Design and synthesis of 3,5-substituted 1,2,4-oxadiazoles as catalytic inhibitors of human DNA topoisomerase IIα

Article abstract of DOI:10.1016/j.bioorg.2020.103828

Cancer constitutes a group of diseases linked to abnormal cell growth that can potentially spread to other parts of the body and is one of the most common causes of death. The molecular motors - DNA topoisomerases - that enable topological changes of the DNA molecule are one of the most established targets of cancer therapies. Due to known limitations of established topo II poisons such as cardiotoxicity, induction of secondary malignancies and recognized cancer cell resistance, an emerging group of catalytic topo II inhibitors attempts to circumvent these challenges. Currently, this approach comprises several subgroups of mechanistically diverse inhibitors, one of which are compounds that act by binding to their ATPase domain. In this study we have designed, synthesized and characterized a new series of 3,5-substituted 1,2,4-oxadiazoles that act as catalytic inhibitors of human topo IIα. The introduction of the substituted rigid substitutions on the oxadiazole backbone was intended to enhance the interactions with the ATP binding site. In the inhibition assays selected compounds revealed a new class of catalytic inhibitors targeting this molecular motor and showed binding to the isolated topo IIα ATPase domain. The predicted inhibitor binding geometries were evaluated in molecular dynamics simulations and subsequently dynophore models were derived, which provided a deeper insight into molecular recognition with its macromolecular target. Selected compounds also displayed in vitro cytotoxicity on the investigated MCF-7 cancer cell line and did not induce double-strand breaks (DSB), thus displaying a mechanism of action diverse from the topo II poisons also on the cellular level. The substituted oxadiazoles thus comprise a chemical class of interesting compounds that are synthetically fully amenable for further optimization to anticancer drugs.

Full text of DOI:10.1016/j.bioorg.2020.103828

Journal Pre-proofs
Design and synthesis of 3,5-substituted 1,2,4-oxadiazoles as catalytic inhibi-
tors of human DNA topoisomerase IIα
Kaja Bergant Loboda, Katja Valjavec, Martina Štampar, Gerhard Wolber,
Bojana Žegura, Metka Filipič, Marija Sollner Dolenc, Andrej Perdih
PII:
S0045-2068(20)30121-8
DOI:
https://doi.org/10.1016/j.bioorg.2020.103828
YBIOO 103828
Reference:
To appear in:
Bioorganic Chemistry
Received Date:
Revised Date:
Accepted Date:
16 January 2020
20 March 2020
5 April 2020
Please cite this article as: K. Bergant Loboda, K. Valjavec, M. Štampar, G. Wolber, B. Žegura, M. Filipič, M.
Sollner Dolenc, A. Perdih, Design and synthesis of 3,5-substituted 1,2,4-oxadiazoles as catalytic inhibitors of
human DNA topoisomerase IIα, Bioorganic Chemistry (2020), doi: https://doi.org/10.1016/j.bioorg.2020.103828
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Inc.

Design and synthesis of 3,5-substituted 1,2,4-oxadiazoles as
catalytic inhibitors of human DNA topoisomerase IIα
Kaja Bergant Loboda†1,2, Katja Valjavec , Martina Štampar , Gerhard Wolber ,
†1
3
4
3
3
2
1
Bojana Žegura , Metka Filipič , Marija Sollner Dolenc, and Andrej Perdih *
1National Institute of Chemistry, Hajdrihova 19, SI 1000 Ljubljana, Slovenia
2
University of Ljubljana, Faculty of Pharmacy, Aškerčeva 7, SI 1000 Ljubljana, Slovenia
3National Institute of Biology, Department of Genetic Toxicology and Cancer Biology, Večna
pot 111, 1000 Ljubljana, Slovenia
4Institute of Pharmacy, Freie Universität Berlin, Königin Luise-Strasse 2+4, 14195 Berlin,
Germany
KEYWORDS: human DNA topoisomerase IIα, catalytic inhibitors, drug design, oxadiazoles,
anticancer agents
†
These two authors contributed equally to this work.
*
Corresponding author:
Andrej Perdih
National Institute of Chemistry
Hajdrihova 19, SI-1001 Ljubljana, Slovenia
Tel.: +386-1-4760-376
e-mail: andrej.perdih@ki.si
1

GRAPHICAL ABSTRACT
2

ABSTRACT
Cancer constitutes a group of diseases linked to abnormal cell growth that can potentially spread
to other parts of the body and is one of the most common causes of death. The molecular motors
-
DNA topoisomerases - that enable topological changes of the DNA molecule are one of the
most established targets of cancer therapies. Due to known limitations of established topo II
poisons such as cardiotoxicity, induction of secondary malignancies and recognized cancer cell
resistance, an emerging group of catalytic topo II inhibitors attempts to circumvent these
challenges. Currently, this approach comprises several subgroups of mechanistically diverse
inhibitors, one of which are compounds that act by binding to their ATPase domain.
In this study we have designed, synthesized and characterized a new series of 3,5-substituted
1
,2,4-oxadiazoles that act as catalytic inhibitors of human topo IIα. The introduction of the
substituted rigid substitutions on the oxadiazole backbone was intended to enhance the
interactions with the ATP binding site. In the inhibition assays selected compounds revealed a
new class of catalytic inhibitors targeting this molecular motor and showed binding to the
isolated topo IIα ATPase domain. The predicted inhibitor binding geometries were evaluated
in molecular dynamics simulations and subsequently dynophore models were derived, which
provided a deeper insight into molecular recognition with its macromolecular target. Selected
compounds also displayed in vitro cytotoxicity on the investigated MCF-7 cancer cell line and
did not induce double-strand breaks (DSB), thus displaying a mechanism of action diverse from
the topo II poisons also on the cellular level. The substituted oxadiazoles thus comprise a
chemical class of interesting compounds that are synthetically fully amenable for further
optimization to anticancer drugs.
1
. INTRODUCTION
3

