Discovery of Small Molecules for Repressing Cap-Independent Translation of Human Vascular Endothelial Growth Factor (hVEGF) as Novel Antitumor Agents

Article abstract of DOI:10.1021/acs.jmedchem.6b01444

Angiogenesis is important in tumorigenesis and tumor progression. Human vascular endothelial growth factor (hVEGF) is an angiogenic growth factor that plays a crucial role in tumor progression. The G-rich region within the 5′-untranslated regions (5′-UTR) of hVEGF-A mRNA can form a "switchable" RNA G-quadruplex structure that is essential for a cap-independent translation initiation. We screened our small-molecule library for binders of this G-tract. One novel quinazoline derivative, compound 1, showed a significant specific interaction with the G-tract and destabilized the G-quadruplex structure. The results of cellular experiments revealed that compound 1 down-regulated hVEGF-A translation and significantly impeded tumor cells migration. We also found that compound 1 exhibited tumor-inhibiting activity in MCF-7 xenograft tumors, which might be related to its ability to reduce hVEGF expression. These findings present a new strategy of hVEGF-A translational control in which small molecules interact with G-quadruplex structure in the 5′UTR.

Full text of DOI:10.1021/acs.jmedchem.6b01444

Subscriber access provided by Binghamton University | Libraries
Article
Discovery of Small Molecules for Repressing Cap-
independent Translation of Human Vascular Endothelial
Growth Factor (hVEGF) as Novel Anti-Tumor Agents
Shi-Ke Wang, Yue Wu, Xiao-Qin Wang, Guo-Tao Kuang, Qi Zhang, Shu-
Ling Lin, Hui-Yun Liu, Jia-Heng Tan, Zhi-Shu Huang, and Tian-Miao Ou
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 22 May 2017
Downloaded from http://pubs.acs.org on May 23, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Small Molecules for Repressing Cap-independent Translation of
Human Vascular Endothelial Growth Factor (hVEGF) as Novel Anti-Tumor
Agents
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Shi-Ke Wang , Yue Wu , Xiao-Qin Wang , Guo-Tao Kuang , Qi Zhang , Shu-Ling
,†
1,†
1,2
1
1
1
1
1
1
Lin , Hui-Yun Liu , Jia-Heng Tan , Zhi-Shu Huang , Tian-Miao Ou
*,1
1
School of Pharmaceutical Sciences, Sun Yat-sen University, 132 Waihuan East Road,
Guangzhou University City, Guangzhou 510006, People’s Republic of China.
2
School of Pharmacy, Guangdong Medical College, 1 Xincheng Avenue, Dongguan
5
23808, People’s Republic of China.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Angiogenesis is important in tumorigenesis and tumor progression. Human vascular
endothelial growth factor (hVEGF) is an angiogenic growth factor that plays a crucial
role in tumor progression. The G-rich region within the 5’-untranslated regions
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(5’-UTR) of hVEGF-A mRNA can form a ‘switchable’ RNA G-quadruplex structure
that is essential for a cap-independent translation initiation. We screened our
small-molecule library for binders of this G-tract. One novel quinazoline derivative,
compound 1, showed a significant specific interaction with the G-tract and
destabilized the G-quadruplex structure. The results of cellular experiments revealed
that compound 1 down-regulated hVEGF-A translation, and significantly impeded
tumor cells migration. We also found that compound 1 exhibited tumor-inhibiting
activity in MCF-7 xenograft tumors, which might be related to its ability to reduce
hVEGF expression. These findings present a new strategy of hVEGF-A translational
control in which small molecules interact with G-quadruplex structure in the 5’UTR.
KEYWORDS: hVEGF-A, translational regulation, G-quadruplex, quinazoline
derivative, antitumor.
ACS Paragon Plus Environment

Page 3 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Angiogenesis is essential for effective wound healing and blood flow restoration in
the impaired tissues of embryos and healthy adults. Angiogenic factors, such as
vascular endothelial growth factor (VEGF), act as mediators of tumor angiogenesis by
stimulating the growth of new blood vessels from nearby capillaries and providing
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tumor cells with oxygen and nutrients. The expression of the human VEGF-A
1
hVEGF-A) is induced by hypoxia and hypoglycaemia, and VEGF-A promotes a
(
2
wide range of functions, including adhesion, survival, migration and invasion.
Moreover, the role of VEGF-A is now well established and forms the basis for using
3
anti-VEGF-A drugs to treat neovascular disorders.
G-quadruplexes are DNA or RNA secondary structures composed of stacked
4
guanine (G) through Hoogsteen hydrogen bonds, and they have been found existing
5
widely in whole genome by genomics and bioinformatics studies. These structures
-7
may play important roles in several biological processes, such as DNA replication,
8
gene transcription and the maintenance of genomic stability. Using the G-quadruplex
antibody BG4, quadruplexes were shown to be more prevalent in cancer cells than
9
non-neoplastic tissues, which implies that cancer cells are more sensitive than
non-neoplastic cells to small G-quadruplex ligands.
Several G-quadruplexes form in the 5’-untranslated regions (5’-UTR) of mRNAs
1
0
11
and act as translational repressors, such as NRAS, Zic-1, MT3-MMP, TRF2, and
12
13
14
cyclin D3 mRNAs. However, a G-quadruplex in the 5’-UTR of hVEGF-A mRNA
1
5-16
may promote the translation via a cap-independent approach.
Specifically, an
18-nucleotide (nt) sequence within the internal ribosome entry site A (IRES-A) of the
hVEGF-A’s 5’-UTR can independently fold into a RNA G-quadruplex and directly
17
recruits a 40S ribosomal subunit via cap-independent translation initiation.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Nevertheless, a recent study showed that this G-quadruplex is optional for
cap-independent function, which made the function of the G-quadruplex in IRES
18
more complicated. Although some G-quadruplex stabilizers were shown to inhibit
1
the cap-independent translation in the previous study, no specific ligands have been
8
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
19
found to date. In addition, a few molecules, such as cationic porphyrins (TMPyP4)
20
and TAPs , showed activity for disrupting G-quadruplexes, but these molecules do
not selectively interact with the hVEGF-A’s IRES. To clarify the function of the
G-quadruplex structure in the IRES, in this study, we aimed to find a specific IRES
G-rich sequence binder and determine whether it exhibits anti-tumour activity. In
addition, a specific IRES G-rich sequence binder can help for clarifying the role of
G-quadruplex in cap-independent translation of hVEGF-A.
We first obtained additional evidence for the existence of the RNA G-quadruplex
in the IRES using the G-quadruplex antibody BG4 and in vitro transcribed
long-strand mRNA (Supplementary results, Fig. S1). Then, we screened our
compound library (built by the School of Pharmaceutical Sciences, Sun Yat-sen
University) containing 144 natural products and related derivatives with diverse
structures using a surface plasmon resonance (SPR) biosensor. Among these
compounds,
we
found
a
novel
quinazoline
derivative,
N-((2-(4-(N-(4-(2-diethylaminoethoxy))phenyl)
amino)-quinazolin-2-yl)
phenyl)-3-(4-methylpiperazin-1-yl) propanamide (compound 1), could specifically
bind to the G-rich sequence in the IRES-A in vitro. We further explored the possible
mechanism of this compound on the G-rich sequence in the IRES-A through in vitro
and cellular experiments, and suggested that this compound could destabilize the
G-quadruplex structure. Additional in vitro and in vivo experiments suggested that the
compound could hinder cell migration and repress tumor growth by decreasing
ACS Paragon Plus Environment

