Imidazothiazole-based potent inhibitors of V600E-B-RAF kinase with promising anti-melanoma activity: biological and computational studies

Article abstract of DOI:10.1080/14756366.2020.1819260

A series of imidazothiazole derivatives possessing potential activity against melanoma cells were investigated for molecular mechanism of action. The target compounds were tested against V600E-B-RAF and RAF1 kinases. Compound 1zb is the most potent agains

Full text of DOI:10.1080/14756366.2020.1819260

Journal of Enzyme Inhibition and Medicinal Chemistry
ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ienz20
Imidazothiazole-based potent inhibitors of V600E-
B-RAF kinase with promising anti-melanoma
activity: biological and computational studies
Hanan S. Anbar , Mohammed I. El-Gamal , Hamadeh Tarazi , Bong S. Lee ,
Hong R. Jeon , Dow Kwon & Chang-Hyun Oh
To cite this article: Hanan S. Anbar , Mohammed I. El-Gamal , Hamadeh Tarazi , Bong S.
Lee , Hong R. Jeon , Dow Kwon & Chang-Hyun Oh (2020) Imidazothiazole-based potent
inhibitors of V600E-B-RAF kinase with promising anti-melanoma activity: biological and
computational studies, Journal of Enzyme Inhibition and Medicinal Chemistry, 35:1, 1712-1726,
DOI: 10.1080/14756366.2020.1819260
To link to this article: https://doi.org/10.1080/14756366.2020.1819260
© 2020 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group.
Published online: 23 Sep 2020.
Submit your article to this journal
Article views: 62
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=ienz20

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
2020, VOL. 35, NO. 1, 1712–1726
https://doi.org/10.1080/14756366.2020.1819260
RESEARCH PAPER
Imidazothiazole-based potent inhibitors of V600E-B-RAF kinase with promising
anti-melanoma activity: biological and computational studies
Hanan S. Anbara, Mohammed I. El-Gamalb,c,d, Hamadeh Tarazib,c, Bong S. Leee, Hong R. Jeonf, Dow Kwone and
Chang-Hyun Ohe,f,g,h
b
aDepartment of Clinical Pharmacy and Pharmacotherapeutics, Dubai Pharmacy College for Girls, Dubai, United Arab Emirates; Department of
c
Medicinal Chemistry, College of Pharmacy, University of Sharjah, Sharjah, United Arab Emirates; Sharjah Institute for Medical Research,
d
University of Sharjah, Sharjah, United Arab Emirates; Department of Medicinal Chemistry, Faculty of Pharmacy, University of Mansoura,
e
f
g
Mansoura, Egypt; CTC SCIENCE, Hwaseong, Gyeonggi-do, Republic of Korea; CTCBIO Inc., Hwaseong, Gyeonggi-do, Republic of Korea; Center
h
for Biomaterials, Korea Institute of Science & Technology (KIST), Seoul, Republic of Korea, Seoul; Department of Biomolecular Science,
University of Science & Technology (UST), Daejeon, Republic of Korea
ABSTRACT
ARTICLE HISTORY
Received 3 August 2020
Revised 26 August 2020
Accepted 31 August 2020
A series of imidazothiazole derivatives possessing potential activity against melanoma cells were investi-
gated for molecular mechanism of action. The target compounds were tested against V600E-B-RAF and
RAF1 kinases. Compound 1zb is the most potent against both kinases with IC50 values 0.978 and 8.2 nM,
respectively. It showed relative selectivity against V600E mutant B-RAF kinase. Compound 1zb was also
tested against four melanoma cell lines and exerted superior potency (IC50 0.18-0.59mM) compared to the
reference standard drug, sorafenib (IC50 1.95-5.45 mM). Compound 1zb demonstrated also prominent
selectivity towards melanoma cells than normal skin cells. It was further tested in whole-cell kinase assay
and showed in-cell V600E-B-RAF kinase inhibition with IC50 of 0.19 mM. Compound 1zb induces apoptosis
not necrosis in the most sensitive melanoma cell line, UACC-62. Furthermore, molecular dynamic and 3D-
QSAR studies were done to investigate the binding mode and understand the pharmacophoric features of
this series of compounds.
KEYWORDS
Apoptosis; imidazothiazole;
melanoma; modelling;
V600E-B-RAF
GRAPHICAL ABSTRACT
1. Introduction
over-expressed in several melanoma patients3,4. Inhibitors of V600E-B-
RAF and RAF1 (C-RAF) kinases have been reported as antiproliferative
agents against melanoma. Among which, vemurafenib5, encorafenib6,
and dabrafenib7,8 (Figure 1) have been marketed as RAF-inhibitory
anti-melanoma therapies. Several clinical trials are currently running
for these drugs either alone or in combination with other drugs. In
addition, sorafenib (Figure 1) is one of the potent inhibitors of RAF
kinases that was utilised as a reference standard in our cell-based
assay9. Moreover, several imidazothiazole derivatives were reported
Melanoma is malignant tumour of melanocytes that are present in
the skin and the eye. People with fair skin are more prone to mel-
anoma than dark-coloured skin individuals. In addition, blue,
green, or hazel-colored eyes are at higher risk to melanoma than
dark-coloured eyes1. Excessive exposure to sun light and family
history of melanoma are also among the risk factors of melanoma.
The World Health Organisation (WHO) indicates about 132,000
new melanoma skin cancer cases are diagnosed annually world-
wide. Moreover, the American Cancer Society estimates diagnosis
of more than 100,000 melanoma cases in USA only in 20202.
Much research efforts are required to develop more efficient and
less toxic anti-melanoma drugs.
as V600E-B-RAF or RAF1 kinase inhibitors10–12
We have previously reported a series imidazothiazole deriva-
tives as potential antiproliferative agents against melanoma13–15
.
.
In our previous reports, we did not investigate their molecular
mechanism of action. In the present study, we tested their
RAS-RAF-MEK-ERK pathway is among the potential targets for
melanoma treatment. V600E mutated B-RAF is over-activated and molecular mechanism of action against kinases in cell-free and
CONTACT Mohammed I. El-Gamal
Arab Emirates; Chang-Hyun Oh
malgamal@sharjah.ac.ae
Department of Medicinal Chemistry, College of Pharmacy, University of Sharjah, Sharjah, United
choh@kist.re.kr Center for Biomaterials, Korea Institute of Science & Technology (KIST), Seoul, Republic of Korea
ß 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1713
O
F
S
O
F
NH
H
N
O
S
N
F
NH
O
F
F
Cl
N
N
Cl
O
N
S
O
F
HN
N
S O
N
O
N
N
HN
O
N
H
NH2
Vemurafenib
Encorafenib
Dabrafenib
O
N
N
S
R
O
Cl
O
N
H
N
N
N
CF3
H
H
N
Sorafenib
N
H
N
X
N
H
Ar
O
The target compounds
R = p-F, m-OMe, m-OH
X = NH or none
Ar = (substituted) phenyl
Figure 1. Structures of vemurafenib, encorafenib, dabrafenib, sorafenib, and the target compounds.
whole-cell assays. The most potent compound was further tested exogenously introduced constitutively active human V600E-B-RAF
fused to the N-terminus of a modified oestrogen receptor (ER). In
this construct, V600E-B-RAF is only active in the presence of the
oestrogen-analogue hydroxytamoxifen. hydroxytamoxifen treat-
ment results in B-RAF-VE600E kinase activation and subsequent
phosphorylation of the direct V600E-B-RAF substrate MEK1 at
Ser218/222. MEF-V600E-B-RAF:ER cells were plated in Dulbecco’s
modified eagle medium (DMEM) supplemented with 10% FCS in
multiwell cell culture plates. Compound incubation (90 min at
37 ꢀC) was done in serum-free medium. Cells treated with 10 mM
sorafenib were used as low control. For stimulation, cells were
treated with 1 mM hydroxytamoxifen for 1 h at room temperature.
After cell lysis, quantification of MEK1 phosphorylation was
assessed in 96-well plates via sandwich-ELISA using a MEK1 spe-
cific capture antibody and a phospho-MEK1-Ser218/222 specific
detection antibody.
against melanoma and normal skin cells to investigate the select-
ivity indexes. The ability of the most potent compound to induce
apoptosis and necrosis were also tested. Moreover, different com-
putational studies such as molecular dynamic simulation and
QSAR were carried out.
2. Experimental
2.1. Synthesis of the target compounds
The methods are reported in our previously published articles13–15
.
2.2. Cell-free kinase profiling
In a final volume of 25 mL, the tested kinase (5–10 mU) was incu-
bated with 25 mM Tris pH 7.5, 0.02 mM EDTA, 0.66 mg/mL myelin
basic protein, 10 mM magnesium acetate and [c33P-ATP] (specific
activity approx. 500 cpm/pmol, concentration as required). The
reaction was initiated by the addition of the Mg-ATP mix. After
incubation for 40 min at room temperature, the reaction was
stopped by the addition of 5 mL of a 3% phosphoric acid solution.
10 mL of the reaction was then spotted onto a P30 filtermat and
washed three times for 5 min in 75 mM phosphoric acid and once
in methanol before drying and scintillation counting.
2.4. MTT assay
The melanoma and normal skin cell lines were maintained in
DMEM including 10% foetal bovine serum (FBS) and 1% penicillin/
streptomycin in a humid atmosphere containing 5% carbon diox-
ide at 37 ꢀC. The cells were taken from culture substrate with
0.05% trypsin-0.02% EDTA and plated at a density of 5 ꢁ 103 cells/
well in 96-well plates followed by incubation at 37 ꢀC for 24 h in a
humid atmosphere with 5% carbon dioxide CO2 before treatment
with various concentrations (3-fold serial dilution starting with
10 mM, 10 points) of the test compound 1zb and sorafenib. The
cells were incubated for 48 h following treatment with the test
2.3. In-cell kinase assay
In the cellular V600E-B-RAF phosphorylation assay the murine
embryonal fibroblast cell line (MEF) was used, which expresses compounds. The cellular viability was assessed by the

