation of concurrent
1
0,16
activation of other bypass pathways.
One promising approach is to disrupt
multiple pathways by the inhibition of histone deacetylase (HDAC). HDAC can
remove acetyl groups from histone and non-histone proteins, condense chromatin, and
1
7
prevent gene transcription. Modulation of activities of HDACs can impact the
activity of a diverse range of proteins, including tubulin, p53, and Hsp90, which
control a variety of cellular processes, such as cell cycle arrest, differentiation, and
1
8
apoptosis. In the past decade, much progress has been made in the development of
HDAC small-molecule inhibitors. Vorinostat (SAHA), romidepsin (FK-228), and
ponobinostat (LBH589) have been licensed by the U.S. FDA as treatments for
1
9-21
cutaneous T cell lymphoma (CTCL) and multiple myeloma.
Many studies have shown that there is synergistic crosstalk between EGFR/HER2
2
2-24
25
and HDAC.
Inhibition of HDAC6 leads to increased acetylation of Hsp90,
resulting in degradation of a number of Hsp90 client proteins that play important roles
2
6,27
in cancer cell proliferation, such as mut-EGFR, HER2, AKT, c-Raf, and ABL.
HDAC6 inhibitors could act in the similar manner as Hsp90 inhibitors. Moreover,
HDAC inhibitors can alter gene expression, increase production of reactive oxygen
1
8
species (ROS), and induce cell cycle arrest, etc., which may also indirectly
contribute to the anti-cancer activity of EGFR/HER2 inhibitors. It is reported that

treatment with vorinostat or panobinostat led to increased Hsp90 acetylation,
EGFR/HER2 degradation, cell death, and synergized with ErbB inhibitors, such as
erlotinib, lapatinib, or trastuzumab in EGFR/HER2 over-expressed cancer cell lines.
Thus, targeting EGFR/HER2 and HDAC simultaneously might offer benefits for
2
2-24
cancer therapy.
Recently, several studies have reported on EGFR/HDAC dual inhibitors and
2
8-29
yielded satisfactory results.
One of the most potent compounds, CUDC-101
(
7-(4-(3-Ethynylphenylamino)-7-methoxyquinazolin-6-yloxy)-N-hydroxyheptanamid
e, 9, Figure 1), which possesses an oxygen atom as a linker between the
pharmacophore of HDAC inhibitors (hydroxamic acid) and the pharmacophore of
EGFR inhibitors (4-phenyl-aminoquinazoline), has already entered Phase I clinical
3
0
investigation. In this study, we utilized a 1,2,3-triazole as a linker between
hydroxamic acid and 4-phenylamino-quinazoline. 1,2,3-triazole is relatively resistant
to metabolic degradation and is exceedingly stable in acidic/basic and
reductive/oxidative conditions, suggesting that it could be used to improve
3
1
pharmacokinetic properties of the desired drugs. Further, triazoles may also be
capable of participation in generating functional bioisosteres to substitute esters and
3
2
amides. Due to these desirable features of this moiety, 1,2,3-triazole has been widely
employed by medicinal chemists in drug discovery processes. In this study, we would
report a novel series of EGFR/HER2/HDAC multitarget inhibitors with 1,2,3-triazole
as the linker, which have potential for cancer therapy.
(
Figure 1 should be listed here)

Figure
1.
Selected
quinazoline-based
EGFR/HER2
inhibitors
1-6,
EGFR/HER2/HDAC multitarget inhibitor 7 and HDAC inhibitor 8.
2
. Design
Among known drugs in clinical use or in clinical trials, 4-aminoquinazoline is the
most common scaffold of selective EGFR and HER2 inhibitors. As shown in the
X-ray cocrystal structures of EGFR/HER2 with selective EGFR inhibitor gefitinb
3
3
34
(
1) and dual EGFR/HER2 inhibitor lapatinib (GW-572016, 3), the quinazoline
moiety occupies the adenine region of the ATP-binding pocket and forms a hydrogen
bond with Met793 (EGFR) or Met801 (HER2) in the hinge at N1 of quinazoline,
while the phenyl-amino group fits into the hydrophobic BP-I and BP-II pockets.
Substituent at C-6 of the quinazoline core extends towards the hydrophilic E
1
region,
which could be utilized to improve the pharmacokinetic properties and cellular
3
5
activity. These key features are also applied to other ErbB inhibitors, such as dual
EGFR/HER2 inhibitors, sapitinib (AZD-8931, 2) and varlitinib (ARRY-334543, 4),

selective HER2 inhibitors CP-724714 (5) and CP-654577 (6) (Figure 1). Since the
C-6 substituent is primarily composed of a hydrophilic part via a flexible linker, such
as furyl, phenyl, or vinyl, we decided to modify this substituent with a triazole linker,
which can mimic the direction of the hydrophilic part without dramatically hampering
EGFR activity.
3
6
Meanwhile, co-crystals of HDACs bound with vorinostat (8) revealed common
binding modes between HDAC and HDAC inhibitors. (1) A cap group, which is a
hydrophobic group, recognizes the surface outside the HDAC active pocket. (2) A
linker that can extend to a certain distance leads to the active site. 3) A zinc-binding
group (ZBG) directly interacts with the zinc ion at the conserved active site. Based on
the structure of vorinostat, we retained the hydroxamic acid functional group as ZBG
and maintained a chain length of five or six carbons (which can exhibit the best
potency against HDAC). However, we replaced the phenyl group with other
hydrophobic groups and used triazole instead of amide since they are bio-isosterisms
in drug design. It is validated that the preference of the cap group is quite flexible in
which variation of this group to a certain extent would not dramatically affect the
binding to HDAC.
Given the above analysis of key determinants for the functional pockets and
molecular recognition of EGFR and HDAC, we merge the critical pharmacophore of
EGFR/HER2 inhibitors and HDAC inhibitors into one compound. As depicted in
Scheme 1, the 4-aminoquinazoline of compounds 9a-9l interacted with EGFR and
HER2 with high potency and selectivity. Meanwhile, the introduction of a long alkyl

chain of hydroxamic acid reached the active site and chelated zinc to inhibit HDAC
enzyme activity. Therefore, compounds 9a-9l may be able to interact with EGFR,
HER2, and HDAC independently and thus block multiple signaling pathways
simultaneously.
(
Scheme 1 should be listed here)
Scheme 1. Design strategy of novel EGFR/HER2/HDAC multitarget inhibitors.
3
. Chemistry
Compounds 9a-l were synthesized according to Scheme 2. The key intermediate
2 was generated from commercially available compound 10 via cyclization and
1
chlorination. Coupling anilines with 12 produced compounds 13a-f. Compounds
5a-f were prepared through a classical Sonogashira reaction and deprotection. The
1
key intermediates 18a and 18b were obtained from commercially available
compounds 16a and 16b respectively via azidation and amidation. Coupling 15a-f
with 18a-b followed by acidification afforded the final products 9a-l.
(
Scheme 2 should be listed here)

