Discovery of 5-aryl-3-thiophen-2-yl-1H-pyrazoles as a new class of Hsp90 inhibitors in hepatocellular carcinoma

Article abstract of DOI:10.1016/j.bioorg.2019.103433

Although hepatocellular carcinoma (HCC)-related mortality has increased over the past decades, treatment options are still very limited, underlining the need for developing new therapeutic strategies. The molecular chaperone heat shock protein 90 (Hsp90)

Full text of DOI:10.1016/j.bioorg.2019.103433

BioorganicChemistryxxx(xxxx)xxxx
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
Discovery of 5-aryl-3-thiophen-2-yl-1H-pyrazoles as a new class of Hsp90
inhibitors in hepatocellular carcinoma
⁎
⁎
Samy Mohamadya, , Muhammad I. Ismaila, Samar M. Mogheithb, Yasmeen M. Attiac,
,
Scott D. Taylord
a Department of Pharmaceutical Chemistry, Faculty of Pharmacy, The British University in Egypt, Al-Sherouk City, Cairo-Suez Desert Road, 11837 Cairo, Egypt
b Faculty of Pharmacy, The British University in Egypt, Egypt
c Department of Pharmacology, Faculty of Pharmacy, The British University in Egypt, Al-Sherouk City, Cairo-Suez Desert Road, 11837 Cairo, Egypt
d Department of Chemistry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G, Canada
A R T I C L E I N F O
A B S T R A C T
Keywords:
Although hepatocellular carcinoma (HCC)-related mortality has increased over the past decades, treatment
options are still very limited, underlining the need for developing new therapeutic strategies. The molecular
chaperone heat shock protein 90 (Hsp90) plays a key role in post-translational maturation of many oncogenic
client proteins that are important for survival and proliferation of cancer cells. Thus, inhibitors of Hsp90 are
promising targets for many cancer types. In this study, 15 diarylpyrazole compounds were screened against
MCF7 and HepG2 cell lines. Compound 8, which contained a thiophene group, demonstrated the highest an-
tiproliferative activity against HepG2 cells having an IC50 of 0.083 μM. Four additional diarylpyrazoles, each
containing a thiophene group, were prepared and screened for antiproliferative activity. None of these four
compounds exhibited superior activity to compound 8 on HepG2 cells. Therefore, compound 8 was selected for
further in vitro assays. Cell cycle arrest was observed at the G2 phase in compound 8-treated cells. Compound 8
also caused a 7.7-fold increase in caspase-3. These results confirm the apoptotic effect of compound 8 on HepG2
Diarylpyrazole
Heat shock proteins
Hsp90
Pyrazole
Client proteins
cells. Moreover, compound 8 inhibited Hsp90 (IC50 = 2.67
0.18 µM) in an in vitro assay and caused a 70.8%
reduction in Hsp90 levels in a HepG2 cell-based assay. Additionally, compound 8 caused significant reduction in
the levels of Hsp90 client proteins (Akt, c-Met, c-Raf, and EGFR) and a 1.57-fold increase in Hsp70. Molecular
docking studies were also performed to predict the binding mode of compound 8 and followed by molecular
dynamics simulations to give further insights into the binding mode of 8.
1. Introduction
synchronize with or control several other downstream molecules im-
plicated in carcinogenesis. Heat shock proteins (HSPs) are molecular
Liver cancer or hepatocellular carcinoma (HCC) has the third
highest mortality rate amongst cancer-related illnesses worldwide [1].
Although treatment modalities have remarkably improved over the past
years, most patients with advanced HCC ultimately require surgery
such as resection or orthotopic liver transplantation. Sorafenib, a non-
selective multiple kinase inhibitor, is the only approved drug for ad-
vanced HCC. However, limited efficacy and de novo resistance to sor-
afenib, emphasize the urgent need for developing new therapeutics to
treat the disease [2].
chaperones that are constitutively expressed cellular proteins and guide
the normal folding, disposition, and turnover of many of the key reg-
ulators of cell growth and survival [3]. These processes are achieved by
multiple chaperone complexes acting in co-operation with several co-
chaperones. The intracellular levels of these molecular chaperones in-
crease in response to various stresses such as heat, electric shock,
chemicals, toxins, and ischemia. These stresses can cause cellular pro-
tein denaturation and cell damage. In response to these stimuli, cells
increase the level of HSPs to mediate protein refolding. Proteins that
fail to refold properly are ubiquitinated and degraded by proteasomes.
Newly synthesized proteins also need HSPs for proper folding and as-
sembly [4]. The molecular chaperone ATP-dependent heat shock pro-
tein 90 (Hsp90) is the most common among the family of HSPs. It plays
Given the intratumor heterogeneity of HCC, working on a single
molecular target in a signaling pathway is not considered a practical
approach and may explain the poor response to many emerging tar-
geted therapies. Hence, a good target to work on is a one that may
⁎ Corresponding authors.
E-mail addresses: samy.mohamady@bue.edu.eg (S. Mohamady), yasmeen.attia@bue.edu.eg (Y.M. Attia).
https://doi.org/10.1016/j.bioorg.2019.103433
Received 22 April 2019; Received in revised form 9 October 2019; Accepted 11 November 2019
0045-2068/©2019ElsevierInc.Allrightsreserved.
Pleasecitethisarticleas:SamyMohamady,etal.,BioorganicChemistry,https://doi.org/10.1016/j.bioorg.2019.103433

