Discovery of pyrrole derivatives for the treatment of glioblastoma and chronic myeloid leukemia

Article abstract of DOI:10.1016/j.ejmech.2021.113532

Long-term survivors of glioblastoma multiforme (GBM) are at high risk of developing second primary neoplasms, including leukemia. For these patients, the use of classic tyrosine kinase inhibitors (TKIs), such as imatinib mesylate, is strongly discouraged,

Full text of DOI:10.1016/j.ejmech.2021.113532

Journal Pre-proof
Discovery of Pyrrole Derivatives for the Treatment of Glioblastoma and Chronic
Myeloid Leukemia
Michela Puxeddu, Hongliang Shen, Ruoli Bai, Antonio Coluccia, Marianna Bufano,
Marianna Nalli, Jessica Sebastiani, Diego Brancaccio, Eleonora Da Pozzo, Chiara
Tremolanti, Claudia Martini, Viviana Orlando, Stefano Biagioni, Maria Stefania
Sinicropi, Jessica Ceramella, Domenico Iacopetta, Addolorata Maria Luce Coluccia,
Ernest Hamel, Te Liu, Romano Silvestri, Giuseppe La Regina
PII:
S0223-5234(21)00381-0
DOI:
https://doi.org/10.1016/j.ejmech.2021.113532
EJMECH 113532
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 6 January 2021
Revised Date: 16 April 2021
Accepted Date: 23 April 2021
Please cite this article as: M. Puxeddu, H. Shen, R. Bai, A. Coluccia, M. Bufano, M. Nalli, J. Sebastiani,
D. Brancaccio, E. Da Pozzo, C. Tremolanti, C. Martini, V. Orlando, S. Biagioni, M.S. Sinicropi, J.
Ceramella, D. Iacopetta, A.M.L. Coluccia, E. Hamel, T. Liu, R. Silvestri, G. La Regina, Discovery of
Pyrrole Derivatives for the Treatment of Glioblastoma and Chronic Myeloid Leukemia, European Journal
of Medicinal Chemistry, https://doi.org/10.1016/j.ejmech.2021.113532.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2021 Elsevier Masson SAS. All rights reserved.

Discovery of Pyrrole Derivatives for the Treatment of Glioblastoma and
Chronic Myeloid Leukemia
Graphical Abstract

Highlighted version with changes shown in “red for addition” and “cyan for deletion”,
Discovery of Pyrrole Derivatives for the Treatment of Glioblastoma and
Chronic Myeloid Leukemia
a,#
b,#
c
a
a
Michela Puxeddu
Hongliang Shen,
Ruoli Bai, Antonio Coluccia, Marianna Bufano,
a
a
d
de
Marianna Nalli, Jessica Sebastiani, Diego Brancaccio, Eleonora Da Pozzo,
Chiara
de de ef ef
Tremolanti,
fg
Claudia Martini,
fg
Viviana Orlando, Stefano Biagioni, Maria Stefania
fg gh
Sinicropi, Jessica Ceramella, Domenico Iacopetta, Addolorata Maria Luce Coluccia,
hi,* a,* a,*
c
Ernest Hamel, Te Liu,
Romano Silvestri, and Giuseppe La Regina,
a
Laboratory affiliated with the Institute Pasteur Italy – Cenci Bolognetti Foundation,
Department of Drug Chemistry and Technologies, Sapienza University of Rome, Piazzale Aldo
Moro 5, I-00185 Roma, Italy
b
Department of Urology, Capital Medical University Beijing Friendship Hospital, Beijing,
China, 100050
c
Molecular Pharmacology Branch, Developmental Therapeutics Program, Division of Cancer
Treatment and Diagnosis, Frederick National Laboratory for Cancer Research, National Cancer
Institute, National Institutes of Health, Frederick, Maryland 21702, United States
d
Department of Pharmacy, University of Naples“Federico II”, Via Domenico Montesano 49,
8
0131, Naples, Italy
de
ef
Department of Pharmacy, University of Pisa, Via Bonanno Pisano 6, I-56126 Pisa, Italy
Department of Biology and Biotechnologies “Charles Darwin”, Sapienza University of Rome,
Piazzale Aldo Moro 5, I-00185 Roma, Italy
fg
Department of Pharmacy, Health and Nutritional Sciences, University of Calabria, I-87036
Rende, Cosenza, Italy
gh
Department of Biological and Environmental Sciences and Technologies, University of
Salento, I-73100 Lecce, Italy

hi
Shanghai Geriatric Institute of Chinese Medicine, Shanghai University of Traditional Chinese
Medicine, 365 South Xiangyang Road, Shanghai, China, 200031

Keywords:
Tubulin
Glioblastoma
Leukemia
Pyrrole
Synthesis
#
First co-authors
M.P. e-mail address: michela.puxeddu@uniroma1.it; phone: +39 06 4991 3404
H. S. e-mail address: 1051486940@qq.com; phone: +86 1391 1859 490
*
Corresponding Authors
G.L.R. e-mail address: giuseppe.laregina@uniroma1.it; phone: +39 06 4991 3893
R.S. e-mail address: romano.silvestri@uniroma1.it; phone: +39 06 4991 3800
T.L. e-mail address: 0721160004@mail.tongji.edu.cn; phone: +86 21 64720010

