Potent Pan-Raf and Receptor Tyrosine Kinase Inhibitors Based on a Cyclopropyl Formamide Fragment Overcome Resistance

Article abstract of DOI:10.1021/acs.jcim.6b00795

While selective BRaf^V600E inhibitors have been proven effective clinically, acquired resistance rapidly develops through reactivation of the mitogen-activated protein kinase (MAPK) pathway. Simultaneous targeting of multiple nodes in the pathway offers the prospect of enhanced efficacy as well as reduced potential for acquired resistance. Replacement pyridine group of Y-1 by a cyclopropyl formamide group afforded I-01 as a novel multitargeted kinase inhibitor template. I-01 displayed enzyme potency against Pan-Raf and receptor tyrosine kinases (RTKs). Based on the binding mode of I-01, analogues I-02-I-18 were designed and synthesized. The most promising compound I-16 potently inhibits all subtypes of Rafs with IC₅₀ values of 3.49 (BRaf^V600E), 8.86 (ARaf), 5.78 (BRaf^WT), and 1.65 nM (CRaf), respectively. I-16 not only exhibit comparable antiproliferative activities with positive control compounds against HepG2, SW579, MV4-11, and COLO205 cell lines, but also suppress the proliferation of melanoma SK-MEL-2 harboring overexpressed BRaf^WT with IC₅₀ values of 0.93 μM. The Western blot results for the ERK inhibition in human melanoma SK-MEL-2 cell lines show that I-16 inhibits the proliferation of SK-MEL-2 cell lines without paradoxical activation of ERK, which support the hypothesis that the inhibition of Pan-Raf and RTKs might be a tractable strategy to overcome the resistance of melanoma induced by the therapy with the current selective BRaf^V600E inhibitors.

Full text of DOI:10.1021/acs.jcim.6b00795

Article
pubs.acs.org/jcim
Potent Pan-Raf and Receptor Tyrosine Kinase Inhibitors Based on a
Cyclopropyl Formamide Fragment Overcome Resistance
Yanmin Zhang,†,∥ Lu Wang,†,∥ Qing Zhang,‡ Gaoyuan Zhu,† Zhimin Zhang,† Xiang Zhou,†
Yadong Chen,† Tao Lu,*,†,§ and Weifang Tang*,†
†School of Basic Science, China Pharmaceutical University, 639 Longmian Avenue, Nanjing, Jiangsu 211198, China
‡School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, 639 Longmian Avenue, Nanjing, Jiangsu 211198,
China
§State Key Laboratory of Natural Medicines, China Pharmaceutical University, 24 Tongjiaxiang, Nanjing, Jiangsu 210009, China
S
* Supporting Information
ABSTRACT: While selective BRafV600E inhibitors have been
proven effective clinically, acquired resistance rapidly develops
through reactivation of the mitogen-activated protein kinase
(MAPK) pathway. Simultaneous targeting of multiple nodes in
the pathway offers the prospect of enhanced efficacy as well as
reduced potential for acquired resistance. Replacement
pyridine group of Y-1 by a cyclopropyl formamide group
afforded I-01 as a novel multitargeted kinase inhibitor
template. I-01 displayed enzyme potency against Pan-Raf
and receptor tyrosine kinases (RTKs). Based on the binding
mode of I-01, analogues I-02−I-18 were designed and
synthesized. The most promising compound I-16 potently inhibits all subtypes of Rafs with IC50 values of 3.49 (BRafV600E),
8.86 (ARaf), 5.78 (BRafWT), and 1.65 nM (CRaf), respectively. I-16 not only exhibit comparable antiproliferative activities with
positive control compounds against HepG2, SW579, MV4-11, and COLO205 cell lines, but also suppress the proliferation of
melanoma SK-MEL-2 harboring overexpressed BRafWT with IC50 values of 0.93 μM. The Western blot results for the ERK
inhibition in human melanoma SK-MEL-2 cell lines show that I-16 inhibits the proliferation of SK-MEL-2 cell lines without
paradoxical activation of ERK, which support the hypothesis that the inhibition of Pan-Raf and RTKs might be a tractable
strategy to overcome the resistance of melanoma induced by the therapy with the current selective BRafV600E inhibitors.
despite impressive initial responses in BRafV600E tumors, most
patients developed acquired resistance (the resistance caused by
1. INTRODUCTION
The upregulation or activation of the Ras-Raf-Mek-Erk
vemurafenib) to vemurafenib after six months, and about 20%
(mitogen-activated protein kinase MAPK) signaling pathway
always leads to abnormal cell proliferation and differentiation.1
of patients have intrinsic resistance and do not respond to the
10
drug despite the presence of BRafV600E
.
In BRafWT tumors,
As a downstream effector of Ras in the MAPK signaling
pathway, Raf plays a crucial role in oncotherapy. BRaf
demonstrates more potent biochemical activity and mutation
rate compared with the other two members of Raf isoforms
ARaf and CRaf.2 Over 60% of melanoma patients or 5% of
colorectal carcinoma patients are associated with mutated BRaf
(BRafV600E, Val600-Glu600, V600E).3 According to the binding
modes of DFG (a conserved amino acid sequence of D594,
F595, and G596), which derived from crystallographic analysis,
Raf kinase inhibitors are categorized into two types: DFG-out
and DFG-in series.4 The DFG-out inhibitors, such as sorafenib5
and LY3009120,6 engage the protein in DFG-out conformation
and inhibit all the subtypes of Raf proteins. The DFG-in
inhibitors, such as vemurafenib,7 bind to the ATP binding site
of the kinase with a DFG-in conformation and inhibit BRafV600E
with high selectivity (Figure 1). Vemurafenib increases overall
survival in patients with BRafV600E melanoma, thus validating
BRafV600E as therapeutic targets in this disease.7−9 However,
vemurafenib even activates the MAPK pathway, which thus
enhances tumor growth in some xenograft models.11
Resistance is generally mediated by pathway reactivation, and
multiple mechanisms have been described. As the repair
mechanism of Path B1 in Figure 2, the addition of DFG-in
BRafV600E inhibitors at a certain concentration range will bind to
BRafV600E and inhibit the kinase, but they lack inhibitory activity
against heterodimers of BRafWT and CRaf, which can induce
the phosphorylation of MEK/ERK leading to acquired
resistance. Raf homodimer of CRafs can directly phosphorylate
MEK despite the absence of BRaf and ARaf as shown in Path
B2.12 In Path B3, ARaf acts as a scaffold to stabilize BRaf-CRaf
heterodimers.13 When there is upstream signaling from RTKs
(receptor tyrosine kinases) amplication, Rafs form dimers
Received: December 31, 2016
Published: May 9, 2017
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 1. Binding mode of BRafV600E (A) DFG-out conformation for PDB 1UWJ (crystallized with sorafenib) and (B) DFG-in conformation for
PDB 3OG7 (crystallized with vemurafenib).
development of Pan-Raf/RTKs MTKIs to overcome the
resistance caused by vemurafenib.
2. MATERIALS AND METHODS
2.1. Computational Methods. 2.1.1. Binding Site
Detection. Two programs including FTMap and SiteMap in
the Schrodinger suite23 were used to detect the active binding
̈
site of BRaf. As a particularly useful and accurate in predicting
the binding site of proteins, FTMap globally samples small
organic molecules as probes on the surface of a target protein
by a library of 16 small organic probe molecules, finds favorable
positions, clusters conformations, and ranks the consensus sites
(CSs) (also called “hot spots”) on the basis of average energy
and the number of probe clusters they included.24 It was
performed through its online server (http://ftmap.bu.edu).
SiteMap24 incorporated in the Schrodinger software is a fast,
Figure 2. Acquired resistance is induced by vemurafenib through paths
A and B.
accurate, and practical binding site identification methods. It
was carried out to detect the binding site of BRaf crystal
structures and predict the druggability of those sites.
shown in Path B.12 These dimers induce MEK/ERK
2.1.2. Molecular Docking. The crystal structures of BRaf
(PDB code 1UWJ, 3IDP, and 4XV2) were downloaded from
Protein Data Bank (PDB) and prepared by the Protein
phosphorylation and increased MAPK pathway output leading
to acquired resistance. This would not happen with Pan-Raf
inhibitors of similar activities against all subtypes of Raf
kinases.14 It is also demonstrated that the feedback activation in
BRafWT or CRaf bearing cancers can be significantly suppressed
by DFG-out Pan-Raf inhibitors.15
Almost all major human cancers seem to harbor not a single
but several concomitant dysregulations of signaling pathways.16
Crosstalk and compensatory mechanism exists in different
pathways.17,18 RTKs, upstream kinases of MAPK and PI3K/
Akt/mTOR pathway, mediate resistance in melanoma suggest-
ing potential upfront therapeutic strategies to prevent resistance
(Path A in Figure 2).19 A compensatory mechanism of acquired
resistance primarily involves Ras activation in response to
upstream RTK signals such as EPHA2,17,18 FGFR,20 and
PDGFRb.21 Drugs, which inhibit the pathway of upstream and
downstream multiple targets, can achieve collaborative treat-
ment to overcome the resistance.22
There is rarely a report on the treatment of acquired resistant
melanoma cancers or intrinsic resistant colorectal cancers by
using a Pan-Raf/RTK MTKI (multitargeted kinase inhibitor).