Cancer is one of the most prevalent diseases, and recent predictions based on 2010-2012 data
suggest that about 39.6% of men and women will be diagnosed with cancer at some point in
their lives [1]. In 2018, 9.6 million deaths were recorded based on the WHO data, corresponding
to about 1 in 6 deaths. As a complex multifunctional disease, the true origins of cancer are still
not fully understood. In their seminal papers Hanahan and Weinberg described the biological
capabilities acquired during the multi-step development of human tumors, known as the
"hallmarks of cancer" [2, 3], one of which is uncontrolled cell division.
DNA topoisomerases are the DNA topology altering enzymes belonging to the group of
molecular motors. Some of its members require the energy of the ATP molecules to direct the
DNA molecular motion [4, 5]. As their primary function they regulate the DNA topology by
disentangling chromosome segments via their ability to cleave, manipulate, and religate DNA
strands [5].
Topoisomerases are divided into two large groups: Type I topoisomerases which can catalyze
topological changes within one DNA molecule and Type II topoisomerases which can catalyze
such changes affecting two segments of the DNA molecule. Type II refers here to the fact that
these enzymes can cleave both strands of one double-stranded DNA segment, in contrast to
Type I, which cleaves only one strand within a double-stranded DNA molecule [4, 6]. In
addition, Type II topoisomerases for its successful catalytic activity require the presence of
2
+
Mg ions along with the mentioned ATP molecules. [7].
The catalytic cycle of human topoisomerase IIα is a complex, multistep process in which ATP
hydrolysis, in conjunction with the controlled association and dissociation of multiple subunits,
leads to the passage of one DNA chain through another by means of cleavage [8, 9]. Despite
intensive research a lot of mechanistic steps remain unclear, especially with the conformational
changes associated with the passage of the T-segment via a cleaved G-segment of the DNA [4,
1
0].
4

Human DNA topoisomerase II α is a well-known target used in anticancer therapy. Inhibitors
of human topo IIα are in the literature divided into two large groups: topo II poisons and
catalytic inhibitors [11, 12]. The more established group, topo II poisons, act by stabilizing the
short-lived covalent complex formed between the topo II enzyme and the T-segment of the
DNA. This in turn leads to permanent breaks in the cell DNA, resulting in chromosome
translocations and finally apoptosis. The examples of topoisomerase poisons that are already
included in standard clinical practice are well-known molecules such as etoposide [13] and
doxorubicin [14]. The development of new topo II poisons that are currently undergoing clinical
trials include various and diverse molecules among them compounds such as Voreloxin, C-
1
311 and R(+)XK469 [15].
Alternatively, a new emerging group of catalytic topo II inhibitors was designed to avoid the
well-known side effects of topo II poisons such as cardiotoxicity and induction of secondary
malignancies [14, 16]. In addition, the increased resistance to topo II cancer therapies is also a
further challenge for successful therapy. Here, diverse and mechanistically different-acting
groups of compounds aim to inhibit the topo II catalytic cycle without the stabilization of the
covalent cleavage complex [12]. The groups can act via an inhibition of binding between the
DNA and topo II (e.g. aclarubicin and suramine) [17, 18], inhibition of the DNA cleavage (e.g.
merbarone) [19], inhibition of the ATP hydrolysis and capturing of topo II in a closed clamp
(e.g. ICRF group of compounds) [20, 21] and finally, via inhibition of binding of the ATP
molecule in its binding site (e.g. gambogic acid) [22-26].
Previously, our research has been directed to explore the potential of the last-mentioned
approach to catalytic topo II inhibition. In our efforts we have characterized several classes of
catalytic inhibitors, that target the ATP binding site including triazin-2(1H)-ones [27, 28], 1,3,5-
triazines [29], 1H-pyrazolo[3,4]pyrimidines [30], 9H-purines [30] and 1H-indazoles [31]. In
the scope of our broader investigation of the type II topoisomerase were also exploring the
5