Abbreviations
GBM
CML
IM
glioblastoma multiforme
chronic myeloid leukemia
imatinib mesylate
TKI
tyrosine kinase inhibitor
long-term survivor
aroyl-diaryl-pyrrole
blood-brain barrier
human topoisomerase I
human topoisomerase II
combretastatin A4
ponatinib
LTS
ARDAP
BBB
hTopoI
hTopoII
CSA4
PN
DMSO
HE
dimethyl sulfoxide
hematoxylin and eosin
protein data bank
PDB
kDNA
XTT
kinetoplast DNA
sodium
3'-[1-[(phenylamino)-carbonyl]-3,4-tetrazolium]-bisbis(4-methoxy-6-
nitro)-benzenesulfonic acid hydrate
Tris-EDTA
TE
TAE
EB
Tris-acetate-EDTA
ethidium bromide
fetal bovine serum
phosphate buffered saline
FBS
PBS

ABSTRACT
Long-term survivors of glioblastoma multiforme (GBM) are at high risk of developing second
primary neoplasms, including leukemia. For these patients, the use of classic tyrosine kinase
inhibitors (TKIs), such as imatinib mesylate, is strongly discouraged, since this treatment causes
a tremendous increase of tumor and stem cell migration and invasion. We aimed to develop
agents useful for the treatment of patients with GBM and chronic myeloid leukemia (CML)
using an alternative mechanism of action from the TKIs, specifically based on the inhibition of
tubulin polymerization. Compounds 7 and 25, as planned, not only inhibited tubulin
polymerization, but also inhibited the proliferation of both GMB and CML cells, including those
expressing the T315I mutation, at nanomolar concentrations. In in vivo experiments in
BALB/cnu/nu mice injected subcutaneously with U87MG cells, in vivo, 7 significantly inhibited
GBM cancer cell proliferation, in vivo tumorigenesis, and tumor growth, tumorigenesis and
angiogenesis. Compound 7 was found to block human topoisomerase II (hTopoII) selectively
and completely, at a concentration of 100 μM.

1
. Introduction
Glioblastoma multiforme (GBM), also called glioblastoma, is one of the most aggressive
malignant brain tumors in humans. GBM is defined as a grade IV glioma by the World Health
Organization because it rapidly invades tissues, frequently relapses due to the presence of stem-
like cells, and shows very poor prognosis and a short life expectancy of 14 months from
diagnosis [1]. GBM incidence is growing: for instance, the incidence rate for glioblastoma more
than doubled in England between 1995 and 2015 [2]. GBM treatment includes surgery to relieve
pressure in the brain, along with adjuvant chemotherapy and postoperative radiation therapy [3].
The development of effective therapies for GBM has been slowed by the complex nature of this
tumor. The three FDA-approved chemotherapeutics include the mustard temozolomide as a first
line drug, which was demonstrated to significantly increase significantly overall survival when
combined with radiotherapy. The anti-VEGF monoclonal antibody bevacizumab, as a second is a
first line drug, was approved by the FDA for the treatment of recurrent glioblastoma.
Bevacizumab can prolong survival complementary to temozolomide and attenuates recurrences
associated with conventional risk factors [4]. The nitrosourea carmustine is in use for the
treatment of both recurrent and newly diagnosed glioblastoma, but it causes severe bone marrow,
liver and kidney toxicity. Studies devoted to developing more effective chemotherapeutic
treatments of glioblastoma are required and expected to evolve in coming years [5,6].
Chronic myeloid leukemia (CML) is a type of cancer that starts in the blood-forming cells of
the bone marrow and invades the blood and other organs. CML accounts for nearly 15% of
newly diagnosed leukemia cases in adults, with an incidence of 1-2 cases per 100,000 adults.
CML is estimated to have occurred in about 8,450 patients in the United States in 2020, with
about 1080 deaths [7]. CML arises from the translocation t(9; 22)(q34; q11) of the Abelson

leukemia Abl1 gene on chromosome 9 with the breakpoint cluster region Bcr gene on
chromosome 22. This forms the Bcr-Abl1 fusion oncogene, with formation of the Philadelphia
chromosome. This oncogene encodes for the expression of an oncoprotein termed Bcr-Abl [8].
The Bcr-Abl is a tyrosine kinase that promotes growth and replication through downstream
signaling pathways such as RAS, RAF, JUNK, MYC and STAT [9]. Imatinib mesylate (IM) has
tremedously greatly improved the outcome for patients with CML. The majority of IM-treated
patients have stable clinical responses with prolonged survival times. IM regimes, however, lead
to drug resistance either because of the occurrence of mutations of the Bcr-Abl oncoprotein [10]
or because of bcr/abl gene amplification [11]. The T315I mutation at the “gatekeeper” position
8
3
15 of the Bcr-Abl gene accounts for drug resistance to IM. Second-generation inhibitors, such
as dasatinib, nilotinib, and bosutinib, are not effective inhibitors of the Bcr-Abl T315I mutation
12]. A third-generation inhibitor, ponatinib (PN), was developed in order to overcome drug
[
resistance and to improve tyrosine kinase inhibitory activity. PN was approved for the second-
+
line treatment of CML and Ph ALL in 2012 in the USA and in 2013 in the European Union
[
13].
Long-term survivors (LTSs) of GBM, defined as surviving >5 years post-diagnosis, represent
% to 10% of all GBM patients. LTSs of cancer are at high risk of developing second primary
2
neoplasms, including leukemia [14]. Second malignancies were also observed in 4% of CML
patients with a median follow-up of 69 months. The relative incidence of overall second
malignancies in CML is slightly higher than that of the general population [15]. Despite its
impressive clinical results in CML, IM has failed clinical trials in GMB, where it did not
significantly inhibit tumor growth, nor did it extend survival time [16,17]. In 2016, it was
reported that IM and nilotinib treatment of GBM and patient-derived GBM stem cells causes an