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
a key role in post-translational maturation, proper folding, and acti-
vation of over a five hundred client proteins [5]. Many of these proteins
are important oncogenic transcription factors, steroid receptors, and
protein kinases. Hsp90 is abundant in normal cells and constitute up to
1–2% of total cell proteins. This percentage increases up to three-fold in
cancer cells [6]. Cancer cells utilize Hsp90 to chaperone mutated and
overexpressed oncogenic client proteins to protect them from mis-
folding and degradation. Thus, inhibition of Hsp90 can result in si-
multaneous and collective disruption of oncogenic signal transduction
pathways and eradication of these oncogenic client proteins via ubi-
quitin-proteasome activation and, ultimately, apoptosis of the cancer
cells [7]. Drugs that target Hsp90 were found to be selective for cancer
cells over normal cells [7,8]. This is attributed to the fact that Hsp90
derived from tumor cells was found to be present in the active ATP
consuming form, while those from normal tissues are in dormant state
[8]. These drugs, whether used as monotherapy or in combination with
other chemotherapeutic agents, are considered as a promising treat-
ment strategy for various types of malignancy.
us to search for novel Hsp90 inhibitor chemotypes and investigate their
anti-HCC potential in vitro. Here we report on the discovery of 5-aryl-3-
thiophen-2-yl-1H-pyrazoles as novel Hsp90 inhibitors possessing potent
Hsp90 inhibitory activity in HepG2 cells.
2. Results and discussion
2.1. Chemistry
Compounds 1, 2, 3, 4, 5, 6, 7, and 8 were previously prepared
according to our reported procedure (Table 1 entries 1–8) [20]. Using
the same reaction methodology, compounds 9, 10, and 11 were pre-
pared by heating a mixture of the corresponding aryl ketone and 1.2
equiv aryl hydrazone in ethanol in presence of catalytic amounts of
hydrochloric acid for 1 h followed by the addition of 4 equiv DMSO and
0.1 equiv iodine and reflux for 16 h. (Table 1 entries 9–11)
The reverse method, whereby the hydrazone of the ketone is con-
densed with an aryl aldehyde was used for the preparation of com-
pounds 12, 13, 14, 15, 16, 17, 18 (Table 2 entries 1–7).
At the molecular level, Hsp90 is a flexible homodimer. Each
monomer consists of three domains, an N-terminal ATP and drug
binding domain, a middle domain for client protein and co-chaperone
binding and ATP-hydrolysis, and a C-terminal domain containing sites
for dimerization. There are two main Hsp90 isoforms, Hsp90α and
Hsp90β. Hsp90α is stress-inducible, while Hsp90β is constitutive [9].
Interest in the development of selective Hsp90 inhibitors has grown
rapidly over the last two decades [10]. More than 20 Hsp90 inhibitors
are currently in various phases of clinical trials. So far, none of these
inhibitors has been commercialized [11]. Most of the Hsp90 inhibitors
target the N-terminal domain at the ATP-binding site, preventing ATP
hydrolysis necessary for proper Hsp90 functioning and ensuing pro-
teasomal degradation of client proteins leading to apoptosis of cancer
cells [12]. These Hsp90 inhibitors are usually classified according to
chronological evolution into two generations. The benzoquinoid ansa-
mycin antibiotic geldanamycin (GA, Fig. 1) is the vanguard of the first
generation and was first described as an Hsp90 inhibitor in mid 1990s
[13]. GA blocks ATP from binding to the N-terminal pocket by mi-
micking the 3D structure that ATP adopts in the active site of the N-
terminal pocket [14]. In an effort to combat GA side effects such as
hepatotoxicity and lower aqueous solubility, three derivatives of GA
were developed; 17-allylamino-demethoxy geldanamycin (17-AAG,
Tanespimycin) was developed to reduce hepatotoxicity, while 17-di-
methylaminoethylamino-17-demethoxy geldanamycin (17-DMAG) and
IPI-504 (retaspimycin) were developed as water soluble GA derivatives
with lower hepatotoxicity. Unfortunately, none of these compounds
passed Phase II clinical trials [15]. The failure of these compounds has
stimulated the quest for other structural scaffolds with better pharma-
cological and safety profiles. This resulted in the development of the
second generation Hsp90 inhibitors; these are synthetic compounds
that can be broadly categorized into three main categories, the purine
family [16], the pyrazole family [17], and the isoxazole family (Fig. 1)
[18].
Compound 19 was prepared according to a variant of our general
synthetic protocol by refluxing 4-amino-acetophenone and 1.2 equiv.
benzalhydrazone in ethanol in presence of catalytic amounts of HCl and
H2SO4 for 100 h (Scheme 1).
3. Biological evaluation
3.1. Effect of compounds on cell viability
Initially, 15 diarylpyrazole compounds were screened for their po-
tential anticancer activity on two cancer cell lines; MCF7 and HepG2.
Table 3 shows the IC50 values of these 15 compounds (1–9, 12–16, and
19) on both MCF7 and HepG2 using Erlotinib as reference. The pre-
sented results showed a promising antiproliferative potential for 8,
particularly on HepG2 cell line which was, therefore, selected as lead
for further optimization.
It was observed from the biological data given in Table 3 that the
most potent analog was compound 8 which carries 2-thienyl moiety. By
comparing the activities of compounds 1, 7, and 8, it was clear that 2-
thienyl moiety contribute well to the activity. It was also observed that
the presence of 3-OCH3, 3-OH, or 4-halogeno substitution contributes to
the biological activity. For example, compound 5 contains 4-CF3 and
3,4-dimethoxy (IC50 = 0.52
and 4-OH (IC50 = 0.57
and 4-CF3 (IC50 = 0.74
0.01 µM) , compound 9 possesses 3-
0.01 µM), and compound 15 possesses 4-Br
0.03 µM) (Table 3). Therefore, four addi-
tional pyrazole compounds were prepared in which the thiophene
group was kept and the other substituents were 3-OH, 4-CF3, 3,4-di-
methoxy, and 2-fluoro-4-OCH3 respectively (Table 4). The four com-
pounds were then tested against HepG2 cells (Table 4). None of the
tested compounds gave superior activity to 8. Thus, further biological
assays were performed on 8.
There is increasing interest in Hsp90 as a target for treating HCC.
The lack of clinical efficacy of current Hsp90 inhibitors [19] prompted
Fig. 1. Structure of some natural and synthetic Hsp90 inhibitors.
2

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Table 1
Synthesis of 3,5-diaryl pyrazoles from aryl ketones and aldehyde hydrazones
.
Entry
1
Ar1
h-
Ar2
Ph-
Product
Yield (%)
85
1
2
3
2
3
4
73
71
77
Ph-
Ph-
4
5
5
6
86
90
Ph-
Ph-
6
7
78
7
8
8
9
74
71
9
10
11
75
82
10
11
3.2. Effect of 8 on cell cycle
and late apoptotic cells were also 3.3- and 21-fold higher in 8-treated
cells, as compared to control, respectively. Additionally, necrotic cells
were 4.2-fold higher in treated cells compared to control. These results
strongly suggest that compound 8 has an antiproliferative effect on
HepG2 cells.
As shown in Fig. 2, cell cycle analysis showed that 8 caused cell
cycle arrest at the G2 phase where the number of cells treated with 8
were 2.3-fold higher in the G2 phase compared to control. Moreover,
treatment with 8 caused a 6.2-fold increase in the cells in pre-G1. Early
3

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Table 2
Synthesis of 3,5-diaryl-pyrazoles using ketone hydrazones and aryl aldehydes.
.
Ar1
Ar2
Product
Yield (%)
81
1
2
3
12
14
81
90
13
4
5
92
92
15
16
6
7
80
80
17
18
3.3. Effect of 8 on caspase-3 levels
3.5. Effect of 8 on Hsp70
In order to further confirm the apoptotic potential of 8 on HepG2
cells, the level of the primary executioner caspase, caspase-3, was
evaluated and compared to that of erlotinib. As shown in Fig. 3, cas-
pase-3 was 7.7-fold higher in 8-treated HepG2 cells compared to con-
trol. Erlotinib, on the other hand, caused almost one-fold higher ex-
pression of caspase-3 compared to 8. These results further confirm the
apoptotic potential of 8.
Hsp90 inhibition is accompanied by the induction of the chaperone
Hsp70 which is considered a molecular signature for both targets.
Hence, the effect of 8 on Hsp70 was investigated in HepG2 cells. As
shown in Fig. 3, Hsp70 level was 1.6-fold higher in 8-treated HepG2
cells compared to untreated cells. These results further confirm in-
hibition of Hsp90 by 8. Despite the fact that a correlation between an
increase in Hsp70 levels and Hsp90 inhibition was reported in many
previous studies [21,22], the study of Wang et al. [23], suggested that
the increase in Hsp70 could be related to stress induced in cells which
may or may not be due to Hsp90 inhibition. Accordingly, Hsp70 alone
might not be a reliable tool to validate Hsp90 inhibition and hence
investigating the effect of 8 on client proteins was also performed.
3.4. Effect of 8 on Hsp90
In order to investigate the inhibitory potential of 8 on Hsp90 (C-
terminal domain), an in vitro enzyme inhibition assay was performed
where the IC50 calculated was found to be 2.67
0.18 µM. Moreover,
3.6. Effect of 8 on Hsp90 client proteins
in order to investigate whether this effect could be extrapolated to
HepG2 cells, a cell-based ELISA assay was performed to estimate the
level of Hsp90 after treatment with 8. As shown in Fig. 3, HepG2-
treated cells showed a 70% reduction in Hsp90, as compared to control
untreated cells.
Hsp90 is responsible for the functioning of a number of client pro-
teins that are known to control cancer development and progression.
So, inhibition of Hsp90 will ultimately cause degradation of the pro-
survival client proteins. Investigation of the effect of 8 on four Hsp90
Scheme 1. Synthesis of 19.
4

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Table 3
In vitro antiproliferative potential of the newly synthesized first series of compounds against MCF7 and HepG2 cell lines. Data are presented as mean
S.D. of
IC50 values.
Compound
Structure
IC50 (µM)
MCF7
HepG2
97.52
1
2
12.85
0.76
0.26
25.36
1.61
1.1
8.57
0.41
3
4
5
19.82
2.64
0.14
1.20
11.62
0.52
0.03
1.2
0.24
0.002
0.06
0.01
0.23
0.02
6
1.21
7.06
7
8
2.11
0.13
0.41
0.87
0.006
0.083
0.003
9
0.11
0.002
0.64
0.57
0.01
12
19.17
119.09
3.7
13
14
15
137.12
1.00
3.8
0.01
0.61
152.29
2.65
5.1
0.11
2.24
0.74
0.03
16
19
16.30
0.94
0.32
0.03
5.20
6.15
0.25
0.21
Erlotinib (Reference)
1.14
0.01
0.87
0.04
5