inhibition of the bacterial type II counterpart, the bacterial DNA gyrase, focusing on its ATP
binding site. Among many of our research activities here we also identified a series of
substituted oxadiazoles that have been designed and synthesized as potential DNA gyrase
inhibitors [32].
In this study we used a structure-based approach to design derivatives of 3,5-substituted 1,2,4-
oxadiazoles that could act as catalytic inhibitors of human topo IIα molecular motor. The
schematic outline of the research process is shown in Scheme 1. The introduction of the
substituted rigid moieties into the initially available flexible, non-active oxadiazoles served to
enhance the interactions with the targeted human topo IIα ATP binding site. After initial SAR
assessment of the two main substituents attached to the core scaffold, a second series of 3,5-
substituted 1,2,4-oxadiazoles was synthesized. We then included a small series of reversely
substituted commercially available 3,5-substituted 1,2,4-oxadiazoles in the inhibitor screening
process to broaden the SAR of our initial synthesized series. We evaluated the inhibition of
topo IIα with a high-throughput screening (HTS) relaxation assay, and for selected active
compounds the mechanism of topo IIα inhibition was further investigated with a cleavage assay
and SPR binding experiments. Predicted inhibitor binding was evaluated in a molecular
dynamics (MD) simulation, and subsequently a dynophore model - dynamic 3D pharmacophore
-
was derived to gain deeper insight into the molecular recognition of this class of compounds
with their macromolecular target. At the cellular level, MTS cytotoxicity measurements and
analysis of the induction of the DNA double-strand breaks (DSB) by γ-H2AX assay were
performed on the MCF-7 breast cancer cell line to evaluate their genotoxic potential.
6

PHASE I
Design, screening and synthesis
STRUCTURE-GUIDED
DESIGN AND SCREENING
N
O
O
H
N
R
N
N
H
N
N
R2
O
R1
SYNTHESIS AND COMPOUND
SELECTION
commercially available compounds
synthetic compounds
PHASE II
Evaluation of compounds
ACTIVITY
STUDIES
Inhibition activity
Molecular dynamics and dynophores
Mechanism
enzymatic level)
Cell studies
(
Scheme 1. Schematic outline of the workflow used in the identification of the 3,5-substituted
1
,2,4-oxadiazoles as catalytic inhibitors of the human DNA topoisomerase IIα.
7

2
. RESULTS AND DISCUSSION
2
.1. Design of 3,5-substituted 1,2,4-oxadiazoles
Our research started with a small subset of the available flexible 3,5-substituted-1,2,4-
oxadiazole compounds A-D, which were previously designed in our research activities as
potential inhibitors of the bacterial type II topoisomerase - DNA gyrase [data not shown].
Unfortunately, when the topo IIα HTS relaxation assay was performed, these compounds were
all classified as inactive (see Supplementary material, Table S1). Not discouraged by this result,
we have decided to try to optimize their structure in order to achieve a better fit into the ATP
binding site of the human topo IIα.
In a first step we docked the oxadiazole compound A into the ATP binding site of the human
topo IIα ATPase domain (PDB: 1ZXM) using GOLD program [33]. Subsequently, we
visualized the results with LigandScout [34], which enabled us to generate the 3D structure-
based pharmacophores thus pinpointing the key intermolecular interactions (Figure 1). In this
process we observed that the acetylated amino group on one phenyl ring attached to the
oxadiazole of compound A forms an interaction with Asn120 in the human topo IIα ATP
binding site, which is the anchoring residue of the native ATP ligand adenine moiety. In
addition, the carboxylic group attached on the flexible aliphatic chain points to the residues
normally occupied by the interactions between the phosphate groups and their binding site, such
as Lys168.
Although the compounds A-D were all rendered inactive, we speculated that the aniline-
substituted 1,2,4-oxadiazole scaffold, after appropriate modification, might allow favorable
interactions with the ATP binding site to such an extent that inhibition of the target enzyme
could be achieved. Therefore, in the first design step, we replaced the flexible chain substituent
attached to the oxadiazole core with a phenyl ring substituent also containing different
functional groups. In this way, we aimed to increase the rigidity of the compounds and introduce
8

such substituents/functional groups that would allow interactions with the "triphosphate" part
of the ATP binding site (Figure 1). The first compound of the 3,5-substituted 1,2,4-oxadiazoles
designed in this way was compound 3a, which contains the carboxyl group in its ester form on
the phenyl ring.
Figure 1. Predicted binding modes of the inactive oxadiazole compound A and the designed
rigid 3,5-subtituted 1,2,4-oxadiazole 3a in the topo IIα ATP binding site (PDB:1ZXM). Bellow:
General structures of the targeted synthesized 3,5-substituted 1,2,4-oxadiazoles 3-6 (left)
selected commercially available substituted 3,5-substituted 1,2,4-oxadiazoles 7-15 (right).
To investigate how such oxadiazole compounds can accommodate the ATP binding site, we
again performed the docking calculations of the designed oxadiazole 3a (Figure 1). When
analyzing the obtained binding models, we again observed that the residue Asn120 forms an H-
bond interaction with the amido group on the 3-phenyl ring, which is bound to the 1,2,4-
oxadiazole scaffold. We also detected the presence of a water-mediated interaction between the
Asn91 residue and the X-ray detected crystal water W924, which was included in the definition
9