1714
H. S. ANBAR ET AL.
conventional 3–(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium within maestro environment and analysed using Desmond inter-
bromide (MTT) reduction assay. MTT assays were carried out with action diagram panel.
R
V
CellTiter 96 (Promega) according to the manufacturer’s instruc-
tions. The absorbance at 590 nm was recorded using EnVision
R
V
2.6.3. Flap 3D-QSAR model construction
2103 (Perkin Elmer; Boston, MA, USA). The IC50 values were calcu-
lated using GraphPad Prism 4.0 software. Triplicate testing
was performed.
The 3D-QSAR models for the RAF1 and V600E-B-RAF active com-
R
pounds were constructed employing the program FlapV (finger-
print for ligands and proteins)20,21
. The program calculates
fingerprints, which are derived from the GRID Molecular
Interaction Fields (MIFs) and are characterised as quadruplets of
pharmacophoric features. Model construction was initiated by
building a compound database for each enzyme. For the RAF1
enzyme, 9 active compounds with their available pIC50 values
were used for this purpose, whereas 22 active compounds with
their corresponding pIC50 values (Table 6) were used for the
V600E-B-RAF database creation. Next, the protonation states for
2.5. Caspase-3/7 and LDH release assays
They were performed following the protocols reported in our pre-
viously published article16.
2.6. Computational studies
2.6.1. Data set
each molecule at pH 7.4 were calculated using an integrated
R
ChemDraw UltraV software (http://www.cambridgesoft.com) was
R
Flap tool called MoKaV22. Subsequently, for each molecule, a
R
V
used for drawing the chemical structures, then the low energy
conformations were achieved by MOPAC2012 program
(MOPAC2012, J. J. P. Stewart, Stewart Computational Chemistry,
version 15.321 web: http://OpenMOPAC.net) embedded within
utilising the Austin Model-1 (AM1) semiemperical
force-field for accurate energy minimisation.
maximum of 50 conformers were generated with an RMSD value
of 0.3 Å. Ligand-based alignment was achieved employing the
most active compound 1zb in its low energy conformation as a
template. Later, for each database, a partial least square (PLS)
model was constructed employing the GRID-probes; H (shape),
DRY (hydrophobic), O (H-bond acceptor) and N1 (H-bond donor).
Model validation was accomplished via leave-one-out cross valid-
ation and the optimal latent-variables for each model were deter-
mined by investigating the R2 vs Q2 plot.
R
VEGA-ZZV
17,18
2.6.2. Docking and molecular dynamic simulation
As a prerequisite for molecular dynamic simulation, docking
experiments were carried-out using the program AutoDock Vina19.
The x-ray crystal structures of V600E-B-RAF kinase (PDB-ID: 3IDP)
and RAF1 kinase (PDB-ID: 3OMV) were retrieved from Protein Data
Bank (http://www.rcsb.org/pdb/). Bound inhibitors and crystalline
water molecules were extracted from the initial structures.
Autodock MGL Tools were used for the addition of polar hydro-
gens and charges. Compound 1zb was treated using the same
procedure. Spacing of 1.0 Å between the grid points was used to
establish grid boxes covering the active site of the studied macro-
molecules, centred towards the coordinates of 5.29 (x), 24.13 (y),
32.82 (z) for V600E-B-RAF and towards the coordinates of 28.61
(x), 39.19 (y), 39.09 (z) for RAF1 kinase. Exhaustiveness was set to
12, while number of poses was set to 10.
R
2.6.4. VolsurfþV QSAR model construction
Two additional QSAR models were constructed; one for the RAF1
enzyme, and another one for the V600E-B-RAF. In each case, the
molecular descriptors for the studied compounds were calculated
R
V
employing the software VolSurfþ 23,24. The software calculates a
set of 1D-3D molecular descriptors (ca. 128) of different classes
including molecular size/shape, hydrophilic/hydrophobic regions
quantification, INTEraction enerGY (INTEGY moments), capacity
factors, amphiphilic moments, hydro-lipo balance, molecular diffu-
sivity, LogP, LogD, pH-dependent solubility, molecular flexibility,
3D pharmacophoric descriptors, etc. During model development,
the calculated descriptors were recruited as independent variables,
whereas the bioactivity values (pIC50) for the compounds under
study were recruited as the dependent variable. Development and
validation of our QSAR models were achieved employing the
Molecular dynamic simulations for the compound 1zb with
respect to V600E-B-RAF and RAF1 were started from the earlier
R
docked complexes. Desmond softwareV (Desmond Molecular
R
QSARINSV software25,26. All subsets were investigated for the first
Dynamics System, version3.8, D. E. Shaw Research, New York, NY,
€
2014) embedded within Maestro interface (Schrodinger Release
two descriptors selection. Then, optimal combinations of descrip-
tors (greater than two) were reached by means of genetic algo-
rithm (GA). In the GA selection phase, the population size,
maximum number of generations and mutation rate were set to
800, 3000 and 0.6, respectively. Multiple linear regression (GA-
MLR) method was adapted for the final model building. The
€
2014–2: Maestro, version 9.8, Schrodinger, LLC, New York, NY,
2014) was used to conduct all-atoms molecular dynamic simula-
tions employing OPLS_2005 force field parameters. Each complex
was subjected to the same dynamic protocol, in which; TIP3P
explicit water molecules as solvent model within an orthorhombic
periodic boundary box of the size 10 Å3 were used to solvate the
protein-ligand complex, then, system neutralisation was accom-
plished by adding appropriate counter-ions followed by adding
0.15 M of salt ions. Prior to the actual dynamic run, system relax-
ation was achieved by performing a series of short (2000 itera-
tions) restrained and non-restrained solute minimizations steps
followed by short 12 ps simulation steps using NVT and NPT
robustness for each model was assessed employing the Q2
(leave one-out) cross validation procedure.
LOO
3. Results and discussion
3.1. Chemistry
ensembles. 50 ns production run was carried out using the NPT The target compounds 1a-zh were prepared utilising the path-
ensemble class integrating the equation of motion every 2 fs and ways illustrated in Schemes 1–313–15. The synthetic strategy
setting the temperature and pressure to 300ꢀK and 1 atmosphere, involved preparation of methyl sulphonyl intermediates 6a,b
respectively. Short-range interactions cut-off was set to 9 Å and (Scheme 1) and amine side chain intermediates 13a-c and 15a-f
the long-range electrostatic interactions were calculated employ- (Scheme 2), followed by interaction of them together to get the
ing the particle mesh Ewald (PME) method. Results were visualised aminopyrimidinyl final compounds (Scheme 3).

