Synthesis and docking study of 2-phenylaminopyrimidine Abl tyrosine kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.09.127

Six analogs of imatinib, an Abl kinase inhibitor clinically used as a first-line therapeutic agent for chronic myeloid leukaemia (CML), have been synthesized and characterized. And their potency as Abl kinase inhibitors have been screened by a robust virt

Full text of DOI:10.1016/j.bmcl.2011.09.127

Bioorganic & Medicinal Chemistry Letters 21 (2011) 6964–6968
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and docking study of 2-phenylaminopyrimidine Abl tyrosine
kinase inhibitors
⇑
Shuang Lü, Qun Luo, Xiang Hao, Xianchan Li, Liyun Ji, Wei Zheng, Fuyi Wang
The CAS Key Laboratory of Analytical Chemistry for Living Biosystems, Beijing National Laboratory for Molecular Sciences, Institute of Chemistry,
Chinese Academy of Sciences, Beijing 100190, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 May 2011
Revised 22 September 2011
Accepted 30 September 2011
Available online 7 October 2011
Six analogs of imatinib, an Abl kinase inhibitor clinically used as a first-line therapeutic agent for chronic
myeloid leukaemia (CML), have been synthesized and characterized. And their potency as Abl kinase
inhibitors have been screened by a robust virtual screening method developed based on the crystal struc-
ture (PDB code 2hyy) of Abl-imatinib complex using Surflex-Docking. The docking results are consistent
with the inhibitory potency of the compounds characterized by MS method. And the H-bonds between
imatinib analogs and Thr315 and Met318 residues in Abl kinase are shown to be crucial for achieving
accurate poses and high binding affinities for the ATP-competitive kinase inhibitors.
Keywords:
Abl kinase
Imatinib analogs
Synthesis
Ó 2011 Elsevier Ltd. All rights reserved.
Docking analysis
Virtual screening
The Abl kinase is normally regulated by an autoinhibitory
mechanism, the disruption of which leads to chronic myelogenous
leukemia (CML).1 CML is a clonal hematopoietic stem cell disorder
characterized by the presence of the Philadelphia (Ph) chromo-
some. A series of inhibitors with the class of 2-phenylaminopyrim-
idine (PAP) pharmacophore, in particular imatinib, have been
identified to have exceptionally high affinity and specificity for
Abl.2 Imatinib (also named STI-571 or gleevec) is the first pro-
tein-tyrosine kinase inhibitor that is used to treat CML clinically.
Imatinib displaces ATP and captures a specific inactive conforma-
tion of the activation loop of Abl in which the loop mimics bound
peptide substrate.1,3–5 This feature of imatinib could reduce the
toxicity to normal cell, but develop drug-resistance due to the
Abl kinase domain mutations.6 To circumvent this problem,
rational design of novel inhibitors exhibiting effectiveness against
Bcr-Abl has attracted increasing interest. Nilotinib (also named
AMN107) is a high affinity aminopyrimidine-based ATP-competi-
tive inhibitor that exhibits superior potency to imatinib as an
inhibitor of wild-type Bcr-Abl and imatinib-resistant Bcr-Abl
mutant-expressing cells in vitro.7,8 Katsoulas et al.9 developed a
novel group of dual targeted agent termed ‘combi-molecules’
capable of not only blocking the Abl kinase active pocket but also
damaging DNA, of which ZRF1 was the first ever multitargeted
combi-molecule exerting a tandem targeting of Bcr-Abl mediated
antiapoptotic signaling and activation of the DNA damage response
pathway. A series of potent dual Src/Abl kinase inhibitors based on
a 9-(arenethenyl)purine core have been synthesized, and several
compounds exhibited modest cellular potency (IC50 = 300–
400 nM) against the Bcr-Abl mutant T315I that is encouraging
given the insensitivity of this mutant to all currently marketed
first- and second-generation Bcr-Abl inhibitors.10
On the other hand, with the understanding on the interactions of
drug candidates and kinases, increasing attention has been paid to
develop virtual screening (VS) for avoiding lots of tedious synthetic
works for pursuing novel kinase inhibitors as potential anticancer
agents. VS, especially docking methodology, is an evolving chal-
lenge and an attractive research proposition.11 However, it still
needs to be studied in depth and transformed into a tool of greater
confidence and utility. Desiraju and co-workers12 carried out a VS
study on 128 selected inhibitors of epidermal growth factor recep-
tor (EGFR) kinase with GOLD and LigandFit. And Wang et al.13
tested 100 protein–ligand complexes to evaluate the abilities of
eleven popular scoring functions to reproduce experimentally
determined structures and binding affinities. The availability of
the crystal of Abl complexes with imatinib facilitates the computa-
tionally structure-based design of inhibitors against this target.3,5
Juan et al. established robust 3D-QSAR model for Bcr-Abl inhibitors,
and revealed novel insights towards inhibition of Bcr-Abl
oncoprotein, providing rational insights into the ligand structural
requirements to achieve better Bcr-Abl inhibitory activity which
can be utilized in the design of new and more potent Bcr-Abl
inhibitors.14 However, arriving at the appropriate docking/scoring
combination is still a challenge because of lacking certain and clear
⇑
Corresponding author. Tel./fax: +86 10 62529069.