of the binding site, as in our previous studies, [23, 27, 30], under the assumption that it could
form an important interaction. Hydrophobic interactions were formed between the two phenyl
substituents and the amino acids Ile125, Ile141, Phe142 and Ala167. The subsequently
generated 3D structure-based pharmacophore detected further favorable interactions with the
amino acids Lys168, Arg162 and Thr165 in the "ribose sugar" part and in the "triphosphate"
part of the ATP binding pocket. Additional comparison of the docked binding mode of active
compound 3a and a non-hydrolysable AMP-PNP ligand in the human DNA topoisomerase IIα
ATP binding site (PDB:1ZXM) is shown in Supplementary material, Figure S2.
In the original oxadiazole series A-D, the acylated amino group substituent was present either
at meta or para position and its carboxylic group was present in both the free acid and ester
forms. To explore this chemical space in our new series, we decided to vary the position of the
carboxyl moiety attached to the phenyl ring as well as the position of the acylated amino group.
In design phase 1 we synthesized the first 9 target compounds 3a, 3b, 4a, 4b, 5a, 5b, 5k, 6a
and 6b (Table 1). After identifying the best position of the substituents in phase 2, further 16
compounds 3-5 (Table 1 and 2) were synthesized to broaden the SAR of this chemical class. In
this step we also investigated the influence of the acyl substitution of the amino group present
on the first phenyl ring on topo IIα function.
In an attempt to broaden the SAR data, we also looked for commercially available 3,5-
substituted 1,2,4-oxadiazoles with a similar interaction pattern (Table S2). Since only
oxadiazoles that showed a reverse phenyl substitution pattern compared to the main oxazole
scaffold were available, we decided to include them in the screening process in the hope of
possibly identifying further active compounds along with obtaining new SAR data. As a result,
9
compounds 7-15 were selected and tested experimentally (Table 3). Proposed docking
binding mode of the active, commercially available compound 7 in human DNA topoisomerase
IIα ATP binding site (PDB:1ZXM) is shown in Supplementary material, Figure S3.
1
0

2
.2. Synthesis of the 3,5-substituted 1,2,4-oxadiazoles
Designed 3,5-substituted 1,2,4-oxadiazoles 3-6 were synthesized using a synthetic route
presented in Scheme 2. From the starting compounds 3-aminobenzonitrile (Ia) and 4-
aminobenzonitrile (Ib), the corresponding acetamide (IIa or IIb) was synthesized by the
procedure already tested [35]. In this way, the amine group was protected from further
reactions, since it exhibits nucleophilic properties, and at the same time the acetamide group is
also important for interaction with the topoisomerase IIα binding site.
Synthesis of compound Ia proceeded rapidly in fairly high yields (about 90%), which were
similar to that of compound Ib. It was synthesized by the same procedure as compound Ia, but
the reaction took longer (48 hours), unlike compound Ia, where the reaction was completed in
half an hour, which is due to the favorable electronic effects of the ciano group. A further step
in the synthesis of the final compounds was the conversion of the cyano group to amidoxime
(Compounds IIIa, IIIb), which was performed according to a standard procedure with
hydroxylammonium chloride in ethanol [36]. It is the nucleophilic addition of the amine group
of hydroxylamine to the electrophilic carbon of the cyano group on the benzene ring of
compounds IIa and IIb, which leads to the formation of amidoxime.
The third step in the synthesis of the final compounds was the O-acylation of the amidoximes
(Supplementary material, Table S3) that we had synthesized in the previous step. The acid
chlorides or the corresponding acids (for compounds 3c, 3j, 5j), previously activated for
nucleophilic substitution with the hydroxyl group of amidoxime, were selected as the acylating
agents (Supplementary material, Table S4). Most often, activation of the selected acids in
mixed anhydride with ethyl chloroformate was used.
The preparation of O-acylated amidoxime was carried out in both ways with appropriate
efficiencies (from 29% to 97%) but was slightly lower in the case of acid activation, since the
activation was carried out in situ. There are several possible synthetic routes for the synthesis
1
1

of 3,5-disubstituted 1,2,4-oxadiazole rings: 1,3-dipolar cycloaddition of nitriles with nitrile
oxides, condensation of amidoxime with carboxylic acids and their derivatives, oxidation of
aldoxime and N-substituted amidoxime [37]. In our case, we chose the cyclization of O-
acylamidoxime, which we were able to isolate in the previous stage of synthesis. Intramolecular
condensation preceded, first a nucleophilic addition between the amine group and the acyl
carbonyl carbon atom, followed by a water removal. The planned 3,5-disubstituted 1,2,4-
oxadiazoles were synthesized in anhydrous toluene at boiling point, and the reaction mixture
was purged with argon to provide anhydrous conditions. The reactions were carried out for
several days, but in this way, we obtained very pure products which, after isolation, had to be
further purified by column chromatography in only a few cases. For the preparation of
compounds 5a, 5b, 6a, 6b, basic hydrolysis (saponification of the ester) was performed, and
the acetamide groups of 5c-k compounds were removed by acid hydrolysis. In most cases, the
compounds were synthesized from the basic building blocks in 3 or 4 steps, and the synthesis
pathways for these compounds are appropriate.
1
2

NH2
H
N
H N
CN
H
N
2
CN
OH
N
O
O
O
O
b
c
Ia
Ib
IIa
IIb
IIIa
a
H
N
H
N
H N
2
N
OH
CN
CN
R
NH2
IIIb
O
N
NH2
N
H
N
H
N
O
N
R
O
O
O
O
O
e
1
a-k
3a-j
IVa-j (Table S4)
H
N
H
N
O
d
N
N
O
NH2
a-b
N
O
4a-b
R
R
2
1
. f
. g
O
O
N
N
2
H
N
H
N
N
N
COOMe
COOH
O
3a-b
O
5a-b
H
N
H
N
1
. f
O
N
2. g
O
N
N
N
O
O
COOMe
COOH
4a-b
6a-b
1
. i
O
N
O
N
N
H
N
2. j
H N
2
N
R
R
O
5c-k
3a, c-j
a = (CH CO) O/MeOH f = NaOH/dioxane
3
2
b = NH OH x HCl
g = H O, HCl/ethyl acetate
2
2
c = TEA/EtOH
i = 5M HCl/THF
o
d = CH Cl , 0 C
j = H O, NaOH/ethyl acetate
2
2
2
o
e = toluen, T=110 C
Scheme 2. General synthetic scheme for the synthesis of the targeted 3,5-substituted 1,2,4-oxadiazoles
-6.
3
1
3