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1715
O
N
S
X
N
S
X
Br
X
S
N
S
N
N
H2N
c
a
b
+
X
N
N
N
N
X = 4-F, 3-OMe
3
4a,b
O
N
N
S
S
2a,b
O
5a,b
Scheme 1. Reagents and conditions: (A) EtOH, reflux, 16 h; (B) 4-iodo-2-(methylthio)pyrimidine, Pd(OAc)2, Cs2CO3, PPh3, DMF, 80ꢀC, 12 h; (C) oxone, MeOH, H2O, rt, 16 h.
6a,b
a
b
H2N
c
d
Cbz
N3
Cbz
OH
Cbz
OMs
OH
N
N
H
N
H
H
8
9
7
H
N
H
H
N
H
N
g
N
Cbz
H2N
R
R
N
R
H
e
O
O
11a-c
12a-c
Cbz
NH2
N
H
f
H
N
10
H
N
g
H2N
Cbz
R
N
O
H
O
14a-f
13a-f
Scheme 2. Reagents and conditions: (A) benzyloxycarbonyl chloride, TEA, CH2Cl2, 0 ꢀC; (B) methanesulfonyl chloride, TEA, CH2Cl2, 0 ꢀC; (C) NaN3, DMSO, 70 ꢀC, 2 h; (D)
PPh3, MeOH, H2O, reflux, 2 h; (E) appropriate isocyanate, THF, rt, 2 h; (F) appropriate carboxylic acid, HOBt, EDCI, TEA, DMF, 80 ꢀC, 8 h; (G) H2/Pd-C, MeOH, rt, 1 h.
HO
X
N
S
N
S
X
N
S
N
N
N
a
c
N
N
N
N
N
H
N
H
N
O
N
N
N
H
N
H
N
S
H
H
O
Ar
Ar
O
O
6a,b
1a-d: X = 4-F
1o-q: X = 3-OMe
1r-t
b
HO
S
N
N
S
X
N
N
c
N
N
N
N
H
H
N
N
H
N
N
H
Ar
Ar
O
O
1e-n: X = 4-F
1u-za: X = 3-OMe
1zb-zh
Scheme 3. Reagents and conditions: (A) 12a-c, DIPEA, DMSO, 80 ꢀC, 8 h; (B) 14a-f, DIPEA, DMSO, 80 ꢀC, 8 h; (C) BBr3, CH2Cl2, –78ꢀC, 30min then rt, 1 h.
Scheme 1 starts with a cyclisation reaction between 2-amino- of compounds 4a,b with 4-iodo-2-(methylthio)pyrimidine using
thiazole (3) and bromoacetophenone derivatives (2a,b) in reflux- palladium(II) acetate, caesium carbonate, and triphenylphosphine
ing ethanol to get imidazothiazole intermediates 4a,b27. Coupling as catalysts led to formation of diarylimidazothiazole intermediates