Scheme 2. Reagents and conditions: (a) NH
reflux; (c) anilines, iso-PrOH, 80C; (d) trimethylsilylacetylene, Pd(PPh
THF/Et N, 80C; (e) TBAF, THF, RT; (f) NaN , DMF, 80C; (g) NH OTHP, EDCI,
DMAP, CHCl , RT; (h) CuSO ·5H O, sodium ascorbate, DMF, 60C; (i) HCl,
dioxane, RT.
2
CHO, reflux; (b) SOCl
2
, DMF(cat),
3
)
2
Cl , CuI,
2
3
3
2
3
4
2
4
4
. Results and discussion
.1. In vitro enzyme inhibition
We first tested the inhibitory activity of the target compounds against EGFR,
HER2, and HDAC enzymes using gefitinib, lapatinib, and vorinostat as the positive
control compounds (Table 1). We were pleased to find that almost all compounds
strongly inhibited EGFR, HDAC1, and HDAC6 with potencies comparable to that of
selective EGFR inhibitor gefitinib and HDAC inhibitor vorinostat. 8 compounds also
exhibited outstanding to moderate inhibitory activity on HER2 with IC50 values less
than 1 µM. Among the 8 compounds, compounds 9c and 9d displayed the best
potencies on EGFR and HDAC. For instance, compound 9d strongly suppressed the
wild-type EGFR and HDAC1 with IC50 values in subnanomolar ranges (0.12 nM for
EGFR and 0.72 nM for HDAC1) and inhibited HDAC6 with IC50 value in single-digit

nanomolar range (3.2 nM), which is 3-80 fold more potent than gefitinib (9.6 nM for
EGFR) and vorinostat (12 and 11 nM for HDAC1 and HDAC6).
(
Table 1 should be listed here)
Table 1 Enzymatic inhibitory activities of compounds 9a-9l against EGFR, HER2,
a
HDAC1, and HDAC6.
EGFR
IC50, nM)
HER-2
(IC50, nM)
HDAC1
(IC50, nM)
HDAC6
(IC50, nM)
Compd
R
n
1
2
(
9
9
a
0.65 ± 0.08
0.69 ± 0.12
165.5 ± 21.6
176.7 ± 10.4
7.0 ± 0.6
2.6 ± 0.4
9.5 ± 0.8
8.4 ± 1.2
b
9
c
1
2
1
2
1
0.27 ± 0.04
0.12 ± 0.03
1.9 ± 0.4
386.1 ± 25.7
174.9 ± 19.6
192.8 ± 25.1
16.4 ± 0.9
1.4 ± 0.1
0.72 ± 0.11
38.6 ± 4.5
8.3 ± 0.7
4.1 ± 1.4
3.2 ± 0.5
9
d
9
e
f
94.2 ± 12.6
19.5 ± 1.1
14.3 ± 2.2
9
4.2 ± 0.3
9
9
g
6.9 ± 1.2
537.4 ± 34.6
7.2 ± 2.5
h
2
3.6 ± 0.5
706.7 ± 109.5
5.6 ± 0.9
13.6 ± 0.9
9
9
i
1
2
6.4 ± 0.5
>1000
>1000
6.9 ± 0.4
5.2 ± 1.8
12.9 ± 4.3
13.8 ± 1.6
j
10.0 ± 1.7
9
k
1
2
5.8 ± 1.2
>1000
>1000
26.6 ± 3.4
12.8 ± 2.5
18.5 ± 3.3
16.0 ± 1.9
9
l
11.6 ± 0.9
b
Gefitinib
Lapatinib
Vorinostat
-
-
-
9.6 ± 0.7
NT
NT
NT
NT
NT
NT
23.9 ± 1.4
NT
NT
10.8 ± 1.4
10.8 ± 0.7
a
b
All the results were obtained from 2 different assays. NT means not tested.
Next, we investigated the influence of chain length and substituents of the
-amino group on the inhibition of enzymes. We first evaluated the influence of chain
4

length on the inhibition of EGFR, HER2, and HDAC. The data suggest that EGFR
and HER2 inhibitions were largely unaffected by the change in carbon chain length
(
basically less than 2-fold increase, or loss of activity was observed between the
-carbon chain length compounds (n=1) and the respective 6-carbon chain length
5
compounds (n=2)) with the exception of compound 9f, which had a 10-fold
enhancement of activity in inhibiting HER2 compared to compound 9e (192.8 nM vs.
1
6.35 nM). Since the triazole-linked long alkyl chains of hydroxamic acid extends
towards the hydrophilic E region, one carbon chain length difference is not expected
to significantly affect the potency of EGFR and HER2. Meanwhile, almost all
1
3
5
6
5
-carbon chain length compounds displayed better potencies than the respective
-carbon chain length compounds against both HDAC1 and HDAC6, which is
3
0
congruent with previous reports. Our next step was to examine the influence of the
phenyl ring on the inhibition of EGFR, HER2, and HDAC. Although the ATP binding
pockets of EGFR and HER2 have high degree of similarity, it is reported that different
3
7
4
-anilino functional groups can alter the selectivity between EGFR and HER2.
Usually, small aniline substituted quinazolines such as gefitinib tend to be strong
3
8
inhibitors of EGFR and moderate inhibitors of HER2. As the aniline motifs become
larger, these quinazoline derivatives such as lapatinb and varlitinib exhibit enhanced
39,40
potency against HER2.
By replacing the aniline 4-position with large lipophilic
4
1,42
phenyl ether, such as in CP-724714 and CP-654577,
the inhibition on EGFR is
abolished, and these compounds become selective HER2 inhibitors (Table S1). Here,
we compared the changes in potency and selectivity of the newly synthesized