Herein, we would like to describe the discovery of a series of
inhibitors based on a cyclopropyl formamide fragment as novel
MTKIs. These compounds potently inhibit the kinase activities
of Pan-Raf and RTKs with low IC50 values. Furthermore, the
compounds also display potent inhibition on the proliferation
of a panel of vemurafenib-resistant cancer cells harboring
overexpressed BRafV600E, representing new leads for further
Preparation Wizard in the Schrodinger suite.23 The small
̈
molecules were initially minimized by the LigPrep module in
Schrodinger. The Glide module (extra precision [XP] mode)
̈
was selected for molecular docking due to its excellent
performance through a self-docking analysis,25 and the 10
best poses of each ligand were minimized by a post docking
program with the best pose saved for further analysis.
2.1.3. Molecular Dynamics Simulations. The protein
complexes after molecular docking were used for MD
simulation. The MD simulations were carried out using
GROMACS26 with AMBER03 force field27 and TIP3P water
model.28 The proteins were simulated in a water filled
dodecahedron box with the distance at least 12 Å between
the complex and the box edges. The charges of system were
neutralized by adding counterions, either Na+ or Cl− by the
genion tool. Then, the system was relaxed through an energy
minization process with steepest descent algorithm first and
then with conjugate gradient algorithm. Afterward, an NVT
simulation was performed with temperature going from 100 to
300 K and followed by an NPT simulation to equilibrate the
pressure. Finally, a production MD was performed for 20 ns at
300 K. Bond lengths were constrained using LINCS
algorithm.29 The last 5 ns of each production simulation
were extracted at every 10 ps interval for calculating the
molecular mechanics/Poisson−Boltzmann surface area binding
energy (MM-PBSA) using the AMBER12 package.30 Average
B
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
a
Scheme 1. Synthetic Route of I-01 to I-17
a
Reagents and conditions: (a) diethyl malonate, NaH, THF, 0 °C, 0.5 h, then reflux, 2.5 h, 86.2%; (b) EtONa, EtOH, reflux, 2.5 h, 92.4%; (c) 1,2-
dibromoethane, NaOH, DMF, rt, 5 h, 90.2%; iodomethane, NaOH, DMF, rt, 5 h, 94.3%; 1,4-dibromobutane, NaOH, DMF, rt, 5 h, 90.2%; (d) A1−
A11, Al(CH3)3, toluene, nitrogen atmosphere, 80 °C, 5 h, 70.2−89.2%; (e) 1 N HCl, CH3OH, rt, 2 h, 72.4−85.6%.
of the energy for the 1000 snapshots with each one for 50 ps
was saved for further analysis.31 The radius of gyration (Rg),
root-mean squared deviation (RMSD), and the number of
hydrogen bond formed was analyzed by g_gyrate, r_rms, and
g_hond tool, respectively, in the GROMACS package.32 In
addition, the program gmx hbond analyzes the hydrogen bonds
(H-bonds) between all possible donors D and acceptors A. To
determine if an H-bond exists, a geometrical criterion (r ≤ rHB
= 3.5 Å, α ≤ rHB = 30 °C) is used, see below:
The value of rHB = 3.5 Å corresponds to the first minimum of
the radial distribution function (RDF) of simple point charge
(SPC) water.
2.2. Synthesis of Compounds. The preparation of the
target purine derivatives I-01−I-17 are shown in Scheme 1.
The reaction of commercially available 6-chloro-9-(tetrahydro-
2H-pyran-2-yl)-9H-purine with diethyl malonate in the
presence of sodium hydride gave diethyl 2-(9-(tetrahydro-2H-
pyran-2-yl)-9H-purin-6-yl) malonate I-a in 86.2%. The
prepared I-a was allowed to react with sodium ethoxide to
afford the desired I-b in 92.4%. Subsequent alkylation of I-b
with 1,2-dibromoethane, iodomethane or 1,4-dibromobutane in
the presence of sodium hydroxide afforded the corresponding
alkylated derivatives I-c-1 to I-c-3 in 90.2%, 93.4%, or 90.2%,
respectively. Amide-forming reactions of I-c-1 to I-c-3 with the
corresponding anilines A1−A11 were carried out in the
C
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
a
Scheme 2. Synthesis of Intermediates A1−A8
a
Reagents and conditions: (a) (COCl)2, DMF, CH2Cl2, rt, 2 h; (b) 4-chloro-3 (trifluoromethyl) aniline, 4-methyl-3-nitroaniline or 3-nitroaniline,
Et3N, CH2Cl2, rt, 5 h; (c) CDI, CH2Cl2, rt, 34 h; (d) Fe, NH4Cl, 75% EtOH, reflux, 3 h.
a
Scheme 3. Synthesis of Intermediates A9−A11
a
Reagents and conditions: (a) morphine, DMSO, 100 °C, 5 h, 94.2%; (b) NBS, DCE, reflux, 5 h, 86.4%; (c) morphine, NEt3, THF, reflux, 2 h; (d)
3-morpholinopropan-1-ol, NaH, DMF, ice bath, 0.5 h, then 100 °C, 5 h; (e) Fe, NH4Cl, 75% EtOH, reflux, 2 h; (f) 4-methyl-3-nitrobenzoyl chloride,
Et3N, CH2Cl2, r.t., 5 h; (g) Fe, NH4Cl, 75% EtOH, reflux, 2 h.
presence of trimethylaluminum (2 M solution of toluene)
under the condition of nitrogen atmosphere to give I-d-01 to I-
d-17 in 70.2−89.2%. Finally, the THP (pyran) group was
deprotected by using 1 N HCl in methanol to provide
compounds I-01 to I-17 in 72.4−85.6%.
A total of 11 important intermediates of amide analogs A1−
A11 were synthesized according to Schemes 2 and 3 via a two/
five-step process.
Intermediates of amide analogs A1−A8 were prepared by the
method shown in Scheme 2. Substituted benzoyl chloride,
prepared in the presence of benzoic acid with various R1 groups
(R1 = CH3, F, Cl, H) and oxalyl chloride, was allowed to react
with 4-chloro-3-(trifluoromethyl) aniline or 4-methyl-3-nitro-
aniline to afford the corresponding nitroaniline. Benzenesul-
fonyl chloride was converted to the desired nitroaniline in the
presence of 3-nitroaniline. Subsequent treatment of the
resulting nitroaniline with Fe/NH4Cl gave the corresponding
anilines A1−A7 in 68.6−80.4%. Urea A8 was generated from
condensation reaction using CDI (N,N-carbonyldiimidazole),
while A1−A7 were obtained by condensation of benzoyl
chloride with their corresponding anilines.
The synthetic route of anilines A9−A11 was outlined in
Scheme 3. The reaction of 1-fluoro-4-nitro-2-(trifluoromethyl)
benzene with morphine through SNAr substitution reaction
provided 1 in 94.2%. Hydrogenation of the nitro group of 1
gave the corresponding aniline. Subsequent treatment of the
resulting aniline with intermediate 4-methyl-3-nitrobenzoyl
chloride provided nitroaniline, which hydrogenated to afford
D
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
a
Scheme 4. Synthetic Route of I-18
a
Reagents and conditions: (a) diethyl carbonate, LDA, THF, rt, 12 h, 70.1%; (b) 1,2-dibromoethane, NaOH, DMF, rt, 1 h, 65.0%; (c) Zn(CN)2,
Pd(PPh3)4, DMF, N2, 90 °C, 2 h, 88.0%; (d) NH3·H2O, H2O2, DMF, 3 h, rt, 3 h, 85.0%; (e) iodomethane, NaH, DMF, rt, 0.5 h, 93.0%; (f) A1,
Al(CH3)3, toluene, 80 °C, 5 h, 60.0%.
the desired intermediate A9 in 85.4%. The desired A10 in
83.8% utilized 1-methyl-4-nitro-2-(trifluoromethyl) benzene as
starting material. With NBS (N-bromosuccinimide), it was
converted to 2 in 86.4%. 3 was prepared by 2 in the presence of
NEt3 as alkali, THF as solvent. The last three steps of A10 were
same as previous reaction (e, f and g) adopted for A9. The
reaction of commercially available 1-fluoro-4-nitro-2-(trifluor-
omethyl) benzene with 3-morpholinopropan-1-ol in the
presence of sodium hydride gave 4. The last three steps were
same as previous reaction (e, f, and g) adopted for A9. Then,
the desired intermediate A11 was afforded in 89.4%.