1716
H. S. ANBAR ET AL.
5a,b. Oxidation of the methylthio group of compounds 5a,b using
oxone yielded the corresponding methyl sulphonyl deriva-
tives 6a,b.
The most potent compound, 1zb, against V600E-B-RAF and
RAF1 was further tested at 1 mM concentration against a 30-kinase
panel belonging to different kinase families. The results are illus-
trated in Table 2. Apart from V600E-B-RAF, it inhibited 73.14% of
FLT3 kinase while its inhibitory effect against all the other kinases
was less than 50%. The IC50 of compound 1zb against FLT3 was
further examined and found to be 838 nM (Table 3). So the com-
pound is 856.85 times more selective towards V600E-B-RAF than
FLT3. It is also 8.38 folds more selective against V600E-B-RAF than
RAF1. Based on these results, compound 1zb can be considered
as a relatively selective V600E-B-RAF kinase inhibitor.
Scheme 2 illustrates the conversion of 2-aminoethanol into the
amino reagents possessing arylamide or arylurea moiety. At the
beginning, the amino group of 2-aminoethanol was Cbz-protected
using benzyloxycarbonyl chloride in presence of triethylamine
(TEA)28. The hydroxyl group of compound 7 was converted into
methanesulfonate using methanesulfonyl chloride and TEA29.
Compound 8 was heated with sodium azide to replace the OMs
group with azido (compound 9). Reduction of azido compound 9
into the corresponding amino was accomplished using triphenyl-
phosphine in refluxing methanol. The free primary amino group
of compound 10 was treated with aryl isocyanates or condensed
with carboxylic acids to form the corresponding urea or amide,
respectively. The last step of Scheme 2 involves deprotection of
Cbz group using hydrogen gas and palladium metal
over charcoal.
In Scheme 3, the mesyl intermediates 6a,b were heated with
the amino side chain reagents 12a-c and 14a-f in presence of
Hunig’s base to produce the target compounds possessing fluoro
or methoxy on the benzene ring at position 6 of the imidazothia-
zole nucleus. Boron tribromide was used to demethylate the
methoxy compounds 1o-q and 1 u-za to form the corresponding
phenolic analogues 1r-t and 1zb-zh, respectively. The structures
of the target compounds 1a-zh are drawn in Table 1.
3.3. Antiproliferative activity against melanoma cell lines
Compound 1zb was tested against four melanoma cell lines and
normal skin epithelial cell line. Its potency was compared with sor-
afenib as a reference standard as it is a potent RAF kinase inhibi-
tor. The results are illustrated in Table 4. Compound 1zb is more
potent than sorafenib against all the four tested melanoma cell
lines. Its IC50 values are within sub-micromolar range, while those
of sorafenib are in one-digit micromolar scale. The selectivity
indexes of compound 1zb against A375, MDA-MB-435, SK-MEL-28,
and UACC-62 melanoma cell lines compared to CCD1106K normal
skin cells are 20.13, 15.69, 24.37, and 51.44, respectively.
Furthermore, the strong potency of compound 1zb against the
tested melanoma cell lines that harbour V600E-B-RAF30 empha-
sises the postulation that V600E-B-RAF could be at least a part of
molecular mechanism(s) of antiproliferative action of com-
pound 1zb.
3.2. Kinase profiling
Based on our previous knowledge of inhibitory effect of imidazo-
thiazole derivatives against V600E-B-RAF and RAF1 11,12, we
decided to start our biological investigation with testing all the 34
final compounds against both kinases. The target compounds
were tested at 1 mM concentration, and the compounds that pro-
duced promising inhibitory effect, more than 75% inhibition, were
further tested in 10-dose mode to measure their IC50 values. The
inhibition percentage and IC50 values are demonstrated in Table
1. The target compounds are generally more active against V600E-
B-RAF than RAF1. A comparative computational study was carried
out to investigate it.
Meta-hydroxyl group on the phenyl ring at position 6 of the
imidazothiazole nucleus is more favourable for activity than meta-
methoxy and para-fluoro substituents against both V600E-B-RAF
and RAF1. This is emphasised by the activity of OH compounds
1zb-zh compared with OMe analogues 1 u-za or F derivatives 1e
and 1j-n. It is obvious that hydrogen bond donor at the this part
of the structure is optimal for stronger affinity with V600E-B-RAF
and RAF1 kinases. This was further investigated by computational
studies. In addition, amide compounds (1e, 1 u, 1x, 1zb, 1zd, and
1ze) are more active and more potent than the corresponding
urea analogues (1a, 1o, and 1q-t). It can be concluded that the
length, hydrophobicity, and electronic properties of amide spacer
are more optimal for activity than those of urea. Furthermore, ter-
minal benzamido moiety with no substitution (e.g. compounds
1 u and 1zb) or ortho-hydroxyl group (e.g. compounds 1v and
1zc) are more favourable for activity than analogues with bulkier
substituents. QSAR study was carried out to further investigate
this issue. Among all the final compounds, compound 1zb pos-
sessing 3-hydroxyphenyl at position 6 of imidazothiazole and
unsubstituted benzamido is the most potent against both V600E-
3.4. In-cell kinase assay
The next step of our study was investigation of ability of com-
pound 1zb to inhibit V600E-B-RAF kinase inside the cells after
penetrating the cell membrane. It was compared with sorafenib as
a reference standard. This assay was done in murine embryonal
fibroblast (MEF) cell line and the ability to inhibit V600E-B-RAF
was investigated through measuring the concentration of phos-
phorylated MEK, the downstream protein of V600E-B-RAF. The
dose-response curves are illustrated in Figure 2. Compound 1zb
could cross the cell membrane and inhibit V600E-B-RAF with IC50
value of 0.19 mM. This concentration is comparable to its IC50 value
against UACC-62 melanoma cell line (0.18 mM) that has over-
expressed V600E-B-RAF30. This result can also prove that V600E-B-
RAF inhibition could be
against melanoma.
a molecular mechanism of action
3.5. Apoptosis and necrosis assays
As per the results shown in Table 4, UACC-62 is the most sensitive
melanoma cell line for compound 1zb. Its IC50 value is 0.18 mM.
We decided to investigate whether this concentration could
induce apoptosis and/or necrosis in UACC-62 cells. Caspase-3/7
were measured as proof of apoptosis induction while lactate
dehydrogenase (LDH) release was measured as an indicator of
necrosis16. At 0.18 mM concentration, compound 1zb increased
caspase-3/7 release by 53.33% but did not show any significant
increase in LDH release (Table 5). This means that compound 1zb
at its IC50 concentration can induce apoptosis of UACC-62 melan-
B-RAF and RAF1. It is the only compound that exerted sub-nano- oma cells but not necrosis. This finding could help us understand
molar IC50 value against V600E-B-RAF. its antiproliferative effect in more details.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1717
Table 1. Structures of the target compounds 1a-zh and their inhibitory effects against RAF1 and V600E-B-RAF kinases. The inhibition percentage results are
expressed as means of duplicate assays SEM
One-dose (1 mM) results
against RAF1
One-dose (1 mM) results
against V600E-B-RAF
(% inhibition)
IC50 value (nM)
against RAF1
IC50 value (nM) against
V600E-B-RAF
Compound no.
1a
Structure
(% inhibition)
N
N
N
N
S
S
S
S
54.04% 5.07%
63.81% 2.51%
59.53% 2.25%
63.44% 1.62%
67.62% 3.61%
ND
ND
ND
ND
ND
F
F
F
F
N
N
N
N
N
N
H
N
H
N
N
H
O
1b
1c
1d
73.06 1.72%
245 nM
N
N
H
N
H
N
N
H
O
O
52.03% 3.61%
ND
N
N
H
N
H
N
N
H
CF
3
O
69.37% 0.05%
ND
N
N
H
N
H
N
N
H
CF
3
O
F C
3
N
S
S
S
S
S
1e
1f
62.63% 0.34%
71.39% 1.32%
13.99% 2.28%
38.67% 5.01%
55.46% 1.61%
90.43% 0.72%
90.99% 0.56%
38.64% 0.21%
77.90% 0.11%
85.45% 0.42%
ND
ND
ND
ND
ND
144 nM
83.8 nM
ND
F
F
F
F
F
N
N
N
H
N
N
H
O
N
N
N
N
N
N
N
H
N
N
H
O
N
N
N
1g
1h
1i
N
N
H
N
N
H
O
O
O
O
191 nM
142 nM
N
N
O
H
N
N
H
O
N
N
O
H
N
N
H
O
(continued)

1718
H. S. ANBAR ET AL.
Table 1. Continued.
One-dose (1 mM) results
against RAF1
One-dose (1 mM) results
against V600E-B-RAF
(% inhibition)
IC50 value (nM)
against RAF1
IC50 value (nM) against
V600E-B-RAF
Compound no.
1j
Structure
(% inhibition)
N
S
69.84% 0.52%
66.14% 2.60%
13.09% 0.12%
75.89% 1.43%
91.40% 0.48%
ND
ND
ND
ND
78.2 nM
F
N
N
HO
N
H
N
N
H
O
N
S
1k
1l
ꢂ5.63 3.11%
ND
F
F
F
N
N
CF
3
N
H
N
N
H
O
N
S
40.59% 1.72%
ND
N
N
CF
3
N
H
N
N
H
O
CF
3
N
S
1m
80.25% 0.61%
82.6 nM
N
N
CF
3
N
H
N
N
H
O
N
N
N
S
1n
1o
75.69% 0.67%
84.82% 0.42%
ND
ND
120 nM
F
N
N
CF
3
N
H
N
H
N
O
N
O
O
O
O
32.50% 1.91%
52.76% 3.66%
ND
N
S
N
N
N
H
N
H
N
N
H
O
1p
1q
60.68% 0.95%
69.65% 0.40%
ND
ND
ND
ND
N
S
N
N
N
H
N
H
N
N
H
CF
O
3
19.33% 2.37%
23.59% 1.99%
N
S
N
N
N
H
N
H
H
N
N
CF
O
3
F C
3
HO
1r
97.45% 0.10%
97.66% 0.16%
45.6 nM
11.2 nM
N
S
N
N
N
H
N
H
N
N
H
O
(continued)