2
.3. Human topo IIα HTS inhibition assay and initial SAR of the synthesized and
commercially available 3,5-substituted 1,2,4-oxadiazoles
The synthesized and commercially available compounds 3-15 were experimentally assayed if
they inhibit catalytic activity of the human topo IIα using a standard topo II high-throughput
screening (HTS) relaxation assay [38]. To validate the assay, etoposide was used as control
compound, for which we experimentally determined the value of 41.6 μM which adequately
compares to the 60.3 μM value, reported in the literature [39]. Results of assay of the two
synthesized series marked by compounds 3-6 are available in Table 1 and Table 2. It should
also be pointed out, that all synthesized final compounds 3-6 as well commercially acquired
oxadiazoles 7-15 were found to possess drug-like properties characterized by the favorable
predicted logP [40] values lower than 5 and topological Polar Surface Area (tPSA) [41] lower
than 120.
After performing the assay of the initial design phase 1 series was gratifying to observe that
some synthesized compounds of the first series compounds showed to be inhibitors of the
human topo IIα. To the best of our knowledge this represents the first compounds of the 3,5-
substituted 1,2,4-oxadiazole that act on this human type II topoisomerase. For the most active
compound 3a an average IC value of 147.7 ± 10.6 μM was determined which is in comparable
5
0
range, although still about 2.5-fold higher than the etoposide reference drug. The initial SAR
of the assay also pointed out that meta substitution of both phenyl rings attached to the core
1
,2,4-oxadiazole is the most optimal substitution position for the inhibition activity.
Table 1. Structures of the synthesized 3,5-substituted 1,2,4-oxadiazoles 3-6 and their IC50
values determined in the topo IIα HTS screening relaxation assay (design phase 1).
1
4

O
N
N
R2
NH
R1
IC50
μM)
3-(COCH ) 3-(COOMe) 147.7
Compound
R1
R2
(
3
5
a
a
3
3-(COCH ) 3-(COOH)
657.8
3-(COOMe) >1000
3-(COCH ) 4-(COOMe) >1000
3
5
3
5
k
b
b
3-H
3
3-(COCH ) 4-(COOH)
>1000
3
4
6
a
a
4-(COCH ) 3-(COOMe) 700.6
3
4-(COCH ) 3-(COOH)
>1000
3
4
6
b
b
4-(COCH ) 4-(COOMe) >1000
3
4-(COCH ) 4-(COOH)
>1000
3
In the design phase II, we kept the meta position of the amide or the appropriate free amine on
of 3-phenyl ring fixed and varied the R2 substituents introduced on the 5-phenyl ring. As a
result, we introduced various meta- and para- substituted phenyls that were synthetically
available (Table 2). Some of the flour- and choro- derivatives were included due to substantial
similarity of the docking placement of such R2 substituted phenyl ring and a corresponding
substituent present in our recently reported 4,6-substituted-1,3,5-triazin-2(1H)-ones occupying
the same position in the ATP binding site (Figure S4) [28]. Inhibitory relaxation assays
identified 7 new compounds possessing minimal topo IIα inhibitory activity with compounds
3
f and 5g possessing the most promising IC values 284.2 µM and 311.3 µM, respectively.
5
0
From these results, we obtained relevant SAR information that both free amide or amine on the
1
5

meta R1 position can lead to topo II inhibition activity. The new substituents also identified
some of the new R2 substituents, with compound 3f possessing the trifluoro substituent on the
5
-phenyl ring and 5g a methoxy substituent which exhibited comparable activity to the initially
synthesized compound 3a.
Table 2. Structures and IC50 values of the synthesized 3,5-substituted oxadiazoles 3-5
determined in the topo IIα HTS screening relaxation assay (Design Phase 2).
O
N
R1
NH
N
R2
IC50
Compound
R1
R2
(
μM)
461.4
>1000
>1000
433.6
657.4
>1000
284.2
>1000
>1000
311.3
>1000
585.1
>1000
3
5
c
c
-(COCH3)
-H
3-(Cl)
3-(Cl)
3
5
d
d
- (COCH3)
-H
4-(Cl)
4-(Cl)
3
5
e
e
-(COCH3)
-H
3-(CF3)
3-(CF3)
4-(CF3)
4-(CF3)
-3(OMe)
-3(OMe)
-4(OMe)
-4(OMe)
-4(NO2)
3
5
f
f
- (COCH3)
-H
3
5
g
g
-(COCH3)
-H
3
5
h
h
-(COCH3)
-H
3
i
-(COCH3)
1
6