10 uCi/uL) (P-ERK in Elmer, NEG302H001 MC) was added
to initiate the reaction, the reactions were carried out at 25 °C
for 120 min. The kinase activities were detected by filter-
binding method. IC50 values and curve fits were obtained by
Prism (GraphPad Software).
2.4. Antiproliferative Assay against Different Cell
Lines. Cell planking: Vi-Cell XR cell counter was used to
count living cells, which collected in exponential phases. After
diluting to 3 × 103−1.5 × 104 cells/mL with the complete
medium, 90 μL of cells suspension was added to each well of
96-well culture plates and cultured in DMEM (Dulbecco
minimum essential medium) or McCoy’s 5a/10% fetal bovine
serum (CrownBio Corporation) for 24 h, with 5% CO2 water
saturated atmosphere at 37 °C.
Compound dispensation: Each selected compounds dis-
solved in DMSO as the storage solution (10 mM). Then with
medium diluted to 10 times solution respectively, each 2 holes
(inhibition ratio) or 3 holes (IC50 value). The final drug
concentration for 10 μM (inhibition ratio) or initial drug
concentration for 10 μM. The compound dispensation, which
volume was 10 μL, was further incubated at 37 °C for another
72 h in a humidified atmosphere with 5% CO2.
Plate detection: 50 μL CTG solution, melt in advance and
balance to room temperature, was added to each hole.
According to the operating instructions of CTG, oscillators
with microporous plate blend 2 min. At room temperature for
10 min, Envision2104 was used to measure the luminescence
signal values.
The target pyridine derivative I-18 was prepared by the
method as shown in Scheme 4. The reaction of commercially
available 2-bromo-4-methylpyridine with diethyl carbonate in
the presence of lithium diisopropylamide gave ethyl 2-(2-
bromopyridin-4-yl) acetate 5 in 70.1%. Subsequent alkylation
of 5 with 1, 2-dibromoethane in the presence of sodium
hydroxide afforded the alkylated derivative 6 in 65.0%. The
prepared 6 was allowed to react with dicyanozinc in the
presence of Pd(PPh3)4 under the condition of nitrogen
atmosphere to afford the desired 7 in 88.0%. A solution of 7
and ammonium hydroxide was added hydrogen peroxide to
achieve 8 in 85.0%. Subsequent methylation of 8 with
iodomethane in the presence of sodium hydride afforded the
methylated derivative 9 in 93.0%. Amide-forming reactions of 9
with the corresponding aniline A1 were carried out in the
presence of Al(CH3)3 (2 M solution of toluene) under the
condition of nitrogen atmosphere to give I-18 in 60.0%.
2.3. BRafV600E Inhibitory Assay. Activity of full length
BRafV600E was determined using Hot-SpotSM kinase assay which
was performed by Reaction Biology Corp. (Malvern PA). A 5
nM portion of human GST-tagged BRafV600E protein (AA416−
766) (Invitrogen, Cat no. PV3894) was mixed with 20 μM of
the substrate His 6-Tagged Full-length Human MEK1 (K97R)
(Reaction Biology Corp.) in reaction buffer (20 mM Hepes pH
7.5, 10 mM MgCl2, 1 mM EGTA, 0.02% Brij35, 0.02 mg/mL
BSA, 0.1 mM Na3VO4, 2 mM DTT, 1% DMSO) at room
temperature, the compounds dissolved in 100% DMSO at
indicated doses (starting at 30 μM with 3-fold dilution) was
delivered into the kinase reaction mixture by Acoustic
technology (Echo550; nanoliter range), incubate for 20 min
at room temperature. After 10 uM 33P-γ-ATP (specific activity
Data processing: inhibition ratio = 1 − Vsample/Vvehicle control
×
100%. Vsample for drug treatment group, Vvehicle control for solvent
control group. Using GraphPad Prism 5.0 software and
nonlinear regression model S type dose draw survival rate
curve and calculate the IC50 value.
2.5. P-ERK Cellular Assay in A375 Cells and SK-MEL-2
Cells. 2.5.1. Recovery, Culture, and Passage of Cells. A375
and SK-MEL-2 human melanoma cell line were both obtained
from American Type Culture Collection (Manassas, VA).
Removed melanoma cells cryopreserved tube from liquid
nitrogen tank, shaking in 37 °C water bath to make it melt
quickly. The centrifugal tube was added cell suspension in
superclean bench, 5 mL EMEM (Eagle’s minimal essential
medium) was added. Centrifugalized the mixture 5 min with
E
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 3. Binding pockets of BRafV600E and the alignment of several potent BRafV600E inhibitors with sorafenib. For A−C, the purple colored
compound is Sorafenib, the greened colored compounds are the cognate ligands for PDB 3IDP, the cognate ligands for PDB 4XV2, and compound
Y-1, accordingly. The yellow dots are the sitemap results and the fragments in different colored lines are binding hot spots detected by FTMap. For
D−F, the green colored compounds are always compound Y-1, the orange compounds are the lead fragment, compound I-01 and compound I-02,
respectively.
Figure 4. Compounds based on cyclopropyl formamide fragment and four dimensional pharmacophore map of BRafV600E
.
F
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 5. Kinome phylogenetic tree of I-01. The left panel shows the inhibition rate of I-01 against all the targets that belong to the Kinome
phylogenetic tree. The atypical protein kinases include ABC1, Alpha, Brd, PDHK, PIKK, RIO, and TIF1 from the top to the bottom accordingly.
The right panel demonstrates the 11 kinases that I-01 obtained the most potent activity (IC50 values).
1000 rpm, discarded supernatant, joined 2 mL EMEM medium
supplemented with 10% serum to suspense cells, vaccinating it
in culture bottle. Put culture bottle in incubator with 37 °C,
saturated humidity, 5% CO2. The next day, discarded the
original medium, added 6 mL fresh medium to remove dead
cells. Passage when cell numbers around 90%. Discarded
supernatant, washed twice with PBS (phosphate buffered
saline), added 2 mL 0.25% pancreatic enzyme to digest 2−3
min. When cells becoming round under microscope, discarded
pancreatic enzyme, added 2 mL EMEM medium to terminate
digestion. Making the adherent cells blow down to a single
suspension cells with a pipet gently. According to the speed of
cell growth, culture the cell from one to three or four.
2.5.2. Plating Cells and Dosing. When the melanoma cells
A375 and SK-MEL-2 in the logarithmic phase, vaccinating 3−5
× 105 cells in 96 cell plate with EMEM medium supplemented
with 10% heat-inactivated FBS (fetal bovine serum), culturing
cells in incubator with 37 °C, saturated humidity, 5% CO2.
When cell confluence reached 80−90%, discarding medium,
diluting drugs with culture medium into 0.4, 0.2, 0.1, 0.05,
0.025, dosing drugs in 96 cell plate, using DMSO as a negative
control.
volume 5× SDS-PAGE loading buffer, a boiling water bath is
created for 10 min to denaturature protein. The specific signals
of bands of interest were quantified by Gel-Pro Analyzer.
2.5.4. Western Blot Detection. Protein samples (30 μg) in
12% SDS-PAGE (polyacrylamide gel electrophoresis) were
used for electrophoresis. The protein was transfered from a
running gel to a PVDF (poly vinylidene fluoride) membrane
with the wet transfer method, and steady flow of 200 mA, with
a 1 kDa/min transfer velocity, according to the interest protein
molecular weight to determine the transfer membrane time
using a 10% nonfat milk blocked PVDF membrane for 1.5 h in
37 °C. Using 5% BSA diluted primary antiphospho-ERK1/2
antibody, 4 °C incubation was conducted overnight. The
membrane was washed with 1× TBST three times; each time
took 10 min. Using 5% nonfat milk diluted secondary
antiphospho-ERK1/2 antibody with HRP tag, room temper-
ature incubation was conducted for 1 h. The membrane was
washed with 1× TBST three times; each time took 10 min.