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1719
Table 1. Continued.
One-dose (1 mM) results
against RAF1
One-dose (1 mM) results
against V600E-B-RAF
(% inhibition)
IC50 value (nM)
against RAF1
IC50 value (nM) against
V600E-B-RAF
Compound no.
Structure
(% inhibition)
HO
1s
97.28% 0.08%
99.54% 0.23%
49.7 nM
18.4 nM
N
S
N
N
N
H
N
H
N
N
H
CF
O
3
HO
1t
67.56% 4.79%
87.74% 0.38%
ND
194 nM
N
S
N
N
N
H
N
H
N
N
H
CF
O
3
F C
3
O
1u
1v
74.00% 0.64%
59.25% 2.53%
56.61% 2.15%
95.03% 0.66%
90.82% 0.25%
83.65% 0.47%
ND
ND
ND
45.1 nM
100 nM
121 nM
N
S
N
N
N
H
N
N
H
O
O
N
S
N
N
HO
N
H
N
N
H
O
O
1w
N
S
N
N
CF
3
N
H
N
N
H
O
O
1x
1y
32.14% 0.08%
56.21% 0.10%
ND
ND
ND
ND
N
S
N
N
CF
3
N
H
N
N
H
O
CF
3
O
57.47% 0.47%
63.23% 0.93%
N
S
N
N
CF
3
N
H
N
N
H
O
N
N
O
1z
32.12% 2.26%
79.41% 0.56%
ND
ND
241 nM
N
S
N
N
CF
3
N
H
O
N
N
N
H
O
O
1za
51.51% 3.74%
71.56% 3.43%
ND
N
S
N
N
CF
3
N
H
N
N
H
O
N
O
(continued)

1720
H. S. ANBAR ET AL.
Table 1. Continued.
One-dose (1 mM) results
against RAF1
One-dose (1 mM) results
against V600E-B-RAF
(% inhibition)
IC50 value (nM)
against RAF1
IC50 value (nM) against
V600E-B-RAF
Compound no.
1zb
Structure
(% inhibition)
HO
99.39% 0.20%
99.32% 0.23%
98.71% 0.23%
86.02% 1.08%
99.48% 0.09%
101.07% 1.53%
100.24% 0.17%
97.81% 1.67%
8.2 nM
0.978 nM
N
S
N
N
N
H
N
N
H
O
HO
1zc
10.5 nM
16.7 nM
468.0 nM
1.9 nM
N
S
N
N
HO
N
H
N
N
H
O
HO
1zd
1ze
3.4 nM
N
S
N
N
CF
3
N
H
N
N
H
O
HO
28.8 nM
N
S
N
N
CF
3
N
H
N
N
H
O
CF
3
HO
1zf
92.71% 0.99%
97.19% 0.32%
151.0 nM
22.2 nM
N
S
N
N
CF
3
N
H
N
N
H
O
N
N
HO
1zg
1zh
90.78% 0.22%
98.97% 0.43%
158.0 nM
62.2 nM
15.8 nM
7.4 nM
N
S
N
N
CF
3
N
H
O
N
N
N
H
O
HO
96.60% 0.18%
100.66% 2.35%
N
S
N
N
CF
3
N
H
N
N
H
O
N
O
system stability and trajectories were assessed in term of root-
mean-squared deviations (RMSD) of protein Ca-atoms as well as
the ligand heavy atoms. Figure 3(A,B) illustrate the RMSD values
(compared to the starting frame-0) along the simulation time,
3.6. Computational studies
3.6.1. Molecular dynamic study
The structure-activity relationships presented in Table 1 were fur-
ther explored via computational techniques in an attempt to pro- where changes of the order of 1–3 Å are perfectly acceptable for
vide plausible explanation for the bioactivity vibrations existed small, globular proteins. Accordingly, in the case of RAF1 enzyme-
among our compound series against both enzymes (RAF1 and ligand complex system equilibration was achieved after 5 ns, while
V600E-B-RAF). Thus, we started our endeavour by performing for the V600E-B-RAF; system equilibration was achieved after 18 ns
molecular docking procedure for the most active compound 1zb and remained stable during the rest of the simulation time.
against the two enzymes. Then, best docked-complexes achieved
Interactions of compound 1zb and the RAF1 kinase revealed a
were utilised to carry-on a 50 nanosecond (ns) molecular dynamic number of important findings. The 3-hydroxyphenyl group (OH)
simulations under explicit hydration environment. In each case, was able to establish hydrogen bonding with the Cys-424 residue

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1721
Table 2. Kinase profiling of compound 1zb (1 mM) against
a 30-kinase panel. The results are expressed as mean
%inhibition SEM
for about 65% of the simulation time (direct and water bridge).
Additional water bridge interaction was observed between the
benzamide carbonyl oxygen (C¼O) and the corresponding gate-
keeper residue Thr-421 for 58% of the simulation time (Figure
4(A,B)). Furthermore, an interesting intra-molecular hydrogen
bonding was observed between the benzamide nitrogen and the
pyrimidine nitrogen last for 41% of time, which is indicative for
conformational flexibility restriction. Other significant interaction
were the hydrophobic interactions with the residues Phe-475, Leu-
406, and Ile-355.
On the other hand, an analysis of compound 1zb within the
V600E-B-RAF kinase revealed another set of interactions. The 3-
hydroxyphenyl group (OH) was able to establish two hydrogen
bonds (direct and water-bridged); one with the hinge Cys-532 resi-
due and another one with the Gln-530 residues for about 83%
and 65% of the simulation time, respectively. Additional direct H-
bond interaction was observed between the benzamide carbonyl
oxygen (C¼O) and the corresponding residue Ser-536 for about
45% of the simulation time (Figure 5(A,B)). The residue Gly-534
was able to form H-bond with the amino group (NH) next to pyr-
imidine ring for about 49% of the simulation time. Other hydro-
phobic interactions were those with the residues Phe-583 and
Phe-595.
Kinase
% inhibition
ABL
ALK
7.57% 2.5%
41.16% 1.5%
9.14% 1.0%
49.77% 5.5%
23.37% 0.9%
16.96% 0.8%
0.84% 0.1%
19.93% 2.0%
2.17% 0.9%
13.67% 2.8%
73.14% 2.0%
12.19% 0%
ꢂ0.26% 0%
3.76% 0.5%
0.76% 0.1%
18.76% 1.0%
9.63% 0%
A-RAF
Aurora-A
B-RAFwt
BTK
DMPK
EGFR
ErbB2
ErbB4
FLT3
FMS
JAK1
JAK2
JNK1(a1)
KDR
cKIT
LCK
12.45% 2.5%
1.50% 0.2%
26.60% 2.5%
7.83% 0.3%
12.54% 1.5%
5.85% 0%
49.99% 2.3%
2.02% 0.3%
1.57% 0.3%
15.65% 0.5%
46.72% 1.5%
3.93% 0.5%
99.48% 0.09%a
MEK1
MEK2
mTOR
PDGFRb
PI3Ka
PI3K-C2a
ROS
Src(T341M)
STK33
TRKA
R
3.6.2. FlapV 3D-QSAR study
At the next level of our investigation, we were interested in gain-
ing more detailed information about the crucial factors affecting
selectivity in each enzyme case, and the contribution of structural
variations to the observed potency, thus we decided to conduct a
VRK1
V600E-B-RAF
aData taken from Table 1.
R
3D-QSAR analysis employing the program FlapV20,21. The program
Table 3. IC50 values of compound 1zb against V600E-B-RAF,
RAF1, and FLT3 kinases.
frequently used for building QSAR models based on molecular
interaction fields (MIFs) generated through the GRID force-field31.
Four different GRID-probes were utilised for calculating the final
interaction fields including; H-probe for computing the shape con-
straints on small ligand or protein cavity, DRY-probe for comput-
ing favourable hydrophobic interactions, O-probe (carbonyl
oxygen) acting as hydrogen bond acceptor (HBA) and N1-probe
(amidic NH) acting as hydrogen bond donor (HBD) probe.
Kinase
IC50 (nM)
V600E-B-RAF
RAF1
0.978a
8.2a
FLT3
838
aData taken from Table 1.
In the case of RAF1 compounds, the best developed model
Table 4. Antiproliferative activity (IC50 value, mM) of compound 1zb and sorafe-
nib against melanoma cell lines and normal skin cells.a
was a two-Latent variables model with an R2 of 0.799 (Figure 6(A))
Cell line
A375
MDA-MB-435
SK-MEL-28
UACC-62
1zb
Sorafenib
Table 5. Caspase-3/7 and LDH release assay results of compound 1zb at
0.18 mM concentration over UACC-62 melanoma cells.
Melanoma cell lines
0.46 0.03
0.59 0.06
0.38 0.05
0.18 0.02
9.26 0.38
5.45 0.53
3.16 0.20
2. 67 0.15
1.95 0.08
11.57 0.46
Assay
Result (compared to control cells treated with DMSO)
Caspase-3/7
LDH release
53.33% increase
0% increase
Normal skin epithelial cell line
CCD1106K
aThe results are expressed as means of triplicate assays SEM
Figure 2. Inhibitory impact of compound 1zb (left) and sorafenib (right) on the cellular kinase activity of V600E-B-RAF in murine embryonal fibroblast (MEF) cell line.