5
i
-H
-4(NO2)
>1000
3
5
j
j
-(COCH3) -3F, -5(CF ) 523.9
3
-H
-3F, -5(CF ) >1000
3
To further investigate the chemically space of this compound class we decided to select and
assay also a small subset of commercially available 3,5-substituted 1,2,4-oxadiazoles 7-15 with
different “reverse” substitutions of the phenyl rings (Table 3). We decided for this search also
because the position of the nitrogen functionality of the 3-phenyl ring was not available in the
pool of commercially available molecules. In this manner, we now varied the functional groups
on the phenyl at the position 3 of the oxadiazole scaffold and fixed the acetylated amino group
on the 5-phenyl ring. We made several searches using the e-molecules search engine and after
molecular docking analysis of the approximately 100 available compounds acquired 9 new
oxadiazoles for inhibition assay.
From the selection, compounds, 10, 11 and 12 showed the IC inhibitory activities of 226.1
5
0
µM, 165.9 µM and 162.7 µM, respectively, these being of comparable potency to the Phase 1
synthesized compound 3a. By determining the inhibition activity of these compounds, we also
showed that reverse 3,5-phenyl substitution pattern of the core oxadiazole ring, with the
acetylated amino group being fixed to the 5-phenyl ring also enables the inhibition activity.
Besides the already investigated methoxy substituent (7) more complex piperidine (10,11) and
morpholine moieties (12), attached to phenyl ring via either an amido or sulfonyl group were
rendered as favorable for the topo IIα inhibition activity broadening the compounds SAR.
Table 3. Determined IC50 values of the commercially available 3,5-substituted 1,2,4-
oxadiazoles 7-15 in the topo IIα HTS screening relaxation assay.
1
7

N
O
H
R
N
N
O
Compound
number
IC50
R
(μM)
7
8
9
OMe
507.7
455.3
>1000
O
S N
O
226.1
165.9
1
1
0
1
O
O
N
N
O
1
2
162.7
N
1
1
3
4
>1000
>1000
N
N
N
N
S
N
1
5
>1000
1
8

To investigate and compare reversely substituted oxadiazoles, with our synthesized
oxadiazoles, docking poses and 3D structure-based pharmacophores of active compounds 10
and 11 (Figure 2) were calculated. Amido group placed on the 5-phenyl ring of the 1,2,4-
oxadiazoles formed a hydrogen bond with Asn120 in the ATP pocket. Such compound
placement can be rendered fully comparable with what we observed previously when the
nitrogen functionality was placed on the 3-phenyl ring. Standard hydrophobic interactions
formed by both phenyl groups and amino acids Ile125, Ile141, Phe142 and Ala167 appear to
contribute important intermolecular interactions in the derived binding model. The docking
modes also detected interactions between the amido (11) or sulfonyl (12) group oxygens and
the amino acids Lys168, Arg162 and Thr165 in the “ribose-sugar” part and” triphosphate” part
of targeted ATP binding pocket.
Figure 2. Predicted binding modes and 3D structure-based pharmacophores of the active
commercially available reversely substituted oxadiazoles 10 and 11.
2
.4. Investigation of the topo IIα inhibition mechanism
After these initial results provided by the HTS topo IIα relaxation assay, we further investigated
the inhibitory mechanism of the 3,5-substituted 1,2,4-oxadiazoles in more detail. This is a vital
important step in characterizing the topo IIα inhibitors, due to the complexity of the topo IIα
catalytic cycle [42].
1
9

2
.4.1 3,5-substituted 1,2,4-oxadiazoles act as catalytic inhibitors of human DNA
topoisomerase IIα
To investigate the mechanism of inhibition, first a topo IIα-mediated cleavage assay was
performed. Human topo IIα enzyme was incubated with the negatively supercoiled plasmid and
four concentrations of the oxadiazole-based inhibitor 3a along with the topo II poison drug
etoposide as a reference compound. The obtained results (Figure 3) clearly displayed the poison
activity of etoposide with an increase in linear DNA with increased compound. However, the
same titration with the compound 3a and did not result in significant increase of linear plasmid
above that of background (DMSO alone) so they were not considered to act as topo II poisons.
The cleavage assay thus confirmed that the 3,5-substituted 1,2,4-oxadiazoles are catalytic topo
II inhibitors and not topo II poisons. Detailed results of the cleavage assay are available in
Supplementary material, Table S6 and Figure S6.
Figure 3. Results of the topo II α cleavage assay for etoposide reference and 3,5-substituted
1
,2,4-oxadiazole 3a.
2
.4.2. 3,5-substituted 1,2,4-oxadiazoles bind to the isolated human topo IIα ATPase
domain
The cleavage assays confirmed that compounds do not act as topo II poisons. Thus, we
subsequently investigated if the inhibitory activity of compound 3a was associated with their
binding to the human topo IIα ATPase domain using surface plasmon resonance (SPR)
experiments. In this binding experiment the truncated ATPase domain of human topo IIα was
2
0

immobilized on the CM5 chip and real-time sensorgrams were recorded and analyzed using
Biacore T100 Evaluation Software and are provided in Supplementary material, Figure S7.
General fitting and steady-state affinity, one-site binding, were calculated using Biacore
Evaluation Software and graph showing response units as a function of the inhibitor
concentration is supplied in Figure 4.
Figure 4: Results of the SPR binding experiment for compound 3a. The binding graph is
showing response units (RU) as a function of compound concentration.
Tested compound 3a was titrated at six different concentrations in three parallel experiments,
and the K values were determined using a steady-state affinity binding model (one-site
D
binding). The maximal theoretical response was calculated according to the molecular masses
of the human ATPase domain and compound 3a. K value for the tested compound was 14.3 ±
D
3
.4 µM. The SPR experiments confirmed that compound 3a binds to the human topo Iiα ATPase
domain where the ATP binding site is located and have consequently provided more confidence
in our in silico models of binding, generated by molecular docking studies. The determined
binding potency of our compounds to the ATPase domain also favorably compares to the
reported binding of the native ATP to this domain for which the K value of 6.15 ± 0.11×10−4
D
2
1