compounds to the original compounds 1-6 against EGFR and HER2. Our results
suggest that the introduction of a vorinostat-like segment to quinazoline basically
improved the inhibition of EGFR but reduced the inhibition of HER2, thus, the
selectivity profiles changed dramatically. For example, sapitinib, lapatinib, and
varlitinib are dual inhibitors of EGFR and HER2, but 9c-h which bearing the
respective 4-anilino functional groups displayed more than 100-fold selectivity
between EGFR and HER2 (with the exception of compound 9f). Similar trends were
also observed in 9i-l in which the potencies on HER2 nearly diminished, and these
four compounds have become EGFR selective inhibitors. These unexpected effects
may result from the introduction of triazole linked long alkyl chains of hydroxamic
acid. It seems that substituents at C-6 of the quinazoline not only affect the
pharmacokinetic properties and cellular activity, but also have an obvious influence
on the activity and selectivity of EGFR and HER2. Different from EGFR/HER2
inhibition, 4-anilino functional groups are not critical for selectivity for HDAC
interaction. However, the substituted groups on the phenyl ring had a significant
impact on the activities of HDAC inhibition. With increasing size of the substituted
groups, the potencies on HDAC decreased markedly. This might be due to the fact
that the substituted groups on the phenyl ring could form steric clash to affect their
potential interaction with the active pocket of HDAC. For instance, Compounds
9
a-9d, which have the smallest substituents on the phenyl ring, displayed the best
potencies of HDAC inhibition (IC50 less than 10 nM). Compounds 9g-9l, which
possess bigger substituents on the phenyl ring, inhibited the HDAC6 with IC50 values

between 10-20 nM. Not surprisingly, compounds 9e and 9f, with the biggest
3
-fluorophenylmethoxy group at 4-position of aniline, displayed the weakest
3
0
potencies of HDAC inhibition.
4
.2. Cytotoxity studies
We next determined the anti-proliferative effects of the synthetic compounds
using an MTT assay against A549 cells (EGFR overexpressed, k-Ras mutation) and
BT-474 cells (HER2 overexpressed) with lapatinib and vorinostat as the positive
control compounds. As shown in Table 2, most of the compounds suppressed the
growth of two cancer cells with IC50 values in the micromolar range. Almost all
6
5
9
-carbon chain length compounds displayed better efficacies than the respective
-carbon chain length compounds against both cells. It is worthy to note that 9b and
d, with small substituents on the phenyl ring, which may possess the better cell
membrane permeability and lower molecule weights, inhibited the proliferation of
A549 cells more effectively than lapatinb and vorinostat (with IC50 values 0.51 and
0
.63 µM respectively). However, the cellular activities against BT-474 cells were
largely unaffected by the changes in phenyl ring substitution. Since most compounds
showed moderate to weak inhibition of HER2, the inhibition of EGFR and HDAC
2
4
may contribute largely to the their anti-proliferative effects against BT-474 cells.
Table 2 should be listed here)
(
Table 2 Cellular inhibitory activities of compounds 9a-9l against A549 and BT-474
c
cells.
A549
IC50, µM)
BT-474
(IC50, µM)
A549
(IC50, µM)
BT-474
(IC50, µM)
Compd
Compd
(
9
9
a
>50
26.01 ± 1.51
3.63 ± 0.31
9h
9i
8.79 ± 0.61
8.46 ± 0.89
2.88 ± 0.26
b
0.51 ± 0.04
14.65 ± 1.02

9
c
>50
>50
9j
9k
16.53 ± 0.95
4.41 ± 0.42
9
d
0.63 ± 0.12
>50
3.88 ± 0.06
2.20 ± 0.20
2.24 ± 0.13
43.45 ± 3.68 23.61 ± 1.28
9
9
e
9l
13.49 ± 0.23
1.74 ± 0.28
2.57 ± 0.37
4.95 ± 0.24
0.10 ± 0.02
2.67 ± 0.38
d
f
3.68 ± 0.55
Lapa
e
9
g
9.96 ± 0.56 20.64 ± 1.21 Saha
c
d
The values are the average of three independent experiments run in triplicate. Lapa
e
means lapatinib, Saha means vorinostat.
Given the excellent enzymatic and cellular activities and related "drug-like"
characteristics of 9b and 9d (with lowest molecule weights which could be more
closed to meet the rule of 5), we next tested their anti-proliferative effects on three
other cancer cell lines with varying levels of EGFR and HER2-A431 (EGFR
overexpressed), SK-BR-3 (HER2 overexpressed), and NCI-H1975 (EGFR
T790M/L858R) in order to determine if these compounds could exhibit broad and
balance antitumor activities against solid tumor cells. As shown in Table 3, both 9b
and 9d possessed modest to high level of inhibitory activities against these cell lines.
Anti-proliferative effects of 9b and 9d on A431 and SK-BR-3 cells are similar to
those against A549 and BT-474 cells. Compounds 9b and 9d inhibited A549 and
A431 (EGFR inhibitor sensitive) cells with potencies comparable to that of lapatinib,
but suppressed BT-474 and SK-BR-3 (HER2 inhibitor sensitive) cells much less
potently than lapatinib. The difference of potencies on HER2 between compounds 9b,
9
d and lapatinib (176.7 and 174.9 versus 23.9 nM) may well explain why lapatinib
exhibited a better anti-proliferative effect than compound 9b and 9d against these two
breast cancer cell lines. Compounds 9b and 9d also modestly inhibited NCI-H1975
cells with potencies comparable to that of lapatinib (with IC50 values 8.76, 8.05 and