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Table 4
These findings suggest a potential degradation of the aforementioned
In vitro antiproliferative potential of the second series of compounds against
client proteins, and further validate Hsp90 inhibition by 8.
HepG2 cell line. Data are presented as mean
S.D. of IC50 values.
Compound
Structure
IC50 (uM)
3.7. Molecular docking
HepG2
5.15
A molecular docking study was carried out in order to predict the
binding mode of compound 8 inside Hsp90 C-Terminal Domain (CTD).
The CTD of human Hsp90 alpha (PDB ID: 3Q6M) [24] was used as a
receptor model for docking. In this crystal structure, missing loop in
Hsp90 alpha (PDB ID: 3Q6M) (616–629) was modeled using Modeller
[25] software which was implemented through “Model Loops/Refine
Structure” utility in UCSF chimera 1.13.1 [26]. Docking search space
was assigned to enclose the interface between the CTD of the two
chains, A and B, (651–698) [27]. Docking simulations were conducted
using two docking engines implementing two different search algo-
rithms (systematic for FRED and stochastic for AutoDock vina) and
scoring functions (Chemgauss4 for FRED [28] and tuned X-score,
combining knowledge-based potentials and empirical scoring functions,
for AutoDock vina [29]). Since the pyrazole hydrogen is tautomeric, in
each docking run, the two ligand pyrazole tautomers, 3-thiophen-2-yl
and 5-thiophen-2-yl, were considered, and are referred to as tautomer1
and tautomer2, respectively, throughout the text. Based on the docking
scores, three poses of docking results were chosen to be subjected to
molecular dynamics (MD) simulations. For FRED results, two types of
binding modes were observed occupying 100% and 60% for tautomer1
and tautomer2, respectively. For vina results, one pose occupied 40%
for tautomer1 which is similar to that of FRED, hence it was cancelled
for being redundant with FRED result for tautomer1, and the other
occupied 30% for tautomer2 and subjected to MD simulations. The
three binding modes are depicted in Fig. 5B. Tautomer1 and tautomer2
poses of FRED, referred to as t1f and t2f, show a hydrogen bond
10
0.21
0.08
0.28
0.11
0.06
11
1.66
9.02
3.25
1.94
17
18
Erlotinib (Reference)
client proteins (Akt, c-Met, c-Raf, and EGFR) was performed in order to
examine whether 8 triggered their degradation in HepG2 cells. Their
levels were all significantly decreased after treatment reaching 87, 90,
77, and 90% for Akt, c-Met, c-Raf, and EGFR, respectively (Fig. 4).
Fig. 2. Effect of 8 on cell cycle progression in HepG2 cell line. (a) Representative histograms for the effect of 8 on cell cycle phases compared to control. (b) Control
and 8-treated HepG2 cells double stained with Annexin V-FITC and propidium iodide (PI) and analyzed by flow cytometry showing living (bottom left), necrotic (top
left), early apoptotic (bottom right), and late apoptotic (top right) cells. (c) Stacked bar chart representing different percentages of cells in each stage of cell cycle
after treatment with 8 compared to control.
6

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
tautomer with more time-evolution stable binding dynamics. The 2D
interaction representations between protein and 8 tautomer1 and tau-
tomer2 are depicted in Fig. S1–4.
3.8. Molecular dynamics
In order to gain more insight into the binding mode of 8 inside
Hsp90 CTD, molecular dynamics simulations for each binding pose
were carried out for 10 ns at NPT ensemble for minimization and
equilibration purposes. Analyses of protein RMSD (Root Mean Square
Deviation), RMSF (Root Mean Square Fluctuation) and Radius of
gyration are depicted in Fig. 6. RMSD is a measure of structure stability.
RMSD of the protein backbone (Cα, C and N) in the 3 dynamics runs
show stability after 7 ns of the simulation with RMSD 3.97
0.73,
4.24
0.99 and 4.34
0.99 Å (mean
SD) for t1f, t2f and t2v,
respectively (Table S1). High RMSD standard deviation (SD) is attrib-
uted to the highly flexible loops present in N-terminal of the protein and
the flexible loops (616–629) modelled by Modeller software (Fig. 5A),
which contributes to the fluctuation of RMSD throughout the simula-
tion.
RMSF is a measure for per residue stability and protein local con-
formational changes during the simulation. Binding site amino acid
residues (651–698 of chain B) show minimal RMSF which is an in-
dication for relatively strong binding with 8 throughout the simulation
and minimal local conformational change in the binding site. Residues
with high RMSF (> 2 Å) showed to be terminal or far away from the
binding site. N-terminal loops of the A and B protein chains (535–554)
and the modelled loops (616–629) show very high RMSF, 6–12 Å,
confirming our observation of their high RMSD SD due to these highly
flexible loops (Fig. 6).
Radius of gyration (Rg) is an indicator of the compactness of the
protein structure and a parameter for protein equilibrium conforma-
tion. As shown in Fig. 6, Rg values of t1f, t2f and t2v, were
43.260
0.182, 43.264
0.184 and 43.249
0.183 Å (mean
SD), respectively (Table S2). Rg is stable for each protein throughout
the simulation, hence, no major conformational changes occur during
the simulation, and this is consistent with our knowledge that major
protein conformational changes rarely occur in the nanosecond time
range.
RMSD analysis of each of the 8 tautomers through evolution of si-
mulation time revealed that t2v (5-thiophen-2-yl tautomer) has less
RMSD than t1f (3-thiophen-2-yl tautomer) and t2f in the three simu-
lations as shown in Fig. 7. The relatively high RMSD in both t1f and t2f
indicates that the ligand abandons its starting position predicted by
docking to a more stable binding pose.
A major finding in the binding interactions was observed that a
hydrogen bond between the pyrazole NeH and Asp661 residue car-
boxylate was formed during the MD simulations that was not predicted
by docking. Fig. 8 shows the total number of hydrogen bonds formed
between the ligand and the protein vs the number of hydrogen bonds
formed between the ligand and Asp661 (Table S3), which indicates the
Asp661 hydrogen bond to be the driving force for such binding.
In Fig. 9, the last snapshot of the dynamics simulations, after 10 ns,
for the three proteins is depicted along with the starting docking pose. It
confirms the stable binding mode of compound 8 inside the binding
pocket with the hydrogen bond between pyrazole NeH of 8 and car-
boxylic group of Asp661 being the main driving force for binding.
In conclusion, the hydrogen bond between 8 and Asp661 carboxylic
group is the main interaction responsible for binding between 8 and
Hsp90 CTD. From molecular dynamics simulations, it is concluded that
Asp661 is responsible for the hydrogen bond with 8 not Arg690 pre-
dicted by docking (Supplementary information, videos 1, 2 and 3).
Video 1, video 2 and video 3 trajectory analysis of t1f, t2f and t2v,
respectively, show the stable binding due to the hydrogen bonding in-
teraction with the key binding residues (Asp661) with less deviation of
t2v.
Fig. 3. Effect of 8 on caspase-3, Hsp90, and Hsp70 levels. (a) Caspase-3, (b)
Hsp90, and (c) Hsp70 protein levels were determined in 8-treated HepG2 cells.
Protein levels were estimated using ELISA. Values are presented as
means
S.D. from three independent experiments performed in triplicates.
*P < 0.05 significant from control untreated or erlotinib-treated cells using
one-way ANOVA followed by Bonferroni post hoc test for caspase-3 and
Student′s t test for Hsp90 and Hsp70.
between the nitrogen of the pyrazole ring, which acts as a hydrogen
bond acceptor, and the guanidine moiety of arginine 690 amino acid of
chain B, which acts as a hydrogen bond donor, with docking scores of
−11.2470 and −10.5955, respectively. Tautomer2 pose predicted by
vina, referred to as t2v, shows no hydrogen bonds with the protein,
with free energy of −6.7 kcal/mol.
Molecular docking was followed by molecular dynamics (MD) si-
mulations to give further insights into the binding mode of 8 and the
7

