A structure-activity relationship study of Forkhead Domain Inhibitors (FDI): The importance of halogen binding interactions

Article abstract of DOI:10.1016/j.bioorg.2019.103269

The Forkhead boX M1 (FOXM1) protein is an essential transcription factor required for the normal activation of human cell cycle. However, increasing evidence supports a correlation between FOXM1 overexpression and the onset of several types of cancer. Based on a previously reported molecular modeling and molecular dynamics simulations (MD) study, we hypothesized the role of an essential halogen-bonding interaction between the 4-fluorophenyl group in the forkhead domain inhibitor-6 (FDI-6) and an Arg297 residue inside the FOXM1-DNA binding domain (DBD). To prove the importance of this binding interaction, we synthesized and screened ten FDI-6 derivatives possessing different groups at the 4-fluorophenyl position of the lead molecule. Briefly, we found that derivatives possessing a 4-chlorophenyl, 4-bromophenyl, or a 4-iodophenyl group, were equipotent to the original 4-fluorophenyl moiety present in FDI-6, whereas derivatives without this 4-halogen moiety were inactive. We also observed that positional isomers in which the halogen was relocated to positions 2- or 3- on the phenyl group were significantly less active. These results provide evidence to support the essential role of a 4-halophenyl bonding interaction, with the Arg297 residue in the FOXM1-DBD, to exert inhibition of transcriptional activity.

Full text of DOI:10.1016/j.bioorg.2019.103269

BioorganicChemistry93(2019)103269
Contents lists available at ScienceDirect
Bioorganic Chemistry
journal homepage: www.elsevier.com/locate/bioorg
A structure-activity relationship study of Forkhead Domain Inhibitors (FDI):
The importance of halogen binding interactions
Seyed Amirhossein Tabatabaei Dakhili, David J. Pérez, Keshav Gopal,
Seyed Yasin Tabatabaei Dakhili, John R. Ussher, Carlos A. Velázquez-Martínez⁎
Faculty of Pharmacy and Pharmaceutical Sciences, University of Alberta, Edmonton, AB, Canada
A R T I C L E I N F O
A B S T R A C T
Keywords:
The Forkhead boX M1 (FOXM1) protein is an essential transcription factor required for the normal activation of
human cell cycle. However, increasing evidence supports a correlation between FOXM1 overexpression and the
onset of several types of cancer. Based on a previously reported molecular modeling and molecular dynamics
simulations (MD) study, we hypothesized the role of an essential halogen-bonding interaction between the 4-
fluorophenyl group in the forkhead domain inhibitor-6 (FDI-6) and an Arg297 residue inside the FOXM1-DNA
binding domain (DBD). To prove the importance of this binding interaction, we synthesized and screened ten
FDI-6 derivatives possessing different groups at the 4-fluorophenyl position of the lead molecule. Briefly, we
found that derivatives possessing a 4-chlorophenyl, 4-bromophenyl, or a 4-iodophenyl group, were equipotent to
the original 4-fluorophenyl moiety present in FDI-6, whereas derivatives without this 4-halogen moiety were
inactive. We also observed that positional isomers in which the halogen was relocated to positions 2- or 3- on the
phenyl group were significantly less active. These results provide evidence to support the essential role of a 4-
halophenyl bonding interaction, with the Arg297 residue in the FOXM1-DBD, to exert inhibition of transcrip-
tional activity.
FOXM1
FDI-6
Direct FOXM1 inhibitors
Transcription factor
1. Introduction
Similar to other FOX proteins, FOXM1 has a conserved DNA Binding
Domain (DBD) that is responsible for binding to the corresponding
The Forkhead Box M1 (FOXM1) protein is an essential transcription
factor required for mitotic progression and cell division. Unlike normal
cells, cancer cells (of practically any tissue origin) undergo changes
leading to overexpression of FOXM1 and the abnormal activation of its
transcriptional cascade [1–3]. Hence, cancer cells sustain a rapid cell
replication pattern. In addition to its role in cell proliferation, FOXM1 is
also involved in cancer initiation [4,5], angiogenesis [6,7], and me-
tastasis [8]. Fig. 1 presents a schematic summary of the different roles
played by FOXM1 in cancer initiation and cancer progression. The
carcinogenic features associated with an overexpression of FOXM1
make this protein an emerging and promising drug target for cancer
treatment [9,10].
promoter regions [11]. Hypothetically, any small molecule capable of
binding to this winged helix could inhibit the FOXM1/DNA complex.
Nevertheless, transcription factors have been regarded as “challenging”
or “inaccessible” using small molecules [12]. This generalization was,
at least in part, due to the large solvent-accessible area observed on the
transcription factor and the lack of well-defined binding pockets on the
protein’s hydrophobic surface [13,14]. In this regard, Gormally et al.
[14] reported a high-throughput screening technique to test more than
50,000 small-molecules, selecting those capable of inhibiting the
binding interaction between FOXM1 and its DBD. Gormaly’s group se-
lected 16 different Forkhead Domain Inhibitors (abbreviated as “FDI”),
among which, the compound FDI-6 was the most potent and selective.
Abbreviations: FOXM1, forkhead box M1; MD, molecular dynamics; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; EMSA, Electrophoretic
Mobility Shift Assay; DBD, DNA binding domain; SEM, standard error of the mean; ANOVA, analysis of variance; TLC, thin layer chromatography; FDI, forkhead
domain inhibitor; NMR, nuclear magnetic resonance; DMSO, dimethyl sulfoxide; TMS, tetramehtylsilane; ATCC, American type culture collection; DMEM, Dulbecco's
Modified Eagle's Medium; RMPI, Roswell Park Memorial Institute medium; FBS, foetal bovine serum; SDS-PAGE, sodium dodecyl sulfate- polyacrylamide gel
electrophoresis; LB, luria broth; IPTG, Isopropyl β-D-1-thiogalactopyranoside; GST, glutathione s-transferase; dsDNA, double stranded DNA; RT, room temperature;
EDTA, Ethylenediaminetetraacetic acid; DTT, Dithiothreitol; Kd, dissociation constant; Ki, inhibition constant; PDB, protein data bank; RMSD, root mean square
fluctuation; MMPBSA, molecular mechanic poisson boltzmann surface area; VDW, van der waals
⁎ Corresponding author.
E-mail address: velazque@ualberta.ca (C.A. Velázquez-Martínez).
https://doi.org/10.1016/j.bioorg.2019.103269
Received 22 March 2019; Received in revised form 8 August 2019; Accepted 9 September 2019
Availableonline12September2019
0045-2068/©2019ElsevierInc.Allrightsreserved.

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Fig. 1. The role of FOXM1 in cancer initiation and cancer progression. FOXM1 is involved in the activation of different genes and the expression of proteins that
control angiogenesis, DNA repair, metastasis, and apoptosis.
Fig. 2. Compounds 7a-7k were prepared by replacing the 4-fluorophenyl group of FDI-6 (7c).