7
.85 µM respectively). It is reported that compared to wild-type EGFR selective
inhibitors, such as gefitinib and erlotinib, NCI-H1975 (EGFR T790M/L858R mutated)
cells are more sensitive to dual EGFR/HER2 inhibitor lapatinib, which could block
the heterodimerization of EGFR and HER2 by HER2 inhibition. In this study,
compound 9d inhibited HER2 less potently than lapatinib. However, the inhibition on
HDAC may compensate the anti-proliferative effect of 9d in a certain extent against
NCI-H1975 cell line, although both 9d and lapatinib are not able to overcome the
resistance caused by EGFR T790M mutation. The anti-proliferative data provide
further evidence that the synthetic compounds may exhibit broad antitumor activities
through inhibiting multiple targets, including EGFR, HER2, and HDAC pathways.
(
Table 3 should be listed here)
Table 3 Cellular inhibitory activities of the representative compounds 9b and 9d
f
against A549, A431, BT-474, SK-BR-3, and NCI-H1975 cells.
A549
IC50, μM)
A431
(IC50, μM)
BT-474
(IC50, μM)
SK-BR-3
(IC50, μM)
NCI-H1975
(IC50, μM)
Compd
(
9
9
b
d
0.51 ± 0.04
0.63 ± 0.12
1.74 ± 0.28
2.57 ± 0.37
1.95 ± 0.13
0.49 ± 0.06
0.15 ± 0.01
2.29 ± 0.04
3.63 ± 0.31
3.88 ± 0.06
0.10 ± 0.02
2.67 ± 0.38
1.27 ± 0.22
0.69 ± 0.03
0.06 ± 0.01
2.58 ± 0.13
8.76 ± 0.26
8.05 ± 1.15
7.25 ± 0.38
1.90 ± 0.09
Lapatinib
Vorinostat
f
The values are the average of three independent experiments run in triplicate.
.3. Compound 9d blocks cellular EGFR and HER2 phosphorylation and
4
induces histone H3 hyperacetylation
Given the highly potent inhibitory activity of synthetic compounds on enzymes
and cancer cell lines, we further tested whether the compounds could inhibit the
phosphorylation of EGFR and HER2 and induce histone H3 hyperacetylation on the

cellular level. Compound 9d was selected for further experimentation. Exposing cells
to 9d led to an obvious suppression of EGFR phosphorylation and activation of H3
hyperacetylation in a dose-dependent manner in A549 cells (Figure 2A), with almost
complete inhibition of p-EGFR (Tyr1068 and Tyr1173) observed at 0.5 μM and
marked upregulation of histone H3 (Lys27) hyperacetylation at 2.5 μM. In addition,
9
d also moderately decreased the levels of HER2 phosphorylation in a
dose-dependent manner in BT-474 cell lines (Figure 2B), with almost complete
inhibition of p-HER2 (Tyr1248) at 1 μM and p-HER2 (Tyr1221/1222) at 25 μM,
which is in consistent with the outstanding inhibitory activity on EGFR, HDAC and
the moderate potency on HER2 of 9d. These results show that 9d displays activities in
a variety of cancer cell types by co-targeting EGFR, HER2, and HDAC pathways.
(
Figure 2 should be listed here)
Figure 2. (A) Western blot analysis of A549 cells treated for 24 h with compound 9d,
B) Western blot analysis of BT-474 cells treated for 24 h with compound 9d.
(
4
.4. Compound 9d induces apoptosis
To further dissect the growth-inhibitory effects of synthetic compounds, we
evaluated the ability of 9d to induce apoptosis in BT-474 cancer cells. We

characterized the proapoptotic activity of 9d by flow cytometry. As shown in Figure 3,
compound 9d induced apoptosis in BT-474 cells in a concentration-dependent manner.
Treatment in BT-474 cells with 5.0 uM 9d resulted induced apoptosis in 63.86% of
cells. This result suggests that 9d is able to promote pronounced anti-proliferative and
proapoptotic effects in BT-474 cells, which may result from concerted inhibition of
EGFR, HER2, and HDAC activity.
(
Figure 3 should be listed here)
Figure 3. In vitro apoptosis analysis of 9d in BT-474 cells.

4
.5. Binding modes of compound 9d with EGFR and HDAC
To explore the binding modes of multitarget compounds with the respective
enzymes, compound 9d, which exhibited the best potency on EGFR and HDAC, was
3
4
docked into active sites of EGFR (PDB ID code: 1XKK ) and HDAC2 (PDB ID
4
3
code: 4LXZ ). Since the structure of HDAC2 (not HDAC1) crystallized with
vorinostat has been solved, in this study, we chose HDAC2 which have identical
active sites around the entrance of the channel with HDAC1 for molecular docking
4
4
simulation study.
The binding modes of compound 9d with EGFR and HDAC2 were shown in
Figure 4A and 4B. As depicted in Figure 4A, the N1 of quinazoline forms a hydrogen
bond with Met793 in the hinge while the phenyl-amino group fits into the
hydrophobic pockets of EGFR and the fluorine atom forms a hydrogen bond with
Thr854. C-6 triazole-linked long alkyl chains of hydroxamic acid extends towards the
hydrophilic region and the hydroxy group forms a hydrogen bond with Asp800.
Figure 4B demonstrates that the 4-aminoquinazoline group occupies the surface
outside the HDAC2 active pocket while the triazole-linked six carbons chain length
extends to a proper distance thus leading the hydroxamic acid group to the active site
and chelating with zinc. The hydroxamic acid group forms four hydrogen bonds with
Tyr308, Asp181, His145, and His146. The above docking results may explain why
compound 9d showed the outstanding effective inhibitory activity against EGFR and
HDAC.
(
Figure 4 should be listed here)

Figure 4. (A) Proposed binding mode of compounds 9d with EGFR (1XKK), (B)
Proposed binding mode of compounds 9d with HDAC2 (4LXZ) (N and O atoms in
blue and red, respectively). These images were generated using the SYBYL program.
5
. Conclusion
A novel series of 4-anilinoquinazolines with C-6 triazole-linked long alkyl chains
of hydroxamic acid were designed, synthesized, and evaluated as multitarget