Page 5 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
VEGF-A protein expression. Taken together, these findings illustrated that the
translational regulatory activity of the G-rich sequence in the IRES repressed
hVEGF-A’s translation and that this regulation could be chemically repressed by a
small molecule.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS
Survey of Small Molecules that Selectively Binding to the IRES-A RNA
We used SPR experiments to screen our compound library containing 144 natural
products and related derivatives with diverse structures (Supplementary Fig. S2),
including quinolone derivatives, quinazoline derivatives, berberine derivatives, and
curcumin derivatives. The oligomers immobilized on a GLH chip included the
biotin-labelled Biotin_IRES-A (WT), Biotin_IRES-A (MU1), and Biotin_hairpin
(Supplementary Table S1).
The SPR data were shown in Fig. 1 and Supplementary Table S2; the y-axis
shows the dissociation constants (K ), indicating the binding affinity between the
D
compounds and the immobilized oligomers. We divided the compounds into three
groups according to the results: group 1 (Fig. 1a) included the compounds with
certain selectivity for IRES-WT vs. hairpin DNA; group 2 (Fig. 1b) included the
compounds exhibiting similar binding to both RNA oligomers and hairpin DNA;
group 3 included the compounds whose binding abilities could not be detected (Table
S2), such as all of the curcumin derivatives.
In group 2, we found some compounds that had been previously identified as
strong
DNA
G-quadruplex
stabilizers,
such
as
the
1
N -(11H-indolo[3,2-b]quinolin-11-yl)-N ,N -dimethylpropane-1,3-diamine (Q73 in
3
3
21
the screen data and SYUIQ-05 in the reference, numbered as 2 in this study) , and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the
1
5
1-((3-(dimethylamino)propyl)amino)-7-fluoro-5-methyl-11H-indolo[3,2-b]quinolin-
22
-ium (Q71 in the screen data and 2r in the reference) , which lent support to our
SPR screening data. Among all the screened compounds, compounds in series a
(quinazoline derivatives) exhibited high affinity and selectivity for the IRES-A (WT)
sequence; compounds of berberine derivatives and quinolone derivatives showed
polarized effects, making the specialized SAR analysis difficult (additional work is
ongoing, but the data are not shown in this manuscript).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Series a appeared to be the best group, although it included only three compounds.
This series was derived from a bi-substituted quinazoline derivative, who possesses
an interaction with the telomeric G-quadruplex DNA, thereby inhibiting tumour cell
23
proliferation and inducing cell senescence. Among them, compound 1 (Scheme 1)
bound to the IRES-A (WT) with a K
D
of 0.928 ꢀM, and bound to the hairpin DNA
with a K
D
of 21.2 ꢀM. Since we aimed to select a compound with specific binding to
the IRES-A sequence, we chose compound 1 for further studies because it exhibited
the highest binding affinity and highest IRES/hairpin selectivity (23 folds) for the
18-nt biotin-labeled IRES-A(WT) oligomer (Fig. 1 and Supplementary Table S2).
To be noted, the selectivity of compound 1 binding to the IRES-A (WT) vs. the
IRES-A (MU1) was not significant (4 folds). This might due to the mutant IRES-A
oligomer we used in the SPR assay still has the ability to form partial G-quadruplex
structure (Supplementary Fig. S1 and Fig. S3a). To exclude the complexity from the
mutant IRES-A, we designed the other mutant sequence, the IRES-A (MU2) sequence,
which has no ability to form a stable G-quadruplex (Supplementary Fig. S3a) for the
following experiments.
ACS Paragon Plus Environment

Page 7 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 1 Binds to the IRES-A RNA G-Quadruplex with High Affinity and
Selectivity.
To verify the binding abilities of compound 1 to RNA in the free state, microscale
thermophoresis (MST) measurements were performed using an 18-nt FAM-labelled
IRES-A (WT) with compound 1 at increasing concentrations (Fig. 2a). After a 5-min
incubation at 25 °C, MST analysis was performed on the Monolith NT.115 instrument
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(NanoTemper Technologies, Germany). The K
D
value of compound 1 for IRES-A
(WT) was 1.29 ꢀM, which was similar to the SPR result. In addition, the binding of
compound 1 to the IRES-A mutant RNA oligomer (IRES-MU1) was weaker than that
to the IRES-A WT (K = 13.4 ꢀM, Fig. 2c), which was consistence with the data in
D
SPR screening. We further used the IRES-A (MU2) RNA that could not form
G-quadruplex structure as a reference, and compound 1 could not reach the binding
platform at the highest concentration of 100 ꢀM (Fig. 2c). In addition, compound 1
bound to DNA G-rich sequence with much weaker affinity, such as the IRES-A
counterpart DNA oligomer and the c-kit1 G-quadruplex DNA (Supplementary Fig.
S4). These findings suggested that compound 1 could selectively bind to the
G-quadruplex-forming sequence in the IRES-A region.
Competition dialysis measurements were performed to further verify the selectivity
of the compound. The IRES-A (WT) oligomer together with several other oligomers
(Supplementary Table S1), including a RNA hairpin, random RNA, CT DNA,
IRES-MU2, flanking sequences of IRES (flank 1 and flank 2, Supplementary Fig.
S3b), DNA G-quadruplexes (TBA, c-kit 1, and c-myc Pu27), and other RNA
G-quadruplexes (ADAM10, TERRA, BCL-2, and NRAS), were placed in
independent dialysis cassettes after the general annealing process. All the dialysis
cassettes were submerged in a solution containing compound 1 and allowed to dialyze
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for 24 h. After spectrophotometric quantitation, compound 1 was found to exhibit the
highest binding with the IRES-WT, and weaker binding with other DNA and RNA
structures (Fig. 2b).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound 1 Reduces the Stability of the IRES-A G-Quadruplex.
Based on the evidence that compound 1 could specifically bind to the
G-quadruplex-forming sequence in the IRES-A in vitro, we further investigated
whether compound 1 affected G-quadruplex’s stability. CD spectra were collected to
determine the secondary structure of the IRES-A (WT) oligomer. The oligomer
formed a typical parallel G-quadruplex in a 10 mM Tris-HCl buffer (pH 7.4, 100 mM
KCl), and the signals representing G-quadruplexes at approximately 260 nm
decreased in a dose-dependent manner after the addition of compound 1 (Fig. 3a). In
contrast, the G-quadruplex signals of the IRES-MU2, IRES-A counterpart DNA and
the G-rich sequence in hVEGF-A’s promoter (VEGF promoter) changed very little in
the presence of compound 1 (Fig. 3a and Supplementary Fig. S5a).
Next, we performed CD melting experiments to evaluate the thermodynamic
stabilization effect of compound 1 on the IRES-A G-quadruplex (Fig. 3b,
Supplementary Fig. S5b and S5c). We used the CD signals at 262 nm to calculate
the melting temperature (T ). The T values of the IRES-A G-quadruplex in the
m
m
absence and presence of 5 ꢀM compound 1 were 80.9 ± 0.8 °C, and 62.2 ± 1.2 °C,
respectively (Fig. 3b), indicating that the thermostability of the G-quadruplex in this
sequence decreased in the presence of compound 1.
To further validate the destabilization effect of compound 1 on the IRES-A
G-quadruplex structure, we performed an electrophoresis mobility shift assay
(EMSA). As shown in Fig. 3c, as the concentrations of compound 1 increased, the
ACS Paragon Plus Environment