We previously conducted and reported a detailed in silico study
[15] to determine FOXM1-DBD/DNA binding interactions exerted by
three structurally-different FOXM1 inhibitors, namely thiostrepton
[16,17], troglitazone [18], and FDI-6 [14]. In our previous study we
proposed a protein-drug-DNA binding model involving a sulfur-His287
interaction. In addition, we also proposed another essential drug
binding interaction involving the fluorine atom at the 4-position of the
phenyl ring present in FDI-6, and the Arg297 residue in FOXM1. To
prove the importance of this halogen binding, in this complementary
study we report (i) a Structure-Activity Relationship (SAR) between the
parent (lead) drug FDI-6, and ten derivatives possessing halogen (Cl, Br,
I) atoms, as well as other substituents at the 4-fluorophenyl (H-, CH3-,
CF3-) position. Furthermore, we also studied the effects of relocating
the 4-fluorine atom to a 2- and 3-position (Fig. 2).
2. Results and discussion
2.1. Design and synthesis
To conduct a structure-activity relationship study on the role of the
4-fluorine atom present in FDI-6, we synthesized ten derivatives pos-
sessing different functional groups at the 4-position of a phenyl group
present in the lead molecule. We adapted the methods previously re-
ported for the synthesis of FDI-6 [14,19] and prepared derivatives 7a-
7k, with overall yields around 80%, using a microwave-assisted
synthesis. The proposed series of test molecules include FDIs devoid of
the 4-fluorophenyl group (7a); replacement of the 4-fluorophenyl by a
simple Ph moiety (7b); 4-bromophenyl (7d), 4-iodophenyl (7e); 4-
chlorophenyl (7f); 4-methylphenyl (7g); 3-fluorophenyl (7h); 2-fluor-
ophenyl (7i); 3,5-difluorophenyl (7k); and a 4-trifluoromethylphenyl,
7j) (Scheme 1).
We report preliminary evidence validating the essential role of a
halogen atom in FDI-6 derivatives, responsible for binding to an Arg297
residue. The results of our experiments validate the hypothesis of a 4-
(halo)phenyl moiety as an essential structural feature in FDIs, as one of
the required drug binding forces at the interface of the FOXM1 protein
and its DBD.
2.2. Cell proliferation inhibition assay
First, we determined the effect exerted by FDI-derivatives 7a-7k on
cancer cell proliferation using two human breast cancer cell lines,
namely the triple negative-breast cancer cell line MDA-MB-231, and the
estrogen receptor positive cell line MCF-7, using the MTT assay. Briefly,
after a 72 h-incubation period of the corresponding cell line with the
2

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Scheme 1. Chemical synthesis of FDI-6 derivatives, 7a-7k. aReagents and conditions: K2CO3, EtOH, 90 °C, μW, 2 h.
addition to the observation that the 2-fluorophenyl compound was in-
active (up to a maximum test drug concentration of 50 μM), we propose
that the halogen-Arg297 binding interaction becomes weaker as the
halogen is relocated farther away from the initial 4-position. This
statement is based on the assumption that the observed cell prolifera-
tion inhibition exerted by the drugs is, at least in part, FOXM1-depen-
Table 1
IC50 values determined for test molecules using the human breast cancer cell
lines MDA-MB-231 and MCF-7; these values were calculated after a 72 h-in-
cubation period with the drug molecules; all values represent the mean
of three different experiments, each one in triplicate.
Compound
SEM
MDA-MB-231
MCF-7
IC50 (µM)
IC50 (µM)
dent. Finally, we observed that the compound having
a 3,5-di-
fluorophenyl moiety (7k) was inactive, likely not because of weaker
binding interactions, but because it was not soluble enough in the cell
media employed in the MTT assay (we observed precipitation at in-
creasingly higher concentrations). Nevertheless, we will need to carry
out additional experiments using different pharmaceutical excipients
(other than DMSO) to increase the water solubility of this compound.
7a
> 50
> 50
> 50
35.27
13.43
7b
6.03
1.82
0.75
7c (FDI-6)
31.1
12.5
9.8
8.7
4.4
7d
7e
7f
3.04
2.2
1.36
0.39
14.6
24.8
35.75
> 50
12.6
> 50
4.2
5.0
2.90
1.15
2.9
7g
7h
7i
11.25
23.35
> 50
10.53
> 50
6.3
6.09
7j
1.6
1.9
2.3. FOXM1 expression (western blot)
7k
Considering that FOXM1 modulates its own transcriptional expres-
sion [21] we determined the effect exerted by the test drugs on the
expression levels of this protein, by western blot analysis (whole cell
lysis), after a 24 h-incubation period with the corresponding drug mo-
lecules, in triple negative-breast cancer cells (MDA-MB-231). We con-
ducted this experiment assuming that a drug-dependent decrease in
FOXM1 at the protein level could be due, at least in part, due to a drug-
induced dissociation of the nuclear FOXM1-DNA complex, which in
turn would suggest transcriptional inhibition. We observed a drug-de-
pendent decrease in FOXM1 protein expression in MDA-MB-231 breast
cancer cells (Fig. 3). Based on a simple structure-activity relationship
study, we can make a few preliminary statements describing the effect
of substituting specific functional groups on protein expression in these
cells. For example, compound 7a (devoid of the 4-fluorophenyl group
present in the lead drug FDI-6) and compound 7b [possessing a simpler
4-(phenyl) moiety], were practically inactive at 40 μM. As expected, the
lead molecule FDI-6 [4-(fluoro)phenyl; 7c] significantly decreased the
expression of FOXM1. Interestingly, compounds possessing a 4-(bromo)
phenyl (7d), 4-(iodo)phenyl (7e) and 4-(chloro)phenyl (7f) were
equipotent to FDI-6 (non-significant differences between them) in this
assay, correlating well with the cancer cell proliferation inhibition
assay. Nevertheless, we observed that compound 7g [4-(methyl)
phenyl] did not exert a significant decrease in FOXM1 expression de-
spite its good cell proliferation inhibitory profile. This suggests that
bioisosteric replacement with a methyl group may not be a suitable
approach to decrease the expression of FOXM1 in triple negative-breast
cancer cells. Finally, extending the SAR study to other derivatives in the
series, we observed that compound 7h [3-(fluoro)phenyl] was active at
test molecules (at increasing concentrations), we observed that drug
potency was significantly reduced for molecules without a 4-(halo)
phenyl group (7a, 7b), as well as for drug 7i in which the fluorine atom
was changed from the 4- to the 2-position; this was also observed for
drug 7k possessing a 3,5-difluorophenyl moiety. Interestingly, re-
placement of the fluorine atom present in FDI-6 (7c) by a 4-bromo (7d),
4-iodo (7e), or 4-chloro (7f) resulted in increased potency (decreased
cell proliferation) in both cell lines. However, the MCF-7 cancer cell
line was more susceptible to proliferation inhibition by the drug mo-
lecules. Interestingly, there was an apparent inverse correlation be-
tween the electronegativity of the halogen atom and the potency of the
corresponding derivative, in which the lower the electronegativity of
the halogen atom, the higher the potency (Table 1). Overall, these re-
sults suggest that the 4-(halo)phenyl ring is essential for cancer cell
proliferation inhibition in vitro.