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Fig. 4. Effect of 8 on HSP90 client proteins. Protein levels of (a) p-Akt, (b) c-Met, (c) c-Raf, and (d) EGFR were estimated in 8-treated HepG2 cells. Protein levels were
estimated using ELISA. Values are presented as means
S.D. from three independent experiments performed in triplicates. *P < 0.05 significant from control
untreated cells using Student′s t test.
4. Experimental
Samples for high resolution positive ion electrospray ionization mass
spectrometry (HRMS-ESI+) were prepared in 1:1 CH3CN/H2O + 0.2%
formic acid. Melting points are measured using Stuart SMP 30 and are
uncorrected.
4.1. General
All reagents and solvents were purchased from commercial sup-
pliers and used without purification unless stated otherwise. Chemical
shifts (δ) for 1H NMR spectra in acetone-d6 are reported in ppm relative
to acetone-d6 residual solvent protons (δ 2.05). Chemical shifts for 13C
NMR spectra run in acetone-d6 are reported in ppm relative to the
solvent residual carbon (δ 30.8). Peaks in NMR spectra are described as
follows: singlet (s), broad singlet (bs), doublet (d) triplet (t) apparent
doublet (appd), apparent triplet (appt), doublet of doublets (dd).
4.2. General method for the synthesis of compounds 17 and 18
To a mixture of 2-acetyl-thiophene hydrazone (1.5 mmol, 210 mg)
and the corresponding aldehyde (1.8 mmol) in ethanol (10 mL), added
HCl (75 µL). The reaction mixture was refluxed for one hour, then
DMSO (6 mmol, 425 μL) and iodine (0.15 mmol, 38 mg) were added.
The reaction mixture was refluxed for 16 h, cooled to room
Fig. 5. (A) CTD of Hsp90 (3q6m) with the flexible N-terminal (red) and the loops modelled with Modeller software (blue). (B) Molecular docking results showing the
two binding poses predicted by FRED (green for t1f and blue for t2f) and AutoDock vina (grey for t2v). The docked pose predicted by FRED shows a hydrogen bond
between the pyrazole nitrogen and Arginine 690 in chain B. (Chain A is depicted in gold and chain B in grey).
8

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Fig. 6. RMSD (top), per residue RMSF (middle) and Radius of gyration (bottom) of the three protein structures (t1f, t2f and t2v). RMSF x-axis depicts the number of
protein chain A 163 residues (535–697) and chain B 165 residues (698–862). Dark green, dark blue and dark orange for t1f, t2f and t2v, respectively. N-terminal (red
braces) and modelled (blue braces) highly flexible loops of chains A and B show high RMSF (middle) and contributes to the RMSD (top) fluctuation. The binding site
of chain B (black braces) shows low RMSF < 2 which is an indication for relatively strong binding with 8 throughout the simulation and minimal local con-
formational change in the binding site.
temperature, quenched with 5% aqueous sodium thiosulfate (ca
30 mL), and extracted with EtOAc (3 × 20 mL). The combined EtOAc
extracts were washed with water (1 × 20 mL), dried over anhydrous
magnesium sulfate and concentrated by rotary evaporation. The crude
product was purified by column chromatography using a gradient of
30–50% EtOAc-Hexane. The desired fractions were pooled, con-
centrated and vacuum dried.
4.4. 5-(2-fluoro-4-methoxyphenyl)-3-(thiophen-2-yl)-1H-pyrazole (18)
Yellow solid, 329 mg, 80%, Mp 138-140oC. 1H NMR (300 MHz,
acetone-d6): δ 12.4 (s, 1H), 7.81 (t, J = 9.2 Hz, 1H), 7.44 (d, J = 4.6 Hz,
1H), 7.39 (d, J = 4.9 Hz, 1H), 7.08 (d, J = 3.6, 5.3 Hz, 1H), 6.84–6.88
(m, 3H); 13C NMR (75 MHz, DMSO-d6):
δ 160.5; 159.2 (d,
J = 243.1 Hz), 146.6, 137.4, 136.8, 128.4, 127.6, 124.6, 123.7, 111.0,
102,1 (d, J = 26.8 Hz); 101.1, 55.7; 19F NMR (282 MHz, acetone-d6): δ-
114.2 HRMS-ESI+ (m/z): [M+H]+ calculated for C14H12ON2SF,
275.0649; found,. 275.0647.
4.3. 5-(3,4-dimethoxyphenyl)-3-(thiophen-2-yl)-1H-pyrazole (17)
Yellowish white solid. 343 mg, 80%, Mp 159-160 °C. 1H NMR (300
MHz, acetone-d6): δ 12.41 (bs, 1H), 7.36-7.44 (m, 4H), 7.09 (t, J = 3.6
Hz, 1H), 7.03 (d, J = 8.2 Hz, 1H), 6.92 (s, 1H), 3.89 (s, 3H), 3.85 (s,
3H); 13C NMR (75 MHz, DMSO-d6): δ 149.4,147.2, 144.0, 137.5, 128.0,
125.0, 123.9, 122.3, 118.1, 112.4, 109.4, 99.2, 55.9; HRMS-ESI+ (m/
z): [M+H]+ calculated for C15H15O2N2S, 287.0849; found, 287.0848.
4.5. General method for the synthesis of compounds 10 and 11
To a mixture of 2-acetyl-thiophene (1.5 mmol, 189 mg) and the
corresponding aldehyde hydrazone (1.8 mmol) in ethanol (10 mL), was
added HCl (75 µL). The reaction mixture was refluxed for one hour,
9

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Fig. 7. RMSD analysis of 8 tautomers with each protein through the simulation time.
then DMSO (6 mmol, 425 μL) and iodine (0.15 mmol, 38 mg) were
added. The mixture was refluxed for 16 hours, cooled to room tem-
perature, quenched with 5% aqueous sodium thiosulfate (ca 30 mL),
and extracted with EtOAc (3 × 20 mL). The combined EtOAc extracts
were washed with water (1 × 20 mL), dried over anhydrous magne-
sium sulfate and concentrated by rotary evaporation. The crude product
was purified by column chromatography using a gradient of 30–50%
EtOAc-Hexane. The desired fractions were pooled, concentrated and
vacuum dried.
4.7. 3-(thiophen-2-yl)-5-(4-(trifluoromethyl)phenyl)-1H-pyrazole (11)
Yellow solid, 353 mg, 80%, Mp 224-226oC. 1H NMR (300 MHz,
acetone-d6): δ 12.76 (bs, 1H), 8.10 (d, J = 6.9 Hz, 1H), 7.80 (d, J = 7.9
Hz, 1H), 7.50 (bs, 1H), 7.13 (bs, 1H); 13C NMR (75 MHz, DMSO-d6): δ
147.7, 142.4, 138.6, 133.3, 128.6, 128.1, 126.7, 126.5, 126.1, 125.4,
125.2, 124.4, 101.1; 19F NMR (282 MHz, acetone-d6): δ – 63.3 HRMS-
ESI+ (m/z): [M+H]+ calculated for C14H10N2SF3, 295.0511 ; found,
295.0512.
4.8. Cell lines and cell culture
4.6. 3-(3-(thiophen-2-yl)-1H-pyrazol-5-yl)phenol (10)
MCF7, estrogen-dependent breast cancer cell line, and HepG2, he-
patocellular carcinoma cell line, were obtained from the American Type
Culture Collection (ATCC, VA, USA). Cells were cultured using
Dulbecco's Modified Eagle Medium (DMEM; Invitrogen, Life
Technologies, CA, USA) supplemented with 10% fetal bovine serum
(FBS; GE Healthcare HyClone, WV, USA), streptomycin (100 μg/ml)
and penicillin (100 units/ml) and incubated in 5% CO2 at a temperature
of 37 °C. Passaging was performed when cells were 80–90% confluent.
Yellow solid, 273 mg, 75%, Mp 179–181 °C. 1H NMR (300 MHz,
acetone-d6): δ 12.45 (bs, 1H), 8.48 (bs, 1H), 7.43 (d, J = 3.6 Hz, 1H),
7.39 (d, J = 4.6 Hz, 1H), 7.22-7.28 (m, 3H), 7.08 (t, J = 4.6 Hz, 1H),
6.91 (s, 1H), 6.83 (d, J = 6.9 Hz, 1H); 13C NMR (75 MHz, DMSO-d6): δ
157.7, 146.8, 143.6, 136.9, 130.0, 127.6, 124.6, 123.7, 116.1, 115.3,
99.3; HRMS-ESI+ (m/z): [M+H]+ calculated for C13H11ON2S,
243.0587; found, 243.0584.
10