Droping an appropriate amount of ECL luminous on a
membrane, a Tanon5200 automatic chemiluminescence image
analysis system was used.
2.5.3. Protein Sample Preparation. The cells were treated
with compounds or DMSO for 24 h. After treatment, cells were
digested with 500 μL 0.25% trypsin, then 500 μL medium were
added to terminate digestion. Cells were blown evenly,
centrifugalized the mixture 5 min with 3000 rpm, discarded
supernatant, put centrifugal tube on the ice, added suitable
amount of protein lysis solution RIPA (Radio-immunoprecipi-
tation assay), protease inhibitor PMSF (Phenylmethanesulfonyl
fluoride) (1 mL RIPA/10 μL PMSF), and phosphatase
inhibitors (1 mL RIPA/10 μL phosphatase inhibitors), cells
blown evenly, 4 °C for 30 min, make cells lysate, every 5 min
vortex once. Centrifugalizing the lysate 15 min with 12 000 rpm
in 4 °C, gathering supernatant, using the BCA protein assay kit
to determine the total protein concentration. Adding 1/4
3. RESULTS AND DISCUSSION
3.1. Inhibitor Design and Molecular Modeling. In a
previous report, we described the synthesis and potency of 3-
((3-(9H-purin-6-yl)pyridine-2-yl)amino)-N-(4-chloro-3-
(trifluoromethyl)phenyl)-4-methylbenzamide Y-1 as a
BRafV600E inhibitor.33 Y-1 shows potent BRafV600E inhibitory
activity (IC50 = 3.15 nM) and potential antiproliferative activity
against cell line A375 (IC50 = 8.79 μM). To improve the
originality of Y-1,34 we performed structure optimization.
Understanding the structure and function of protein active sites
is a cornerstone of drug design. By aligning sorafenib (BRafWT
IC50 = 22 nM and BRafV600E IC50 = 38 nM)35 with three potent
BRafV600E inhibitors which include the cognate ligand of PDB
G
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Table 1. SAR Exploration of Compounds against BRafV600E Activity
a
b
Under 0.5 μM concentration of compounds. Not determination.
3IDP (BRafV600E IC50 = 1.6 nM),34 cognate ligand of PDB
4XV2 (i.e., dabrafenib, BRafV600E IC50 = 5.4 nM),36 and the lead
compound Y-1 (BRafV600E IC50 = 3.15 nM),33 we found that all
of them occupied an extra hydrophobic pocket (Figure 3A−C)
that is confirmed by two binding sites detection methods
FTmap24 and Sitemap37 (binding site detection methods).
Then, three fragment generation methods38 were employed to
decompose several databases,39 and the resulted fragments were
docked to the active binding sites. A N-phenyl-1-(pyrimidin-4-
yl) cyclopropane-1-carboxamide fragment (the red colored part
of I-01 as shown in Figure 4) was found to be in great
alignment with the N-phenyl-3-(9H-purin-6-yl)pyridin-2-amine
part of compound Y-1 (Figure 3D). Thus, by using a scaffold
hopping method, we designed I-01 which adopted a similar
binding conformation as sorafenib with the original essential
hydrogen bonds kept (two hydrogen bond formed with Cys532
H
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Moreover, I-18 have better physicochemical property (tPSA =
99.56 Å2, LogP = 4.56) than I-01 (tPSA = 107.31 Å2, LogP =
4.34). However, the enzymatic potency was lost as shown in
Figure 4. The possible reasons were presented in binding
mechanism analysis. Based on the above information, a total of
18 novel compounds were designed and synthesized.
Table 2. Pan-Raf Activity of the Chosen Compounds
IC50 (nM)
compd
I-01
BRafV600E
ARaf
BRafWT
CRaf
1.91
3.41
1.12
4.56
8.86
13.3
2.14
7.13
2.41
5.11
5.78
16.1
6.92
22.65
0.49
1.91
I-15
I-16
3.49
1.65
3.2. Chemistry. As summarized in Schemes 1 and 4, 18
compounds were synthesized in total. The synthetic routes are
illustrated in Schemes 1−4. The chemical structures of these
I-17
5.31
3.02
vemurafenib
sorafenib
31.64
38.12
135.81
6.28
1
compounds were confirmed by H NMR, 13C NMR, HRMS
spectra, and the results are presented in the Supporting
Information.
in the hinge region and hydrogen bonds formed with Asp594
and Glu501 in the DFG-out pocket) and a better occupation of
the other important binding pockets (Figure 3E). However, by
moving the aromatic amide functional group from meta-
position of the central benzene ring (I-01) to the para-position
(I-02) led to the loss of enzymatic activity which can be initially
explained by its almost opposite binding conformation of I-02
(Figure 3F and Figure 4) compared with I-01 (Figure 3E).
Before optimizing the lead I-01, we determined the structural
features essential for enzymatic activity and identified possible
modification sites. The binding mode of I-01 and BRafV600E
kinase revealed that I-01 binds to the BRafV600E ATP binding
pocket with the purine core forming two essential hydrogen
bonds with the hinge region Cys532 (Figures 3 and 4). In
addition to these hydrogen bonds, the amide group forms two
hydrogen bonds with Asp594 and Glu501 in the DFG-out
region.
Structure-based optimization of I-01 focused on potency at
enzymatic as well as cellular levels. L as the linkage using
different groups was attempted. R1 group with different sizes
occupied a small hydrophobic pocket typically closed the
gatekeeper region. Fragment Q was investigated on rigid and
steric factors. R2 substituent in the terminal phenyl ring was
directed toward the solvent accessible region. R3 substituent at
terminal phenyl ring was exposed to another hydrophobic back
pocket, which was formed by a rearrangement of the activation
loop and subsequent movement of a phenylalanine side chain
of DFG loop.40 4-Substitute-N-methylpicolinamide, the hinge
region structure of sorafenib, was introduced to replace purine
to afford I-18. I-18 maintained the binding mode of I-01.
3.3. Investigation of the Enzymatic Activity of
Compounds. To further determine the efficacy, the broad
kinase spectrum of I-01 was evaluated at 0.5 μM concentration
against an RBC (Reaction Biology Corporation) screen panel
containing 349 therapeutically important kinases as shown in
Figure 5. The kinome phylogenetic tree was generated using
KinomeReader.41 The top hits were subjected to follow-up IC50
generation. I-01 was found to inhibit 11 kinases (Raf and
RTKs) with IC50 values range from 0.47 to 40.91 nM, including
ARaf IC50 = 0.97 nM, BRaf IC50 = 0.99 nM, CRaf IC50 = 0.47
nM, DDR1 IC50 = 2.96 nM, DDR2 IC50 = 2.57 nM, EPHA2
IC50 = 13.50 nM, FGFR2 IC50 = 39.21 nM, FMS IC50 = 3.94
nM, LCK IC50 = 40.91 nM, PDGFRb IC50 = 25.34 nM, and Ret
IC50 = 5.24 nM, showing I-01 to be a potent multitargeted
kinase inhibitor (right panel of Figure 5). I-01, highly potent
against the Pan-Raf and other oncogenic kinase members of the
RTKs group with low IC50 values, exhibits excellent target
specificity in a selectivity profiling investigation against 349
kinases (Figures 2 and 5). Resistance to BRafV600E and RTKs
which involved bypass or restoration of persistent MAPK
activation might be overcome by I-01.42
3.3.1. Compounds Screened for Enzymatic Activity
against BRafV600E. Compounds were evaluated the potency
against BRafV600E using vemurafenib and sorafenib as positive
control compounds. Disruption of hydrogen bond interactions
by replacement of the L of I-01 with “reverse” amide I-03 or
urea I-04 led to inapparent change at enzymatic activity.
Structurally, sulfonyl group has similar properties in molecular
size and charge distribution with carbonyl, so sulfonyl group
Figure 6. RMSD values fluctuation of the compounds (Y-1, I-01, sorafenib, vemurafenib) during the 20 ns MD simulation.