E-mail address: fuyi.wang@iccas.ac.cn (F. Wang).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.09.127

S. Lü et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6964–6968
6965
relationship between docking and ranking accuracies.15 In this
present work, we have established a molecular model based on
the crystal structure of Abl-imitanib complex and developed a
robust VS method using the Surflex-Docking module for rapid
screening of imatinib analogs synthesized in our laboratory. The
docking analysis revealed that the formation of two hydrogen
bonds between the ligands and the kinase are essential for
inhibitory activity, providing rational concepts for the design and
synthesis of novel Abl inhibitors.
Firstly, benzoyl chloride intermediates 1b, 2b, 3b, 4b, 5b and 6b
were synthesized by reacting respective benzoic acids (1a–5a,
Scheme 1) or picolinic acid (6a, Scheme 1) with thionyl chloride.
In general, 1 g of benzoic acid compounds 1a, 2a, 3a, 4a, 5a or
picolinic acid (6a) was added to thionyl chloride (12 mL) and then
7.494 (d, 1H), 7.412 (d, 2H), 7.237 (d, 1H), 2.525 (s, 3H, –SCH3),
2.245 (s, 3H, –CH3); 13C NMR (600 MHz, DMSO-d6) d: 165.19,
161.07, 160.09, 159.67, 145.65, 143.35, 143.02, 141.87, 137.61,
137.58, 135.12, 131.25, 130.54, 128.59, 127.85, 126.91, 125.25,
117.65, 117.46, 108.28, 17.94, 14.54; MS (ESI, m/z): 428.5 (M+H+,
C
24H22N5OS requires 428.5), 450.5 (M+Na+); mp: 185–186 °C; Anal.
Calcd for C24H21N5OSÁ3HCl: C, 53.69; H, 4.51; N, 13.04. Found: C,
55.95; H, 4.78; N, 12.88.
4-nitro-N-(4-methyl-3(4-(pyridine-3-yl)pyrimidin-2-ylami-
no)phenyl)benzamide (3): 1H NMR (400 MHz, DMSO-d6): 10.590
(s, 1H), 9.430 (d, 1H), 9.155 (s, 1H), 8.918 (d, 1H), 8.875 (d, 1H),
8.592 (d, 1H), 8.339 (d, 2H), 8.196 (d, 2H), 8.160 (s, 1H), 7.910
(m, 1H), 7.543 (d, 1H), 7.457 (d, 1H), 7.224 (d, 1H), 2.234 (s, 3H,
–CH3); 13C NMR (600 MHz, DMSO-d6) d: 164.17, 161.02, 160.08,
159.68, 149.44, 145.27, 142.65, 141.76, 141.02, 137.72, 137.16,
135.12, 130.60, 129.76, 128.27, 126.94, 123.79, 117.54, 117.46,
108.40, 18.01; MS (ESI, m/z): 427.2 (M+H+, C23H19N6O3 requires
427.4), 449.2 (M+Na+); mp: 237–238 °C; Anal. Calcd for
C23H18N6O3ÁHCl: C, 59.68; H, 4.14; N, 18.16. Found: C, 59.29; H,
4.12; N, 17.85.
stirred with addition of 50 lL pyridine as a catalyst. The resulting
mixture was refluxed until no gas overflowed. Then the excess
thionyl chloride was removed off in vacuo, and the residue was di-
rectly used for the next step synthesis.