Page 9 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
lower bands representing the IRES-A (WT) G-quadruplex weakened, and the upper
bands representing ssRNA increased. In contrast, compound 1 could not induce
significant changes to the reference sequence IRES-A (MU2).
To obtain additional evidence relating to compound 1’s destabilizing effect,
experiments using the G-quadruplex antibody BG4 were applied. A recent study
demonstrated the competitive binding of BG4 and a G-quadruplex stabilizer,
pyridostatin, to the TERRA G-quadruplex and subsequent rearrangement into a more
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
stable BG4-PDS-RNA complex. However, in the presence of compound 1, the
situation was different. A filter binding assay (Fig. 4a), a pull-down assay (Fig. 4b),
and an EMSA (Supplementary Fig. S6) experiment with the BG4 were performed to
detect the influence of compound 1 on BG4’s binding to the oligomer. Both of them
showed that the RNA-BG4 complex reduced. We also performed MST measurements
in a ternary system containing BG4, in vitro transcribed IRES-A mRNAs and the
compound 1. When the in vitro transcribed mRNA annealed with compound 1, the
binding ability of BG4 antibody to IRES-A G-rich mRNA dramatically decreased,
and the K
D
value changed from 10.8 nM to 4.64 ꢀM, corresponding to a nearly
4
30-fold reduction (Fig. 4c and 4d). In contrast, the BG4 antibody could not
specifically bind to the IRES-A (MU2) sequence in the absence or presence of
compound 1 (Fig. 4e and 4f). These findings further implied that, unlike the stabilizer
24
pyridostatin, compound 1 destabilized the RNA G-quadruplex in the hVEGF-A
mRNA, and BG4 could not recognize the single-stranded RNA anymore.
In addition, we used a RNA immunoprecipitation (RIP) assay to detect the binding
of BG4 to the IRES-A region after compound 1 treatment in MCF-7 human breast
cancer cells. After transfecting the pcDNA3.1-BG4 plasmid into the MCF-7 cells and
incubating them with compound 1 (2.5 or 5 ꢀM), complexes of the BG4 antibody
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with its RNA targets were obtained using protein A/G plus agarose beads. The
complexes were detected via dot blotting, reverse transcription-polymerase chain
reaction (RT-PCR), and quantitative real-time RT-PCR (qRT-PCR). The amount of
BG4-bound IRES-A sequences reduced by compound 1 in a dose-dependent manner
according to the dot blotting assay (Fig. 5a). Furthermore, RT-PCR (Fig. 5b) and
qRT-PCR (Fig. 5c) were employed to quantitatively detect the amount of BG4-bound
IRES-A using the IRES-A sequence-specific primers RT_IRES-A and RT_IRES-S
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Supplementary Table S1). The results were in good agreement with those of dot
blotting and showed that the bound IRES-A RNA immunoprecipitated by BG4
significantly decreased after the addition of compound 1. In addition, compound 1
could not influence on the expression of BG4 protein (Supplementary Fig. S7).
Taken together, these results demonstrated that compound 1 might destabilize the
G-quadruplex structure of IRES-A (WT) RNA, thereby decrease the recognition and
binding of the BG4 antibody to the IRES-A (WT) G-quadruplex in a dose-dependent
manner.
Compound 1 Represses Cap-Independent Translation Initiation Mediated via
G-Quadruplex within the IRES-A.
The G-rich sequence G774-G790 within the IRES-A of hVEGF-A’s 5’-UTR has
17
been shown to be critical for the translation initiation activity of IRES-A. To explore
whether compound 1 affected IRES-A’s function, we constructed a dual-luciferase
reporter assay for detecting the translation level of hVEGF-A. For constructing the
recombinant plasmids, the 293-nt IRES-A (WT), IRES-A (MU2), or IRES-A (MU3)
cDNA was inserted upstream of the Met-luciferase gene (MetLuc) (Fig. 6a and
Supplementary Fig. S1a). To be noted, sequence MU2 contains an alternative start
ACS Paragon Plus Environment

Page 11 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
codon AUG that might influence on the transcription, and we designed the other
mutant sequence MU3 (Supplementary Table S1) as another negative reference in
the luciferase reporter experiment. By measuring the expression level of luciferase
and using the Renilla-luciferase gene (RLuc) as an internal control, this
dual-luciferase reporter system allowed for the quantification of the effect of
compound 1 on the translation-initiation activity of IRES-A.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the dual-luciferase experiments, 1.5 ꢀM was the highest concentration of
compound 1 used because the cytotoxicity IC50 of compound 1 in MCF-7 cells was
1
2.5 ꢀM, and the compound could be uptake ranged from 65.5% (24 h) to 82.7% (48
h). As shown in Fig. 6b, the MCF-7 cells treated with 1.5-ꢀM compound 1 exhibited
an 80% decrease in secreted MetLuc activity for the WT plasmid, whereas no
significant change was observed for all the mutant plasmids.
Subsequently, we performed an enzyme-linked immunosorbent assay (ELISA) of
secreted VEGF-A protein and a Western blot of cellular VEGF-A protein in order to
test the influence of compound 1 on IRES-A downstream protein VEGF-A expression
in MCF-7 cells. The ELISA results showed a decrease of approximately 75% in
VEGF-A protein expression when cells were treated with 1.5-ꢀM compound 1 (Fig.
6
c). In addition, the Western blot results revealed that VEGF-A protein also decreased
after treatment with increasing concentrations of compound 1 (Fig. 6d). In contrast,
no effect was observed on the transcription process, as shown by RT-PCR and
qRT-PCR assays of the IRES-A cDNA in the luciferase plasmids (Fig. 6e) and the
hVEGF-A gene in the MCF-7 cells genome (Fig. 6f and 6g). Therefore, compound 1
could down-regulate hVEGF-A expression at the translational level but not at the
transcriptional level, indicating that this down-regulation likely occurred via
interactions with the RNA G-quadruplex structure within the IRES-A region of the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hVEGF-A’s 5’-UTR. To further support this finding, a PCR stop assay was performed
to confirm that compound 1 did not affect the transcription process, even at a high
concentration (10 ꢀM) (Supplementary Fig. S8).
To strengthen the data on compound 1’s destabilizing activity and because most
G-quadruplex ligands are known to stabilize G-quadruplex structures, it was
important to compare the effect of compound 1 on IRES-A G-quadruplex function
with that of the G-quadruplex ligands that were previously reported to stabilize this
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
structure.
We
used
compound
2
and
the
1
-(2-aminoethoxy)-N ,N -bis(4-(2-aminoethoxy)quinolin-2-yl)isophthalamide (PDS)
3
5
2
as reference compounds. As shown in Supplementary Fig. S9a, compound 2 could
5
o
stabilize the quadruplex in IRES-A; the melting temperature was over 81 C since the
curve could not reach a platform at the highest detected temperatures. In the
dual-luciferase assay (Supplementary Fig. S9b), compound 2 showed an opposite
effect on IRES-A-mediated luciferase expression: increasing the expression of
luciferase instead of decreasing the expression by compound 1. However, compound
2
also increased the expression of luciferase inserted by the IRES-MU3 sequence,
which might come from its poor selectivity between G-quadruplex and duplex DNA.
On the other hand, compound 3 could also stabilize the quadruplex in IRES-A with a
o
melting temperature of over 76 C (Supplementary Fig. S9c). Luciferase results
turned out that this compound could increase the expression in the WT plasmid, but
showed no significant effect in the MU3 plasmid (Supplementary Fig. S9d). This
was well supported our hypothesis, that destabilizing the quadruplex in the IRES-A
region of hVEGF-A can promote the expression of this gene.
Compound 1 Impedes the Biological Function of the VEGF-A Protein.
ACS Paragon Plus Environment