I
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Table 3. Time-Average and Standard Deviation of H-Bond, RMSD, Rg of protein, and Predicted Delta G for the Last 5 ns MD
average RMSD (Å)
compd
Y-1
num of Hbonds
Rg protein (Å)
complex
protein
ligand
average delta G (kcal/mol)
RafV600E IC50 (nM)
4
4
2
3
4
1
3
1
4
4
0
1
1
1
1
1
1
1
1
1
19.2 0.1
19.2 0.1
19.2 0.1
19.4 0.1
19.3 0.1
19.3 0.1
19.3 0.1
19.2 0.1
19.3 1.3
19.0 0.2
11.2 0.1
11.2 0.1
11.3 0.2
10.4 0.1
2.6 0.3
11.3 0.1
11.6 0.1
11.3 0.1
3.3 0.5
2.9 0.5
2.4 0.3
2.5 0.3
2.4 0.5
2.5 0.3
2.6 0.3
2.1 0.2
2.6 0.2
2.5 0.4
3.3 0.5
2.9 0.5
0.8 0.2
0.8 0.2
0.9 0.2
1.0 0.2
1.0 0.2
1.4 0.2
1.1 0.2
0.9 0.3
0.7 0.2
0.6 0.1
−53.2 4.0
−51.9 4.7
−31.9 4.6
−37.2 4.06
−52.1 3.8
−32.3 7.2
−47.3 5.1
−31.9 4.1
−50.8 3.4
−53.1 3.0
3.15
1.91
NA
NA
3.89
NA
3.41
NA
38
I-01
I-02
I-05
I-09
I-12
I-15
I-18
sorafenib
vemurafenib
31
hydrophobic pocket hadn’t enough space to accommodate
cyclopentyl.
In order to get a better occupation of the solvent accessible
region leaning the terminal phenyl ring, the solubilizing
morpholine functionality with increasing length of the linker
carbon chain was introduced which gave I-15 to I-17. This
modification could sustain the BRafV600E inhibitory activity
compared to I-01, which is consistent with the hypothesis that
the R2 group could occupy a solvent accessible region in
BRafV600E
.
Most of compounds show IC50 values at nanomolar ranges,
superior to positive control compounds vemurafenib and
sorafenib tosylate. I-01, using amide group as L and cyclopropyl
as Q group, shows optimal enzymatic activity.
3.3.2. Pan-Raf Activity of the Chosen Compounds. As
mentioned earlier, Pan-Raf inhibitors with similar activity
against all subtypes of Raf kinases, might have potency to
overcome the resistance of vemurafenib. The chosen
compounds I-01 and I-15 to I-17 were subjected to follow-
up Pan-Raf IC50 generation. The results are shown in Table 2.
I-15 to I-17 sustained the enzymatic activity compared to I-01.
3.3.3. Binding Mechanism Analysis. Due to the different
enzymatic activity of compounds with only a slight difference in
structure, molecular dynamics (MD) was conducted to simulate
the binding process and calculate the binding energy of these
compounds. A total of 10 compounds including sorafenib,
vemurafenib, Y-1, I-01, I-02, I-05, I-09, I-12, I-15, and I-18
were investigated. RMSD (root mean standard deviation)
variation of protein atoms along time is one of the standard
ways to measure the stability of the protein. From Figure 6 and
Figure S1 (see the Supporting Information), the blue lines
represent the fluctuation of protein while the orange lines
demonstrate the change of the small compounds. Both the
protein and small compounds reached certain equilibrium
during the 20 ns simulation with the protein obtained a higher
RMSD than its corresponding ligands. Combined with the
time-averaged value of RMSD values with standard deviation
are listed in Table 3. The average RMSD for protein were about
2.6 Å (2.1 Å for compound I-12 ∼3.3 Å for sorafenib) with a
standard deviation of about 0.4 Å. Similarly, the ligands
obtained a smaller RMSD of about 0.9 Å (0.7 Å for
vemurafenib ∼1.4 Å for I-12) with a standard deviation of
0.2 Å. This means during the MD simulation process, the small
compounds are more stable than the protein. Nonetheless, the
radius of gyration (Rg) of the protein were quite steady (in the
range of 19.0 to 19.4 from Table 3), indicating the compactness
of the protein which is stably folded.
Figure 7. Predicted Delta G and the average number of hydrogen
bonds for several compounds.
can be introduced into the drug molecules as the bioisostere of
carbonyl to remain or improve activity. Replacement of the
amide of I-01 with sulfonyl led to I-05 loss of enzymatic activity
(Table 1), which confirmed the importance of amide group for
enzymatic activity.
Subsequently, the influence of R1 substituent at central
phenyl ring was investigated. Halogen atoms, including fluorine
and chloride, were employed to afford compounds I-06 and I-
07. Hydrogen was explored to afford compound I-08. All
modifications sustained the BRafV600E inhibitory activity
compared to I-01. This result is consistent with the hypothesis
that the R1 group would occupy a small hydrophobic pocket
typically closed the gatekeeper region of BRafV600E
.
Q group of different rigidity was investigated. I-09 to I-11
showed comparable BRafV600E inhibitory activity with I-01. This
result suggested that the flexible group dimethyl could sustain
the BRafV600E inhibition compared to rigid cyclopropyl group.
Subsequently, Q of different steric sizes was explored.
Expanding the cyclopropyl of I-01 to cyclopentyl in I-12 to
I-14 resulted in loss of potency due to steric clash with protein.
This result indicates that the Q group occupying a small
J
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
Figure 8. Binding free energy (Delta G, kcal/mol) fluctuation of the compounds (Y-1, I-01, sorafenib, vemurafenib) during the last 5 ns of
production MD simulation.
We focused on the number of hydrogen bonds and binding
free energy calculation to analyze the reason why some
compounds (e.g., I-02, I-05, I-12, and I-18) do not exhibit any
enzymatic activity although they are only slightly different with
other compounds with very high enzymatic activity. From
Table 3 and Figure 7a, we could observe that most of the
compounds achieved more than three hydrogen bonds with
BRafV600E protein with three of them (i.e., sorafenib,
vemurafenib, and I-01) have four hydrogen bonds. However,
the other three compounds (i.e., I-02, I-12, and I-18) without
BRafV600E enzymatic activity only obtain an average of two, one,
and one hydrogen bond(s), respectively. Although I-05 also
does not show any BRafV600E enzymatic activity, it got an
average of three hydrogen bonds. The average binding free
energy from the last 5 ns MD production process is shown in
Table 3 and Figure 7b, and the fluctuation of binding free
energy along time is depicted in Figure 8 and Figure S2 (see the
Supporting Information). The binding free energy values for
the compounds with high enzymatic activity are relatively low
(less than −50.00 kcal/mol except I-15 of −47.34 kcal/mol);
however, those for the compounds without enzymatic activity
are higher than −40.00 kcal/mol (i.e., −31.90, −37.19, −32.34,
and −31.89 kcal/mol for I-02, I-12, I-15, and I-18,
respectively). In total, the difference of the number of hydrogen
bonds and binding free energy can explain why those
Table 4. Compounds Screened for Cellular Potency against
Different Melanoma Cells
a
A375 activity %
A375 activity
SK-MEL-2
compd
IC50 (μM)
compd
I-15
% IC50 (μM)
IC50 (μM)
I-01
I-03
I-04
I-06
I-07
I-08
58.4
98.7
13.2
80.5
14.7
14.6
17.2
11.1
29.6
5.4
4.2
3.4
0.9
I-16
62.3
25.8
I-17
6.4
1.2
b
100.6
102.6
85.1
ND
vemurafenib
sorafenib
0.7
5.6
ND
58.7
13.6
11.4
a
Cellular activity of compounds under the concentration of 10 μM.
Not determination.
b
Figure 9. Cellular activity of compounds under the concentration of
10 μM.
Figure 10. (A) ERK kinase inhibition in A375. (B) ERK kinase inhibition in SK-MEL-2.
K
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
compounds although has similar binding modes with the
actives but do not show any enzymatic activity to some extent.
3.4. Antiproliferative Activity against Various Cell
Lines. As described above, BRafV600E and BRafWT are important
targets in developing small-molecular inhibitors for cancer
therapies, especially melanoma, colon cancer, leukemia,
hepatoma, and thyroid carcinoma.43 A375,44 SK-MEL-2,45
and COLO-20546 cell lines have extraordinary expression of
BRafV600E. MV4-11,47 HepG2, and SW57948 cell lines have
extraordinary expression of BRafWT. Although other substituted
compounds display similar enzyme inhibitory performances to
I-01, their cell activity is completely different. Replacement of
the amide function of I-01 with either a “reverse” amide or urea
linkage, which lead to decreased potency at cellular level (Table
4). Compared to I-01, I-03 and I-04 showed a 6/2−fold loss of
A375 inhibition. This result shows that the amide as the L
would be better suited to antiproliferative activity. Subse-
quently, the influence of R1 substituent on central phenyl ring
was researched. I-08 shows loss of potency in a range similar to
halogen-substituted compounds I-06 to I-07. I-01 with a
methyl substituent displayed optimal A375 antiproliferative
activity.