EGFR/HER2/HDAC inhibitors. Most of the compounds displayed excellent inhibition
against these enzymes. Furthermore, compounds 9b and 9d potently inhibited
proliferation of five human cancer cells. Compound 9d also obviously inhibited the
phosphorylation of EGFR and HER2 and induced histone H3 hyperacetylation on the
cellular level. Moreover, compound 9d induced remarkable apoptosis in BT-474 cells.
4
5-47
This system network-based
multitarget inhibition approach, which is able to
simultaneously interact with multiple targets (also called polypharmacology), offers
enhanced benefits compared to single-targeted therapies, including lower probability
of developing resistance, larger therapeutic windows, and more durable
1
6,48,49
responses.
This system also proves to minimize several liabilities of drug-drug
combination approaches, including the less predictable pharmacokinetic profiles, the
potential for additive or synergistic toxicities, possibilities of drug-drug interactions,
50,51
poor patient compliance, and high treatment costs.
This work provides a novel
approach to be considered for cancer therapies, acknowledging that
polypharmacology has been gaining much attention in drug discovery and synthesis
5
2-55
as of recent.
6
6
. Experimentals
.1. Chemistry
The reagents (chemicals) were commercially available and used without further
purification. Nuclear magnetic resonance spectra were obtained using a Bruker 400
400 MHz) spectrometer using TMS as an internal standard. Chemical shifts were
reported in a d (ppm) and spin–spin coupling constants as J (Hz) values. The mass
(

spectra were obtained on a Waters Micromass Q-TOF Premier Mass Spectrometer.
Melting points were determined with a SGW X-4 digital apparatus, uncorrected and
reported in degrees Centigrade.
6
.1.1. General procedure for compounds 9a-l.
M HCl in dioxane (1 mL) was added dropwise to a solution of the 19a-l
2
intermediate (0.5 mmol) in dioxane (5 mL) and the mixture was stirred at room
temperature for 1 h. The precipitate was filtered off, washed with ethyl ether, and
dried to give the desired hydroxamic acid 9a-l.
6
.1.1.1. 6-(4-(4-((3-chloro-4-fluorophenyl)amino)quinazolin-6-yl)-1H-1,2,3-triazol
1
-
1-yl)-N-hydroxyhexanamide (9a) Yield 85%; mp 216-218C; H NMR (400 MHz,
DMSO) δ 12.07 (s, 1H), 10.43 (s, 1H), 9.63 (s, 1H), 8.99 (s, 1H), 8.96 (s, 1H), 8.56 (d,
J = 8.8 Hz, 1H), 8.12 (dd, J = 6.8, 2.5 Hz, 1H), 8.03 (d, J = 8.7 Hz, 1H), 7.85 (ddd, J
=
8.8, 4.2, 2.7 Hz, 1H), 7.57 (t, J = 9.1 Hz, 1H), 4.46 (t, J = 7.0 Hz, 2H), 2.05 – 1.81
13
(
m, 4H), 1.64 – 1.43 (m, 2H), 1.35 – 1.18 (m, 2H). C NMR (101 MHz, DMSO) δ
1
1
69.41, 160.31, 156.98, 154.53, 151.29, 145.34, 138.75, 134.51, 134.48, 133.39,
31.71, 127.09, 125.84, 125.77, 123.40, 121.18, 119.79, 119.60, 117.50, 117.28,
+
114.63, 50.16, 32.51, 29.77, 25.92, 24.97. HR-MS(ESI) m/z: Calcd for [M+H]
4
70.1508; Found: 470.1501.
6
.1.1.2. 7-(4-(4-((3-chloro-4-fluorophenyl)amino)quinazolin-6-yl)-1H-1,2,3-triazol
1
-
1-yl)-N-hydroxyheptanamide (9b) Yield 94%; mp 228-230C; H NMR (400 MHz,
DMSO) δ 12.06 (s, 1H), 10.40 (s, 1H), 9.62 (s, 1H), 8.99 (s, 1H), 8.95 (s, 1H), 8.56
dd, J = 8.7, 1.3 Hz, 1H), 8.13 (dd, J = 6.8, 2.5 Hz, 1H), 8.03 (d, J = 8.7 Hz, 1H), 7.85
(

(
ddd, J = 8.9, 4.3, 2.5 Hz, 1H), 7.58 (t, J = 9.1 Hz, 1H), 4.46 (t, J = 7.0 Hz, 2H), 2.03
1
3
–
1.75 (m, 4H), 1.56 – 1.40 (m, 2H), 1.36 – 1.20 (m, 4H). C NMR (101 MHz,
DMSO) δ 169.53, 160.30, 156.97, 154.52, 151.36, 145.35, 138.93, 134.55, 134.52,
33.40, 131.69, 127.07, 125.82, 125.75, 123.37, 121.14, 119.79, 119.61, 117.51,
17.29, 114.67, 50.24, 32.64, 29.96, 28.44, 26.07, 25.40. HR-MS(ESI) m/z: Calcd for
1
1
+
[
M+H] 484.1664; Found: 484.1661.
6
6
.1.1.3.
-(4-(4-((3-chloro-2-fluorophenyl)amino)quinazolin-6-yl)-1H-1,2,3-triazol-1-yl)-N
1
-
hydroxyhexanamide (9c) Yield 79%; mp 179-181C; H NMR (400 MHz, DMSO)
δ 12.10 (s, 1H), 10.38 (s, 1H), 9.46 (s, 1H), 8.95 (s, 1H), 8.84 (s, 1H), 8.58 (d, J = 8.6
Hz, 1H), 8.05 (d, J = 8.7 Hz, 1H), 7.68 (t, J = 6.8 Hz, 1H), 7.59 (t, J = 6.7 Hz, 1H),
7
2
1
1
2
6
7
.40 (t, J = 8.1 Hz, 1H), 4.48 (t, J = 6.9 Hz, 2H), 2.02 – 1.81 (m, 4H), 1.65 – 1.45 (m,
13
H), 1.37 – 1.17 (m, 2H). C NMR (101 MHz, DMSO) δ 169.40, 161.21, 154.30,
51.81, 151.53, 145.31, 139.13, 133.70, 131.82, 130.11, 128.07, 126.40, 125.89,
25.83, 123.31, 121.54, 121.23, 121.06, 120.89, 114.29, 50.18, 32.53, 29.80, 25.93,
+
4.97. HR-MS(ESI) m/z: Calcd for [M+H] 470.1508; Found: 470.1513.
.1.1.4.
-(4-(4-((3-chloro-2-fluorophenyl)amino)quinazolin-6-yl)-1H-1,2,3-triazol-1-yl)-N
1
-
hydroxyheptanamide (9d) Yield 91%; mp 194-196C; H NMR (400 MHz, DMSO)
δ 12.25 (s, 1H), 10.39 (s, 1H), 9.56 (s, 1H), 8.95 (s, 1H), 8.88 (s, 1H), 8.60 (dd, J =
8
7
.7, 1.4 Hz, 1H), 8.07 (d, J = 8.7 Hz, 1H), 7.72 – 7.63 (m, 1H), 7.63 – 7.52 (m, 1H),
.39 (t, J = 8.1 Hz, 1H), 4.47 (t, J = 7.0 Hz, 2H), 2.01 – 1.79 (m, 4H), 1.60 – 1.40 (m,