1722
H. S. ANBAR ET AL.
Figure 3. The 50 nanoseconds simulation RMSD values for the proteins and ligands involved; (A) RAF1 kinase Ca-atoms (blue) and compound 1zb heavy atoms (red),
(B) V600E-B-RAF kinase Ca-atoms (blue) and compound 1zb heavy atoms (red).
Figure 4. 50 nanoseconds simulation interaction diagram panel for 1zb and RAF1 kinase; (A) Fractions of interaction between 1zb and RAF1 kinase active site residues,
(B) 2 D- interaction diagram of 1zb within the RAF1 kinase active site.
Figure 5. 50 nanoseconds simulation interaction diagram panel for 1zb and V600E-B-RAF; (A) Fractions of interaction between 1zb and V600E-B-RAF kinase active site
residues, (B) 2 D-interaction diagram of 1zb within the V600E-B-RAF kinase active site.
consisted of three interaction fields encoding shape, hydrophobic the shape constraints would have negative impact on the RAF1
and HBA regions. Alignment of the most potent compound (1zb, bioactivity and only small compacted substituents could
IC50 ¼ 8.2 nM) to the model (Figure 7(A)) revealed a good match be tolerated.
In the case of V600E-B-RAF compounds, the best developed
model was a three-Latent variables model with an R2 of 0.857
(Figure 6(B)) consisted of three interaction fields encoding shape,
hydrophobic and HBA regions. Again, alignment of the most
potent compound (1zb, IC50 ¼ 0.978 nM) to the model (Figure
between its structural components and the given fields, where
the amino group (NH) next to pyrimidine ring is matching the
HBA field (red) and the benzamide ring is matching well with the
hydrophobic field (yellow) and there is no shape violation
observed at the benzamide ring. On the other hand, the least
potent compounds 1ze, 1zg and 1zf exhibited the same structural 8(A)) revealed a good match between its structural components
components match except for the shape field. The trifluoromethyl and the given fields, where the 3-hydroxyphenyl group (OH) is
groups on the benzamide ring of compound 1ze clearly violating matching the HBA (red) and the shape (blue mesh) fields. The
the shape constrains (Figure 7(B)). The same phenomenon was benzamide ring is matching well with the hydrophobic field (yel-
observed for the 4-morpholino substitution of 1zg (Figure 7(C)) low) and there is no shape violation observed at the benzamide
and the 5-methylimidazole ring substitution of 1zf (Figure 7(D)). ring. The least potent compound (1 b, IC50 ¼ 245 nM) suffers from
Accordingly, it seems that any benzamide substitution violating of many drawbacks as observed in Figure 8(B). The 4-fluorophenyl

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1723
R
Figure 6. Scatter plots of experimental vs. predicted bioactivities (expressed as pIC50) derived from; (A) RAF1 FlapV 3 D-QSAR model (2-Latent variables, R2¼0.799), (B)
R
V600E-B-RAF FlapV 3 D-QSAR model (3-Latent variables, R2¼0.857).
R
V
Figure 7. Flap 3 D-QSAR model generated for the RAF1 compounds; (A) alignment of the most potent compound 1zb, (B) alignment of compound 1ze, (C) align-
ment of compound 1zg, (D) alignment of compound 1zf. The GRID molecular interaction fields are shown as; cyan (Shape), yellow (hydrophobic) and red (HBA).
group deviates from the perfect shape match and provide no linker. Furthermore, small compacted substituents at the ortho
complementarity with the HBA field. Additionally, the presence of and meta positions of the benzamide ring are of more positive
urea linker appeared inferior leading to misalignment of the phe- impact for the bioactivity.
nylurea ring from the perfect hydrophobic ring match. In com-
pound 1z (Figure 8(C)), the presence of 3-methoxyphenyl (HBA) in
R
front of HBA field and the shape constraints violation by the ben- 3.6.3. VolsurfþV QSAR study
zamide ring substitutions (3-trifluoromethyl and 4-morpholino) Later, we turned our focus towards developing QSAR models for
appeared inferior to the bioactivity. The lower potency observed the RAF1 and V600E-B-RAF enzymes based on multiple linear
for compound 1t (Figure 8(D)) could be attributed to the presence regressions (MLR) analysis of molecular descriptors (ca. 128) gener-
R
V
of urea linker and the phenylurea ring misalignment. In light of ated by VolSurfþ software23,24 so that to gain more insight of
this information, it seems that the presence of small HBD group structural trends responsible for activity discrimination. Two statis-
such as (OH) at meta-position of the phenyl ring is highly recom- tically significant models were achieved and found to fit well with
mended, and the amide linker is more preferred over the urea the experimental biodata (Figures 9(A,B)) as indicated by their