increase of phosphorylation of Tyr of several key proteins involved in cell motility, including
p130Cas, FAK, and PXN; such an effect produces a tremendous increase of tumor and stem cell
migration and invasion [18].
GBM treatment has proved elusive, despite decades of research. The complexity of this
disease has hampered these efforts, although a number of new therapeutic approaches have been
tried. In an effort to try bridging this gap, we planned to synthesize and evaluate as inhibitors of
both GBM and CML cells a series of new compounds whose mechanism is based on the
inhibition of tubulin polymerization and without activity as kinase inhibitors. GBM, as other
astrocytic brain tumors, are characterized by dynamic ultra-long microtubes which serve as
routes for brain invasion and proliferation and as a network for long distance interconnections,
enabling these cells to act as a single, organism-like unit. Tumor microtubes are enriched in actin
and microtubules. Interferring with the actin or microtubule cytoskeleton appears to impair
microtube formation and/or maintenance [19,20].
We previously reported aroyl-diaryl-pyrrole (ARDAP) derivatives characterized by the
presence of one or two aminophenyl groups at positions 1 and 4 [21]. Metabolism is a major
cause of low bioavailability because of increased clearance and reduced exposure to
pharmacologically active concentrations. First-pass metabolism, which can occur in the gut and
in the liver, can sharply reduce the concentration of drug before it reaches the systemic blood
circulation. Amino groups could rapidly undergo phase II metabolic glucuronidation or
acetylation reactions [22]. We synthesized carboxamido derivatives of the 4-aminophenyl group,
but these derivatives were found to be less effective than the parent compounds in
antiproliferative assays (data not shown). In the studies described here, we replaced the 4-
aminophenyl group with five- and six-membered bioisosteric heterocycles, while retaining the

important methyl and fluoro substituents on the 1-phenyl ring (Chart 1). Moreover, we increased
lipophilicity, since for GBM treatment this strategy often helps therapeutic drugs, including
small molecules, in reaching the tumor target [23,24]. Compounds 7 and 25 inhibited the growth
of both GMB and CML cells, including those expressing the T315I mutation, at nanomolar
concentrations. In experiments in BALB/cnu/nu mice injected subcutaneously with U87MG
cells, 7 administered intraperitoneally at 25 mg/kg significantly inhibited GBM cancer cell
proliferation, in vivo tumorigenesis, and tumor growth and angiogenesis.
Microtubule inhibitors and human topoisomerase II (hTopoII) inhibitors are frequently used
in synergistic combination in anticancer therapy, but few single agents have been reported to
inhibit both microtubules and hTopoII [25]. Compound YCH337 was reported to inhibit both
microtubule polymerization by binding to the colchicine site and hTopoII, eliciting anticancer
activity and overcoming drug resistance [26]. Since YCH337 has structural similarity with the
ARDAP series, we aimed to evaluate the efficacy of compound 7 as an inhibitor of
topoisomerases. In addition to inhibition of tubulin assembly, the primary target, compound 7
proved to block selectively and completely hTopoII at a concentration of 100 μM. These features
led us to determine that, after accumulation in cells, compound 7 can behave as an effective dual-
target antitumor agent.
Compound 7 showed strong inhibition of cancer cell growth, and, compared to the previously
reported lead 1, had a calculated improvement of metabolic stability to phase II metabolism and
higher lipophilicity.

H N
2
OMe
OMe
OMe
OMe
OMe
OMe
O
O
R2
N
N
R1
NH2
1
2-25
H N
2
OMe
OMe
OMe
OMe
OMe
OMe
OMe
O
O
O
R2
R2
OMe
OMe
N
N
N
R1
-16
R1
NH2
2
17-25
1
Chart 1. General structure of 2-25 (see Table 1) and reference compound 1. R : 3-Me-, 4-Me- or 4-F-Ph (2-
1
2
5); R : five (2-16) or six (17-25) member heterocycle.
2
2
. Chemistry
Compounds 2-25 were prepared by reacting pyrrole derivatives 35-42 with the appropriate
boronic acid in the presence of copper(II) acetate and triethylamine in dichloromethane at 40 °C
for 18 h. Pyrrole derivatives 35-42 were synthesized by reacting chalcones 27-34 with 4-
toluenesulfonylmethyl isocyanide in the presence of sodium hydride in diethyl ether/dimethyl
sulfoxide (DMSO) at 25 °C for 2 h. Reaction of 3′,4′,5′-trimethoxyacetophenone (26) with the
appropriate carboxaldehyde in the presence of sodium hydroxide or lithium hydroxide in ethanol
at 25 °C for 20 min gave chalcones 27-34 (Scheme 1).