1
3
2
1
1
2
6
6
H), 1.38 – 1.20 (m, 4H). C NMR (101 MHz, DMSO) δ 169.53, 161.18, 154.29,
51.80, 151.59, 145.30, 139.21, 133.72, 131.80, 130.10, 128.05, 126.52, 125.91,
25.86, 123.28, 121.62, 121.14, 121.05, 120.89, 114.30, 50.25, 32.63, 29.99, 28.44,
+
6.06, 25.40. HR-MS(ESI) m/z: Calcd for [M+H] 484.1664; Found: 484.1660.
.1.1.5.
-(4-(4-((3-chloro-4-((3-fluorobenzyl)oxy)phenyl)amino)quinazolin-6-yl)-1H-1,2,3
1
-
triazol-1-yl)-N-hydroxyhexanamide (9e) Yield 75%; mp 183-185C; H NMR (400
MHz, DMSO) δ 11.90 (s, 1H), 10.41 (s, 1H), 9.55 (s, 1H), 8.91 (s, 2H), 8.51 (d, J =
8
7
1
.7 Hz, 1H), 8.04 – 7.90 (m, 2H), 7.71 (dd, J = 8.9, 2.3 Hz, 1H), 7.49 – 7.42 (m, 1H),
.35 – 7.26 (m, 3H), 7.21 – 7.12 (m, 1H), 5.28 (s, 2H), 4.43 (t, J = 6.8 Hz, 2H), 2.02 –
13
.78 (m, 4H), 1.61 – 1.42 (m, 2H), 1.32 – 1.15 (m, 2H). C NMR (101 MHz, DMSO)
δ 169.41, 163.90, 161.48, 159.98, 152.19, 145.37, 139.92, 139.84, 133.25, 131.57,
31.18, 131.09, 130.90, 126.74, 125.05, 123.87, 123.34, 121.55, 121.08, 115.41,
15.20, 114.68, 114.53, 69.85, 50.14, 32.50, 29.79, 25.90, 24.97. HR-MS(ESI) m/z:
1
1
+
Calcd for [M+H] 576.1926; Found: 576.1919.
6
7
.1.1.6.
-(4-(4-((3-chloro-4-((3-fluorobenzyl)oxy)phenyl)amino)quinazolin-6-yl)-1H-1,2,3
1
-
triazol-1-yl)-N-hydroxyheptanamide (9f) Yield 47%; mp 228-229C; H NMR
(
400 MHz, DMSO) δ 12.03 (s, 1H), 10.31 (s, 1H), 9.62 (s, 1H), 8.95 (s, 2H), 8.55 (d,
J = 8.6 Hz, 1H), 8.10 – 7.91 (m, 2H), 7.74 (d, J = 8.9 Hz, 1H), 7.54 – 7.43 (m, 1H),
7
1
.38 – 7.28 (m, 3H), 7.21 – 7.12 (m, 1H), 5.30 (s, 2H), 4.45 (t, J = 6.1 Hz, 2H), 2.23 –
13
.78 (m, 4H), 1.61 – 1.42 (m, 2H), 1.33 – 1.13 (m, 4H). C NMR (101 MHz, DMSO)

δ 170.96, 163.90, 161.47, 160.07, 152.26, 145.32, 139.91, 139.83, 133.31, 131.65,
31.18, 131.09, 130.80, 126.83, 125.15, 123.87, 123.38, 121.55, 119.36, 115.41,
15.20, 114.67, 114.52, 69.83, 50.21, 32.62, 29.96, 28.39, 26.06, 25.41. HR-MS(ESI)
1
1
+
m/z: Calcd for [M+H] 590.2083; Found: 590.2073.
6
6
3
.1.1.7.
-(4-(4-((3-chloro-4-(thiazol-2-ylmethoxy)phenyl)amino)quinazolin-6-yl)-1H-1,2,
1
-triazol-1-yl)-N-hydroxyhexanamide (9g) Yield 89%; mp 144-146C; H NMR
(
400 MHz, DMSO) δ 12.05 (s, 1H), 10.41 (s, 1H), 9.61 (s, 1H), 9.03 – 8.83 (m, 2H),
.57 (dd, J = 8.7, 1.4 Hz, 1H), 8.03 (d, J = 8.7 Hz, 1H), 7.99 (d, J = 2.5 Hz, 1H), 7.89
d, J = 3.2 Hz, 1H), 7.83 (d, J = 3.2 Hz, 1H), 7.75 (dd, J = 8.9, 2.5 Hz, 1H), 7.45 (d, J
9.1 Hz, 1H), 5.63 (s, 2H), 4.47 (t, J = 6.9 Hz, 2H), 2.03 – 1.81 (m, 4H), 1.63 – 1.46
8
(
=
13
(
m, 2H), 1.36 – 1.19 (m, 2H). C NMR (101 MHz, DMSO) δ 169.39, 165.92, 160.22,
1
1
2
6
7
3
51.92, 151.09, 145.32, 143.16, 137.98, 133.52, 131.79, 131.28, 126.94, 125.21,
23.39, 121.93, 121.69, 121.21, 120.70, 114.94, 114.55, 68.17, 50.17, 32.51, 29.79,
+
5.92, 24.96. HR-MS(ESI) m/z: Calcd for [M+H] 565.1537; Found: 565.1531.
.1.1.8.
-(4-(4-((3-chloro-4-(thiazol-2-ylmethoxy)phenyl)amino)quinazolin-6-yl)-1H-1,2,
1
-triazol-1-yl)-N-hydroxyheptanamide (9h) Yield 92%; mp 158-160C; H NMR
(
400 MHz, DMSO) δ 12.02 (s, 1H), 10.42 (s, 1H), 9.61 (s, 1H), 9.07 – 8.83 (m, 2H),
.56 (dd, J = 8.7, 1.4 Hz, 1H), 8.06 – 7.97 (m, 2H), 7.89 (d, J = 3.2 Hz, 1H), 7.84 (d, J
3.2 Hz, 1H), 7.78 (dd, J = 8.9, 2.4 Hz, 1H), 7.44 (d, J = 9.1 Hz, 1H), 5.63 (s, 2H),
.47 (t, J = 6.9 Hz, 2H), 2.01 – 1.80 (m, 4H), 1.62 – 1.41 (m, 2H), 1.39 – 1.17 (m, 4H).
8
=
4