Page 13 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Based on our results, we attempted to detect the impact of compound 1 on cell
migration, which is the main biological function of VEGF-A protein, by performing a
26 27
cell scratch test and a trans-well test . First, all the MCF-7 cells were treated with
1
00-ꢀM CoCl2 to create a chemically hypoxic environment for VEGF-A
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
8-29
cap-independent expression; the blank control was not treated.
The cells were
treated with increasing concentrations of compound 1 or 0.03% dimethylsulfoxide
DMSO) at the same time. On the other hand, small interfering RNA (siRNA)
(
targeting VEGF-A was used as a negative control, and VEGF-A protein was used as
another control. Some of the cells treated with compound 1 and incubated for 48 h
were further treated with VEGF-A protein and incubated for another 12 h (Fig. 7a) to
re-trigger the migration. Then, we quantified the results with Image J software to
analyze the wound closure of each well. The results showed that the relative wound
closure (compared to the DMSO control group) of migrated MCF-7 cells significantly
decreased to ~35% in the presence of 3-ꢀM compound 1; this value was similar to the
~
30% decrease observed for cells transfected with hVEGF-A siRNA group (Fig. 7b).
Meanwhile, Western blot experiments were conducted to detect the VEGF-A protein
expression of corresponding cell samples. This analysis also revealed a drastic
reduction when cells were treated with 3-ꢀM compound 1, similar to the result
obtained for cells in which the VEGF-A protein was knocked down by siRNA (Fig.
7
c). These findings indicated that MCF-7 cells migration depended on the amount of
VEGF-A protein; and this function of VEGF-A protein could be hindered by
compound 1 in a dose-dependent manner. The cells were able to migrate again after
treatment with exogenous VEGF-A protein during a 12-h incubation at 37 °C, which
further supported that the effect of compound 1 come from its interaction with
VEGF-A.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The trans-well tests were conducted in another human breast tumor cell line,
MDA-MB-231, which expresses more VEGF-A protein and has higher invasive and
2
migration capacities than MCF-7 cells. The cells were treated with compound 1 at
concentrations ranging from 0.375 ꢀM to 3 ꢀM. The results were in good agreement
with those of the cell scratch test. MDA-MB-231 cell migration was reduced to
approximately 25% when cells were treated with 3-ꢀM compound 1 (Fig. 7d). To
continuously monitor migration, we employed a real-time cell analyzer (RTCA)
system to monitor and record the MDA-MB-231 cell migration rate over 24 h after
cells were seeded into a 16-well CIM-Plate in the presence of increasing
concentrations of compound 1. Each well consisted of a top chamber, a bottom
chamber, a microporous membrane between the two compartments, and
microelectrodes as cell-migration sensors. The cell index indicated the amount of
cells migrating from the upper compartment to the lower compartment in each well.
The results showed that compound 1 suppressed cell migration after 4 h and that this
effect plateaued after 10 h (Fig. 7e).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Therefore, in tumor cells, we observed that compound 1 not only down-regulated
hVEGF-A expression but also significantly impeded cell migration in a
dose-dependent manner.
Compound 1 Inhibits Tumor Growth in a Human Breast Tumor Xenograft.
Considering the importance of VEGF-A for tumor growth, we evaluated compound
1
’s antitumor activity in vivo against a human breast tumor xenograft (MCF-7) in
athymic nude mice. The results were presented in Fig. 8. An experimental group of
five tumor-bearing mice was treated with compound 1 once a day at a dose of 7.5
mg/Kg for 20 days. We measured the body weight (Fig. 8a) and the tumor diameter
ACS Paragon Plus Environment

Page 15 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(Fig. 8b) every day. After 20 days, all the mice were executed and the tumors were
dissected and weighed (Fig. 8c). In this experiment, doxurubicin, a first-line
anti-neoplastic agent for treating breast cancer, was used as a positive control with a
dose of 1 mg/Kg once a day.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Firstly, no significant change in body weight was observed during the treatment
(
Fig. 8a), indicating less side effects than the positive control doxurubicin. The
negative control group, which was treated with saline only, had an average tumor
3
volume >700 mm after 20 days. In contrast, the tumor-bearing mice treated with
3
compound 1 had an average tumor volume of < 300 mm (Fig. 8b). The tumor weight
in the presence of compound 1 reduced around 60.1% to a final weight of 0.18 g (Fig.
8
c). On the other hand, the positive control group showed an effect similar to that of
3
compound 1 with an average tumor volume of < 300 mm ; and the tumor weight
reduced to 0.14 g (Fig. 8c). These data suggested that compound 1 at a dose of 7.5
mg/Kg showed anti-tumor activity on the MCF-7 xenografted mice.
To determine whether the expression of hVEGF-A was consistently affected within
tumors by treatment with compound 1, we assessed the VEGF-A level (brown area in
Fig. 8d) in tumor tissue using immunohistochemistry (IHC) evaluation. The
experimental group exhibited significantly decreased expression relative to the
negative control group. The results indicated hVEGF-A down-regulation by
compound 1 might contribute to tumor inhibition in vivo.
DISCUSSION AND CONCLUSIONS
G-quadruplexes readily formed a stable structure in mRNAs despite the routine role
30
of the putative quadruplex sequence (PQS) in DNA. Applying luciferase assays and
nonspecific G-quadruplex binders to reveal the function of the G-quadruplex in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
8
cap-independent translation initiation, as in previous studies, was not effective
enough for obtaining convincing conclusion. Finding a real specific ligand for this
structure would help to reliably illustrate the role of translational regulation in the
quadruplex.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
31
Although several DNA G-quadruplex small ligands, such as pyridostatins,
1
9
13, 32
cationic porphyrins, and bisquinoliniums
can modulate translation both in vitro
or cellular assays and are presumed to target RNA G-quadruplexes, little evidence
indicated a compound that could selectively decrease the stability of a RNA
G-quadruplex structure. Therefore, we performed a small-scale screening via SPR and
identified a new quinazoline compound 1, with the ability to selectively interact with
the RNA-quadruplex-forming mRNA sequence.
One of the major goals of this study was to determine whether this compound could
destabilize the G-quadruplex structure. In addition to CD and EMSA experiments, we
also tried to use BG4 to verify this hypothesis. It turned out that the recognition and
binding of BG4 to this G-quadruplex structure in the IRES region weakened in the
presence of compound 1. To be noted, considering the complexity of G-quadruplexes’
conformation, there exists another possibility that compound 1 induces a
conformational transition or influences on the recognition of transcription repressor,
which is hard to distinguish from the destabilizing effect.
Our present results supported the hypothesis that a small molecule can recognize
the G-rich sequence in the IRES region of hVEGF-A mRNA and destabilize the
secondary structure, and subsequently impede the biological functions of VEGF-A
protein to inhibit tumor growth. Determining why this compound has selectivity for
the RNA G-quadruplex is very important for the further development of this molecule.
However, due to the limitation of cellular detection, it’s hard to confirm that
ACS Paragon Plus Environment