It should be noted that I-15 to I-17 introduced different
hydrophilic group R2 to the solvent accessible region
successfully improved the cellular potency against various cell
lines, which is probably due to enhanced water solubility and
improved cellular permeability. The antiproliferative activities
of I-15 to I-17 are not only with potency against A375 and SK-
MEL-2 cell lines as shown in Table 4 but also against HepG2,
SW579, MV4-11, and COLO205 cell lines as shown in Figure
9.
I-16 displays potent antiproliferative activities against A375
and SK-MEL-2 cell lines with IC50 values as 4.25 or 0.93 μM,
respectively. The result proves that I-16 is superior than
sorafenib against this two cell lines. Although it is not
comparable with vemurafenib against A375, it is superior to
vemurafenib against SK-MEL-2. As mentioned earlier, vemur-
afenib shows paradoxical activation of MAPK pathway in wild-
type BRaf cells, I-16 might have potency to overcome the
resistance of vemurafenib. I-16 also exhibits comparable
antiproliferative activities against HepG2, SW579, MV4-11,
and COLO205 cell lines to positive control compounds. Thus,
I-16 may be a promising compound to be developed as a
candidate for therapy.
3.5. Western Blot: ERK Kinase Inhibition in A375 and
SK-MEL-2 Cells. We tested I-16 for its ability to inhibit P-ERK
in cancer cells by Western blot analysis (Figure 10). I-16
effectively and dose-dependently inhibits P-ERK in the A375
cell line with BRafV600E at concentration as low as 400 nM,
which is as potent as vemurafenib. In the SK-MEL-2 cell line
with BRafwt, I-16 effectively and dose-dependently inhibits P-
ERK at concentration as low as 400 nM. However, vemurafenib
activates P-ERK with the increase of its concentration. A key
finding from this class of compounds is that, unlike
vemurafenib, they dramatically inhibit the proliferation of
melanoma A375 cells through ERK pathway, without para-
doxical activation of the MAPK in SK-MEL-2 cells. Our results
support the hypothesis that the Pan-Raf and RTKs inhibition
might be a tractable strategy to overcome the intrinsic
resistance of colon cancer and the acquired resistance of
melanoma caused by the current BRafV600E inhibitor therapy.
4. CONCLUSION
Depending on FTmap and Sitemap, we found that all of DFG-
out conformation Raf inhibitors occupied an extra hydrophobic
pocket. Through fragment-based approach, we determined
cyclopropyl formamide fragment could occupy this pocket
properly like other positive control compounds. The obtained
compound I-01 with DFG-out conformation inhibits all
subtypes of Raf proteins. Based on the structure homology of
Raf and RTKs, we evaluated the broad kinase spectrum of I-01.
The results displayed that I-01 also inhibits eight oncogenic
kinase members of the RTKs groups in the low nanomolar
range. Since I-01 can avoid the repair and compensatory
mechanism leading to resistance, we develop derivatives I-01 to
I-18 with activity against multiple targets of Pan-Raf and RTKs.
I-16 inhibits all subtypes of Raf proteins with IC50 value of 3.49
nM (BRafV600E), 8.86 nM (ARaf), 5.78 nM (BRafWT), and 1.65
nM (CRaf), respectively, it was more potent than positive
controls (vemurafenib and sorafenib). In particular, I-16
exhibits the lower IC50 values 4.25 μM, 0.93 μM against
A375 and SK-MEL-2, which also displays better antiprolifer-
ative activity against four cell lines HepG2, SW579, MV4-11,
and COLO-205 compared to positive control (these cell lines
are extraordinarily expressed BRAFV600E or BRafWT). Then,
further mechanism investigation in terms of the ERK inhibition
in human melanoma A375 and SK-MEL-2 cell lines by Western
blot was investigated. The results reveal that our compounds
inhibit the proliferation of melanoma A375 cells through ERK
pathway, without paradoxical activation of ERK in BRafWT type
melanoma SK-MEL-2 cells. Our results support the hypothesis
that the Pan-Raf and RTKs inhibition might be a tractable
strategy to overcome the intrinsic and acquired resistance of
melanoma caused by the current BRafV600E inhibitor therapy.
Moreover, with this multitargeted profile, the compounds could
be used as a therapeutic aid for patients developing resistance
while being treated with currently marketed kinase inhibitors.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jcim.6b00795.
■
S
Figure S1. RMSD value fluctuation of the compounds
during the 20 ns MD simulation. Figure S2. Distribution
of predicted binding energy (Delta G, kcal/mol) for the
5 ns production MD. Figure S3. The IC50 curves for
compound I-01. Figure S4. The IC50 curves for
compound I-16 and the chemical synthesis and structure
identification of compounds (PDF)
AUTHOR INFORMATION
Corresponding Authors
*Tel.: +86-25-86185201. Fax: +86-25-86185182. E-mail: twf@
cpu.edu.cn (W.T.).
■
*Tel.: +86-25-86185180. Fax: +86-25-86185179. E-mail:
lut163@163.com (T.L.).
ORCID
Weifang Tang: 0000-0002-5527-897X
Author Contributions
∥Y.Z. and L.W. contributed equally to this work and should be
considered as cofirst authors.
Notes
The authors declare no competing financial interest.
L
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
P.; Jaiswal, B. S.; Seshagiri, S.; Koeppen, H.; Belvin, M.; Friedman, L.
S.; Malek, S. RAF inhibitors prime wild-type RAF to activate the
MAPK pathway and enhance growth. Nature 2010, 464, 431−5.
(12) Rebocho, A. P.; Marais, R. New insight puts CRAF in sight as a
therapeutic target. Cancer Discovery 2011, 1, 98−9.
(13) Mooz, J.; Oberoi-Khanuja, T. K.; Harms, G. S.; Wang, W.;
Jaiswal, B. S.; Seshagiri, S.; Tikkanen, R.; Rajalingam, K. Dimerization
of the kinase ARAF promotes MAPK pathway activation and cell
migration. Sci. Signaling 2014, 7, ra73.
ACKNOWLEDGMENTS
■
This work was financially supported by National Natural
Science Foundation of China (No. 21572273/B020601). A
Project Funded by the Priority Academic Program Develop-
ment of Jiangsu Higher Education Institutions and College
Student Innovation Project for the R&D of Novel Drugs (No.
J1030830). We also express our gratitude to Dr. Xiazhong Ren,
Dr. Huajun Yang, Dr. Chunlan Dong (Crown Bioscience
Corporation), and Dr. Jamie Planck (RBC, Pennsylvania, USA)
for their helpful support in biological evaluation.
(14) Ribas, A.; Flaherty, K. T. BRAF targeted therapy changes the
treatment paradigm in melanoma. Nat. Rev. Clin. Oncol. 2011, 8, 426−
33.
(15) Arora, R.; Di Michele, M.; Stes, E.; Vandermarliere, E.; Martens,
L.; Gevaert, K.; Van Heerde, E.; Linders, J. T.; Brehmer, D.; Jacoby, E.;
Bonnet, P. Structural investigation of B-Raf paradox breaker and
inducer inhibitors. J. Med. Chem. 2015, 58, 1818−31.
REFERENCES
■
(1) Caronia, L. M.; Phay, J. E.; Shah, M. H. Role of BRAF in thyroid
oncogenesis. Clin. Cancer Res. 2011, 17, 7511−7.
(16) Klein, C. A. Selection and adaptation during metastatic cancer
progression. Nature 2013, 501, 365−372.
(2) Rebocho, A. P.; Marais, R. ARAF acts as a scaffold to stabilize
BRAF:CRAF heterodimers. Oncogene 2013, 32, 3207−3212.
(3) Wan, P. T. C.; Garnett, M. J.; Roe, S. M.; Lee, S.; Niculescu-
Duvaz, D.; Good, V. M.; Project, C. G.; Jones, C. M.; Marshall, C. J.;
Springer, C. J.; Barford, D.; Marais, R. Mechanism of activation of the
RAF-ERK signaling pathway by oncogenic mutations of B-RAF. Cell
2004, 116, 855−867.
(17) Huang, J.; Hu, W.; Bottsford-Miller, J.; Liu, T.; Han, H. D.;
Zand, B.; Pradeep, S.; Roh, J. W.; Thanapprapasr, D.; Dalton, H. J.;
Pecot, C. V.; Rupaimoole, R.; Lu, C.; Fellman, B.; Urbauer, D.; Kang,
Y.; Jennings, N. B.; Huang, L.; Deavers, M. T.; Broaddus, R.; Coleman,
R. L.; Sood, A. K. Cross-talk between EphA2 and BRaf/CRaf is a key
determinant of response to dasatinib. Clin. Cancer Res. 2014, 20,
1846−55.