Assuming that a methyl group could be a suitable bioisosteric re-
placement for some halogen atoms [20], we synthesized and screened
compounds 7g [4-(CH3) phenyl] and 7j [4-(CF3)phenyl]. In this regard,
we observed an equipotent potency for both drugs compared to the
analogs possessing a 4-(halo)phenyl group, including the lead drug FDI-
6. Consequently, these observations strongly suggest that the 4-sub-
stituted phenyl ring in FDI is essential to exert significant cell pro-
liferation inhibition of these two breast cancer cell lines.
We also investigated the effect produced by moving the fluorine
atom from the 4- to 3- position (compound 7h) and we observed a
minor decrease in potency compared to the lead molecule, which in
3

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Fig. 3. Expression levels of the FOXM1 pro-
tein and the concentration-dependent in-
hibitory effect produced by the test drugs;
incubation time: 24 h; cell line: triple nega-
tive-breast cancer (MDA-MB-231); drug
concentration = 40 μM; the bars represent
the corresponding average values obtained
after carrying out three independent
experiments
SEM; one-way ANOVA was
used to determine significance (* = P ≤ 0.05,
** = P ≤ 0.01, **** = P ≤ 0.0001) com-
pared to DMSO.
40 μM, but compound 7i [2-(fluoro)phenyl] was inactive, and this ob-
servation provides a “fine-tuning” of the hypothesis described above in
the sense that, the farther apart from position 4-, the weaker the
binding interaction. We observed that compound 7k (3,5-di-
fluorophenyl) and 7j [4-(trifluoromethyl)phenyl] exerted only a mod-
erate decrease in FOXM1 expression levels, likely due to low water
solubility.
provide one more piece of evidence confirming our initial hypothesis, in
which we proposed the need for a 4-halo substituted phenyl ring in the
FDI scaffold.
Compound 7g possessing a 4-CH3 bioisosteric replacement was
much less active (IC50 = 228.9 μM) than 7c (FDI-6) and compared to
other halogen-containing derivatives. Nevertheless, we observed that a
4-(trifluoromethyl)phenyl group (7j) exerted higher activity
(IC50 = 27.5 μM), supporting the need of having a halogen atom at the
4-phenyl position, essential to exert a binding interaction with Arg297
in FOXM1. Finally, moving the fluorine atom from the 4-position (7c)
to 3-position (7h) did not decrease the activity (IC50 = 42.6 μM), but
when the halogen is farther away (2-position; 7i) the potency decreased
(IC50 = 69.6 μM).
These results are in accordance with our hypothesis highlighting the
role of a 4-(halo)phenyl moiety for the drug binding process resulting in
inhibition of FOXM1′s transcriptional activity. The results obtained in
the western blot assay support the assumption that drugs possessing the
4-(halo)phenyl moieties decrease the expression of FOXM1 in breast
cancer cells, and provide evidence in favor of a 4-(halo)phenyl group/
Arg297 binding interaction in the FOXM1 protein.
Finally, when we screened compounds 7j and 7k, possessing a [4-
(trifluoromethyl)phenyl] and a 3,5-(difluoro)phenyl moiety, respec-
tively, we observed a good inhibitory profile of the FOXM1-DNA
complex (IC50 = 27.6 and 27.5 μM respectively). It should be noted that
in contrast to the results obtained in the protein immunoblot assay
described previously, in which compound 7j and 7k were relatively
weak, in the cell-free EMSA assay the binding interactions exerted by
these molecules may be more significant (cell membrane permeability
is not required). Considering that the soluble fraction of the drug is in
direct contact with the FOXM1 protein, it is possible that this is good
enough to exert inhibition of the FOXM1/DNA complex. This ob-
servation should be considered when trying to extrapolate results from
one assay to the other, and therefore, to assess the complete profile of
any given drug molecule as a transcription factor “inhibitor”, one must
consider the results from several screening assays. Moreover, to com-
pare the activity of different inhibitors, the Ki values (Fig. 4) should be
used as a complementary evaluation parameter, because IC50 values can
be altered by changing the protein concentration and, hence, could be
misleading.
2.4. Electrophoretic Mobility Shift Assay (EMSA)
This cell-free assay was selected based on its ability to measure the
concentration-dependent effect produced by drug molecules on the
FOXM1-DNA complex in vitro. This screening method was first reported
by Gormally et al. [14] as part of a high-throughput screening approach
to find new FOXM1 inhibitors, in which the drug FDI-6 was identified
as the most potent molecule. Based on their model, we incubated the
recombinant FOXM1-DBD reported to bind to DNA (also called FOXM1-
DNA binding domain, or FOXM1-DBD), with the target DNA, in the
presence of the corresponding FDI derivatives, and then running the
mixture on native polyacrylamide gel. As shown in Fig. 4, we observed
that compounds 7a (devoid of the 4-(halo)phenyl group) and 7b
(having a simple phenyl) were weak drugs (IC50 values = 128.2 and
121.7 μM respectively), whereas compounds 7c [4-(fluoro)phenyl], 7d
[4-(bromo)phenyl], and 7e [4-(iodo)phenyl] were significantly more
active (IC50 values around 40 μM). Compound 7f [4-(chloro)phenyl],
the most active molecule in this series (IC50 value = 27.2 μM), was
about 2-fold more potent than the lead drug FDI-6. These observations
4

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Fig. 4. Electrophoretic Mobility Shift Assay. (A) Titration curve of FOXM1-DBD protein with the target DNA (5-/IRD700/-AAACAAACAAACAATCAAACAAAC
AAACAATC-3′); (B to L) concentration-dependent effects of the drug molecules on the FOXM1-DNA complex presenting the calculated IC50 values, and their
corresponding Ki’s; all data are reported as average of three replicates (n = 3) and the error bars represent the SEM.
5

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Fig. 4. (continued)
2.5. Molecular modeling
drugs, we carried out a series of complementary docking and molecular
dynamic simulations with compounds 7a, 7b, 7c, 7g, 7h, 7i, 7j and 7k
comparing their binding profile with that of other molecules in the
same series (Table 2). In this regard, we observed a direct relationship
To further support the experimental observations suggesting the
need for a halogen atom at the 4-position of the phenyl moiety of FDI
6

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
Table 2
Free binding energies calculated for eight FDI derivatives. The total binding energies represent the sum of Van der waals, electrostatic, solvent accessible surface area
(SASA), and polar solvation energy, during the last nanosecond of the corresponding molecular dynamic simulation.