The imatinib analogs 3,4-R1,R2-N-(4-methyl-3(4-(pyridine-3-
yl)pyrimidin-2-ylamino)phenyl)benzamide
(R1 = R2 = H
(1);
R1 = H, R2 = SCH3 (2); R1 = NO2, R2 = H (3); R1 = NH2, R2 = H (4);
R1 = R2 = NO (5)) and N-(4-methyl-3-(4-(pyridin-3-yl) pyrimidin-
2-ylamino)phenyl)picolinamide (6) was synthesized following a
modified procedure described in the literature16 as shown in
Scheme 1. Briefly, an excess amount of benzoyl chloride compound
1b, 2b, 3b, 4b, 5b or 6b was added to a suspension of 6-methyl-N-
4-amino-N-(4-methyl-3(4-(pyridine-3-yl)pyrimidin-2-ylami-
no)phenyl)benzamide (4): 1H NMR (400 MHz, DMSO-d6): 9.303
(s, 1H), 8.722 (s, 1H), 8.652 (s, 1H), 8.480 (d, 3H), 8.290 (d, 3H),
8.043 (t, 4H), 7.460 (s, 1H), 7.231 (s, 1H), 7.196 (d, 1H), 7.097 (s,
1H), 2.381 (s, 3H, –CH3); 13C NMR (600 MHz, DMSO-d6) d:
164.07, 162.00, 161.51, 159.84, 151.76, 149.46, 148.57, 141.11,
138.30, 137.06, 134.78, 132.56, 130.53, 129.53, 128.48, 124.15,
123.90, 117.57, 117.15, 107.99, 18.05; MS (ESI, m/z): 397.3
(M+H+, C23H20N6O requires 396.4), 419.2 (M+Na+), 435.3 (M+K+);
mp: 238–239 °C. Anal. Calcd for C23H20N6OÁHCl: C, 63.81; H,
4.89; N, 19.41. Found: C, 63.05; H, 4.30; N, 19.14.
3,4-dinitroso-N-(4-methyl-3(4-(pyridine-3-yl)pyrimidin-2-
ylamino)phenyl)benzamide (5): 1H NMR (400 MHz, DMSO-d6):
10.585 (s, 1H), 9.360 (s, 1H), 9.092 (s, 1H), 8.762 (d, 2H), 8.674
(d, 1H), 8.564 (d, 1H), 8.179 (m, 3H), 7.684 (d, 1H), 7.492 (m,
2H), 7.246 (d, 1H), 2.250 (s, 3H, –CH3); 13C NMR (600 MHz,
DMSO-d6) d: 164.84, 162.09, 161.62, 159.96, 155.64, 154.26,
151.86, 148.68, 138.37, 137.36, 136.43, 134.89, 132.66, 130.62,
129.34, 128.42, 124.25, 121.77, 121.14, 117.62, 117.17, 108.06,
18.15; MS (ESI, m/z): 440.1 (M+H+, C23H18N7O3 requires 439.4);
mp: 231–233 °C; Anal. Calcd for C23H17N7O3ÁHCl: C, 58.05; H,
3.81; N, 20.60. Found: C, 58.20; H, 3.85; N, 20.92.
(4-(pyridin-3-yl)pyrimidin-2-yl)
benzene-1,3-diamine
(0.5 g,
2 mmol) in dichloromethane (10 mL) containing TEA (0.58 mL,
4 mmol), and the mixture was reacted at 0 °C for 4 h, and TLC of
reaction mass indicated the absence of starting compound. The
solution was then filtered and washed with dichloromethane, the
excess dichloromethane was removed in vacuo, and the residue
was purified by recrystallization from ethanol.