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Fig. 8. Number of hydrogen bonds between 8 tautomers, and the three protein structures (left), and Asp661 (right), as a function of simulation time.
4.9. Cell viability assay
overnight and on the next day, cells were treated with serial dilutions of
the compounds besides DMSO. Then after seventy-two hours, 20 μl of
5 mg/ml MTT was added to each well then incubated for 2 h. The ab-
sorbance was then measured at 570 nm. The percentage of the absor-
bance at 570 nm in treated cells against control represented the cell
viability. All experiments were performed in triplicates. Half-maximal
inhibitory concentration (IC50) was calculated using GraphPad Prism
software, version 5.00 (GraphPad Software, CA, USA). For further cell
cycle analysis and cell-based assays, HepG2 cells were treated with the
IC50 of compound 8.
The first series of compounds were tested against both MCF7 and
HepG2 cells. Following this screening step, optimization of compounds
was performed where the second series was tested against HepG2 cells
only. MCF7 and HepG2 cells were seeded in 96-well plates at a density
of 1 × 104 cells/well. One day after seeding plates, MCF7 and HepG2
cells were treated with a 10 fold serial dilution (0.01–100 μM) of the
first series; (compounds 1–9, 12–16, and 19) or Dimethyl sulfoxide
(DMSO; Sigma-Aldrich, St. Louis, MO, USA) as a vehicle representing
the negative control. Erlotinib was used as a reference. As for the
second optimized series of compounds (10, 11, 17, 18), their cytotoxic
effect was investigated against HepG2 cells only using their 10-fold
serial dilutions (0.01–100 μM). Cytotoxicity assay of compounds be-
longing to both series was assessed using 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (MTT) reagent (Sigma-Aldrich, St.
Louis, MO, USA). Briefly, cells seeded in 96-well plates were cultured
4.10. Apoptosis assay using Annexin V/Propidium iodide staining
HepG2 cells cultured in six-well plates were treated with 8 for 48
hours. Then, the cells were fixed in cold ethanol for half an hour and
stained with 5 μl Annexin V-FITC and 5 μl propidium iodide (PI) using
an Annexin VFITC Apoptosis Detection Kit (BioVision, CA, USA). The
11

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
Fig. 9. Comparison between the starting protein structure (chain A in dark grey and chain B in white) with starting ligand pose (khaki) and the last snapshot after
10 ns molecular dynamics run (chain A in dark green and chain B in lime green) with ligand pose (salmon) for t1f (top), t2f (middle) and t2v (bottom). The three runs
show a conserved hydrogen bond with Asp661 which is considered to be the driving force for protein ligand interaction.
cells were then placed at room temperature for 15 min in the dark then
analyzed by a FACScan flow cytometer (Beckman Coulter, CA, USA).
Apoptosis was evaluated in terms of the FITC-positive cells.
were resuspended in PBS and filtered. Cell cycle analysis was then
performed using flow cytometer.
4.11. Cell cycle analysis
4.12. Determination of caspase-3 in 8-treated HepG2 cells
Forty-eight hours following the treatment of HepG2 with 8, cells
were trypsinized and washed with phosphate buffered saline (PBS),
resuspended in cold methanol, and kept overnight at 4 °C. Collected
cells were then resuspended in sodium citrate buffer together with
RNase, and incubated at 37 °C for 30 min. After centrifugation, cells
In order to determine the 8 apoptotic potential, active caspase-3
level was measured in cell lysates using human caspase-3 ELISA kit
obtained from Invitrogen (CA, USA), according to the manufacturer’s
instructions. Erlotinib was included in the assay as a reference.
12

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
4.13. In vitro enzyme inhibition assay against Hsp90
rcsb.org/) [33]. Receptors and ligands were prepared using pre-
pare_receptor4.py and prepare_ligands4.py scripts, respectively, im-
plemented through AutoDock Tools 1.5.6 [34]. Protein and ligand
preparations involved adding polar hydrogen atoms and Gasteiger
partial charges. Single bonds of the ligands were considered rotatable
while the binding site residues were considered rigid during the
docking calculations. The search box was centered around the co-
crystallized ligands and assigned 60 Å box side and 0.375 Å grid spa-
cing. Box dimensions determination, writing AutoDock vina config-
uration files and docking automation were carried out using AutoDock/
vina pymol plugin [35]. Docking calculations were carried out using
AutoDock vina docking engine [29]. Crystallized water molecules were
removed from protein before docking. AutoDock implements La-
marckian Genetic search Algorithm (LGA) where LGA runs were set to a
value of 100 with 150 population size, 2,500,000 evaluations and
27,000 generations returning the best binding pose in the largest
cluster. Binding energies are evaluated in terms of difference in Gibbs
free energy (∆G).
The most promising compound, 8, was selected for in vitro enzyme
inhibition screening assay against Hsp90 C-terminal domain using the
Hsp90 (C-terminal) inhibitor screening assay kit (BPS Bioscience, USA),
according to the manufacturer’s instructions, in order to determine the
IC50 of 8. Briefly, wells of a microtiter plate were designated to sub-
strate control, positive control, test inhibitor and blank. Aliquots of
Hsp90 and its protein target (cyclophilin D), provided in the kit, were
thawed on ice and diluted with assay buffers. Then, 4 µl of diluted Hsp
α protein were added to each well designated for the positive control,
test inhibitor and blank. To the wells designated for substrate control, 4
µl of Hsp90 assay buffer were added. Then, 2 µl of geldanamycin so-
lution were added to wells designated for test inhibitor. For positive
and substrate control wells, 2 µl of the same solution without inhibitor
(inhibitor buffer) were added. Moreover, 4 µl of HspP90 assay buffer
were added to blank wells. In order to initiate the reaction, cyclophilin
D was diluted in assay buffer from which 4 µl were added to wells
designated for substrate and positive controls as well as for test in-
hibitor. The plate was then incubated at room temperature for 30 min.
Acceptor beads were added, then streptavidin-conjugated donor beads,
followed by reading alpha-counts.
4.17.3. FRED Docking
Receptors were prepared using Make Receptor graphical utility and
pdb2receptor utility program of OEDOCKING 3.2.0.2 of OpenEye
Scientific Software [28]. Search box was determined around the co-
crystallized ligands after cavity detection step using a molecular probe
implemented through Make Receptor followed by a step of site shape
potential detection. FRED docking engine requires a multiconformer
ligand OpenEye binary file, so ligands were prepared using OpenEye
OMEGA 3.0.1.2 conformer generation program [36]. FRED applies a
systematic exhaustive search algorithm through rotational and trans-
lational degrees of freedom and evaluates the binding interactions using
consensus Chemgauss4 scoring function.
4.14. Determination of Hsp90 and Hsp90 client proteins in 8-treated
HepG2 cells
The effect of 8 on Hsp90 and its client proteins; Akt, the hepatocyte
growth factor (c-Met), c-Raf, and epidermal growth factor receptor
(EGFR) was investigated using ELISA kits for human Hsp90 (Cloud-
Clone Corp., USA), phospho-Akt (DRG International, USA), c-Met
(Sigma Aldrich, USA), phospho-c-Raf (Cisbio, France), and EGFR
(Raybiotech, USA), according to the manufacturers’ instructions. For
preparing cell lysates, cells were trypsinized and centrifuged following
72 h of treatment with 8. The cell pellet was then washed with PBS and
RIPA lysis buffer was used for cell lysis. Centrifugation was then per-
formed to remove cell debris. Cell lysates were then collected and
stored at −880 °C for performing the aforementioned ELISA assays.
All ligands and protein visualizations and figures generation were
carried out using UCSF Chimera [26]. 2D interaction representation is
generated by OpenEye toolkits [37].
4.17.4. Molecular Dynamics Simulations
The NAMD 2.12 package [38] was used for all the molecular dy-
namics simulation experiments. All system preparations were carried
out using VMD 1.9.3 graphics software [39]. First, the protein molecule
was stripped of any non-protein atoms (ligands, water and ions). Pro-
tein and ligand topology files were generated using VMD psfgen 1.6.4
package and SwissParam web server [40], respectively. The simulation
systems were constructed manually by incorporating the topology files,
solvating the protein-ligand complex in a box of TIP3P water solvent
model with a 12 Å margin in all directions, using VMD solvate 1.7
package, and neutralizing the system charge with NaCl molecules, using
VMD autoionize 1.4 package, ending up with systems of approximately
35,000 atoms. Periodic boundary conditions were considered during
the simulations, wrapping all system atoms; by translating atoms that
pass the periodic boundary to its mirror position on the opposite side of
the cell.
4.15. Determination of Hsp70 in 8-treated HepG2 cells
To determine the effect of 8 on Hsp70, human Hsp70 ELISA kit
(Cloud-Clone Corp., USA) was used according to the manufacturer’s
instructions. The assay was performed on cell lysates that were pre-
pared, as explained earlier.
4.16. Statistical analysis
All values are presented as means
standard deviation (S.D.) from
three independent experiments performed in triplicates. Statistical
analysis was performed by Student t test for unpaired data and One-
Way Analysis of variance (ANOVA) followed by Bonferroni post hoc test
for multiple comparisons using GraphPad Prism, version 5.0 (GraphPad
Software, CA, USA) (v5). Statistical significance was determined at
P < 0.05.
100,000 minimization steps were carried out coupling the NAMD
rigorous conjugate gradient algorithm with line search algorithm, fol-
lowed by 10 ns NPT equilibration.
CHARMM36 Force Field [41] (July 2017 release) was used for
providing protein parameters, whereas SwissParam parameter file was
considered for ligands. The SwissParam derives the ligand atoms data
from MMFF (Merck Molecular Force Field) [42], while Van de Waals
parameters are taken from the closest atom type in CHARMM22 [41].
During dynamics calculations, any 1–2 and 1–3 atom interactions
were neglected, and 1–4 interactions are fully considered. Van der
Waals and electrostatic interactions are searched for in the system
within a pair list distance of 13.5 Å, which is restarted once every in-
tegration cycle, and calculated with a 12 Å cut-off implementing
function smoothing at 10 Å switch distance. For all runs, the velocity
Verlet integration method was used for velocity and position
4.17. Molecular Modeling
4.17.1. Ligands preparation
The two 8 tautomers were constructed using Marvin Sketch 2D
sketcher [30] and minimized using OpenBabel 2.4.1 [31] via Steepest
Descent minimization algorithm for 100,000 steps and a convergence
criterion of 10−6 kcal/mol/Å implementing MMFF94s (Merck Mole-
cular Force Field static variant) [32] for stepwise energy calculations.
4.17.2. AutoDock vina docking
Receptors were fetched from PDB online database (http://www.
13