1724
H. S. ANBAR ET AL.
R
V
Figure 8. Flap 3 D-QSAR model generated for the V600E-B-RAF compounds; (A) alignment of the most potent compound 1zb, (B) alignment of compound 1b, (C)
alignment of compound 1z, (D) alignment of compound 1t. The GRID molecular interaction fields are shown as; cyan (Shape), yellow (hydrophobic) and red (HBA).
Figure 9. Experimental versus predicted bioactivities scatter graphs derived from (A) The developed QSAR equation for the RAF1 kinase. (B) The developed QSAR equa-
tion for the V600E-B-RAF kinase. In each graph, solid line is the regression line, green dashed-lines representing the 95% confidence limit, while the red dashed-lines
representing the 95% prediction limits.
corresponding Fischer’s value (F), squared correlation coefficients logarithmic averaged differences between maximum and minimum
(R2), and the standard errors of estimate (s).
distances of atom “i” in a selected conformer.
The two descriptors found to negatively influence the bioactiv-
ity as demonstrated by their corresponding negative coefficients,
hence, for more improved RAF1 inhibitory activity, the model sug-
gested lesser molecular flexibility which is in agreement with the
intra-molecular HB observed during molecular dynamic simulation.
Furthermore, distribution of hydrophilic moieties towards the ter-
minal ends of the structure would greatly improve the activity
(Table 6).
The best developed model capable of describing the inhibitory
pattern exerted by our compounds against the C-RAF enzyme is
illustrated by the following equation;
pIC50 ¼ ꢂ35:647 ꢁ IW1 ꢂ 1:367 ꢁ FLEX þ 13:505
n ¼ 9, F-Statistic ¼ 102.77, R2 ¼ 0.972, Q2
¼ 0.946, s ¼
LOO
0.114
On the other hand, the best developed model for the V600E-B-
RAF enzyme is illustrated by the following equation.
where in this model; IW1 is the INTEGY (INTEraction enerGY)
moment presented as vector pointing from centre of mass to the
centre of the hydrophilic region, Low INTEGY descriptors such as IW1
indicates the hydrophilic moieties are either close to the centre of
mass or they balance at an opposite ends of the molecule. Flex is
the maximum flexibility of the molecule estimated after the random
generation of 50 conformers; such descriptor is calculated as the 0.238
pIC50 ¼ 0:111 ꢁ ACDODO ꢂ 17:193 ꢁ IW2 ꢂ 0:339 ꢁ DD8 ꢂ 0:062
ꢁ PSA þ 14:780
n ¼ 22, F-Statistic ¼ 41.33, R2 ¼ 0.907, Q2
¼ 0.852, s ¼
LOO

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1725
R
V
Table 6. The calculated VolSurfþ descriptors employed for the QSAR models
substituted with a small group. The most promising compound
that possesses all these pharmacophoric features is compound
1zb. It showed extremely high potency and relative selectivity
towards V600E-B-RAF. Moreover, compound 1zb exerted high
potency against different melanoma cell lines with sub-micromolar
IC50 values. It also showed high selectivity towards melanoma cells
than normal skin cells. Furthermore, it could penetrate the cell
membrane and inhibit V600E-B-RAF kinase with IC50 value of
0.19 mM that is comparable to its IC50 value against UACC-62 mel-
anoma cells (0.18 mM). At 0.18 mM concentration, it induced cas-
pase-3/7 release that lead to apoptosis. These promising results of
compound 1zb especially makes it a promising drug candidate for
further consideration and optimisation to end up with a new
promising anti-melanoma agent.
construction.
pIC50
(V600E- pIC50
Compound B-RAF) (RAF1) IW1
IW2 FLEX
PSA
ACDODO DD8
1b
1e
1f
1h
1i
1j
1m
1n
1r
1s
1t
1u
1v
1w
1z
1zb
1zc
1zd
1ze
1zf
1zg
1zh
6.611
6.842
7.077
6.719
6.848
7.107
7.083
6.921
7.951
7.735
6.712
7.346
7.000
6.917
6.618
9.010
8.721
8.469
7.541
7.654
7.801
8.131
–
–
–
–
–
–
–
–
0.031 0.081 3.740 122.660 12.117 0.250
0.045 0.112 2.772 99.200
0.049 0.126 2.834 99.200
0.040 0.092 3.433 122.060
0.024 0.054 3.577 122.060
7.005 1.500
7.135 0.125
7.896 0.000
7.773 0.375
0.033 0.077 3.048 119.430 14.874 0.875
0.012 0.028 4.289 119.940
0.011 0.021 4.106 119.840
6.937 0.625
8.935 2.875
7.341 0.033 0.087 3.748 126.060 26.359 0.000
7.304 0.017 0.056 4.217 131.160 25.881 1.750
–
–
–
–
–
0.046 0.118 4.984 136.260 23.780 0.250
0.027 0.078 3.353 105.230
0.024 0.066 3.592 116.660
0.037 0.095 4.087 110.330
0.025 0.065 4.241 125.870
6.998 0.250
6.230 0.500
5.902 0.250
6.422 0.250
Acknowledgements
8.086 0.027 0.075 3.237 114.030 23.318 1.000
7.979 0.018 0.047 3.510 134.260 25.265 0.000
7.777 0.013 0.044 3.874 119.130 16.949 0.000
6.330 0.048 0.108 3.975 124.230 25.509 1.625
6.821 0.021 0.061 4.241 134.770 23.943 0.250
6.801 0.020 0.045 4.350 134.670 28.806 2.750
7.206 0.018 0.041 4.085 134.670 23.955 0.500
The authors thank CTCBIO Inc., Korea Institution of Science and
Technology (KIST) (project #2E27930), and University of Sharjah
(project #16011101018-P & operational fund of Drug Design and
Discovery research group) for financial support.
Disclosure statement
The authors have declared no conflict of interest
where, ACDODO is a 3D-pharmacophoric descriptor based on the
triplets of pharmacophoric points, where it represents an
Acceptor-Donor-Donor triplet. IW2 is the INTEGY (INTEraction
enerGY) moment presented as vector pointing from centre of
mass to the centre of the hydrophilic region, Low INTEGY descrip-
tors such as IW2 indicates the hydrophilic moieties are either close
to the centre of mass or they balance at an opposite ends of the
molecule. DD8 is the difference between the maximum hydropho-
bic volumes (obtained upon variation of ligand conformations)
and the hydrophobic volume (D8) of the imported 3D structure
calculated at the 8 levels of energy. Furthermore, the hydrophobic
volume (D8) is the molecular envelope generating attractive
hydrophobic interactions calculated using the DRY-probe at an
energy range of ꢂ1.6 kcal/mol. PSA is the Polar Surface Area cal-
culated via the sum of polar region contributions.
Among the emerged descriptors, three were found to nega-
tively influence activity (IW2, DD8 and PSA), while one descriptor
(ACDODO) found to have positive impact. Thus, the model sug-
gested the needed requirements for improved V600E-B-RAF bio-
activity as follow; 1) the presence of three pharmacophoric points
(Acceptor-Donor-Donor) is greatly preferred, which is consistent
with the dynamic profile discussed earlier, 2) the proper position-
ing of the small hydrophilic moieties towards the terminal rings is
highly advised and 3) the proper hydrophilic-lipophilic balance is
highly desirable (Table 6).
Funding
This work was supported by the Korea Institute of Science and
Technology (KIST), Seoul, Republic of Korea (grant #2E27930) and
University of Sharjah, Sharjah, United Arab Emirates (grant
#16011101018-P).
References
1. https://www.who.int/news-room/q-a-detail/ultraviolet-(uv)-
radiation-and-skin-cancer. Accessed on Aug 3, 2020.
2. https://www.cancer.org/cancer/melanoma-skin-cancer/about/
key-statistics.html. Accessed on Aug. 3rd, 2020.
3. Rajagopalan H, Bardelli A, Lengauer C, et al. Tumorigenesis:
RAF/RAS oncogenes and mismatch-repair status. Nature
2002;418:934.
4. Abdel-Maksoud M, El-Gamal M, Gamal El-Din M, et al.
Design, synthesis, in vitro anticancer evaluation, kinase
inhibitory effects, and pharmacokinetic profile of new 1,3,4-
triarylpyrazole derivatives possessing terminal sulfonamide
moiety. J Enzyme Inhib Med Chem 2019;34:97–109.
5. Bollag G, Hirth P, Tsai J, et al. Clinical efficacy of a RAF
inhibitor needs broad target blockade in BRAF-mutant mel-
anoma. Nature 2010;467:596–9.
6. Honndorf VS, Coudevylle N, Laufer S, et al. Dynamics in the
p38alpha MAP kinase-SB203580 complex observed by
liquid-state NMR spectroscopy. Angew Chem Int Ed Engl
2008;47:3548–51.
7. Rheault TR, Stellwagen JC, Adjabeng GM, et al. Discovery of
Dabrafenib: a selective inhibitor of Raf kinases with antitu-
mor activity against B-Raf-driven tumors. ACS Med Chem
Lett 2013;4:358–62.
4. Conclusion
In this study, we studied the molecular mechanism of action of a
promising series of imidazothiazole derivatives that are potent
against melanoma cells. The target compounds were tested
against V600E-B-RAF and RAF1 kinases. The structure-activity rela-
tionship study revealed the importance of the following structural
moieties for activity; meta-hydroxyphenyl moiety at position 6 of
the imidazothiazole nucleus, amide linker, and terminal benzene
ring attached to the amide spacer either unsubstituted or
8. Karoulia Z, Gavathiotis E, Poulikakos PI. New perspectives for
targeting RAF kinase in human cancer. Nat. Rev. Cancer
2017;17:676–91.