Page 17 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound 1 directly targets the hVEGF-A mRNA within the whole transcriptome.
Taken together, the ability of compound 1 down-regulating the translation of the
hVEGF-A and inhibiting tumour activity, opens up new experimental strategies to
screen for novel compounds from natural compounds and derivatives that interfere
with tumour cells and have therapeutic applications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
1. Materials. The oligomers and primers used in present work were listed in
Supplementary Table S1. Primers were purchased from Invitrogen (China), Generay
(China), and Integrated Device Technology (USA). All other chemicals and reagents
were commercial available.
2. Chemistry. The synthetic pathway for compound 1 was shown in Scheme 1.
Intermediate 4, 5, and 6 were prepared mainly according to the previously reported
38
procedure. Compound 7 was prepared by mixing 6 and p-methoxyaniline in EtOH
under reflux condition, compound 7 reacted with BBr in a dichloromethane solution
3
at -78 °C. The resulting solution was stirred overnight at ambient temperature and
gained compound 8. Compound 9 was prepared by mixing 8 and 1, 2-dibromoethane
in an acetone solution under a basic condition and refluxed for 24 h，then coupling
with diethylamine in tetrahydrofuran with refluxing to obtain compound 10. In order
to obtain the amine functionality, the conventional procedure for the reduction of the
nitro group was used, with 80% hydrazine hydrate in the presence of 10% Pd/C and
isopropanol as a solvent. Reaction of amine 11 with chloropropionyl chloride and
potassium carbonate in CHCl resulted compound 12. The product was directly used
3
in the next reaction without purification. Finally, the target compound 1 was obtained
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
through aminolysis of compound 12 using 4-methyl piperazine with refluxing.
1 13
Synthesis and characterization. H and C NMR spectra were recorded using TMS as
the internal standard in DMSO-d or CDCl with a Bruker BioSpin GmbH
3
6
spectrometer at 400 MHz and 100 MHz, respectively; Mass spectra (MS) were
recorded on a Shimadzu LCMS-2010A instrument with an ESI or ACPI mass
selective detector, and high resolution mass spectra (HRMS) were recorded on a
Shimadzu LCMS-IT-TOF instrument. Melting points (m.p.) were determined using
an SRS-OptiMelt automated melting point instrument without correction. Flash
column chromatography was performed with silica gel (200-300 mesh) purchased
from Qingdao Haiyang Chemical Co. Ltd. The purity of synthesized compound was
confirmed to be higher than 95% by analytical HPLC performed with a dual pump
Shimadzu LC-20AB system equipped with an Ultimate XB-C18 column and eluted
with methanol-water (70:30) containing 0.1% TFA at a flow rate of 0.3 mL/min.
Preparation of N-((2-(4-(N-(4-(2-diethylaminoethoxy))phenyl) amino)-quinazolin-2-yl)
phenyl) - 3-(4-methylpiperazin-1-yl) propanamide (compound 1). 4-methylpiperazine
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(2.0 mL, 18 mmol) was added to a stirred refluxing suspension of the chloride
compounds 12 (0.5 mmol) in MeOH (15 mL). The mixture was stirred under reflux
for 10 h, and then cooled to room temperature. After concentration in vacuum, 30 mL
of water was added to the residue, the precipitate formed was filtered and washed
with water. The crude solid was purified by using chromatography with
CHCl
3
3 2
/MeOH/NH .H O (40:1:0.1) elution to obtain compound 1 as a white solid.
1
Yield: 0.42g (66%). m.p. 173.5-175.1 °C. H NMR (400 MHz, CDCl
3
): δ 13.51 (s,
1H), 8.70 (d, J = 8.3 Hz, 1H), 8.55 (d, J = 8.0 Hz, 1H), 7.94-7.81 (m, 3H), 7.67 (d, J =
8
.9 Hz, 2H), 7.58 (t, J = 7.1 Hz, 1H), 7.52 (s, 1H), 7.45 (t, J = 7.2 Hz, 1H), 7.15 (t, J =
Hz, 1H), 7.02 (d, J = 8.9 Hz, 2H), 4.16 (t, J = 8.0 Hz, 2H), 2.97 (t, J = 6.1 Hz, 2H),
8
ACS Paragon Plus Environment

Page 19 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2.89 (t, J = 8.0Hz, 2H), 2.73 (q, J = 8.1Hz, 4H), 2.63 (t, J = 8.0 Hz, 2H), 2.58 (s, 4H),
13
2
1
1
3
.45 (s, 4H), 2.28 (s, 3H), 1.14 (t, J = 7.1 Hz, 6H). C NMR (101 MHz, CDCl ): δ
70.45, 161.06, 157.15, 156.14, 149.09, 140.02, 133.30, 131.33, 130.88, 128.11,
26.52, 124.02, 123.98, 123.60, 122.76, 120.86, 120.67, 114.93, 113.30, 66.78, 54.95,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
54.18, 52.80, 51.81, 47.87, 45.86, 36.17, 11.68. Purity: 99.8% by HPLC. HRMS (ESI)
+
+
calculated for C34
Preparation of N-(2-carbamoylphenyl)-2-nitrobenzamide (4). A mixture of 2-nitro
benzoic acid (2.0 g, 12.0 mmol) and SOCl (15 mL) was heated at 80 °C for 3 h. After
43 7 2
H N O ([M H] ) m/z: 582.35510, found 582.35434.
2
completion of the reaction, excess of SOCl was distilled off. The obtained yellow
2
residue was added dropwise to the mixture of anthranilamide (2.3 g, 16.9 mmol),
triethylamine (6.3 mL) in chloroform (50 mL) at 0 °C, then stirred at room
temperature for 10 h. The precipitated solid was collected through filtration, washed
with ethanol, dried and recrystallized from ethanol to afford the product 4 as a white
1
solid. Yield: 2.9g (86%). m.p. 210.2-213.5 °C. H NMR (400 MHz, DMSO-d ): δ
6
1
2.57 (s, 1H), 8.53 (d, J = 8.2 Hz, 1H), 8.43 (d, J = 4 Hz, 1H), 8.13 (d, J = 8.3 Hz,
H), 7.95-7.78 (m, 5H), 7.62 (d, J = 11.4, 4.0 Hz, 1H), 7.29-7.22 (m, 1H).
1
Preparation of 2-(2-Nitrophenyl) quinazolin-4(3H)-one (5). A mixture of compound 4
3.0 g, 10.5 mmol) in 10% aqueous KOH (50 mL) and EtOH (40 mL) was heated
(
under reflux for 2 h. Then ethanol was removed in vacuum, and the aqueous layer was
acidified with diluted hydrochloric acid to pH 3 to give a solid residue. The solid was
filtered and washed with water to give white product 5. Yield: 2.6g (93%). m.p.
1
16.8-218.1 °C. H NMR (400 MHz, DMSO-d
2
6
): δ 12.86 (s, 1H), 8.27-8.16 (m, 2H),
7.95-7.81(m, 4H), 7.67(t, J = 7.8 Hz, 1H), 7.62-7.56(m, 1H).
Preparation of 4-Chloro-2-(2-nitrophenyl) quinazoline (6). A mixture of 5 (3.0 g,
1
1.2 mmol), POCl (60 mL), DMF (1 mL) and dioxane (50 mL) was heated at 120 °C
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for 6 h. The reaction was monitored by TLC. After completion of the reaction, excess
3
of POCl was distilled off. The residue was stirred with ice water for 15 min and the
pH was adjusted to 7 using aqueous NaOH. Then the compound was collected by
filtration and washed with water. The crude product was purified by using flash
column chromatography with CH Cl /MeOH (20:1) to obtain a white solid 6. Yield:
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
1
.7g (52%). m.p, 185.1-186.9 °C. H NMR (400 MHz, DMSO-d
1
6
): δ 8.23-8.19 (m,
2
H), 7.96-7.83 (m, 4H), 7.69 (t, J = 4.1 Hz, 1H), 7.62-7.58 (m, 1H).
Preparation of 4-(N-4-methoxyphenyl) amino -2-(2-nitrophenyl) quinazoline (7). Five
drops of concentrated hydrochloric acid were added to a suspension of 6 (2.0 g, 7.0
mmol) and p-methoxyaniline (1.0 g, 8.2 mmol) in 50 mL EtOH. The mixture was
heated under reflux for 5 h, and TLC analysis indicates completion of the reaction.
After cooling to room temperature, the precipitated solid was collected through
filtration and washed by ethanol, the recrystallization from ethanol afforded the
1
yellow solid as 7. Yield: 2.0g (78%). m.p. 165.5-168.1 °C. H NMR (400 MHz,
CDCl ): δ 8.07 (d, J = 8 Hz, 1H), 7.97 (d, J = 8.0 Hz, 1H), 7.87 (d, J = 8Hz, 1H), 7.80
3
(t, J = 8 Hz, 1H), 7.79 (d, J = 8.0 Hz, 1H), 7.63 (t, J = 8 Hz, 1H), 7.57 (t, J = 8 Hz,
1
3
1
H), 7.54-7.49 (m, 3H), 7.44 (s, 1H), 6.94 (d, J = 8 Hz, 2H), 3.83 (s, 3H). C NMR
OD): δ 159.33, 157.56, 156.89, 150.38, 150.07, 134.61, 133.09,
132.04, 131.79, 130.59, 129.61, 129.10, 126.85, 123.96, 123.80, 120.43, 114.28,
13.58, 55.50.
Preparation of 4-(N-4-hydroxyphenyl) amino -2-(2-nitrophenyl) quinazoline (8). BBr
1.5 mL) was slowly added to a suspension of 7 (2 g, 5.4 mmol) in 100 mL CH Cl at
(101 MHz, CD
3
1
3
(
2
2
-78 °C, then stirred at room temperature for 5 h, and TLC analysis indicates
completion of the reaction. The resulting mixture was poured into MeOH (40 mL),
the organic solvent was removed in vacuum and 50 mL of water was added to the
ACS Paragon Plus Environment