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
calculations with Particle Mesh Ewald Sum (PME) method for electro-
statics calculations [43], and a 2 fs time step. 1 step size (2 fs) was set as
frequency for non-bonded and full electrostatic interactions calcula-
tions. System temperature was kept constant at 310 K (37 °C) by con-
trolling Kinetic Energy of the system using Langevin Dynamics with
damping coefficient 1 ps−1 applying it to only heavy atoms. Pressure
was also kept constant at 1 atm (1.01325 bar) using Langevin piston at
310 Kelvin with oscillation and damping periods of 200 and 50 fs, re-
spectively, coupled with Nosé-Hoover method to control barostat
fluctuations [44].
[5] (a) For a complete list of Hsp90 interactoms see online comprehensive database of
the Picard lab available at http://www.picard.ch/downloads/HSP90interactors.
pdf. (b) See also, Advances in Cancer Research, Hsp90 in Cancer: Beyond the usual
suspects. J. Isaacs, L. Whitesell 129 (2016), pp 1-20, Academic press.
[6] (a) I. Koca, A. Ozgur, M. Er, M. Gumus, K.A. Cosun, Y. Tutar, Design and synthesis
of pyrimidinyl acyl thioureas as novel Hsp90 inhibitors in invasive ductal breast
cancer and its bone metastasis, Eur. J. Med. Chem. 122 (2016) 280–290;
(b) C. Didelot, D. Lanneau, M. Brunet, A.L. Joly, A. De Theonel, G. Chiosis,
C. Garrido, Anticancer therapeutic approaches based on intracellular and extra-
cellular heat shock proteins, Curr. Med. Chem. 14 (2007) 2839–2847.
[7] (a) P. Workman, Combinatorial attack on multistep oncogenesis by inhibiting the
Hsp90 molecular chaperone, Cancer Lett. 206 (2004) 149–157;
(b) J.R. Peter, C.C. Fritz, K.M. Depew, Discovery and development of Hsp90 in-
hibitors: a promising pathway for cancer therapy, Curr. Opin. Chem. Biol. 14 (2010)
412–420.
All dynamics simulations analysis was carried out using ProDy
Python package [45] and VMD RMSD trajectory tool [39]. All analysis
charts were constructed using Matplotlib 3.0.2 Python plotting library
[46].
[8] A. Kamal, L. Thao, J. Sensintaffar, L. Zhang, M.F. Boehm, L.C. Fritz, F.J. Burrows, A
high affinity conformation of Hsp90 confers tumour selectivity on Hsp90 inhibitors,
Nature 425 (2003) 407–410.
[9] (a) C. Prodromou, The active life of Hsp90 complexes, Biochem. Biophys. Acta
2012 (1823) 614–623;
5. Conclusions
(b) A. Rohl, J. Rohrberg, J. Buchner, The chaperone Hsp90: changing partners for
demanding clients, Trends Biochem. Sci. 38 (2013) 253–262.
[10] (a) T. Taldone, W. Sun, G. Chiosis, Discovery and development of heat shock
protein 90 inhibitors, Biorg. Med. Chem. 17 (2009) 2225–2235;
(b) J. Trepel, M. Mollapur, G. Giaccone, L. Neckers, Targeting the dynamic Hsp90
complex in cancer, Nat. Rev. Cancer 10 (2010) 537–549;
A series of 3,5-diarylpyrazoles has been synthesized and biologically
evaluated against HepG2 and MCF7 cancer cell lines. Compound 8
exhibited cytotoxic potential against both cell lines yet a lower IC50 was
reported on HepG2 cells. Hence, this compound was selected for further
structure optimization and biological evaluation. Compound 8 caused
cell cycle arrest at the G2 phase and showed apoptotic potential as
evidenced by the increase in caspase-3 levels compared to control.
Moreover, compound 8 exhibited potent inhibition of Hsp90 which was
further validated by the degradation of the pro-survival client proteins
(Akt, c-Met, c-Raf, and EGFR) and the increase in Hsp70 levels.
Molecular docking and dynamics simulations were employed to predict
the binding mode of 8.
(c) J.H. Jeong, Y.J. Oh, Y. Lho, S.Y. Park, K.H. Liu, E. Ha, Y.H. Seo, Targeting the
entry region of Hsp90’s ATP binding pocket with a novel 6,7-dihydrothieno [3,2-c]
pyridine-5(4H)-yl amide, Eur. J. Med. Chem. 124 (2016) 1069–1080;
(d) A. Bargioti, L. Musso, S. Dallavalle, L. Merlini, G. Gallo, A. Ciacci, G. Giannini,
W. Cabri, S. Penco, L. Vesci, M. Castorina, F.M. Milazzo, M.L. Cervoni,
M. Barbarino, C. Pisano, C. Giommarelli, V. Zuco, M. De Cesare, F. Zunino,
Isoxazolo-(aza)-naphthaquinones: a new class of cytotoxic Hsp90 inhibitors, Eur. J.
Med. Chem. 53 (2012) 64–75;
(e) D. Chen, A. Shen, J. Li, F. Shi, W. Chen, J. Ren, H. Liu, Y. Xu, X. Wang, X. Yang,
Y. Sun, M. Yang, J. He, Y. Wang, L. Zhang, M. Huang, M. Geng, B. Xiong, J. Shen,
Discovery of potent N-(isoxazolyl-5-yl)amides as Hsp90 inhibitors, Eur. J. Med.
Chem. 87 (2014) 765–781;
(f) D. Montoir, S. Barilli-Nion, A. Tonnerre, P. Juin, M. Duflos, M.A. Bazin, Novel
1,6-napthyridine-2(1H)-ones as potential anticancer agents targeting Hsp90, Eur. J.
Med. Chem. 119 (2016) 17–33;
Declaration of Competing Interest
The authors have declared no conflict of interest.
Acknowledgements