OMe
OMe
O
O
CHO
a
OMe
OMe
R2
+
X
Me
OMe
X = S,O
OMe
OMe
2
7-29
26
O
N
OMe
OMe
b
CHO
R2
or
2
6
+
CHO
X = S,O
X
N
30-34
OMe
OMe
O
R2
OH
B
OH
c
R1
27-34
+
OMe
N
H
35-42
OMe
O
R2
OMe
OMe
d
N
R1
2-25
Scheme 1. Synthesis of 2-25 and 27-42. For structures 2-25: see Table 1; 27, 35: R = thiophen-3-yl; 28, 36:
2
R = furan-2-yl; 29, 37: R = furan-3-yl; 30, 38: R = thiazol-2-yl; 31, 39: R = oxazol-2-yl; 32, 40: R =
2
2
2
2
2
pyridin-2-yl; 33, 41: R = pyridin-3-yl; 34, 42: R = pyridin-4-yl. Reagents and reaction conditions: (a) 27-29:
2
2
appropriate carboxaldehyde, sodium hydroxide, ethanol, 25 °C, 20 min; (b) 30-34: (i) lithium hydroxide
monohydrate, ethanol, 25 °C, 10 min; (ii) appropriate carboxaldehyde, 25 °C, 10 min; (c) 35-42: p-
toluenesulfonylmethyl isocyanide, sodium hydride, anhydrous diethyl ether/DMSO, 25 °C, 2 h, argon stream;
(d) 2-25: appropriate boronic acid, copper(II) acetate, triethylamine, dichloromethane, 40 °C, 18 h.
3
. Results and discussion
3
.1. Compound selection
As a first step, a series of ARDAP derivatives bearing different substituents at positions 1 and
of the pyrrole nucleus were evaluated by the XLOGP3 method to predict their LogP values
4
(Table 1SD, Supplementary data). We decided to exclude from these studies compounds having
predicted LogP values lower than that of 1 (except 16, upon completion of SAR studies).
Moreover, compounds bearing an amino group were not examined further because of their being
potentially being substrates of phase II metabolism. Hence, we synthesized compounds 2-25 in
search of effective inhibitors of tubulin polymerization to be evaluated as GBM and CML cell

growth inhibitors (Table 1).
3
.2. Inhibition of tubulin polymerization and colchicine binding
Compounds 2-25 inhibited tubulin polymerization with IC50 values at submicromolar
concentrations, as compared with combretastatin A-4 (CSA4, IC50 = 0.54-0.73 μM range),
except for derivative 16, which had an IC of 1.4 µM. Introduction of the 3- and 4-methyl- or 4-
5
0
fluorophenyl groups at position 1 of the pyrrole provided strong inhibition of tubulin
polymerization. Compounds 7 and 14 were strong inhibitors of tubulin polymerization, with IC50
values of 0.39 and 0.38 M, respectively. Pyridinyl derivatives 17-25 with the 4-(pyridin-4-yl)
group tended to be stronger inhibitors of tubulin polymerization as compared to the isomeric
pyridin-2-yl and pyridin-3-yl compounds. In general, independent of the nature of the
heterocyclic ring at position 4 of the pyrrole, 3-methyl (2-6 and 17-19), 4-methyl (7-11 and 20-
2
2) and 4-fluoro (12-16 and 23-25) substituents on the 1-phenyl ring provided robust inhibition
of tubulin polymerization.
Table 1.
Inhibition of tubulin polymerization, growth of MCF-7 human breast carcinoma cells and colchicine binding to
a
tubulin by compounds 2-25 and reference compound 1.
OMe
OMe
OMe
OMe
OMe
OMe
O
O
R2
R2
N
N
R1
R1
2
-16
17-25
Compd
R1
R2
IC ± SD (M)
50
IC ± SD (nM)
50
(% ± SD)
b
c
d,e
Tubulin
MCF-7
Inhib. colch. bind
2
S
0.52 ± 0.03
65 ± 10
74 ± 0.2
Me

3
4
5
6
7
8
9
0.56 ± 0.06
0.42 ± 0.03
0.49 ± 0.08
0.72 ± 0.1
23 ± 6
21 ± 2
20 ± 1
75 ± 0
80 ± 1
87 ± 1
O
Me
Me
Me
Me
O
N
N
83 ± 0.8
77 ± 3
S
O
S
0.39 ± 0.03
0.46 ± 0.04
0.47 ± 0.03
0.56 ± 0.09
0.72 ± 0.05
0.55 ± 0.04
0.47 ± 0.01
0.38 ± 0.06
0.78 ± 0.06
4.0 ± 1
22 ± 8
18 ± 3
14 ± 2
42 ± 0.8
15 ± 4
17 ± 2
15 ± 0.6
22 ± 4
90 ± 0.9
92 ± 0.5
95 ± 0.6
91 ± 4
Me
Me
Me
Me
Me
F
O
O
N
N
1
0
1
2
3
4
5
S
1
1
1
1
1
90 ± 1
O
S
76 ± 0.5
82 ± 0.007
87 ± 2
O
F
O
F
N
69 ± 4
S
F

N
1
1
1
1
2
2
2
2
2
2
6
7
8
9
0
1
2
3
4
1.4 ± 0.2
93 ± 20
140 ± 20
52 ± 9
65 ± 10
36 ± 4
19 ± 4
9.0 ± 1
73 ± 4
39 ± 1
9.0 ± 1
14 ± 2
68 ± 2
66 ± 0.8
78 ± 0.9
88 ± 3
O
F
0.78 ± 0.07
0.74 ± 0.04
0.59 ± 0.04
0.46 ± 0.05
0.64 ± 0.08
0.44 ± 0.04
0.82 ± 0.04
0.56 ± 0.08
0.58 ± 0.1
0.47 ± 0.03
N
Me
Me
Me
N
N
83 ± 4
N
Me
Me
Me
F
92 ± 1
N
N
94 ± 0.8
44 ± 3
N
80 ± 3
N
N
F
5
87 ± 1
F
H N
2
f
1
92 ± 0.07
NH2
a
Experiments were performed in duplicate or triplicate.
Inhibition of tubulin polymerization. Tubulin was at 10 μM in the assembly assay. CSA4 as a reference
b
compound yielded IC ’s in the 0.54-0.73 M range.
5
0
c
d
e
Inhibition of growth of MCF-7 human breast carcinoma cells.
Inhibition of [ H]colchicine binding: tubulin, [ H]colchicine, inhibitor at 1:5:5 M.
Inhibition of [ H]colchicine binding: tubulin, [ H]colchicine, inhibitor at 1:5:1 M: 4, 64 ± 0.8%; 7, 71 ± 2%;
, 71 ± 3%; 9, 71 ± 3%; 10, 63 ± 2%; 11, 45 ± 0.5%; 14, 71 ± 0.4%; 21, 58 ± 0.9%; 22, 70 ± 2%. Values for
3
3
3
3
8
CSA4 were 99 ± 0.3 % at 5 and 90 ± 3 % at 1 
f
Lit.[20].