Page 21 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
residue. The yellow solid was collected through filtration, washed with water to give
1
. Yield: 1.7g (88%). m.p. 184.8-186.1 °C. H NMR (400 MHz, DMSO-d
8
1
7
6
): δ 9.88 (s,
H), 9.35 (s, 1H), 8.54 (d, J = 8.0 Hz, 1H), 8.01 (d, J = 8 Hz, 1H), 7.93-7.87 (m, 2H),
.80-7.75 (m, 2H), 7.72-7.65 (m, 2H), 7.44 (d, J = 8 Hz, 2H), 6.78 (d, J = 8.3 Hz, 2H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
3
C NMR (101 MHz, CD OD): δ 160.31, 157.00, 150.08, 149.92, 147.11, 136.18,
3
1
1
34.29, 134.16, 132.61, 132.40, 130.07, 129.26, 128.86, 126.53, 125.74, 124.42,
16.41, 114.47.
Preparation of 4-(N-4-(2-bromoethoxy) phenyl) amino-2-(2-nitrophenyl) quinazoline
9). A mixture of compound 8 (1.0 g, 2.8 mmol), Cs CO (1.8 g, 5.5 mmol) and
(
2
3
1,2-dibromo-ethane (2 mL, 23.1 mmol) in acetone (30 mL) was heated under reflux
for 24 h, then cooled to room temperature, the organic solvent was removed in
vacuum and 40 mL of water was added to the residue, the resulting solution was
2 2
extracted with CH Cl for three times, and the organic layer was then dried over
anhydrous MgSO and evaporated, the residue was chromatographed on silica gel
4
eluting with CH Cl /Petroleum ether (1:1) to obtain 9 as yellow solid. Yield: 0.7g
2
2
1
(54%). m.p. 190.2-194.7 °C. H NMR (400 MHz, CDCl
3
): δ 8.09 (d, J = 8 Hz, 1H),
.01 (d, J = 8 Hz, 1H), 7.93 (d, J =8 Hz, 1H), 7.83 (t, J = 8 Hz, 1H), 7.82(d, J = 8 Hz,
H), 7.66 (t, J = 8 Hz, 2H), 7.60 (d, J = 8 Hz, 1H), 7.58-7.54 (m, 2H), 7.53 (s, 1H),
8
1
1
3
6.97 (d, J = 8 Hz, 2H), 4.33 (t, J = 6 Hz, 2H), 3.66 (t, J = 6 Hz, 2H). C NMR (101
MHz, CDCl ): δ 159.23, 157.47, 155.03, 150.29, 149.95, 134.67, 133.04, 132.14,
3
1
6
31.74, 131.37, 129.62, 128.81, 126.82, 123.94, 123.64, 120.74 , 115.06, 113.58,
8.12, 29.30.
Preparation of 4-(N-4-(2-diethylaminoethoxy) phenyl) amino-2-(2-nitrophenyl)
quinazoline (10). A mixture of compound 9 (0.50 g, 1.1 mmol), K CO (0.60 g, 4.3
2
3
mmol) and diethyl amine (0.25 mL, 2.2 mmol) in tetrahydrofuran (50mL) was heated
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
under reflux for 24 h. After cooling to room temperature, the organic solvent was
removed in vacuum, and 25 mL of water was added to the residue, the precipitate
formed was filtered out, and further purified by using flash column chromatography
with CH Cl /Petroleum ether (2:3) elution to give compound 10 as yellow solid. Yield:
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
0.36g (73%). m.p. 186.3-189.1 °C. H NMR (400 MHz, CDCl ): δ 8.15 (s, 1H), 8.06
3
(
d, J = 8.3 Hz, 1H), 7.99 (d, J = 7.6 Hz, 1H), 7.85 (d, J = 8.3 Hz, 1H), 7.75 (d, J = 4
Hz, 1H), 7.70 (t, J = 8.1 Hz, 1H), 7.58 (t, J = 7.6 Hz, 1H), 7.51 (d, J = 8.7 Hz, 2H),
.48-7.41 (m, 2H), 6.82 (d, J = 7.9 Hz, 2H), 4.00 (t, J = 6.1 Hz, 2H), 2.85 (t, J = 6.2
7
1
Hz, 2H), 2.62 (q, J = 7.0 Hz, 4H), 1.06 (t, J = 7.0 Hz, 6H). C NMR (400 MHz,
3
CDCl ): δ 159.36, 157.57, 155.97, 150.49, 150.06, 134.74, 133.03, 132.04, 131.77,
3
1
4
30.79, 129.56, 129.09, 126.78, 123.94, 123.75, 120.58, 114.97, 113.64, 66.22, 54.41,
9.49, 13.95.
Preparation
of
4-(N-4-(2-diethylaminoethoxy)
phenyl)
amino-2-(2-aminophenyl)-quinazoline (11). A mixture of compound 10 (0.50 g, 1.1
mmol) and 80% hydrazine hydrate (4.5 mL) in the presence of 10 % Pd/C (0.10 g)
and isopropanol (20 mL) was heated under reflux for 2 h, and TLC analysis indicated
completion of the reaction. Pd/C was removed by filtration. After concentration in
vacuum, 25 mL of water was added to the residue. The precipitate formed was then
filtered off. The product was used directly in the next reaction without purification.
Preparation of 3-chloro- N-((2-(4- (N-(4-(2-diethylaminoethoxy)) phenyl)
amino)-quinazolin-2-yl) phenyl) propanamide (12). A solution of chloropropionyl
chloride (0.1 mL, 1.04 mmol) in CHCl
mixture of compound 11 (0.20 g, 0.52 mmol) and anhydrous potassium carbonate
0.30 g, 2.2 mmol) in CHCl (35 mL) at 0 °C. The reaction was allowed to warm up to
3
(10 mL) was slowly added to a well-stirred
(
3
room temperature, and then stirred for 24 h. After concentration in vacuum, and
ACS Paragon Plus Environment