Compound Van Der Waal Energy (kJ/
Electrostatic Energy (kJ/ Polar Solvation Energy A
SASA Energy (kJ/
Binding Energy (kJ/mol) Binding Energy (kcal/
mol)
mol)
(kJ/mol)
mol)
mol)
7a
7b
7c
−89.7
−94.8
−147.6
5.3
−14.7
−7.7
−3.3
3.7
5.5
2.9
55.9
40.6
59.0
35.3
−9.09
−11.5
−13.7
1.1
1.9
0.6
−57.6
−73.4
−105.6
38.1
11.5
−13.7
−17.5
−25.2
9.1
2.7
8.1
16.9
16.6
31.9
16.2
34.2
(FDI-6)
7g
7h
7i
−62.9
−122.4
−69.2
−5.7
41.6
12.0
34.6
25.5
14.8
0.5
2.8
31.3
55.7
41.0
−3.5
40.5
31.6
19.5
18.3
31.5
17.0
−7.9
−12.9
−8.6
−0.9
−11.6
5.1
0.8
4.3
2.8
1.5
−38.9
−72.9
−41.7
−11.0
−74.8
44.7
24.8
30.9
15.3
10.0
−9.2
−17.4
−9.9
−2.6
−17.8
10.6
5.9
7.3
3.6
2.3
−3.1
−4.8
−0.8
−7.6
3.3
4.8
3.9
6.2
7j
7k
−96.2
between the position of the halogen (fluorine) atom and the total
binding energy calculated for the corresponding drug molecule: 4-
flouro (−25.2 kcal/mol), 3-flouro, −17.4 kcal/mol, and 2-flouro
(−9.9 kcal/mol). With these results, we think it is reasonable to pro-
pose that the strength of the binding force between the drug molecule
and the FOXM1 protein is determined, at least in part, by the relative
position of the fluorine atom, in which 4-phenyl > 3-phenyl > 2-
phenyl. Compound 7c (FDI-6) showed the strongest binding energy
(−25.2 kcal/mol), with a low ligand Positional Root Media Square
Deviation (RMSD), compared to compounds 7b (phenyl), 7h [3-(fluoro)
phenyl], and 7i [2-(fluoro)phenyl] (Fig. S23), Supplementary in-
formation). The low RMSD values observed for 7c, along with the low
Van Der Waal energy (VDW, −147.6 Kj/mol), suggests that the 4-
(flouro)phenyl moiety increases the stability of a possible active con-
formation inside the binding pocket during the molecular dynamic si-
mulation.
As shown in Table 2, with the exception of compound 7j we could
correlate the binding energies for all compounds with the screening
assays described above. In other words, the unusually high theoretical
binding energy (low binding force = −2.6 Kcal/mol) calculated for 7j,
would suggest a weak activity profile, but this was not the case. Com-
pound 7j decreased cell proliferation in MDA-MB-231 cells, it sig-
nificantly decreased the FOXM1 protein level of FOXM1, and it showed
significant activity in the EMSA.
Finally, in Fig. 5 we present a graphical representation of the
binding conformations observed for compounds 7c [4-(fluoro)phenyl]
and 7h [3-(fluoro)phenyl] and their relative distance from Arg297 in
which we observed practically the same binding interactions between
the drug molecules and the Arg297.
3. Summary and conclusions
We present a series of in vitro and in silico experiments supporting
the essential role played by a 4-(halo)phenyl moiety in the chemical
structure of FDIs involving an Arg297 amino acid residue within the
DBD of the FOXM1 protein. Despite a few minor differences in the re-
lative potency of individual compounds among individual assays, we
propose that (1) the halogen binding interaction is equipotent for 4-
fluoro, 4-chloro, 4-bromo, or 4-iodo groups; (2) a bioisosteric replace-
ment involving a 4-(methyl)phenyl group (7j) did not result in sig-
nificant binding interactions or an improved activity profile. However,
the use of a 4-(trifluoromethyl)phenyl moiety is a suitable strategy to
improve the potency of FDIs (by about two-fold) as shown in the EMSA,
and it maintained the drug’s cancer cell proliferation inhibitory profile
in triple negative-breast cancer (MDA-MB-231) cells.
One major limitation associated with this work is the fact that we
are not considering the effects produced by the test drugs on other
(potential) targets that would affect cell proliferation or FOXM1 protein
expression. However, the EMSA assay is relatively selective to evaluate
the effects produced by “direct” FOXM1 inhibitors, distinguishing be-
tween inactive and active drugs that bind to (and interfere with) the
Fig. 5. The graphical representation of the relative distance [Å] of fluorine
atom of 7c, 2.7 Å (A), 7h, 5.1 Å (B) and 7i, 6.5 Å (C) to Arg297.
7

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
protein’s DNA binding domain. In this regard, we used structurally
unrelated molecules as negative controls (ranolazine and andro-
grapholide) and we did not observe dissociation (at any drug con-
centration) of the protein-DNA complex (Fig. S25; supplementary in-
formation).
(151 MHz, DMSO) δ 164.49, 160.99, 152.15, 142.72, 132.16, 131.91,
131.55, 131.38, 129.49, 129.23, 125.93, 124.26, 124.12, 122.30,
120.49, 118.74, 112.86, 112.82. Anal. Calc. for (%) C19H11BrF3N3OS2,
C 45.79; H 2.23, N 8.43; S 12.87; found C 45.72, H 2.44, N 8.10, S
12.51.
In summary, we provide evidence validating the essential role of a
4-(halo)phenyl–Arg297 binding interaction as part of the overall me-
chanism of action exerted by FDIs, postulated by our group, in a pre-
vious publication [15]. In this report, we also propose a specific binding
interaction to fine-tune the design of FOXM1 inhibitors based on the
chemical scaffold of the lead FDI-6 molecule first described by Gormally
et al. [14] and we submit that this model could also be used in the
design of small-molecule drugs possessing a 4-(halo)phenyl moiety.
4.2.2. 3-Amino-N-(4-iodophenyl)-6-(thiophen-2-yl)-4-(trifluoromethyl)
thieno[2,3-b]pyridine-2-carboxamide (7e)
6e (55 mg, 0.18 mmol), 3 (54 mg, 0.18 mmol) and K2CO3 (25 mg,
0.18 mmol) in 5 mL of ethanol, yellow (flocculent) crystals, 92% yield
(90 mg, 0.19 mmol); mp: 226–228 °C;1H NMR (600 MHz, DMSO‑d6) δ
9.87 (s, 1H), 8.35 (s, 1H), 8.28 (dd, J = 3.8, 1.1 Hz, 1H), 7.91 (dd,
J = 5.0, 1.1 Hz, 1H), 7.77–7.71 (m, 2H), 7.63–7.57 (m, 2H), 7.32 (dd,
J = 5.0, 3.7 Hz, 1H), 6.85 (s, 2H). 13C NMR (151 MHz, DMSO‑d6) δ
164.09, 161.04, 152.78, 145.55, 142.54, 137.59, 132.47, 132.25,
131.67, 129.57, 129.54, 125.84, 124.12, 122.21, 118.27, 113.16,
113.12, 88.03. Anal. Calc. for (%) C19H11F3IN3OS2, C 41.85, H 2.03, N
7.71, S 11.76; found C 41.31, H 2.10, N 7.45, S 11.69.
4. Experimental section
4.1. General information
All the reagents and solvents were purchased from Sigma-Aldrich
and were used without further purification. All reactions were mon-
itored by thin-layer chromatography (RediSep® TLC plates) and visua-
lized using UV light. Melting points were measured with an
Electrothermal melting point apparatus (Thermofisher, USA) and were
uncorrected. 1H NMR and 13C NMR spectra were determined on a
Bruker FT-600 MHz instrument (600 MHz and 150 MHz, respectively)
using DMSO‑d6 as the solvent and TMS as a reference. Chemical shifts
(δ) and coupling constants (J) are expressed in parts per million and
Hertz, respectively. Signal multiplicity is expressed as s (singlet), d
(doublet), t (triplet), q (quartet), m (multiplet) and br (broad singlet).