3
All compounds were strong inhibitors of the binding of [ H]colchicine to tubulin, but none
were as active as CSA4. The best inhibition, with tubulin and inhibitor at 1.0 µM and the
colchicine at 5.0 µM, was observed with compounds 7, 8, 9, 14 and 22 (see Table 1, note e).
3
.3. MCF-7 cancer cell growth
Most compounds inhibited MCF-7 cancer cell growth at nanomolar concentrations, and, in
general, the IC values showed good correlation with inhibition of tubulin polymerization. Three
5
0
compounds, 7, 22 and 25, were the most potent MCF-7 cell growth inhibitors within the series,
with IC values of 4.0, 9.0 and 9.0 nM, respectively. These compounds were characterized by
5
0
the presence of the thiophen-3-yl (7) or pyridin-4-yl (22, 25) group at position 4 of the pyrrole.
Compounds 5, 9, 10, 12-14 and 21 inhibited MCF-7 cell growth with IC values in the 10−20
5
0
nM range, and 3, 4, 8, 15, 20 and 24 had IC values in the 21−50 nM range. ARDAPs 3-6
5
0
bearing a 1-(3-methylphenyl) group were almost equipotent in this assay, with IC ’s close to 20
5
0
nM. ARDAPs 8-10 and 20-22, characterized by the presence of the 1-(4-methylphenyl) group,
showed strong inhibition of the MCF-7 cell growth independent of the character of the
heterocyclic ring. 1-(4-Fluorophenyl) derivatives 12-14, bearing a 4-thiophenyl or 4-furanyl
group (IC = 15-17 nM), and 25 with a pyridin-4-yl ring (IC = 9.0 nM) were also strong
5
0
50
inhibitors of MCF-7 cell growth.
Table 2
Growth Inhibition of U343MG and U87MG Cells by Compounds 7 and 25.

a
a
Compd
U343MG
U87MG
b
c
b
c
IC ± SD (nM)
Emax (%)
48.5
IC ± SD (nM)
Emax (%)
55.6
5
0
50
7
18.6 ± 2.6
432 ± 61
19.4 ± 2.0
39.1 ± 1.6
2
5
75.1
56.0
a
b
c
U343MG cells, incubation time was 72 h; U87MG cells, incubation time was 48 h.
Cytotoxic concentrations for the indicated cell lines. Experiments were performed in triplicate.
Maximal efficacy of compound in inhibiting cell viability.
3
.4. Glioblastoma U87MG and U343MG cell lines
We selected two human glioblastoma cell lines to evaluate the potential of two of the most
potent compounds, 7 and 25, as therapeutic agents for this disease. The cell lines were U343MG,
derived from an adult male, and U87MG, derived from a child. The two compounds yielded a
dose-dependent inhibition of cell viability. The IC values obtained were in the nanomolar range
5
0
for each cell line, as shown in Table 2 (see also Fig. 1-4 SD, Supplementary data). Compounds 7
and 25 inhibited the U87MG cells with IC values of 19.4 nM (7) and 39.1 nM (25). Inhibition
5
0
of the U343MG cells displayed a wider IC50 range, of 18.6 nM (7) and 432 nM (25). The
stronger inhibitor of the U343MG cells was characterized by the presence of the thiophen-3-yl
ring at position 4 of the pyrrole (compare 7 and 25), but this substituent was not important for
inhibition of U87MG cell growth. The weak inhibition of the U343MG cells by 25 may be due to
an interaction of the pyridine nitrogen with different cell proteins [27].
3
.5. Activity of compound 7 in a murine glioblastoma model
Tumors were observed on the backs of mice injected subcutaneously with U87MG cells in the
saline treatment group on Day 18, but no tumors were found at this time on the backs of the
group treated intraperitoneally with 7 at 25 mg/kg. Mice were euthanized on day 35, and the
tumors were collected for measurement of volume and weight (Fig. 1A). Tumors from the 7-

treated group were significantly smaller than those of the control group (saline treatment) in both
volume and weight (Fig. 1B). Hematoxylin and eosin (HE) staining showed that the tumors
formed in all mice were human gliomas (Fig. 1C). Immunofluorescence staining (Fig. 2) showed
that the tumors from the 7-treated group had significantly lower expression of proliferation factor
Ki67, of CD31 (a marker for tumor angiogenesis), of mitochondrial stability protein BCL-2, of
GFAP (a glioma marker) and of stem cell marker CD133 as compared with the control group.
The treated tumors also displayed significantly higher expression of the apoptotic factor
activated caspase 3 (ΔCaspase 3) as compared with the control group (Fig. 2). Thus, in vivo
experiments showed that compound 7 significantly inhibited cancer cell proliferation, in vivo
tumorigenesis, and tumor angiogenesis.
Fig. 1. Compound 7 inhibited the in vivo tumorigenicity of the human glioma cell line U87MG. (A) Comparison of
tumors formed on the backs of nude mice (day 35 after implantation of tumor cells) . (B) The weight and volume of
tumors following treatment with 7 for 35 days were significantly reduced as compared with those following
treatment with saline.**p<0.01 vs. saline group. *p<0.05 vs. saline group. t test. n=3. (C) Type of tumors formed in
the two groups from U87MG-injected mice. HE staining revealed that tumors from both groups were human
glioblastomas. Magnification 200×.