(18) Macrae, M.; Neve, R. M.; Rodriguez-Viciana, P.; Haqq, C.; Yeh,
J.; Chen, C.; Gray, J. W.; McCormick, F. A conditional feedback loop
regulates Ras activity through EphA2. Cancer Cell 2005, 8, 111−8.
(19) Lo, R. S. Receptor tyrosine kinases in cancer escape from BRAF
inhibitors. Cell Res. 2012, 22, 945−7.
(20) Yadav, V.; Zhang, X.; Liu, J.; Estrem, S.; Li, S.; Gong, X. Q.;
Buchanan, S.; Henry, J. R.; Starling, J. J.; Peng, S. B. Reactivation of
mitogen-activated protein kinase (MAPK) pathway by FGF receptor 3
(FGFR3)/Ras mediates resistance to vemurafenib in human B-RAF
V600E mutant melanoma. J. Biol. Chem. 2012, 287, 28087−98.
(21) Fedorenko, I. V.; Paraiso, K. H.; Smalley, K. S. Acquired and
intrinsic BRAF inhibitor resistance in BRAF V600E mutant melanoma.
Biochem. Pharmacol. 2011, 82, 201−9.
(4) Wang, X.; Kim, J. Conformation-specific effects of Raf kinase
inhibitors. J. Med. Chem. 2012, 55, 7332−41.
(5) Backes, A.; Zech, B.; Felber, B.; Klebl, B.; Muller, G. Small-
molecule inhibitors binding to protein kinase. Part II: the novel
pharmacophore approach of type II and type III inhibition. Expert
Opin. Drug Discovery 2008, 3, 1427−49.
(6) Henry, J. R.; Kaufman, M. D.; Peng, S. B.; Ahn, Y. M.; Caldwell,
T. M.; Vogeti, L.; Telikepalli, H.; Lu, W. P.; Hood, M. M.; Rutkoski, T.
J.; Smith, B. D.; Vogeti, S.; Miller, D.; Wise, S. C.; Chun, L.; Zhang, X.;
Zhang, Y.; Kays, L.; Hipskind, P. A.; Wrobleski, A. D.; Lobb, K. L.;
Clay, J. M.; Cohen, J. D.; Walgren, J. L.; McCann, D.; Patel, P.;
Clawson, D. K.; Guo, S.; Manglicmot, D.; Groshong, C.; Logan, C.;
Starling, J. J.; Flynn, D. L. Discovery of 1-(3,3-dimethylbutyl)-3-(2-
fluoro-4-methyl-5-(7-methyl-2-(methylamino)pyrido[2,3- d]-
pyrimidin-6-yl)phenyl)urea (LY3009120) as a pan-RAF inhibitor
with minimal paradoxical activation and activity against BRAF or
RAS mutant tumor cells. J. Med. Chem. 2015, 58, 4165−79.
(7) McArthur, G. A.; Chapman, P. B.; Robert, C.; Larkin, J.; Haanen,
J. B.; Dummer, R.; Ribas, A.; Hogg, D.; Hamid, O.; Ascierto, P. A.;
Garbe, C.; Testori, A.; Maio, M.; Lorigan, P.; Lebbe,
Schadendorf, D.; O’Day, S. J.; Kirkwood, J. M.; Eggermont, A. M.;
Dreno, B.; Sosman, J. A.; Flaherty, K. T.; Yin, M.; Caro, I.; Cheng, S.;
Trunzer, K.; Hauschild, A. Safety and efficacy of vemurafenib in
BRAFV600E and BRAFV600K mutation-positive melanoma (BRIM-
3): extended follow-up of a phase 3, randomised, open-label study.
Lancet Oncol. 2014, 15, 323−332.
(8) Flaherty, K. T.; Puzanov, I.; Kim, K. B.; Ribas, A.; McArthur, G.
A.; Sosman, J. A.; O’Dwyer, P. J.; Lee, R. J.; Grippo, J. F.; Nolop, K.;
Chapman, P. B. Inhibition of mutated, activated BRAF in metastatic
melanoma. N. Engl. J. Med. 2010, 363, 809−819.
(9) Sosman, J. A.; Kim, K. B.; Schuchter, L.; Gonzalez, R.; Pavlick, A.
C.; Weber, J. S.; McArthur, G. A.; Hutson, T. E.; Moschos, S. J.;
Flaherty, K. T.; Hersey, P.; Kefford, R.; Lawrence, D.; Puzanov, I.;
Lewis, K. D.; Amaravadi, R. K.; Chmielowski, B.; Lawrence, H. J.; Shyr,
Y.; Ye, F.; Li, J.; Nolop, K. B.; Lee, R. J.; Joe, A. K.; Ribas, A. Survival in
BRAF V600−mutant advanced melanoma treated with vemurafenib.
N. Engl. J. Med. 2012, 366, 707−714.
(22) Spagnolo, F.; Ghiorzo, P.; Queirolo, P. Overcoming resistance
to BRAF inhibition in BRAF-mutated metastatic melanoma.
Oncotarget 2014, 5, 10206−10221.
(23) Schrodinger, L. Schrodinger software suite; Schrodinger, LLC:
̈
́
C.; Jouary, T.;
New York 2011.
(24) Ngan, C. H.; Bohnuud, T.; Mottarella, S. E.; Beglov, D.; Villar,
E. A.; Hall, D. R.; Kozakov, D.; Vajda, S. FTMAP: extended protein
mapping with user-selected probe molecules. Nucleic Acids Res. 2012,
40, W271−5.
(25) Zhang, Y.; Yang, S.; Jiao, Y.; Liu, H.; Yuan, H.; Lu, S.; Ran, T.;
Yao, S.; Ke, Z.; Xu, J.; Xiong, X.; Chen, Y.; Lu, T. An integrated virtual
screening approach for VEGFR-2 inhibitors. J. Chem. Inf. Model. 2013,
53, 3163−3177.
(26) Van Der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A.
E.; Berendsen, H. J. GROMACS: fast, flexible, and free. J. Comput.
Chem. 2005, 26, 1701−18.
(27) Duan, Y.; Chowdhury, S.; Lee, M. C.; Xiong, G.; Zhang, W.;
Yang, R.; Luo, R.; Lee, T.; Caldwell, J.; Wang, J.; Kollman, P.; Wu, C.;
Cieplak, P. A point-charge force field for molecular mechanicssimu-
lations of proteins based on condensed-phase quantum mechanical
calculations. J. Comput. Chem. 2003, 24, 1999−2012.
(28) Jorgensen, W. L. Quantum and statistical mechanical studies of
liquids. 10. Transferable intermolecular potential functions for water,
alcohols, and ethers. Application to liquid water. J. Am. Chem. Soc.
1981, 103, 335−340.
(29) Hess, B. P-LINCS: a parallel linear constraint solver for
molecular simulation. J. Chem. Theory Comput. 2008, 4, 116−122.
(30) Miller, B. R.; McGee, T. D.; Swails, J. M.; Homeyer, N.; Gohlke,
H.; Roitberg, A. E. MMPBSA.py: an efficient program for end-state
free energy calculations. J. Chem. Theory Comput. 2012, 8, 3314−3321.
́
(10) Sanchez-Laorden, B.; Viros, A.; Girotti, M. R.; Pedersen, M.;
Saturno, G.; Zambon, A.; Niculescu-Duvaz, D.; Turajlic, S.; Hayes, A.;
Gore, M.; Larkin, J.; Lorigan, P.; Cook, M.; Springer, C.; Marais, R.
BRAF inhibitors induce metastasis in RAS mutant or inhibitor-
resistant melanoma cells by reactivating MEK and ERK signaling. Sci.
Signaling 2014, 7, ra30.
(11) Hatzivassiliou, G.; Song, K.; Yen, I.; Brandhuber, B. J.;
Anderson, D. J.; Alvarado, R.; Ludlam, M. J.; Stokoe, D.; Gloor, S.
L.; Vigers, G.; Morales, T.; Aliagas, I.; Liu, B.; Sideris, S.; Hoeflich, K.
M
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX

Journal of Chemical Information and Modeling
Article
(31) Cher
́
on, N.; Jasty, N.; Shakhnovich, E. I. OpenGrowth: an
Antitumor activity of the selective pan-RAF inhibitor TAK-632 in
BRAF inhibitor-resistant melanoma. Cancer Res. 2013, 73, 7043−55.