Elemental analyses were performed on a Carlo Erba EA1108 Elemental
Analyzer and the results are within 0.4% of the theoretical values.
The synthesis of 6-(thiophen-2-yl)-2-thioxo-4-(trifluoromethyl)-1,2-di-
hydropyridine-3-carbonitrile (3) was carried out following the method
reported [19], and confirmed by 1H, 13C and 19F NMR. 2-Chlor-
oacetamide (6a) was purchased from Sigma Aldrich; compounds 6b-6k
were synthesized based on protocols previously reported [22–24] and
confirmed by 1H NMR (more details in the supplementary information).
FDI derivatives 7a-7c, and 7f, were confirmed by 1H and 13C spectra,
both in agreement to those reported in literature [19] (see supporting
information, Figs. S1-S22). The microwave-assisted synthesis of deri-
vatives 7a-7k was carried out using a Biotage Initiator Reactor. All test
compounds were purified by flash column chromatography using a
Combi Flash EZ prep (Teledyne isco), using prepacked silica cartridges
(RediSep Rf® Gold Resolution) and a gradient of hexane-ethyl acetate as
mobile phase.
4.2.3. 3-Amino-N-(4-methylphenyl)-6-(thiophen-2-yl)-4-(trifluoromethyl)
thieno[2,3-b]pyridine-2-carboxamide (7g)
6 g (40 mg, 0.2 mmol), 3 (60 mg, 0.2 mmol) and K2CO3 (30 mg,
0.18 mmol) in 5 mL of Ethanol, yellow (flocculent) crystals yellow
powder, 90% yield (80 mg, 0.18 mmol), mp: 233–235 °C;1H NMR
(600 MHz, DMSO‑d6) δ 9.69 (s, 1H), 8.31 (s, 1H), 8.23 (dd, J = 3.8,
1.1 Hz, 1H), 7.86 (dd, J = 5.0, 1.1 Hz, 1H), 7.58–7.53 (m, 2H), 7.28
(dd, J = 5.0, 3.8 Hz, 1H), 7.19–7.15 (m, 2H), 6.75 (s, 2H), 2.30 (s, 3H).
13C NMR (151 MHz, DMSO‑d6) δ 163.91, 160.95, 152.70, 145.33,
142.56, 136.31, 133.53, 132.46, 132.24, 131.64, 129.54, 129.38,
124.04, 122.21, 118.37, 113.15, 113.11, 101.61, 20.99. Anal. Calc. for
(%) C20H14F3N3OS2, C 55.42, H 3.26, N 9.69, S 14.79, found C 55.42, H
3.31, N 9.49, S 15.08
4.2.4. 3-Amino-N-(3-fluorophenyl)-6-(thiophen-2-yl)-4-(trifluoromethyl)
thieno[2,3-b]pyridine-2-carboxamide (7h)
6 h (45 mg, 0.23 mmol), 3 (65 mg, 0.23 mmol) and K2CO3 (30 mg,
0.23 mmol) in 5 mL of ethanol, yellow (flocculent) crystals yellow
powder, 92% yield (90 mg, 0.21 mmol), mp: 218–220 °C; 1H NMR
(600 MHz, DMSO‑d6) δ 9.95 (s, 1H), 8.35 (s, 1H), 8.27 (dd, J = 3.8,
1.1 Hz, 1H), 7.90 (dd, J = 5.0, 1.1 Hz, 1H), 7.71 (ddd, J = 11.8, 2.6,
1.9 Hz, 1H), 7.57 (ddd, J = 8.2, 1.9, 0.9 Hz, 1H), 7.44 (td, J = 8.2,
6.9 Hz, 1H), 7.32 (dd, J = 5.0, 3.7 Hz, 1H), 6.99 (tdd, J = 8.5, 2.6,
0.9 Hz, 1H), 6.87 (s, 2H). 13C NMR (151 MHz, DMSO): δ 164.21,
163.21, 161.61, 161.07, 152.91, 145.83, 142.50, 140.95, 132.53,
132.31, 131.72, 130.54, 130.48, 129.63, 129.55, 125.83, 124.02,
122.20, 118.18, 117.49, 117.48, 113.20, 113.16, 110.76, 110.62,
108.56, 108.39, Anal. Calc. for (%) C19H11F4N3OS2, C 52.17, H 2.53, N
9.61, S 14.66 found C 51.80, H 2.63, N 9.28, S 14.88.
4.2. General procedure for synthesis of compounds 7a-7k
The corresponding chloroacetamide 6a-6 k (1 equiv.), compound 3
(suppl. info.; 1 equiv.) was mixed with K2CO3 (1 equiv.), and ethanol
(5 mL) in a microwave reaction vessel. This mixture was stirred at 90 °C
for 2 h using a “high energy absorption” setting. The crude product was
filtered-off, washed with water and fixed onto silica gel powder before
running a solvent gradient (flash column chromatography). Combined
organic fractions were dried under vacuum and the corresponding final
product was recrystallized from ethanol (when needed).
4.2.5. 3-Amino-N-(2-fluorophenyl)-6-(thiophen-2-yl)-4-(trifluoromethyl)
thieno[2,3-b]pyridine-2-carboxamide (7i)
6i (45 mg, 0.23 mmol), 3 (65 mg, 0.23 mmol) and K2CO3 (30 mg,
0.23 mmol) in 5 mL of ethanol, yellow (flocculent) crystals, 95% yield
(93 mg, 0.22 mmol); mp: 248–250 °C;1H NMR (600 MHz, DMSO‑d6) δ
9.73 (s, 1H), 8.32 (s, 1H), 8.24 (dd, J = 3.8, 1.1 Hz, 1H), 7.86 (dd,
J = 5.0, 1.1 Hz, 1H), 7.51 (t, J = 7.8 Hz, 1H), 7.36–7.29 (m, 2H), 7.28
(dd, J = 5.0, 3.7 Hz, 1H), 7.27–7.20 (m, 1H), 6.76 (s, 2H). 13C NMR
4.2.1. 3-Amino-N-(4-bromophenyl)-6-(thiophen-2-yl)-4-(trifluoromethyl)
thieno[2,3-b]pyridine-2-carboxamide (7d)
6d (65 mg, 0.22 mmol), 3 (63 mg, 0.22 mmol) and K2CO3 (30 mg,
0.22 mmol) in 5 mL of ethanol, yellow (flocculent) crystals, 90% yield
(100 mg, 0.2 mmol), mp: 225–227 °C; 1H NMR (600 MHz, DMSO‑d6) δ
9.78 (s, 1H), 8.33 (s, 1H), 8.25 (dd, J = 3.8, 1.1 Hz, 1H), 7.89 (dd,
J = 5.0, 1.1 Hz, 1H), 7.75–7.68 (m, 2H), 7.56–7.51 (m, 2H), 7.32 (dd,
J = 5.0, 3.7 Hz, 1H), 6.91 (s, 2H). 13C NMR (151 MHz, DMSO‑d6) δ
131.55, 131.38, 129.49, 129.23, 124.26, 40.52, 40.24. 13C NMR
(151 MHz, DMSO‑d6)
δ 164.08, 161.10, 157.77, 156.13, 152.87,
145.57, 142.53, 132.61, 132.39, 131.72, 129.64, 129.56, 128.47,
127.97, 127.92, 125.74, 125.66, 124.78, 124.00, 122.19, 118.29,
116.38, 116.25, 113.19, 113.15, 100.96. Anal. Calc. for (%)
C19H11F4N3OS2 C 52.17, H 2.53, N 9.61, S 14.66 found C 52.19, H 2.71,
N 9.36, S 15.02.