Page 23 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
addition of 25 mL of water to the residue, the precipitate formed was filtered off. The
product was used directly in the next reaction without purification.
3
. Microscale thermophoresis (MST) measurements
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The BG4 antibody was labelled with commercially available NT-647 dye using a
Protein Labelling Kit RED (NanoTemper Technologies, Germany). The concentration
of the labelled protein was kept at 100 nM, and the concentration of the in vitro
transcribed mRNA was diluted 1:2 from 8 ꢀM 11 times in 10-mM Tris-HCl buffer
(
100-mM KCl and 0.1-mM ethylenediaminetetraacetic acid [EDTA]). The samples
were loaded into standard treated MST-grade glass capillaries. The thermophoresis
movements of the fluorescently labelled protein in complex with IRES-A mRNA
were measured by monitoring the fluorescence distribution inside the capillary using a
MST system (Nano Temper Technologies, Germany). The light intensities of the
light-emitting diode (LED) and laser were 20% and 40%, respectively. For the
experiments evaluating the binding between RNAs and compound 1, all the
procedures and parameters were the same, except the RNA oligomers were labelled
with FAM, and the concentration was maintained at 0.5 ꢀM. Additionally, compound
1
was diluted 1:2 from 20 ꢀM 14 times in 10-mM Tris-HCl buffer (100-mM KCl and
0.1-mM EDTA). The experimental thermograms were analysed by Monolith NT.115
NanoTemper Technologies, Germany).
(
4
. Surface Plasmon Resonance (SPR) measurements
SPR experiments were performed on a ProteOn XPR36 Protein Interaction Array
system (Bio-Rad Laboratories, Hercules, CA) with a GLH chip. In a typical
experiment, biotin-labelled IRES-A (WT), IRES-A (MU1), and hairpin DNA were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
immobilized in flow cells 1, 2 and 4, leaving flow cell 3 as a blank. Compounds were
prepared by 1:2 dilutions in running buffer (50-mM Tris, 150-mM KCl, 0.005%
Tween-20, pH 7.4) and injected at a flow rate of 50 ꢀL/min for a 300-s association
phase, followed by a 200-s disassociation phase at 25 °C. Data were analysed with
ProteOn manager software using the Langmuir model for fitting kinetic data.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
. Pull-down assay
ꢀM of in vitro transcribed IRES-A mRNAs or synthesized oligomers in 10 mM
5
Tris-HCl buffer (pH 7.4) containing 100 mM KCl, were first heated at 95 °C for 5
min and slowly cooled to 25 °C. 20 ꢀL of magnetic beads labelled with anti-FLAG
label (SIGMA, USA) were incubated for 1.5 h with 200 nM Flag-labelled BG4
antibody or 200 nM BSA protein (as negative control) in a 50 ꢀL TBS buffer
(containing 50 mM Tris-HCl and 150 mM NaCl, pH 7.4) at 37 °C. 10 ꢀL of 5 ꢀM
annealed RNAs or RNA-compound mixture were added to the bead-protein mix, the
beads were captured with a magnet followed by removal of unbound oligomers. Then
the bound RNAs were eluted by the elution buffer containing 1% SDS and 0.1 M
NaHCO . The eluted RNAs were then separated on a 1.5% agarose gel or a 16%
3

native polyacrylamide gel, stained with the GelRed (Biotium, China) and visualized

on a Tanon 4500sf Imager System (Tanon, China).
6
. RNA immunoprecipitation assay (RIP)
The MCF-7 cells were treated with 2.5-ꢀM or 5-ꢀM compound 1 or normal
Dulbecco’s Modified Eagle Medium (DMEM) for 24 h at 37 °C with CO . At the end
2
of the incubation, 18.5% formaldehyde was added to each dish to obtain a final
concentration of 1% formaldehyde, and the treated cells were incubated at 25 °C for
ACS Paragon Plus Environment

Page 25 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10 min, followed by the addition of glycine solution. The cells were detached by
scraping after adding 1 mL of ice-cold phosphate-buffered saline (PBS) containing 10
ꢀ
L of the halt cocktail (Thermo Scientific, USA). Then, 100 ꢀL of RIP Lysis buffer
(25-mM Tris-HCl, 150-mM KCl, 2-mM EDTA, 1-mM NaF, 1-mM dithiothreitol
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
[DTT], 0.5% NP-40, pH 7.5) containing the halt cocktail and 100-U/mL RNasin
Ribonuclease Inhibitor (Takara, China) were added to the cell pellet and incubated on
ice for 10 min. Next, 100 ꢀL of the supernatant was transferred to 400 ꢀL of
immunoprecipitation (IP) dilution buffer (Thermo Scientific, USA) containing
protease inhibitors and the RNasin ribonuclease inhibitor. For each IP, 500 ꢀL of
diluted lysate was added to protein A/G plus agarose beads with 5 ꢀg of BG4
antibody and incubated at 4 °C for 12 h. After centrifugation at 2,000 g for 30 s, the
beads were washed with low-salt washing buffer (50-mM Tris-HCl, 150-mM NaCl,
1
2
-mM MgCl , 2-mM EDTA, 1-mM DTT, 100-U/mL RNase inhibitor, pH 7.4) twice
and high-salt washing buffer (50-mM Tris-HCl, 300-mM NaCl, 1-mM MgCl , 2-mM
2
EDTA, 1-mM DTT, 100-U/mL RNase inhibitor, pH 7.4) twice. Then, 150 ꢀL of IP
elution buffer (Thermo, USA) with 6 ꢀL of 5-M NaCl and 2 ꢀL of 20-mg/mL
proteinase K were added and incubated at 65 °C for 30 min. The obtained total RNA
samples were used for the dot blotting assay, RT-PCR, and qRT-PCR.
7
. Dot blotting assay
Dot blotting assays were performed as follows: The nylon membrane was washed
with 2 × saline sodium citrate (SSC) buffer for 10 min at 4 °C prior to use. RNA
samples obtained from the RIP assay were diluted to 100 pmol in 100 ꢀL using 2 ×
SSC buffer, and then, the samples were applied to the membrane under vacuum. The
2
crosslinking was conducted using 265-nm ultraviolet (UV) irradiation (1,000 ꢀJ/cm )
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
for 5 min. Then, the membrane was placed into a denaturing buffer (1.5 M NaCl and
0
.5 M NaOH) for 20 min, followed by a neutralizing buffer (1 M NaCl and 0.5 M
Tris-HCl, pH 7.0) for 20 min. A pre-hybridization step was subsequently performed
to prevent unspecific binding, in which the membrane was submerged in a buffer
containing 10 × Denhardt’s solution, 5 × SSC buffer, and 0.5% sodium dodecyl
sulfate (SDS) at 42 °C for 2 h. Hybridization was conducted with 100 pmol
biotin-labelled reverse-complement DNA (Biotin-rc DNA) to the IRES-A oligomer at
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
7 °C for 12 h. Biotin-labelled DNA was detected using the Chemiluminescent
Nucleic Acid Detection Module Kit (Thermo Scientific, USA) according to the
manufacturer’s protocol.
8
. Dual-luciferase reporter assay
For this assay, 200 ng of constructed pMetLuciferase plasmid (Clontech, USA)
containing WT IRES-A or mutant IRES-A and 200 ng of pRL-TK (Promega, USA)
was co-transfected into MCF-7 cells using Lipofectamine 2000 (Invitrogen, USA).
After 4 h, compounds 1, 2, and 3 were added to the cells at concentrations ranging
2
from 0.1875 to 1.5 ꢀM. The cells were incubated at 37 °C with CO for another 24 h,
and the transfected cells were first washed with ice-cold PBS to reduce the
background signals from the medium. Luciferase assays were subsequently performed
according to the manufacturer’s instructions using the Ready-To-Glow Secreted
Luciferase Reporter System (Clontech, USA) and the Renilla Luciferase Assay
System (Promega, USA). After a 3-s delay, secreted luciferase signals were collected
for 10 s using a microplate reader (Molecular Devices, Flex Station 3, USA), and
Renilla signals were collected for 10 s as an internal control. The quantification was
performed using a multimode reader (Molecular Devices). The secreted Metluciferase
ACS Paragon Plus Environment