1
3
C NMR (101 MHz, DMSO) δ 169.56, 165.92, 160.08, 151.80, 151.15, 145.35,
1
1
43.16, 138.36, 133.32, 131.67, 131.40, 126.81, 125.08, 123.36, 121.91, 121.67,
21.20, 120.94, 114.89, 114.57, 68.17, 50.23, 32.64, 29.96, 28.43, 26.06, 25.40.
+
HR-MS(ESI) m/z: Calcd for [M+H] 579.1694; Found: 579.1675.
6
.1.1.9.
N-hydroxy-6-(4-(4-((3-methyl-4-((6-methylpyridin-3-yl)oxy)phenyl)amino)quinaz
1
olin-6-yl)-1H-1,2,3-triazol-1-yl)hexanamide (9i) Yield 55%; mp 133-135C; H
NMR (400 MHz, DMSO) δ 12.16 (s, 1H), 10.33 (s, 1H), 9.75 (s, 1H), 9.02 (d, J = 6.0
Hz, 1H), 8.96 (s, 1H), 8.59 (d, J = 8.6 Hz, 1H), 8.50 (d, J = 2.3 Hz, 1H), 8.08 (d, J =
8
8
3
.7 Hz, 1H), 8.01 – 7.91 (m, 1H), 7.84 (s, 1H), 7.79 (d, J = 8.8 Hz, 1H), 7.74 (d, J =
.7 Hz, 1H), 7.18 (d, J = 8.7 Hz, 1H), 4.47 (t, J = 6.6 Hz, 2H), 2.67 (s, 3H), 2.27 (s,
13
H), 2.04 – 1.78 (m, 4H), 1.61 – 1.44 (m, 2H), 1.37 – 1.18 (m, 2H). C NMR (101
MHz, DMSO) δ 174.85, 169.42, 160.19, 153.83, 151.61, 150.96, 149.70, 145.39,
38.05, 134.11, 133.29, 131.69, 130.02, 128.53, 124.83, 123.49, 120.66, 119.99,
14.55, 50.16, 32.50, 29.77, 25.89, 24.97, 24.39, 16.42. HR-MS(ESI) m/z: Calcd for
1
1
+
[
M+H] 539.2519; Found: 539.2510.
.1.1.10.
6
N-hydroxy-7-(4-(4-((3-methyl-4-((6-methylpyridin-3-yl)oxy)phenyl)amino)quinaz
olin-6-yl)-1H-1,2,3-triazol-1-yl)heptanamide (9j) Yield 42%; mp 112-114C; 1H
NMR (400 MHz, DMSO) δ 12.19 (s, 1H), 10.31 (s, 1H), 9.76 (d, J = 6.6 Hz, 1H),
9
.03 (d, J = 6.0 Hz, 1H), 8.97 (s, 1H), 8.60 (d, J = 8.7 Hz, 1H), 8.53 (s, 1H), 8.09 (d, J
8.7 Hz, 1H), 8.05 – 7.95 (m, 1H), 7.84 (s, 1H), 7.80 (d, J = 8.8 Hz, 1H), 7.74 (d, J =
=

8
3
.7 Hz, 1H), 7.20 (d, J = 8.7 Hz, 1H), 4.47 (t, J = 6.9 Hz, 2H), 2.69 (s, 3H), 2.27 (s,
13
H), 2.00 – 1.77 (m, 4H), 1.60 – 1.40 (m, 2H), 1.37 – 1.20 (m, 4H). C NMR (101
MHz, DMSO) δ 174.93, 170.45, 160.23, 151.88, 151.56, 150.98, 148.78, 145.39,
38.04, 134.18, 133.33, 131.72, 130.05, 128.55, 124.87, 123.48, 120.66, 120.04,
14.58, 50.23, 32.62, 29.97, 28.43, 26.06, 25.42, 24.77, 16.42. HR-MS(ESI) m/z:
1
1
+
Calcd for [M+H] 553.2676; Found: 553.2682.
.1.1.11.
6
N-hydroxy-6-(4-(4-((3-methoxy-4-phenoxyphenyl)amino)quinazolin-6-yl)-1H-1,2,
1
3
-triazol-1-yl)hexanamide (9k) Yield 77%; mp 209-211C; H NMR (400 MHz,
DMSO) δ 11.94 (s, 1H), 10.43 (s, 1H), 9.63 (s, 1H), 8.97 (s, 1H), 8.96 (s, 1H), 8.57 (d,
J = 8.6 Hz, 1H), 8.04 (d, J = 8.7 Hz, 1H), 7.72 (d, J = 2.1 Hz, 1H), 7.50 (dd, J = 8.6,
2
6
1
1
1
2
6
.2 Hz, 1H), 7.40 – 7.26 (m, 2H), 7.16 (d, J = 8.6 Hz, 1H), 7.06 (t, J = 7.3 Hz, 1H),
.90 (d, J = 7.9 Hz, 2H), 4.47 (t, J = 6.8 Hz, 2H), 3.78 (s, 3H), 2.04 – 1.80 (m, 4H),
13
.62 – 1.36 (m, 2H), 1.38 – 1.17 (m, 2H). C NMR (101 MHz, DMSO) δ 169.40,
60.03, 158.20, 151.55, 151.23, 145.39, 142.24, 134.48, 133.30, 131.63, 130.30,
23.40, 122.83, 121.88, 121.11, 117.77, 116.60, 114.65, 110.77, 56.39, 50.14, 32.51,
+
9.80, 25.90, 24.97. HR-MS(ESI) m/z: Calcd for [M+H] 540.2359; Found: 540.2362.
.1.1.12.
N-hydroxy-7-(4-(4-((3-methoxy-4-phenoxyphenyl)amino)quinazolin-6-yl)-1H-1,2,
1
3
-triazol-1-yl)heptanamide (9l) Yield 88%; mp 193-195C; H NMR (400 MHz,
DMSO) δ 11.95 (s, 1H), 10.39 (s, 1H), 10.24 (s, 1H), 9.59 (s, 1H), 8.98 (s, 1H), 8.93
s, 1H), 8.57 (d, J = 8.8 Hz, 1H), 8.03 (d, J = 8.7 Hz, 1H), 7.68 (d, J = 2.2 Hz, 1H),
(