8

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
4.2.6. 3-Amino-6-(thiophen-2-yl)-4-(trifluoromethyl)-N-
using Odyssey scanner (LI-COR Biosciences). The REVERT was then
reversed and the membrane was blocked with 10% fat-free milk in
TBST for 1 h. The membrane was incubated with the primary antibody
(1:1000 dilution) at 4 °C overnight. Then, the membrane was washed
three times with TBST, incubated with the corresponding Li-Cor sec-
ondary antibody and incubated again at room temperature for 1 h. The
membrane was washed three times (15 min total) with TBST. The blots
were visualized using Odyssey scanner (LI-COR Biosciences). The
quantification was carried out for all proteins relative to total protein
(REVERT) using ImageJ for each lane.
[4(trifluoromethyl)phenyl]thieno[2,3-b]pyridine-2-carboxamide (7j)
6j (55 mg, 0.22 mmol), 3 (65 mg, 0.22 mmol) and K2CO3 (30 mg,
0.22 mmol) in 5 mL of ethanol, yellow (flocculent) crystals, 95% yield
(100 mg, 0.21 mmol), mp: 183–185 °C; 1H NMR (600 MHz, DMSO‑d6) δ
10.05 (s, 1H), 8.30 (s, 1H), 8.23 (dd, J = 3.8, 1.1 Hz, 1H), 7.94 (d,
J = 8.5 Hz, 2H), 7.86 (dd, J = 5.0, 1.1 Hz, 1H), 7.70 (d, J = 8.5 Hz,
2H), 7.27 (dd, J = 5.0, 3.7 Hz, 1H), 6.86 (s, 2H): 13C NMR (151 MHz,
DMSO‑d6) δ 164.52, 161.14, 152.85, 142.54, 132.50, 132.28, 131.70,
129.60, 129.54, 127.62, 126.15, 126.13, 126.10, 125.82, 124.03,
122.22, 121.65, 121.63, 120.40, 118.22, 113.16, 113.12. Anal. Calc. for
(%) C20H11F6N3OS2 C 49.28, H 2.27, N 8.62, S 13.15 found C 49.09, H
2.30, N 8.33, S 12.93.
4.7. Protein expression and purification
We used the PEX-N-GST-FOXM1-DBD plasmid (OriGene
Technologies, USA) transformed into BL21(DE3) E. Coli cells; positive
colonies were selected on LB agar plates with ampicillin (100 μg/mL).
Then, these cells were grown in LB media with ampicillin (100 μg/ml)
at 37 °C with orbital shaking at 220 rpm until reaching an optical
density (OD600) of 0.8; protein expression was induced by adding
1 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 6 hr at 37 °C.
The GST protein and GST-FOXM1 protein from soluble fractions were
purified using glutathione resin (GenScript, USA), following the man-
ufacturer’s instructions. Please see Fig. S24 for the representative gel
image of purified protein.
4.2.7. 3-Amino-N-(3,5-difluorophenyl)-6-(thiophen-2-yl)-4-
(trifluoromethyl)thieno[2,3-b]pyridine-2-carboxamide (7k)
6 k (50 mg, 0.23 mmol), 3 (65 mg, 0.23 mmol) and K2CO3 (35 mg,
0.23 mmol) in 5 mL of ethanol, red needle crystals, 95% yield (100 mg,
0.22 mmol), mp: 208–210 °C; 1H NMR (600 MHz, DMSO‑d6) δ 10.02 (s,
1H), 8.30 (s, 1H), 8.23 (dd, J = 3.8, 1.1 Hz, 1H), 7.86 (dd, J = 5.0,
1.1 Hz, 1H), 7.54–7.46 (m, 2H), 7.27 (dd, J = 5.0, 3.7 Hz, 1H), 6.92 (t,
J = 9.3 Hz, 1H), 6.86 (s, 2H). 13C NMR (151 MHz, DMSO‑d6) δ 163.53,
163.43, 161.92, 161.82, 161.15, 153.15, 148.79, 146.32, 146.31,
132.64, 132.42, 132.03, 131.85, 129.92, 129.77, 129.57, 125.80,
123.99, 122.17, 118.02, 113.30, 113.26, 104.31, 104.27, 104.15,
104.11, 99.37, 99.20, 99.03. Anal. Calc. for (%) C19H10F5N3OS2
C
4.8. Electromobility shift assay
50.11, H 2.21, N 9.23, S 14.08 found C 49.78, H 2.39, N 8.91, S 13.66.
All values of a titration (binding) curve of recombinant FOXM1-DBD
with its target double strand DNA oligo (Forward strand: 5′-/IRD700/-
AAACAAACAAACAATCAAACAAACAAACAATC-3′), were recorded
using EMSA by the method previously reported by Gormally et al. [14]
Briefly, dsDNA and an increasing concentration of the FOXM1 protein
were incubated at RT for 30 min in a buffer solution containing 20 mM
Tris (pH 7.5), 100 mM KCl, 1 mM EDTA, 0.1 mM DTT, and 10% gly-
cerol, before running the samples on 6% native gel for 30 min at 120 V.
The dissociation constant of protein DNA complex (Kd) was calculated
using Graph pad Prism 6.2. The displacement EMSA experiments were
carried out by incubating each test compound with the recombinant
FOXM1-DBD protein, for 1.5 h, at room temperature, followed by a
second incubation with DNA, for 20 min, before conducting the elec-
trophoresis. The concentration of FOXM1-DBD and DNA in each reac-
tion was 480 nM and 12.8 nM, respectively. The Ki values were calcu-
lated for each compound (7a-7k) using Eq. (1) [25]:
4.3. Cell culture
The MDA-MB-231 and MCF-7 cells were a generous gift from Dr.
Frank Wuest (Cross Cancer Institute; Edmonton, Alberta, Canada). The
cells were cultured in RPMI and DMEM media respectively, supple-
mented with 10% fetal bovine serum (FBS) in a 5% CO2 atmosphere at
37 °C.