Page 27 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
activity was normalized to the Renilla luciferase activity.
9
. Cell scrape test
Approximately 2 × 10 MCF-7 cells were seeded and grown to 80% confluence in
5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6-well plates, and a scrape was made through the confluent monolayers with a plastic
pipette tip. Then, all the cells were treated with 100-ꢀM hypoxia mimetic cobalt
2
chloride (CoCl ) (Invitrogen, USA) to create a chemically hypoxic environment for
VEGF-A protein cap-independent expression; the blank control cells were not treated.
The cells were divided into control, compound-treated, and compound-treated
followed by VEGF-A protein-treated groups. In the control group, the cells were all
treated with 0.03% DMSO and transfected with VEGF-A-targeted siRNAs in
Opti-MEM reduced serum medium at 37 °C for 48 h or treated with 20-ng/mL
VEGF-A protein (Sigma-Aldrich, USA) in Opti-MEM reduced serum medium for 12
h. In the compound-treated group, the cells were treated with increasing
concentrations of compound 1 from 0.375 ꢀM to 3 ꢀM in medium for 48 h. In the
compound-treated followed by the VEGF-A protein-treated group, the cells were
treated with compound 1 as mentioned above and then with 20-ng/mL VEGF-A
protein for another 12 h in Opti-MEM reduced serum medium. All the cells were
photographed via phase-contrast microscopy (QIMAGING, Canada) and quantified
with Image J software.
1
0. Transwell assay
4
MDA-MB-231 cells (5 × 10 cells, 100 ꢀL) were plated in the top chambers of
0
.8-ꢀm-pore trans-wells (Millipore, USA) in Opti-MEM reduced serum medium in
the presence or absence of compound 1. Meanwhile, 600 ꢀL of DMEM containing 10%
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 51
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
foetal bovine serum (FBS) and 100-ꢀM CoCl
2
were added to the lower chambers. The
cells were allowed to migrate for 24 h. At the end of the assay, the cells in the top
chamber were removed, and the cells at the bottom of the filter were treated by adding
5
00
3-(4,5)-dimethylthiazol(-z-y1)-3,5-diphenyl-tetrazoliumbromide (MTT) to each well.
After incubating at 37 °C with 5% CO for 4 h, 500 ꢀL of DMSO was added to each
ꢀL
of
DMEM
containing
2.5-mg/mL
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
well, and the plate was gently rotated for 10 min. Absorbance (570 nm) was measured
using a microplate reader (BioTek, USA).
11. xCELLigence real-time cell analysis (RTCA)
Cell migration experiments were performed using CIM-16 16-well plates (Roche
Diagnostics GmbH, Mannheim, Germany), in which each well consisted of a top
chamber and a bottom chamber separated by a microporous membrane containing
randomly distributed 8-µm pores. Prior to each experiment, cells were deprived of
FBS and incubated for 24 h. Initially, 165 µL (various concentrations of compound 1
in medium with 100-ꢀM CoCl
and upper chambers, and the CIM-16 plate was locked in the RTCA DP device
Roche Diagnostics GmbH, Mannheim, Germany) at 37 °C with 5% CO for 60 min
2
) and 30 µL of medium were added to both the lower
(
2
to achieve equilibrium according to the manufacturer's guidelines. The lower
chambers contained 1 × DMEM (Gibco, USA) with or without FBS and were used to
assess cell migration when cells were exposed to FBS. Background migration to
serum-free (OptiMEM medium) medium was used as a negative control. After the
incubation period, a measurement step was performed to determine the background
signal generated by cell-free media. To initiate an experiment, cells were detached
using 0.25% trypsin and 0.02% EDTA and re-suspended in OptiMEM medium. The
ACS Paragon Plus Environment

Page 29 of 51
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
cells were then counted and seeded in the upper chamber at a concentration of 5 × 10
cells in 100 µL without or with various concentrations of compound 1 that matched
the corresponding lower chambers. After adding the cells, the CIM-16 plates were
incubated for 30 min at 25 °C in a laminar flow hood to allow the cells to settle onto
the membrane according to the manufacturer's guidelines. Each condition was
repeated four times with a programmed signal detection schedule with 15-min
intervals for 24 h.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2. Xenograft animal model and drug treatments
BALB/c female nude mice were obtained from the laboratory animal centre of Sun
Yat-sen University. MCF-7 cells were harvested, pelleted by centrifugation at 800 g
for 5 min, and re-suspended in sterile serum-free medium without EDTA. The cells
7
1×10 in 100 µl) were then subcutaneously implanted into the underarm regions of
(
3
two mice. After the tumours grew to almost 1000 mm , the tumour tissues were
removed and divided. Then, the divided tissues were implanted into the underarm
regions of 15 mice. These mice were separated into three groups: negative control,
compound 1-treated, and positive control (doxorubicin-treated). Compound 1,
doxorubicin, and saline were administered by i.p. injection to athymic nude mice with
human tumour xenografts established using MCF-7 breast cancer cells. Mice were
injected i.p. once a day for 20 days. Negative controls were injected with 150 µL of
saline. The positive control group received doxorubicin by i.p. injection at a dose of 1
mg/Kg. Compound 1 was similarly administered to mice at a dose of 7.5 mg/Kg.
After treating the animals for 20 days, the tumour tissues were collected, and IHC
assays were conducted using an anti-VEGF-A antibody according to a general
procedure.
ACS Paragon Plus Environment