Fig. 2. Low expression of proliferation and angiogenic proteins in U87MG tumors from mice treated with 7.
Immunofluorescence staining results showed that the tumors from the 7-treated group had significantly lower
expression of Ki67, BCL-2, GFAP, CD133 and CD31 as compared with the control group and a higher expression
of ΔCaspase 3. Magnification 200×.
Table 3
Growth inhibition of KU812, KBM5-wt and KBM5-T315I CML cell lines by
compounds 7 and 25 and by IM and PN as reference compounds.
a
Compd
IC ± SD (nM)
50
KU812
9.5 ± 1
KBM5-wt
8.2 ± 1
KBM5-T315I
8.8 ± 1
7
2
5
12.4 ± 1
311.2 ± 1
8.5 ± 1
12.2 ± 1
152.2 ± 1
10.2 ± 1
13.1 ± 1
IM
PN
310.6 ± 1
10.6 ± 1
a
Cytotoxic concentrations for the indicated cell lines. Experiments were performed in
triplicate. Incubation time was 48 h.
3
.6. KBM5-wt, KBM5-T315I and KU812 leukemia cell lines
ARDAP compounds 7 and 25 were compared with the TKIs IM and third-generation PN in
the Bcr-Abl-expressing KU812 cells, in the IM-sensitive KBM5 cells and in the IM-resistant

KBM5-T315I cells expressing the threonine for isoleucine mutation at position 315 of the
oncoprotein. The IC values of both compounds were at least one order of magnitude superior to
5
0
IM. Compound 7 showed strong inhibition at single digit nanomolar concentrations of the
KBM5-wt (IC = 8.2 nM) and KBM5-T315I (IC = 8.8 nM) cell lines, and 7 was 1.2 fold
5
0
50
superior to PN against both cell lines (IC = 10.2 and 10.6 nM, respectively) (Table 3).
5
0
Combination of 7 with PN showed a large increase in the percentage of apoptotic cells,
greater than in the combinations of PN with either IM or nilotinib (data not shown), in the three
CML lines examined (Fig. 3). Most importantly, in the KBM5-T315I cells, the PN+7
combination yielded 70% apoptotic cells, as compared with 35% for PN alone.
Fig. 3. Compound 7 potentiates TKI mediated CML cell death, including cells that express the T315I mutation.
Cells were treated for 48 h with DMSO as vehicle, 7 at 5 nM as a single agent or in combination with 100 nM IM or
5
nM PN. The cells then underwent annexin V staining as a cell death assay and were quantitated by flow cytometric
analysis. Experiments were performed in triplicate, and SD’s are indicated by the error bars.
Compound 7 was compared with IM and PN alone and in combination in the KBM5-T315I

cells for mitochondrial membrane depolarization. In this assay, 7 was superior to IM and slightly
inferior to PN. Combining PN+7 strongly increased mitochondrial membrane depolarization by
augmenting the PN-mediated effects. These effects were maximal after 48 h (Fig. 4A and 4B).
Fig. 4. Compound 7 promotes mitochondrial membrane depolarization by augmenting PN-mediated effects in
KBM5-T315I cells. (A) KBM5-T315I cells were treated for the indicated times with 5 nM 7, 100 nM IM or 5 nM
PN as single agents or IM or PN in combination with 7. At the indicated times, cells were stained with 5 μg/mL of
JC-1 (5,5’,6,6’-tetrachloro-1,1’,3,3’-tetraethylbenzimidazolyl-arbocyanine iodide) at 37 °C for 15 min and then
washed twice with ice-cold PBS. The intracellular red (590 nm, JC-1 red) and green (525 nm, JC-1 green)
fluorescence emissions of JC-1 were then analyzed by flow cytometry for mitochondrial potential. (B) Panels show
a set of representative plots of KBM5-T315I cells at 48 h. The percentages represents the cell population of the
gated region inside the polygon. **P < 0.01; ***P < 0.0001, compared with control, one-way ANOVA, post hoc
intergroup comparisons.

The synergy of PN in combination with 7 was evaluated for effects on the phosphorylation of
Bcr-Abl and for the expression of apoptosis/mitochondrial damage-related proteins. Compound 7
altered mitochondrial functionality even when used alone, affecting the expression of Mcl-1 and
the amount of cytoplasmatic cytochrome C, which are major effectors of the cellular apoptotic
machinery (Fig. 5). Mcl-1 is an antiapoptotic protein of the Bcl2 family originally isolated from
human myeloid leukemia cells, and Mcl-1 plays a key role in regulating the mitochondrial
pathway leading to apoptosis in a wide variety of tissues. These proteins manage mitochondrial
permeability to cytochrome c and hence activate the initiation of a caspase cascade that leads to
programmed cell death [28]. Inhibition of tyrosine phospholylation of Bcr-Abl was boosted by
the combination of 7 with PN, while the total level of the Bcr-Abl oncoprotein was little affected.
These results indicate that 7 used as a single agent did not have a significant effect on Bcr-Abl
activation, suggesting that an upstream inhibition of a “Bcr-Abl-addicted pathway” is required to
prompt the pro-apoptotic effect.
Fig. 5. Compound 7 synergizes with PN in reducing the phosphorylation of Bcr-Abl and the expression of
apoptosis/mitochondrial damage-related proteins. KBM5-T315I cells were exposed to vehicle alone (-), PN and 7 at
1
, 5 or 10 nM for 24 h and then lysed. Western blot analysis of whole cell lysates with antibodies against phospho-