4.4. Cell proliferation inhibition (MTT) assay
Cells were seeded in 96-well plates (approx. 4000 cells/well for
MDA-MB-231 and approx. 5000 for MCF-7), then we added the test
compounds at different concentrations and we incubated all plates for
72 h. We added 30 μL of MTT solution (3 mg/ml) and continued the
incubation for 3 h at 37 °C. The precipitate crystals were dissolved using
DMSO, and the absorbance of the resulting solution was recorded at
570 nm using a Synergy H1-Hybrid Multi-Mode Reader (BioTek). We
analyzed the data using GraphPad Prism. All experiments were carried
out in triplicate.
[I]50
Ki =
([l]
50
)
[P]
0
d
(
+
+ 1)
(1)
K
K
d
where [I]50 = IC50 of the inhibitor; [L]50 = concentration of IR-labelled
DNA at 50% inhibition; [P] = concentration of the FOXM1 protein; and
Kd = dissociation constant calculated from the initial titration curve.
4.5. Antibodies
We used
a FOXM1 mouse monoclonal antibody (Santa Cruz
Biotechnology) and IRDye® 800CW Goat anti-Mouse IgM (Li-Cor
4.9. Molecular modeling and dynamics simulations
Biosciences).
The crystal structure of FOXM1-DBD was acquired from the protein
data bank (PDB_ID: 3G73) [11]. Using Pymol v.1.8 [26], we removed
Chain A followed by a short minimization using CHIMERA V 1.10.2
[27], the missing sidechains were added and the protonated group was
equilibrated to the biological pH 7.0 using PROPKA [28].
4.6. Western blot
After treatment with test compounds for 24 h at different con-
centrations, the cells were trypsinized; the FOXM1 protein was isolated
with RIPA lysis and extraction buffer (Thermo fisher) according to the
manufacturer’s protocol. The protein levels in the supernatant were
measured using the Bradford assay. Then the protein (40 μg/lane) was
All the 3D format structures were prepared for docking using the
dock prep tool of UCSF CHIMERA V1.10.2 in the framework of
AMBER99SB force field. All the compounds were docked in a previously
identified binding pocket [15], into a grid of 40 × 40 × 50 Å with a
spacing of 0.375 Å using Autodock vina [29]; 12 runs per docking were
performed with the exhaustiveness of 40 for each ligand. Before per-
forming the molecular dynamic simulation on the Protein-ligand com-
plexes, we performed a molecular dynamic simulation on the FOXM1-
loaded into
a
4–20% SDS-PAGE (Sodium dodecyl sulfate
Polyacrylamide gel electrophoresis). After completion of the run, the
protein was transferred from the gel to a nitrocellulose membrane
(Thermofisher), and stained with REVERT (Li-Cor Biosciences) total
protein stain. The membrane was then detected in the 700 nm channel
9

S.A. Tabatabaei Dakhili, et al.
BioorganicChemistry93(2019)103269
DBD using GROMACS 4.5.6 package [30]. We used the TIP3P water
models to solvate the protein with 1 nm marginal cushion on each side.
The box was then neutralized using NaCl, and the system was mini-
mized using the AMBER99SB0ILDN force field. The system was heated
to 300 K and equilibrated for 500 ps using the Berendsen Thermostat.
Using the Isothermal-isobaric ensemble at 1 bar with the Parrinello-
Rahman barostat an additional equilibration was also performed. A
20 ns production run was performed using the periodic boundary con-
dition. The Lenard-Jones, the Coulomb (Cut-off = 1.0 nm), and the
Particle Mesh Ewald (PME) were used to calculate the VDW and elec-
trostatic interactions. The FOXM1-DBD/Ligand complexes were per-
formed using the same condition. The AnteChamber Python Parser in-
terfacE (ACPYPE) [31] was used for ligand parameterization. The Root
Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF)
and ligand positional RMSD were calculated using GROMACS tools. All
the data were plotted using the GraphPad Prism 6.0.7. Discovery Studio
Visualizer [32] and the Schrodinger’s PyMOL package were used as the
visualization tools.
is expressed in actively dividing lymphocytes, Immunobiology 198 (1997) 157–161.
[2] K.-M. Yao, M. Sha, Z. Lu, G.G. Wong, Molecular Analysis of a Novel Winged Helix Protein,
win: expression pattern, DNA binding property, and alternative splicing within the DNA
binding domain, J. Biol. Chem. 272 (1997) 19827–19836.
[3] H. Ye, T.F. Kelly, U. Samadani, L. Lim, S. Rubio, D.G. Overdier, K.A. Roebuck, R.H. Costa,
Hepatocyte nuclear factor 3/fork head homolog 11 is expressed in proliferating epithelial
and mesenchymal cells of embryonic and adult tissues, Mol. Cell. Biol. 17 (1997)
1626–1641.
[4] M.-T. Teh, E. Gemenetzidis, T. Chaplin, B.D. Young, M.P. Philpott, Upregulation of
FOXM1 induces genomic instability in human epidermal keratinocytes, Mol. Cancer 9
(2010) 45.
[5] Z. Wang, A. Ahmad, Y. Li, S. Banerjee, D. Kong, F.H. Sarkar, Forkhead box M1 tran-
scription factor: A novel target for cancer therapy, Cancer Treat. Rev. 36 (2010) 151–156.
[6] H. Chen, Y. Zou, H. Yang, J. Wang, H. Pan, Downregulation of FoxM1 inhibits pro-
liferation, invasion and angiogenesis of HeLa cells in vitro and in vivo. International
Journal of Oncology, Int. J. Oncol. 45 (2014) 2355–2364.
[7] Y. Zhang, N. Zhang, B. Dai, M. Liu, R. Sawaya, K. Xie, S. Huang, FoxM1B transcriptionally
regulates vascular endothelial growth factor expression and promotes the angiogenesis
and growth of glioma cells, Cancer Res. 68 (2008) 8733–8742.
[8] B.Y. Fei, X. He, Jie Ma, M. Zhang, R. Chai, FoxM1 is associated with metastasis in col-
orectal cancer through induction of the epithelial-mesenchymal transition, Oncol Lett. 14
(2017) 6553–6561.
[9] A. Bergamaschi, Z. Madak-Erdogan, Y.J. Kim, Y.-L. Choi, H. Lu, B.S. Katzenellenbogen,
The forkhead transcription factor FOXM1 promotes endocrine resistance and invasiveness
in estrogen receptor-positive breast cancer by expansion of stem-like cancer cells, Breast
Cancer Res. 16 (2014) 436.
The free energy of interaction between each ligand and FOXM1-
DBD was calculated using the g_mmpbsa Gromacs tool [33]. Using the
Molecular Mechanic Poisson-Boltzmann Surface Area (MM-PBSA), this
program calculates the binding free energy based on the electrostatic
and VDW interactions besides polar and non-polar solvation energies.
The G_mmpbsa module solves the following Eqs. (2) and (3) to calculate
the binding energy of ligands:
[10] X. Song, S.S. Fiati Kenston, J. Zhao, D. Yang, Y. Gu, Roles of FoxM1 in cell regulation and
breast cancer targeting therapy, Med. Oncol. 34 (2017) 41.
[11] D.R. Littler, M. Alvarez-Fernández, A. Stein, R.G. Hibbert, T. Heidebrecht, P. Aloy,
R.H. Medema, A. Perrakis, Structure of the FoxM1 DNA-recognition domain bound to a
promoter sequence, Nucleic Acids Res. 38 (2010) 4527–4538.