(46) Mathieu, S.; Gradl, S. N.; Ren, L.; Wen, Z.; Aliagas, I.; Gunzner-
Toste, J.; Lee, W.; Pulk, R.; Zhao, G.; Alicke, B.; Boggs, J. W.;
Buckmelter, A. J.; Choo, E. F.; Dinkel, V.; Gloor, S. L.; Gould, S. E.;
Hansen, J. D.; Hastings, G.; Hatzivassiliou, G.; Laird, E. R.; Moreno,
D.; Ran, Y.; Voegtli, W. C.; Wenglowsky, S.; Grina, J.; Rudolph, J.
Potent and selective aminopyrimidine-based B-Raf inhibitors with
favorable physicochemical and pharmacokinetic properties. J. Med.
Chem. 2012, 55, 2869−81.
(47) Tiacci, E.; Trifonov, V.; Schiavoni, G.; Holmes, A.; Kern, W.;
Martelli, M. P.; Pucciarini, A.; Bigerna, B.; Pacini, R.; Wells, V. A.;
Sportoletti, P.; Pettirossi, V.; Mannucci, R.; Elliott, O.; Liso, A.;
Ambrosetti, A.; Pulsoni, A.; Forconi, F.; Trentin, L.; Semenzato, G.;
Inghirami, G.; Capponi, M.; Di Raimondo, F.; Patti, C.; Arcaini, L.;
automated and rational algorithm for finding new protein ligands. J.
Med. Chem. 2016, 59, 4171−4188.
(32) Swetha, R. G.; Ramaiah, S.; Anbarasu, A. Molecular dynamics
studies on D835N mutation in FLT3-its impact on FLT3 protein
structure. J. Cell. Biochem. 2016, 117, 1439−45.
(33) Yang, W.; Chen, Y.; Zhou, X.; Gu, Y.; Qian, W.; Zhang, F.; Han,
W.; Lu, T.; Tang, W. Design, synthesis and biological evaluation of bis-
aryl ureas and amides based on 2-amino-3-purinylpyridine scaffold as
DFG-out B-Raf kinase inhibitors. Eur. J. Med. Chem. 2015, 89, 581−96.
(34) Smith, A. L.; DeMorin, F. F.; Paras, N. A.; Huang, Q.; Petkus, J.
K.; Doherty, E. M.; Nixey, T.; Kim, J. L.; Whittington, D. A.; Epstein,
L. F.; Lee, M. R.; Rose, M. J.; Babij, C.; Fernando, M.; Hess, K.; Le, Q.;
Beltran, P.; Carnahan, J. Selective inhibitors of the mutant B-Raf
pathway: discovery of a potent and orally bioavailable aminoisoquino-
line. J. Med. Chem. 2009, 52, 6189−6192.
(35) Wilhelm, S. M.; Carter, C.; Tang, L.; Wilkie, D.; McNabola, A.;
Rong, H.; Chen, C.; Zhang, X.; Vincent, P.; McHugh, M.; Cao, Y.;
Shujath, J.; Gawlak, S.; Eveleigh, D.; Rowley, B.; Liu, L.; Adnane, L.;
Lynch, M.; Auclair, D.; Taylor, I.; Gedrich, R.; Voznesensky, A.; Riedl,
B.; Post, L. E.; Bollag, G.; Trail, P. A. BAY 43−9006 exhibits 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, 7099−7109.
̀
Musto, P.; Pileri, S.; Haferlach, C.; Schnittger, S.; Pizzolo, G.; Foa, R.;
Farinelli, L.; Haferlach, T.; Pasqualucci, L.; Rabadan, R.; Falini, B.
BRAF mutations in hairy-cell leukemia. N. Engl. J. Med. 2011, 364,
2305−2315.
(48) Xing, M.; Westra, W. H.; Tufano, R. P.; Cohen, Y.; Rosenbaum,
E.; Rhoden, K. J.; Carson, K. A.; Vasko, V.; Larin, A.; Tallini, G.;
Tolaney, S.; Holt, E. H.; Hui, P.; Umbricht, C. B.; Basaria, S.; Ewertz,
M.; Tufaro, A. P.; Califano, J. A.; Ringel, M. D.; Zeiger, M. A.;
Sidransky, D.; Ladenson, P. W. BRAF mutation predicts a poorer
clinical prognosis for papillary thyroid cancer. J. Clin. Endocrinol.
Metab. 2005, 90, 6373−6379.
(36) Zhang, C.; Spevak, W.; Zhang, Y.; Burton, E. A.; Ma, Y.; Habets,
G.; Zhang, J.; Lin, J.; Ewing, T.; Matusow, B.; Tsang, G.; Marimuthu,
A.; Cho, H.; Wu, G.; Wang, W.; Fong, D.; Nguyen, H.; Shi, S.;
Womack, P.; Nespi, M.; Shellooe, R.; Carias, H.; Powell, B.; Light, E.;
Sanftner, L.; Walters, J.; Tsai, J.; West, B. L.; Visor, G.; Rezaei, H.; Lin,
P. S.; Nolop, K.; Ibrahim, P. N.; Hirth, P.; Bollag, G. RAF inhibitors
that evade paradoxical MAPK pathway activation. Nature 2015, 526,
583−586.
(37) Yuan, Y.; Pei, J.; Lai, L. Binding site detection and druggability
prediction of protein targets for structure- based drug design. Curr.
Pharm. Des. 2013, 19, 2326−2333.
(38) Zhang, Y.; Jiao, Y.; Xiong, X.; Liu, H.; Ran, T.; Xu, J.; Lu, S.; Xu,
A.; Pan, J.; Qiao, X.; Shi, Z.; Lu, T.; Chen, Y. Fragment virtual
screening based on Bayesian categorization for discovering novel
VEGFR-2 scaffolds. Mol. Diversity 2015, 19, 895−913.
(39) van Linden, O. P. J.; Kooistra, A. J.; Leurs, R.; de Esch, I. J. P.;
de Graaf, C. KLIFS: a knowledge-based structural database to navigate
kinase−ligand interaction space. J. Med. Chem. 2014, 57, 249−277.
(40) Anforth, R. M.; Blumetti, T. C.; Kefford, R. F.; Sharma, R.;
Scolyer, R. A.; Kossard, S.; Long, G. V.; Fernandez-Penas, P.
Cutaneous manifestations of dabrafenib (GSK2118436): a selective
inhibitor of mutant BRAF in patients with metastatic melanoma. Br. J.
Dermatol. 2012, 167, 1153−60.
(41) Chartier, M.; Chenard, T.; Barker, J.; Najmanovich, R. Kinome
render: a stand-alone and web-accessible tool to annotate the human
protein kinome tree. PeerJ 2013, 1, e126.
(42) Chapman, P. B.; Solit, D. B.; Rosen, N. Combination of RAF
and MEK inhibition for the treatment of BRAF-mutated melanoma:
feedback is not encouraged. Cancer Cell 2014, 26, 603−4.
(43) Maurer, G.; Tarkowski, B.; Baccarini, M. Raf kinases in cancer-
roles and therapeutic opportunities. Oncogene 2011, 30, 3477−88.
(44) Gould, A. E.; Adams, R.; Adhikari, S.; Aertgeerts, K.; Afroze, R.;
Blackburn, C.; Calderwood, E. F.; Chau, R.; Chouitar, J.; Duffey, M.
O.; England, D. B.; Farrer, C.; Forsyth, N.; Garcia, K.; Gaulin, J.;
Greenspan, P. D.; Guo, R.; Harrison, S. J.; Huang, S. C.; Iartchouk, N.;
Janowick, D.; Kim, M. S.; Kulkarni, B.; Langston, S. P.; Liu, J. X.; Ma,
L. T.; Menon, S.; Mizutani, H.; Paske, E.; Renou, C. C.; Rezaei, M.;
Rowland, R. S.; Sintchak, M. D.; Smith, M. D.; Stroud, S. G.; Tregay,
M.; Tian, Y.; Veiby, O. P.; Vos, T. J.; Vyskocil, S.; Williams, J.; Xu, T.;
Yang, J. J.; Yano, J.; Zeng, H.; Zhang, D. M.; Zhang, Q.; Galvin, K. M.
Design and optimization of potent and orally bioavailable tetrahy-
dronaphthalene Raf inhibitors. J. Med. Chem. 2011, 54, 1836−46.
(45) Nakamura, A.; Arita, T.; Tsuchiya, S.; Donelan, J.; Chouitar, J.;
Carideo, E.; Galvin, K.; Okaniwa, M.; Ishikawa, T.; Yoshida, S.
N
DOI: 10.1021/acs.jcim.6b00795
J. Chem. Inf. Model. XXXX, XXX, XXX−XXX