Bcr-Abl (p-Bcr-Abl at tyrosine Y245), Bcr-Abl, cytochrome c (Cyto c), Mcl-1 and Bcl-2 are shown.
3
.7. Human colon cancer SW480 and SW6220 cell lines
Turcot syndrome (TS) is a disorder characterized by the occurrence of both primary tumors of
GBM and colon adenocarcinoma [29]. Hence, we evaluated compounds 7 and 25 as inhibitors of
the human colon cancer primary cells SW480 and metastatic cells SW620. Compounds 7 and 25
inhibited the growth of the colon cancer cell lines, with IC values ranging from 66.5 nM (25,
5
0
SW620 cells) to 933.7 nM (25, SW480 cells). In these assays, compound 7 was a stronger
inhibitor of the SW620 cells than of the SW480 cells, and 25 was superior to 7 as an inhibitor of
the SW620 cells, with an IC50 value 66.5 nM (Table 4). These results are interesting since
metastatic cells SW620 have shown greater drug resistance than the primary tumor cells SW480,
due to activation of cancer stem cells with drug resistance proteins [30] and abnormal activation
of the WNT/β-catenin signaling pathway [31].
Table 4
Growth inhibition of human colon cancer primary SW480 and
metastatic SW620 cells by compounds 7 and 25.
a,b
Compd
IC ± SD (nM)
50
SW480
SW620
111.4 ± 1.5
66.5 ± 1.4
7
656.1 ± 1.5
933.7 ± 1.5
2
5
a
Cytotoxic concentrations for the indicated cell lines.
Experiments were performed in triplicate. Incubation time was
b
4
8 h.
3
.8. Inhibition of human topoisomerases I and II

Compound 7 was evaluated for its ability to inhibit two vital enzymes involved in cell
metabolism, human topoisomerases I and II (hTopoI and II), which manage DNA topology
during replication and transcription [32]. The in vitro hTopoI relaxation assay was performed by
exposing hTopoI to the vehicle (DMSO) or to 7 at 10, 50 or 100 μM. At all compound
concentrations, hTopoI activity was detected, and no inhibitory effect occurred with 7. Multiple
bands, due to DNA cleavage by hTopoI, were observed upon exposure to 100 μM 7 (Fig. 6, left
panel, lane 3), as well as in the vehicle control (left panel, lane 2) (lane 1 shows the uncut
pHOT1 DNA plasmid bands).
In the hTopoII decatenation assay, hTopoII was exposed to 7 at 10, 50 or 100 μM or to
vehicle. No inhibition occurred at 10 or 50 μM of either compound (data not shown), but, at 100
μM, compound 7 totally inhibited hTopoII activity. Fig. 6, right panel, lane 3 shows a band at the
top of the agarose gel. This band represents the catenated circles of kinetoplast DNA (kDNA),
used as substrate, unable to penetrate the agarose gel because of its large size. In contrast, in the
control reaction (vehicle only, Fig. 6, right panel, lane 2), two bands corresponding to kDNA
decatenation products are visible at the bottom of the gel, indicating the full activity of hTopoII.
Thus, compound 7 completely blocked the activity of hTopoII at 100 μM, while it failed to
inhibit hTopoI, at least under the conditions used in these assays.

Fig. 6. Human topoisomerase I and II assays. Left panel. hTopoI relaxation assay: pHOT1 DNA was incubated with
hTopoI in the presence of vehicle or compound 7 at 100 μM. Lane 1: pHOT1 DNA (upper band: open circular
plasmid; lower band: supercoiled plasmid); lane 2: vehicle (DMSO) + hTopoI; lane 3: 7 + hTopoI. Right panel.
hTopoII decatenation assay: kinetoplast DNA (kDNA) was incubated with hTopoII in the presence of vehicle or
compound 7. Lane 1: kDNA only; lane 2: vehicle (DMSO) + hTopoII; lane 3: 100 µM 7 + hTopoII.
4
. Computational studies
4
.1. Colchicine site
The computations were run using tubulin structure 1SA0 from the Protein Data Bank [33] and
Plants as the docking engine [34]. Inspection of the binding modes of 7 and 25 into the colchicine
site of tubulin led to the identification of key contacts: (i) the trimethoxyphenyl ring formed
hydrophobic interactions with the Leu248 and Leu255 side chains and polar contacts with
Cys241; (ii) the heterocycle (pyridine, thiophene, furan or thiazole) formed hydrophobic
contacts with Lys352. These rings were at bond distance to form a π-cation interaction with the
ε-nitrogen of Lys352 or a H-bond with the nitrogen atom of the Asn101 side chain; (iii) the
substituted phenyl ring was involved in extensive hydrophobic contacts with the Met259β,
Lys353β, Ala180α and Val181α side chains (Fig. 7).

Fig. 7. Docking of ARDAPs 7 and 25 into the colchicine site on tubulin. Proposed binding modes for 7 (cyan) and
5 (green). Colchicine is shown in magenta. The residues involved in the interactions described in the text are
2
shown as grey lines. Tubulin is depicted as a cartoon: α-subunit (light blue), β-subunit (green).
Fig. 8. Docking pose of 7 into the hTopoII etoposide site. Proposed binding mode for 7 (cyan). Etoposide is shown
in green. The residues involved in the interactions described in the text are shown as grey lines. DNA bases involved
in interactions are shown as magenta lines. hTopoII is depicted as a cartoon (light blue). For the sake of clarity the
diagram shows only derivative 7 and only one hTopoII monomer. Closest DNA bases not involved in interactions
are shown as a yellow ribbon.