[12] F. Fontaine, J. Overman, M. François, Pharmacological manipulation of transcription
factor protein-protein interactions: opportunities and obstacles, Cell Regener. 4 (4)
(2015) 2.
[13] A.M. Redmond, J.S. Carroll, Defining and targeting transcription factors in cancer,
Genome Biol. 10 (2009) 311.
Gbidning = Gcomplex (Gprotein + Gligand
)
(2)
[14] M.V. Gormally, T.S. Dexheimer, G. Marsico, D.A. Sanders, C. Lowe, D. Matak-Vinković,
S. Michael, A. Jadhav, G. Rai, D.J. Maloney, A. Simeonov, S. Balasubramanian,
Suppression of the FOXM1 transcriptional programme via novel small molecule inhibi-
tion, Nat. Commun. 5 (2014) 5165.
where the Gcomplex is the protein-ligand complex total free energy, Gprotein
is the total free energy of protein and Gligand is the total energy of ligand
in solvent. The free energy of protein ligand complex, isolated protein
and isolated ligand (G) can be given by:
[15] S.A. Tabatabaei-Dakhili, R. Aguayo-Ortiz, L. Domínguez, C.A. Velázquez-Martínez,
Untying the knot of transcription factor druggability: Molecular modeling study of
FOXM1 inhibitors, J. Mol. Graphics Modell. 80 (2018) 197–210.
[16] N.S. Hegde, D.A. Sanders, R.L. Rodriguez, S. Balasubramanian, The transcription factor
FOXM1 is a cellular target of the natural product thiostrepton, Nat. Chem. 3 (2011)
725–731.
G = EMM TS + G
(3)
solvation
where EMM is the average of molecular mechanics potential in vacuum,
[17] A.L. Gartel, Thiostrepton, proteasome inhibitors and FOXM1, Cell Cycle 10 (2011)
4341–4342.
TSis the Temprature and Entropy respectively and
G
solvation
is the sol-
[18] V. Petrovic, R.H. Costa, L.F. Lau, P. Raychaudhuri, A. Tyner, Negative regulation of the
oncogenic transcription factor FoxM1 by thiazolidinediones and mithramycin, Cancer
Biol. Ther. 9 (2010) 1008–1016.
vation free energy [33].
Declaration of Competing Interest
[19] M.I. Abdel-Monem, O.S. Mohamed, E.A. Bakhite, Fluorine-containing heterocycles:
Synthesis and some reactions of new 3-amino-2-functionalized-6-(2′-thienyl)-4-tri-
fluoromethylthieno [2,3-b]pyridines, Pharmazie 56 (2001) 41–44.
[20] G.A. Patani, E.J. LaVoie, Bioisosterism: a rational approach in drug design, Chem. Rev. 96
(1996) 3147–3176.
The authors declare no competing financial interest.
[21] A.L. Gartel, Targeting FOXM1 auto-regulation in cancer, Cancer Biol. Ther. 16 (2015)
185–186.
Acknowledgments
[22] M.A. Matulenko, A.A. Hakeem, T. Kolasa, M. Nakane, M.A. Terranova, M.E. Uchic,
L.N. Miller, R. Chang, D.L. Donnelly-Roberts, M.T. Namovic, R.B. Moreland, J.D. Brioni,
A.O. Stewart, Synthesis and functional activity of (2-aryl-1-piperazinyl)-N-(3-methyl-
phenyl)acetamides: selective dopamine D4 receptor agonists, Bioorg. Med. Chem. 12
(2004) 3471–3483.
Authors acknowledge the support from the Noujaim Fund for
Strategic Initiatives, Faculty of Pharmacy and Pharmaceutical Sciences,
University of Alberta. A portion of the described studies were also
supported from a Discovery Grant (RGPIN-2015-04946) from the
Natural Sciences and Engineering Research Council of Canada to J.R.U;
S.A.T.D is the recipient of Antoine Noujaim graduate scholarship. This
research was enabled in part by Compute Canada and its regional
partner WestGrid. D.J.P is grateful for the Postdoctoral Fellowship
(287012) granted by the National Council for Science and Technology
(CONACYT, Mexico).
[23] R.V. Patel, P.K. Patel, P. Kumari, D.P. Rajani, K.H. Chikhalia, Synthesis of benzimidazolyl-
1,3,4-oxadiazol-2ylthio-N-phenyl (benzothiazolyl) acetamides as antibacterial, antifungal
and antituberculosis agents, Eur. J. Med. Chem. 53 (2012) 41–51.
[24] W. Zhang, J. Ai, D. Shi, X. Peng, Y. Ji, J. Liu, M. Geng, Y. Li, Discovery of novel type II c-
Met inhibitors based on BMS-777607, Eur. J. Med. Chem. 80 (2014) 254–266.
[25] Z. Nikolovska-Coleska, R. Wang, X. Fang, H. Pan, Y. Tomita, P. Li, P.P. Roller,
K. Krajewski, N.G. Saito, J.A. Stuckey, S. Wang, Development and optimization of a
binding assay for the XIAP BIR3 domain using fluorescence polarization, Anal. Biochem.
332 (2004) 261–273.
[26] Schrödinger; LLC. The {PyMOL} Molecular Graphics System, Version 1.8, 2015.
[27] 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.
Appendix A. Supplementary material
[28] C.R. Søndergaard, M.H.M. Olsson, M. Rostkowski, J.H. Jensen, Improved treatment of
ligands and coupling effects in empirical calculation and rationalization of pKa values, J.
Chem. Theory Comput. 7 (2011) 2284–2295.
General procedure for the synthesis of compounds 6b-6k and 3,
NMR spectra for compounds 7a-7k (Figs. S1–S21), the ligand positional
RMSD calculated for compounds 7b, 7c, 7h and 7i (Fig. S23), and a
representative gel image of the purified GST-FOXM1-DBD (Fig. S24).
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.bioorg.2019.103269.
[29] O. Trott, A.J. Olson, AutoDock Vina: Improving the speed and accuracy of docking with a
new scoring function, efficient optimization, and multithreading, J. Comput. Chem. 31
(2010) 455–461.
[30] M.J. Abraham, T. Murtola, R. Schulz, S. Páll, J.C. Smith, B. Hess, E. Lindahl, GROMACS:
High performance molecular simulations through multi-level parallelism from laptops to
supercomputers, SoftwareX 1–2 (2015) 19–25.
[31] A.W. Sousa da Silva, W.F. Vranken, ACPYPE - AnteChamber PYthon Parser interfacE.
BMC Res, Notes 5 (2012) 367.
References
[32] D. Systèmes, BIOVIA Discovery Studio Modeling Environment, 2015.
[33] R. Kumari, R. Kumar, A. Lynn, g_mmpbsa−GROMACS Tool for High-Throughput MM-
PBSA Calculations, J. Chem. Inf. Model. 54 (2014) 1951–1962.
[1] W. Korver, J. Roose, A. Wilson, H. Clevers, The Winged-Helix transcription factor trident
10