1726
H. S. ANBAR ET AL.
9. Wilhelm SM, Carter C, Tang L, et al. BAY 43-9006 exhibits 20. Baroni M, Cruciani G, Sciabola S, et al. A common reference
broad spectrum oral antitumor activity and targets the RAF/
MEK/ERK pathway and receptor tyrosine kinases involved in
tumor progression and angiogenesis. Cancer Res 2004;64:
7099–109.
framework for analyzing/comparing proteins and ligands.
Fingerprints for Ligands and Proteins (FLAP): theory and
application. J Chem Inf Model 2007;47:279–94.
21. Cross S, Baroni M, Goracci L, et al. GRID-Based Three-
10. Lapierre J-M, Namdev ND, Ashwell MA, et al. RAF inhibitors
and their uses. PCT Pat Appl WO 2007/123892.
11. Abdel-Maksoud MS, Kim M-R, El-Gamal MI, et al. Design, syn-
thesis, in vitro antiproliferative evaluation, and kinase inhibi-
Dimensional Pharmacophores I: FLAPpharm,
a
Novel
Approach for Pharmacophore Elucidation. J Chem Inf Model
2012;52:2587–98.
22. Milletti F, Storchi L, Sforna G, et al. New and original pKa
prediction method using grid molecular interaction fields. J
Chem Inf Model 2007;47:2172–81.
23. Crivori P, Cruciani G, Carrupt P-A, et al. Predicting blood-
brain barrier permeation from three-dimensional molecular
structure. J Med Chem 2000;43:2204–16.
24. Cruciani G, Crivori P, Carrupt P-A, et al. Molecular fields in
quantitative structure-permeation relationships: the VolSurf
approach. Theochem 2000;503:17–30.
25. Gramatica P, Chirico N, Papa N, et al. QSARINS: a new soft-
ware for the development, analysis, and validation of QSAR
MLR models. J Comput Chem 2013;34:2121–32.
26. Gramatica P, Cassani S, Chirico N. QSARINS-chem: Insubria
datasets and new QSAR/QSPR models for environmental
pollutants in QSARINS. J Comput Chem 2014;35:1036–44.
27. Ashwell M, Tandon M, Lapierre J-M. Synthesis of imidazoox-
azole and imidazothiazole inhibitors of p38 map kinase. PCT
Patent Application WO 2006044869, 2006.
28. Tully DC, Liu H, Chatterjee AK, et al. Synthesis and SAR of
arylaminoethyl amides as noncovalent inhibitors of cathe-
psin S: P3 cyclic ethers. Bioorg Med Chem Lett 2006;16:
5112–7.
29. Townsend CA, Basak A. Experiments and speculations on
the role of oxidative cyclization chemistry in natural product
biosynthesis. Tetrahedron 1991;47:2591–602.
30. Sumimoto H, Imabayashi F, Iwata T, et al. The BRAF-MAPK
signaling pathway is essential for cancer-immune evasion in
human melanoma cells. J Exp Med 2006;203:1651–6.
31. Carosati E, Sciabola S, Cruciani G. Hydrogen bonding interac-
tions of covalently bonded fluorine atoms: From crystallo-
graphic data to a new angular function in the GRID force
field. J Med Chem 2004;47:5114–25.
tory effects of
a new series of imidazo[2,1-b]thiazole
derivatives. Eur J Med Chem 2015;95:453–63.
12. Ammar UM, Abdel-Maksoud MS, Ali EMH, et al. Structural
optimization of imidazothiazole derivatives affords a new
promising series as B-Raf V600E inhibitors; synthesis, in vitro
assay and in silico screening. Bioorg Chem 2020;100:103967
13. Park J-H, Oh C-H. Synthesis of new 6-(4-fluorophenyl)-5-(2-
substituted pyrimidin-4-yl)imidazo[2,1-b]thiazole derivatives
and their antiproliferative activity against melanoma cell
line. Bull Korean Chem Soc 2010;31:2854–60.
14. Park J-H, El-Gamal MI, Lee YS, et al. New imidazo[2,1-b]thia-
zole derivatives: Synthesis, in vitro anticancer evaluation,
and in silico studies. Eur J Med Chem 2011;46:5769–77.
15. Abdel-Maksoud MS, El-Gamal MI, Gamal El-Din MM, et al.
Broad-spectrum antiproliferative activity of a series of 6-(4-
fluorophenyl)-5-(2-substituted
pyrimidin-4-yl)imidazo[2,1-
b]thiazole derivatives. Med Chem Res 2016;25:824–33.
16. El-Gamal MI, Abdel-Maksoud MS, Gamal El-Din MM, et al.
Synthesis, in vitro antiproliferative and antiinflammatory
activities, and kinase inhibitory effects of new 1,3,4-triaryl-
pyrazole derivatives. Anticancer Agents Med Chem 2017;17:
75–84.
17. Pedretti A, Villa L, Vistoli G. Atom-type description language:
a universal language to recognize atom types implemented
in the VEGA program. Theor Chem Acc 2003;109:229–32.
18. Pedretti A, Villa L, Vistoli G. VEGA – An open platform to
develop chemo-bio-informatics applications, using plug-in
architecture and script programming. J Comput Aid Mol Des
2004;18:167–73.
19. Trott O, Olson AJ. AutoDock Vina: improving the speed and
accuracy of docking with a new scoring function, efficient
optimization, and multithreading. J Comput Chem 2010;31:
455–61.