7
.47 (dd, J = 8.6, 2.4 Hz, 1H), 7.39 – 7.32 (m, 2H), 7.17 (d, J = 8.6 Hz, 1H), 7.07 (t, J
7.4 Hz, 1H), 6.90 (d, J = 7.8 Hz, 2H), 4.48 (t, J = 6.9 Hz, 2H), 3.79 (s, 3H), 1.98 –
=
13
1
.82 (m, 4H), 1.57 – 1.41 (m, 2H), 1.37 – 1.22 (m, 4H). C NMR (101 MHz, DMSO)
δ 169.53, 160.13, 158.19, 151.59, 151.21, 145.36, 142.45, 134.38, 133.48, 131.73,
1
3
30.29, 123.34, 122.87, 121.85, 121.06, 117.87, 116.67, 114.64, 110.90, 56.44, 50.24,
+
2.63, 29.98, 28.43, 26.07, 25.40. HR-MS(ESI) m/z: Calcd for [M+H] 554.2516;
Found: 554.2512.
6
6
.2. Bioassay
.2.1. EGFR and HER2 inhibition assay
In vitro kinase assays were carried out by Medicilon Co., Ltd in Shanghai, China.
The general procedures were as the following: 1×kinase buffer was prepared,
compounds were transferred to assay plate by echo with 3-fold dilution in
2
.5%DMSO/kinase buffer. Then kinases, substrates and compounds in the reaction
buffer were mixed. The assay plate was incubated at RT for 15 minutes. ATP was then
added to each well to start reaction (positive controls contained gefitinib or lapatinib
and all the above components except the inhibitor. The negative controls contained
neither enzyme nor inhibitor). The assay plate was incubated at RT for 30 minutes and
reaction was stopped by the addition of Streptavidin-XL665 and TK antibody
europium cryptate (1:100) solution. The luminescence was monitored at 320 nm
(
(
excitation) and 665/615 nm (emission) using an EnVision multilabel plate reader
PerkinElmer Life Sciences, Boston, MA, USA). The signal was correlated with the
amount of ATP remaining in the reaction and was inversely correlated with the kinase

activity. The Emission Ratio was calculated with the formula: 665 nm Emission
signal/615 nm Emission signal. IC50 values were calculated from the inhibitory
curves.
6
.2.2. HDAC inhibition assay
In vitro HDAC assays were carried out by Shanghai Chempartner Co., Ltd in
Shanghai, China. The general procedures were as the following: 1×assay buffer
modified Tris Buffer) was prepared, compounds were transferred to assay plate by
(
echo with 3-fold dilution in 100% DMSO. Substrate solution was made by preparing
enzyme solution in 1×assay buffer and adding trypsin and Ac-peptide substrate in
1
×assay buffer. Enzyme solution or 1×assay buffer were transferred to assay plate or
for low control and incubated at RT for 15 minutes. Substrate solution was then added
to each well to start reaction (positive controls contained vorinostat and all the above
components except the inhibitor. The negative controls contained neither enzyme nor
inhibitor). The plates were incubated for 60 minutes at room temperature to allow the
fluorescence signal to develop. The fluorescence generated was monitored at 355 nm
(
excitation) and 460 nm (emission) using a Synergy MX plate reader (PerkinElmer
Life Sciences, Boston, MA, USA). IC50 values were calculated from the inhibitory
curves.
6
.2.3. Cell culture
The human cancer cell lines, A549, A431, BT-474, SK-BR-3 and NCI-H1975
were purchased from American Type Culture Collection (Manassas, VA, USA) and
cultured according to the supplier's instructions.

6
.2.4. Cytotoxicity assay
Cell viability was assessed by the MTT assay. Briefly, 3-8×10 cells were seeded
3
into 96-well plates for 12h, followed by incubation with various doses of compounds
for 48 h. After adding 10 μL per well of MTT (5 mg/ml) solution, the formazan
crystals were dissolved in 100 μL per well DMSO. The absorbance was measured
using Multimode Detector (Beckman Coulter, Fullerton, CA, USA). Three
independent experiments were performed.
6
.2.5. Western blot analysis
0-40 μg protein was separated by 7.5%-15% SDS-PAGE and transferred to
PVDF membrane (Roche). Membranes were blocked for 2h in 1×TBST containing 5%
w/v) BSA, and then incubated with a primary specific antibody in 5% of BSA for
-12h at 4°C. Followed by a HRP-conjugated anti-mouse or anti-rabbit second
2
(
8
antibodies, proteins were visualized with the SuperSignal West Pico
Chemiluminescent Substrate kit (Pierce). Antibodies against the following were used:
p-EGFR (Tyr1068), p-EGFR (Tyr1173), EGFR, p-HER2 (Tyr1221/1222), p-HER2
(
Tyr1248), HER2, Acetyl-Histone H3 (Lys27) were obtained from Cell Signaling
Technologies (Cambridge, MA).
6
.2.6. Flow cytometry assay
Phosphatidylserine externalization was measured by Annexin V-FITC/PI
apoptosis detection kit (Beyotime Company) according to the manufacturer's
instructions.
6
.2.7. Docking studies