(g) F. Jiang, A.P. Guo, J.C. Xu, H.J. Wang, X.F. Mo, Q.D. You, X.L. Xu,
Identification and optimization of novel 6-acylamino-2-aminoquinolines as poten-
tial Hsp90 C-terminal inhibitors, Eur. J. Med. Chem. 141 (2017) 1–14;
(h) C. Liang, X. Wu, Z. Li, J. Zhu, C. Lu, Y. Shen, Design, synthesis, and pharma-
cological evaluation of N-(5-chloro-2,4-dihydroxybenzoyl)-(R)-N-arylmethyl-
1,2,3,4-tetrahydro-3-isoquinoline carboxamides as potent Hsp90 inhibitors, Eur. J.
Med. Chem. 143 (2018) 85–96;
This work was supported by The Center for Drug Research and
Development (CDRD), Faculty of Pharmacy, The British University in
Egypt, Grant 2019, and via a Natural Sciences and Engineering
Research Council (NSERC) of Canada Discovery Grant (RGPIN-2017-
04233) to SDT.
(i) S.Y. Park, Y.J. Oh, Y. Lho, J.H. Jeong, K.H. Liu, J. Song, S.H. Kim, E. Ha,
Y.H. Seo, Design, synthesis, and biological evaluation of a series of resorcinol-based
N-benzyl benzamide derivatives as potent Hsp90 inhibitors, Eur. J. Med. Chem. 143
(2018) 390–401;
(j) Q. Wei, J.Y. Ning, X. Dai, Y.D. Gao, L. Su, B.X. Zhao, Discovery of novel Hsp90
inhibitors that induced apoptosis and impaired autophagic flux in A549 lung cancer
cells, Eur. J. Med. Chem. 145 (2018) 551–558;
We gratefully acknowledge the support of OpenEye Scientific
Software Inc. for offering an academic license.
(k) R. Ojha, H.L. Hung, W.C. Huangfu, Y.W. Wu, K. Nepali, M.J. Lai, C.J. Su,
T.Y. Sung, Y.L. Chen, S.L. Pan, J.P. liou, 1-Aroylindoline-hydroxamic acids as an-
ticancer agents, inhibitors of Hsp90 and HDAC, Eur. J. Med. Chem. 150 (2018)
667–677;
We also gratefully acknowledge the support of NVIDIA Corporation
with the donation of the Titan Xp GPU used for molecular dynamics
calculations.
(l) X. Chen, P. Liu, Q. Wang, Y. Li, L. Fu, H. Fu, J. Zhu, Z. Chen, W. Zhu, C. Xie,
L. lou, DCZ, 3112, a novel Hsp90 inhibitor, exerts potent antitumor activity against
HER2-positive breast cancer through disruption of Hsp90-Cdc37 interaction, Cancer
Lett. 434 (2018) 70–80;
Appendix A. Supplementary material
(m) P.A. Brough, X. Barril, M. Beswick, B.W. Dymock, M.J. Drysdale, L. Wright,
K. Grant, A. Massey, A. Surgenor, P. Workman, 3-(5-chloro-2,4-dihydroxyphenyl)-
pyrazol-4-carboxamides as inhibitors of the Hsp90 molecular chaperone, Biorg.
Med. Chem. Lett. 15 (2005) 5197–5201.
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bioorg.2019.103433.
[11] K. Sidera, E. Patsavoudi, Hsp90 inhibitors: current development and potential in
cancer therapy, Recent Pat. Anticancer Drug Discov. 9 (2014) 1–20.
[12] A. Ozgur, Y. Tutar, Heat shock protein 90 inhibitors in oncology, Curr. Proteomics
11 (2014) 2–16.
References
[1] A. Forner, M. Reig, J. Bruix, Hepatocellular carcinoma, Lancet 391 (2018)
1301–1314.
[13] L. Whitesell, E.G. Mimnaugh, B. De Costa, C.E. Myers, L.M. Neckers, Inhibition of
heat shock protein Hsp90-pp60-src heteroprotein complex formation by benzo-
quinone ansamycins: essential role for stress proteins in oncogenic transformation,
PNAS 91 (1994) 8324–8328.
[2] J.M. Llovet, S. Ricci, V. Mazzaferro, P. Hilgard, E. Gane, J.F. Blanc, A.C. de Oliveira,
A. Santoro, J.L. Raoul, A. Forner, M. Schwartz, C. Porta, S. Zeuzem, L. Bolondi,
T.F. Greten, P.R. Galle, J.F. Seitz, I. Borbath, D. Häussinger, T. Giannaris, M. Shan,
M. Moscovici, D. Voliotis, J. Bruix, Sorafenib in advanced hepatocellular carcinoma,
N. Engl. J. Med. 359 (2008) 378–390.
[14] S.M. Roe, C. Prodromou, R. O’ Brien, J.E. Ladburry, P.W. Piper, L.H. Pearl,
Structural basis for inhibition of the Hsp90 molecular chaperone by the antitumor
antibiotics radicicol and geldanamycin, J. Med. Chem. 42 (1999) 260–266.
[15] L. Neckers, P. Workman, Hsp90 molecular chaperone inhibitors: are we there yet?
Clin. Cancer Res. 18 (2012) 64–76.
[3] L. Whitesell, S.L. Lindquist, Hsp90 and the chaperoning of cancer Nat, Rev. Cancer
5 (2005) 761–1722.
[4] (a) A. Maloney, P. Workmann, Hsp90 as a new therapeutic target for cancer
therapy: the story unfolds, Expert Opin. Biol. Ther. 2 (2002) 3–24;
(b) R.E. Wang, Targeting heat shock proteins 70/90 and proteasome for cancer
therapy, Curr. Med. Chem. 18 (2011) 4250–4264;
[16] S.R. Kasibhatla, K. Hong, M.A. Biamonte, D.J. Busch, P.L. Karjian, J.L. Sensintaffar,
A. Kamal, R.E. Lough, J. Brekken, K. Lundgren, R. Grecko, G.A. Timony, Y. Ran,
R. Mansfield, L.C. Fritz, E. Ulm, F.J. Burrows, M.F. Boehm, Rationally designed
high-affinity 2-amino-6-halopurine heat shock protein 90 inhibitors that exhibit
potent antitumor activity, J. Med. Chem. 50 (2007) 2767–2778.
[17] B.W. Dymock, X. Barril, P.A. Brough, J.E. Cansfield, A. Massey, E. McDonald,
(c) F.U. Hartl, Molecular chaperones in cellular protein folding, Nature 381 (1996)
571–579.
14