M was determined using the same SPR methodology [43]. The binding graphs of other two
parallels are represented in Supplementary material, Figure S8.
It is important to note that in the cells many existing enzymes and molecular machines similar
to topo IIα exploit the energy of the ATP molecule to support various biological processes
among others citric acid cycle, beta oxidation, proteins synthesis and muscle contraction. In
addition, there are numerous protein kinases possessing the ATP binding site that play a vital
role in the regulation of most cellular pathways [44, 45]. Thus, the issue of selectivity will be
one of the aspects to explore in further optimization efforts. Interestingly, recent findings
suggested that in certain cases, targeting other relevant targets (e.g. protein kinases)
concurrently with topo IIα ATP inhibition might even be desired, due to resulting synergistic
effects during chemotherapy [46]. All these results indicate that exciting further development
of this class of topo II catalytic inhibitors is still ahead.
2
.5. Analysis of the interactions of the 3,5-substituted 1,2,4-oxadiazoles in the ATP binding
site using molecular dynamics (MD) and dynophore analysis
Our SPR data suggested that optimized series could bind into the ATP binding site located on
the human topo IIα ATPase domain. In addition, the performed cleavage experiment confirmed
that these compounds do act as topo II catalytic inhibitors. Having a static binding pose of the
target-ligand complex obtained by molecular docking, we initiated molecular dynamics
simulations in order to acquire insights into the flexibility [47] of the designed oxadiazole
compound 3a in its targeted ATP binding site. It is important to note that to the best of our
knowledge so far, no co-crystal structure of a small molecule inhibitor bound in the human topo
IIα ATP site has been reported. Using CHARMM we constructed a solvated target topo IIα-
ligand system that was then equilibrated and simulated in the 20 ns MD simulation as described
in the Experimental section. Animation of the observed ligand-protein conformational space is
2
2

available in the Supplementary material and a representative MD snapshot is presented in
Figure 5A.
The MD trajectory was first analyzed with the RMSD parameter. Relatively low value of
RMSD = 2.56 ± 0.54 Å indicates the geometric stability of the predicted docked binging pose
in the human topo IIα binding site (Figure S5). Considering the calculated GOLD docking
model, we first analyzed the stability of designed compound 3a in the adenine part of the
binding pocket, which, according to our previous knowledge and simulations, serves as the
compound’s anchor, with the amino acid Asn120 acting as the main hydrogen bond interaction
(Figure 5B). Proposed interaction between the carbonyl oxygen of Asn120 and oxadiazole’s
ring nitrogen proved to remain fully stable during the simulation, with the average distance
value of 2.90 ± 0.30 Å. Next, we analyzed the proposed interactions in the triphosphate part of
the ATP binding site. Interaction with amino acid residue Ala167 remained reasonably stable
at 2.90 ± 0.20 Å, regardless of relatively high flexibility of the ethylene tail of the molecule.
Despite the observed interaction in the static docking pose, H-bond interaction between the
compound’s tail oxygen and Lys168 (d = 6.1 ± 0.8 Å) did not prove to be stable enough during
MD simulation, as the residue moved away from the compound during the simulation (Figure
5
A, position of Lys168).
Subsequently, we upgraded the geometrical MD analysis with a dynophore model, a powerful
new method of the analysis of MD trajectories developed at the Freie Universität Berlin. Such
produced dynophores aim at supplementing information provided by the classical 3D
pharmacophores with the statistical and sequential information as they represent probability
density clouds for each occurring intermolecular interaction during the MD simulation. In this
way, the analysis of MD trajectories does not depend solely on the geometrical parameters such
as intermolecular distances [48-50].
2
3

Such insight into the flexibility of the interaction pattern confirmed our observations from the
geometrical analysis (Figure 5C) and provided new important insights and clues concerning the
ligand molecular recognition. The hydrogen bond between the amido nitrogen and the amino
acid Asn120 is present 95.0% of the time. Also, interaction between the ethylene tail oxygen
and the amino acid Ala167 remains present 96.9% of the simulation time. Additionally, we
observed a formation of a hydrogen bond between the oxadiazole’s ring nitrogen and the amino
acid Asn91, present 88.4% of the simulation time. This interaction might be complementing
the water-mediated interaction of Asn91 observed earlier. It is important to mention that this
interaction was not detected by the classical geometrical analysis of the MD trajectory, which
verifies the valuable support of dynophore models to classical analysis of MD simulations.
Moreover, the obtained dynophore further substantiated the importance of hydrophobic
interactions that could not be so easily addressed by utilizing classical parameters of geometry-
based analysis (Figure 5C). Both benzene rings form hydrophobic interactions with the amino
acid environment, which is present throughout the whole simulation. First benzene ring majorly
interacts with amino acids Thr215 and Ile125. Second ring forms interactions with Ile14,
Ala167, Phe142, and Tyr151. This type of interaction is most probably crucial for our
compounds to interact with the binding site of topo IIα. In addition, the inherent characteristics
of the ATP binding site imply that non-polar interactions (such as dispersion energy)
importantly contribute to the stabilization of these protein-ligand complexes. To study these
aspects further, application of more rigid computational techniques, such as for example Linear
Interaction Energy (LIE) method developed by Åqvist and co-workers, where the electrostatic
and non-polar contributions to binding free energy can be evaluated in more details would be
valuable [51]. It should be also noted that ligands´ pKa values can change when compounds
bind to the protein and thus influence the molecular recognition process [52, 53].
2
4