4
.2. Human topoisomerase II
hTopoII inhibitors are categorized into two broad classes, etoposide like (poison) and ATP
like (catalytic inhibitors), corresponding to two distinct binding sites [35]. Derivative 7 was
docked into both binding sites. The best binding energies (docking scores) were predicted for the
etoposide pocket (data not shown), and etoposide shared pharmacophoric moieties with 7. The
key role played by a methoxy-substituted phenyl ring for binding to this site was reported by Wu
et al. [36]. Thus, we assumed the etoposide site as the putative binding pocket for compound 7.
The docking poses revealed much superimposition of the methoxy-substituted phenyl rings: (i)
the trimethoxyphenyl moiety of 7 formed hydrophobic contacts with the Asp479, Leu502 and
Arg503 backbones; (ii) the thiophene of 7 was involved in hydrophobic interactions with Ala777
and Gln778; (iii) the trimethoxyphenyl and heterocycle formed hydrophobic interactions with
thymine (T ); (iv) the substituted phenyl ring formed -cation contacts with Arg503 and
+
1
hydrophobic interactions with guanine (G ); and (v) the pyrrole ring showed contacts with G
+5
+
5
(Fig. 8).
4
.3. Drug-like properties
The drug-like properties of compounds 7 and 25 were predicted through representative
descriptors. The compounds do not violate the Lipinski [37], Veber [38] and Egan [39] rules and
may have good absorption after oral administration. The predicted Caco-2 and MDCK cell
permeability was in the range of recommended values (Papp >500 nm/s) [40] (Table 5).
Table 5
Predicted Physicochemical Properties of Compounds 7 and 25.
a
b
c
d
e
f
g
h
i
j
k
Comp
MW
H don
H Acc
logP
logS
tPSA
Lr
Vr
Er
Caco
MDCK

7
433.52
432.44
0
0
4
6
5.25
4.23
-5.90
-5.26
77.9
62.6
1
0
0
0
0
0
5003
2902
5697
2819
2
5
Physicochemical properties predicted by FAF-drug4 server [41,42].
a
Molecular Wweight
Number of H-bond donors
Number of H-bond acceptors
Octanol-water partition coefficient predictor by XLOGP3 method [43]
Logarithm of compound water solubility by ESOL method [44]
Topological polar surface area
Lr: Lipinsky Rule deviation (log P <5, H-bond donors ≤5, H-bond acceptors ≤10, and a molecular weight <500)
36]
Vr: Veber rule deviation (rotable bonds ≤ 10, tPSA ≤ 140) [38]
Er: Egan rule deviation, bad or good bioavailability (0  tPSA  132 and -1  logP  6) [39]
b
c
d
e
f
I
[
h
I
Physicochemical properties predicted by QikProp [34,45]
j
Caco-Apparent Caco-2 permeability (nm/sec) (<25 poor, >500 excellent)
MDCK - Apparent MDCK permeability (nm/sec) (<25 poor, >500 excellent)
k
5
. Conclusions
LTS of GBM patients are at high risk of developing second primary neoplasms, including
leukemia. For these patients, the use of classic TKIs, such as IM, is strongly discouraged, since
this treatment causes an increase of phosphorylation of Tyr of several key proteins and,
ultimately, results in a large increase in stimulation of GBM tumor and stem cell migration and
invasion. We aimed to developed potentially useful agents useful for the treatement of for
patients with GBM and CM, working with an alternative mechanism of action from the TKIs
based on inhibition of tubulin polymerization. Starting from the previouly reported compound 1,
wWe replaced the 4-aminophenyl group of 1 with five- and six-membered bioisosteric
heterocycles while retaining the important methyl and fluoro substituents on the 1-phenyl ring.
Compounds 7 and 25 inhibited both GMB and CML cells, including those expressing the CML
T315I mutation, at nanomolar concentrations. Compound 7 was also found to inhibit hTopoII
selectively and completely at 100 μM, but it was inactive against hTopoI. Beyond inhibition of
tubulin polymeration as a primary target, 7 can in addition selectively block hTopoII activity. In

experiments in BALB/cnu/nu mice injected subcutaneously with U87MG cells, 7 administered
intraperitoneally at 25 mg/kg significantly inhibited GBM cancer cell proliferation, in vivo
tumorigenesis, and tumor angiogenesis. Compound 7 showed strong inhibition of cancer cell
growth and had a calculated improvement of metabolic stability to phase II metabolism and
higher lipophilicity compared to the previously reported lead 1. Compound 7 thus has potential
as a dual-target antitumor agent for the treatment of both GBM and CML. These results pave the
way for further development of this class of compounds, and additional advances will be
reported in due course.
6
. Experimental section
6
.1. Chemistry
All reagents and solvents were handled according to the material safety data sheet of the
supplier and were used as purchased without further purification. 3′,4′,5′-
Trimethoxyacetophenone (26) is commercially available. ꢀrganic solutions were dried over
anhydrous sodium sulfate. ꢁvaporation of solvents was carried out on a Bꢂchi Rotavapor R-210
equipped with a Bꢂchi V-850 vacuum controller and a Büchi V-700 vacuum pump. Column
chromatography was performed on columns packed with silica gel from Merck (70−230 mesh).
Silica gel thin layer chromatography (TLC) cards from Merck (silica gel precoated aluminium
cards with fluorescent indicator visualizable at 254 nm) were used for TLC. Developed plates
were visualized with a Spectroline ENF 260C/FE UV apparatus. Melting points (mp) were
determined on a Stuart Scientific SMP1 apparatus and are uncorrected. Infrared (IR) spectra
were recorded on a PerkinElmer Spectrum 100 FT-IR spectrophotometer equipped with a
universal attenuated total reflectance accessory. IR data were acquired and processed by