S. Mohamady, et al.
BioorganicChemistryxxx(xxxx)xxxx
R.E. Hubbard, A. Surgenor, S.D. Roughley, P. Webb, P. Workman, L. Wright,
[33] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig,
M.J. Drysdale, Novel, potent small-molecule inhibitors of the molecular chaperone
Hsp90 discovered through structure-based design, J. Med. Chem. 48 (2005)
4212–4215.
I.N. Shindyalov, P.E. Bourne, The protein data bank, Nucl. Acids Res. 28 (2000)
235–242.
[34] (a) M.F. Sanner, Python: a programming language for software integration and
development, J. Mol. Graphics Mod. 17 (1999) 57–61;
[18] (a) P.A. Brough, W. Aherne, X. Barril, J. Borgognoni, K. Boxall, J.E. Cansfield,
K.M.J. Cheung, I. Collins, N.G.M. Davies, M.J. Drysdale, B. Dymock, S.A. Eccles,
H. Finch, A. Fink, A. Hayes, R. Howes, R.E. Hubbard, K. James, A.M. Jordan,
A. Lockie, V. Martins, A. Massey, T.P. Mattews, E. McDonald, C.J. Northfield,
L.H. Pearl, C. Prodromou, S. Ray, F.I. Raynaud, S.D. Roughley, S.Y. Sharp,
A. Surgenor, D.L. Walmsey, P. Webb, M. Wood, P. Workman, L. Wright, 4,5-
Diarylisoxazole Hsp90 chaperone inhibitors: potential therapeutic agents for the
treatment of cancer, J. Med. Chem. 51 (2008) 196–218.
(b) G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,
A.J. Olson, Autodock4 and autodocktools4: automated docking with selective re-
ceptor flexiblity, J. Comput. Chem. 16 (2009) 2785–2791.
[35] D. Seeliger, B.L. de Groot, Ligand docking and binding site analysis with PyMOL
and autodock/vina, J. Comput. Aided Mol. Des. 24 (2010) 417–422.
[36] OMEGA 3.0.1.2: OpenEye scientific software, Santa Fe, NM. (http://www.
eyesopen.com). P.C.D. Hawkins, A.G. Skillman, G.L. Warren, B.A. Ellingson, M.T.
Stahl, Conformer generation with OMEGA: algorithm and validation using high
quality structures from the protein databank and the Cambridge structural database
J. Chem. Inf. Model. 50 (2010) 572-584.
[19] K. Jhaveri, T. Taldone, S. Modi, G. Chiosis, Advances in the clinical development of
heat shock protein 90 (Hsp90) inhibitors in cancers, Biochem. Biophys. Acta 2012
(1825) 742–755.
[20] S. Mohamady, B. Kralt, S.K. Samwel, S.D. Taylor, Efficient one-pot, two-component
modular synthesis of 3,5-disubstituted pyrazoles, ACS Omega 3 (2018)
15566–15574.
[37] OpenEye Toolkits 2017.Oct.1 OpenEye scientific software, Santa Fe, NM, http://
www.eyesopen.com.
[38] J.C. Phillips, R. Braun, W. Wang, J. Gumbart, E. Tajkhorshid, E. Villa, C. Chipot,
R.D. Skeel, L. Kale, K. Schulten, Scalable molecular dynamics with NAMD, J.
Comput. Chem. 26 (2005) 1781–1802.
[21] J.H. Jeong, Y.J. Oh, Y. Lho, S.Y. Park, K.H. Liu, E. Ha, Y.H. Seo, Targeting the entry
region of Hsp90's ATP binding pocket with a novel 6,7-dihydrothieno[3,2-c]pyr-
idin-5(4H)-yl amide, Eur. J. Med. Chem. 124 (2016) 1069–1080.
[22] R. Ojha, H.L. Huang, W.C. HuangFu, Y.W. Wu, K. Nepali, M.J. Lai, C.J. Su,
T.Y. Sung, Y.L. Chen, S.L. Pan, J.P. Liou, 1-Aroylindoline-hydroxamic acids as an-
ticancer agents, inhibitors of HSP90 and HDAC, Eur. J. Med. Chem. 150 (2018)
667–677.
[39] W. Humphrey, A. Dalke, K. Schulten, VMD – visual molecular dynamics, J. Molec.
Graphics 14 (1996) 33–38.
[40] V. Zoete, M.A. Cuendet, A. Grosdidier, O. Michielin, SwissParam, a fast force field
generation tool for small organic molecules, J. Comput. Chem. 32 (2011)
2359–2368.
[23] Y. Wang, Y.C. Koay, S.R. McAlpine, Redefining the phenotype of heat shock protein
90 (Hsp90) inhibitors, Chem. Eur. J. 23 (2017) 2010–2013.
[41] (a) R.B. Best, X. Zhu, J. Shim, P.E.M. Lopes, J. Mittal, M. Feig, A.D. MacKerell Jr.,
Optimization of the additive CHARMM all-atom protein force field targeting im-
proved sampling of the backbone phi, psi and side-chain chi1 and chi2 dihedral
angles, J. Chem. Theory Comput. 8 (2012) 3257–3273;
[24] C.-C. Lee, T.-W. Lin, T.-P. Ko, A.H.-J. Wang, The hexameric structures of human
heat shock protein 90, PLoS ONE 6 (2011) e19961.
[25] A. Šali, T.L. Blundell, Comparative protein modelling by satisfaction of spatial re-
straints, J. Mol. Biol. 234 (1993) 779–815.
(b) A.D. MacKerell Jr., M. Feig, C.L. Brooks III, Extending the treatment of back-
bone energetics in protein force fields: limitations of gas-phase quantum mechanics
in reproducing protein conformational distributions in molecular dynamics simu-
lations, J. Comput. Chem. 25 (2004) 1400–1415;
[26] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng,
T.E. Ferrin, UCSF chimera – a visualization system for exploratory research and
analysis, J. Comput. Chem. 25 (2004) 1605–1612.
(c) A.D. MacKerell Jr., D. Bashford, M. Bellott, R.L. Dunbrack Jr., J.D. Evanseck,
M.J. Field, S. Fischer, J. Gao, H. Guo, S. Ha, D. Joseph-McCarthy, L. Kuchnir,
K. Kuczera, F.T.K. Lau, C. Mattos, S. Michnick, T. Ngo, D.T. Nguyen, B. Prodhom,
W.E. Reiher III, B. Roux, M. Schlenkrich, J.C. Smith, R. Stote, J. Straub,
M. Watanabe, J. Wiorkiewicz-Kuczera, D. Yin, M. Karplus, All-atom empirical po-
tential for molecular modeling and dynamics studies of proteins, J. Phys. Chem. B
102 (1998) 3586–3616.
[27] S. Terracciano, A. Russo, M.G. Chini, M.C. Vaccaro, M. Potenza, A. Vassallo,
R. Riccio, G. Bifulco, I. Bruno, Discovery of new molecular entities able to strongly
interfere with Hsp90 C-terminal domain, Sci. Rep. 8 (2018) 1709.
[28] (a) OEDOCKING 3.2.0.2: OpenEye scientific software, Santa Fe, NM. (http://www.
eyesopen.com). (b) B.P. Kelley, S.P. Brown, G.L. Warren, S.W. Muchmore, POSIT:
Flexible shape-guided docking for pose prediction. J. Chem. Inf. Model. 55 (2015)
1771-1780. (c) M. McGann, FRED Pose prediction and virtual screening accuracy, J.
Chem. Inf. Model. 51 (2011) 578-596. (d) M. McGann, FRED and HYBRID docking
performance on standardized datasets. J. Comput. Aided Mol. Des. 26 (2012) 897-
906.
[42] T.A. Halgren, Merck molecular force field, I. J. Comput. Chem. 17 (1996) 490–519.
[43] U. Essmann, L. Perera, M.L. Berkowitz, T. Darden, H. Leem, L.G. Pedersen, A
smooth particle mesh Ewald method, J. Chem. Phys. 103 (1995) 8577–8593.
[44] D.J. Evans, B.L. Holian, The nose-hoover thermostat, J. Chem. Phys. 83 (1985)
4069–4074.
[29] O. Trott, A.J. Olson, AutoDockVina: improving the speed and accuracy of docking
with a new scoring function, efficient optimization and multithreading, J. Comput.
Chem. 31 (2010) 455–461.
[45] (a) A. Bakan, L.M. Meireles, I. Bahar, ProDy: protein dynamics inferred from theory
and experiments, Bioinformatics 27 (2011) 1575–1577;
[30] Marvin 18.10, 2018, ChemAxon, http://www.chemaxon.com.
[31] N.M. O'Boyle, M. Banck, C.A. James, C. Morley, T. Vandermeersch, G.R. Hutchison,
OpenBabel: an open chemical toolbox, J. Cheminf. 3 (2011) 33–46.
[32] T.A. Halgren, MMFF VI. MMFF94s option for energy minimization studies, J.
Comput. Chem. 20 (1998) 720–729.
(b) A. Bakan, A. Dutta, W. Mao, Y. Liu, C. Chennubhotla, T.R. Lezon, I. Bahar, Evol
and ProDy for bridging protein sequence evolution and structural dynamics,
Bioinformatics 30 (2014) 2681–2683.
[46] J.D. Hunter, Matplotlib: a 2D graphics environment, Comput. Sci. Eng. 9 (2007)
90–95.
15