Finally, it is important also to note that MD simulation although reasonably long, still does not
enable an extensive coverage of the conformational space for an unambiguous quantitative
ligand stability assessment. However, it should be emphasized that in the absence of a complex
human topo IIα crystal structure this approach still provides some initial clues of the ligand
dynamics, especially since the SPR experiments confirmed the interaction with the ATPase
domain [54].
2
5

Figure 5. (A) Representative MD snapshot of compound 3a in the ATP binding site of the
human topo IIα. (B) Time-dependence graph for the distance between the OD1 atom of Asn120
and the N9 nitrogen of compound 3a (C). An overview of the interaction pattern obtained with
2
6

the dynophore analysis: (Left): Calculated percentage of occurrence of a dynamic
pharmacophore element on the basis of all frames broken into element-interacting amino acids
pairs; (Right): Full calculated dynophore model of compound 3a presented with pharmacophore
feature types – hydrogen bond donor, HBD (green), hydrogen bond acceptors, HBA (red) and
hydrophobic area, H (yellow); (Bellow) Separately presented main pharmacophore features of
3
a – topo II molecular recognition with the corresponding residues involved in this interaction
and their interaction occurrence.
2
.7. Cytotoxicity measurements of the 3,5-subtituted 1,2,4-oxadiazoles on the MCF-7
human breast cancer cell line
Encouraging results of the in vitro characterization of our oxadiazole compounds on the
enzymatic topo IIα level stimulated us to investigate the cytotoxicity of the active synthesized
compounds using the MTS assay and the MCF-7 human breast cancer cell line [42]. Selected
human cancer cell line is representative and well-established systems for cell-based evaluation
of potential anticancer compounds [42].
The initial screening of the synthesized derivatives was performed by exposing exponentially
growing cells to a 100 µM concentration of the compounds for 72 h for compound 3a, 3c, 3e,
3
f, 3j, 4a, 5a, 5d, 5g, 5h, 12 and to a 50 µM for compounds 7, 8, 10 and 11. As a positive
control, etoposide topo II poison (PC) was used. It was gratifying to observe that the first results
revealed significant decrease of viability for four compounds 3a, 7 and 11 along with the
expected cytotoxicity of the PC. Results of the initial screening are represented in Figure 6.
2
7

Figure 6. Results of the initial screening for cytotoxic activity of active 3,5-subtituted 1,2,4-
oxadiazoles on the MCF-7 breast cancer cell line after 72h treatment. Compounds were tested
at 100 µM, except compounds 7, 8, 10 and 11 at 50 µM. Significant differences between the
solvent control and treated cells were calculated using ANOVA (****p<0.0001).
Based on the promising initial MTS screening results, we selected compounds 3a and 7 that all
reduced the cell viability by more than 40% relative to the control for the EC determination.
5
0
Obtained EC values were 38.5 μM for compound 3a and 29.4 μM for compound 7, while the
5
0
determined EC value for etoposide on the MCF-7 cell line after 72h treatment was 12.6 μM.
5
0
Obtained dose-response curves of log concentrations versus the percent of cell viability are
represented on Figure 7 for investigated 3,5-subtituted 1,2,4-oxadiazoles. For etoposide
reference these data are available in Figure S9. This result renders these compounds from the
3
,5-subtituted 1,2,4-oxadiazoles in the comparable range of cytotoxic activity as the etoposide
reference.
2
8

Figure 7. Dose-response curves of log concentrations versus percent viability for the 3,5-
subtituted 1,2,4-oxadiazoles 3a and 7 on the MCF-7 breast cancer cell line (72h treatment).
Experiments were performed in five parallels and SD values were calculated. Significant
differences between the solvent control and treated cells were calculated using ANOVA
(****p<0.0001).
2
.8. Analyses of the induction of the DNA double strand breaks by γ-H2AX assay
As we sought to gain more insight into the mechanism of action also on the molecular level we
further investigated if this class of compounds causes the formation of the DNA double strand
breaks (DSB) for the selective most characterized active oxadiazole compound 3a. DSB are
one form of the DNA damage that can lead to chromosomal breakage and rearrangement [55]
and is related to severe side effects of topo II poisons, such as cardiotoxicity and induction of
secondary malignancies [14, 16, 56]. With topo IIα-mediated cleavage assay presented in
Section 2.4., we confirmed that the active compound 3a acts via the catalytic inhibition and not
as topo II poison on the molecular level. With the γ-H2AX assay developed to assess the
compounds ability to induce the DNA double strand breaks [42], we aimed to confirm the same,
but on the next molecular level of investigation.
2
9