N-(4-methyl-3(4-(pyridine-3-yl)pyrimidin-2-ylamino)
phenyl)benzamide (1): 1H NMR (400 MHz, DMSO-d6): 10.190 (s,
1H), 9.266 (s, 1H), 8.954 (s, 1H), 8.673 (s, 1H), 8.458 (d, 2H),
8.082 (s, 1H), 7.938 (m, 2H), 7.471 (m, 5H), 7.416 (s, 1H), 7.189
(d, 1H), 2.215 (s, 3H, –CH3); 13C NMR (600 MHz, DMSO-d6) d:
165.77, 161.98, 161.56, 159.84, 151.75, 148.57, 138.19, 137.54,
135.48, 134.78, 132.59, 131.80, 130.41, 128.72, 127.98, 124.14,
117.60, 117.13, 107.90, 18.02; MS (ESI, m/z): 382.3 (M+H+,
C
23H20N5O requires 382.4), 404.3 (M+Na+), 420.3 (M+K+); mp:
190–191 °C; Anal. Calcd for C23H19N5O: C, 72.42; H, 5.02; N,
18.36. Found: C, 71.86; H, 4.97; N, 18.18.
3-(sulfurmethyl)-N-(4-methyl-3(4-(pyridine-3-yl)pyrimidin -
2-ylamino)phenyl)benzamide (2): 1H NMR (400 MHz, DMSO-d6):
10.160 (s, 1H), 9.300 (s, 1H), 8.996 (s, 1H), 8.713 (d, 1H), 8.543–
8.492 (m, 2H), 8.100 (s, 1H), 7.951 (d, 2H), 7.600–7.511 (m, 2H),
N-(4-methyl-3-(4-(pyridin-3-yl)pyrimidin-2-ylamino)
phenyl)picolinamide (6): 1H NMR (400 MHz, DMSO-d6): 10.566 (s,
1H), 9.281 (s, 1H), 8.987 (s, 1H), 8.744 (d, 1H), 8.688 (d, 1H), 8.527–
8.477 (m, 2H), 8.256 (s, 1H), 8.163 (d, 1H), 8.089 (t, 1H), 7.678 (m,
1H), 7.580–7.507 (m, 2H), 7.446 (d, 1H), 7.234 (d, 1H), 2.245 (s, 3H,
–CH3); 13C NMR (600 MHz, DMSO-d6) d: 158.58, 136.73, 136.55,
H
N
N
NH2
R1
R2
N
H
N
H
N
N
X
R1
R2
N
R1
R2
N
N
O
X
COOH
X
COCl
i
ii
1a X=C R1= R2=H
1b X=C R1= R2=H
1 X=C R1= R2=H
2a X=C R1=H R2=SCH3
3a X=C R1=NO2 R2=H
4a X=C R1=NH2 R2=H
5a X=C R1= R2=NH2
6a X=N R1= R2=H
2b X=C R1=H R2=SCH3
3b X=C R1=NO2 R2=H
4b X=C R1=NH2 R2=H
5b X=C R1= R2=NH2
6b X=N R1= R2=H
2 X=C R1=H R2=SCH3
3 X=C R1=NO2 R2=H
4 X=C R1=NH2 R2=H
5 X=C R1= R2=NH2
6 X=N R1= R2=H
Scheme 1. Synthesis route of imatinib analogs. (i) SOCl2, py, N2, rt, 70 °C; (ii) CH2Cl2, Et3N, 0 °C, 4 h.

6966
S. Lü et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6964–6968
132.88, 131.44, 130.80, 130.07, 128.53, 126.56, 126.16, 125.44,
114.50, 56.05, 49.65; MS (ESI, m/z): 383.1 (M+H+, C22H19N6O re-
quires 382.4); mp: 192–193 °C; Anal. Calcd for C22H18N6O: C,
69.10; H, 4.74; N, 21.98. Found: C, 68.68; H, 4.68; N, 21.35.
Crystal of compound 1 was grown from its saturated DMF solu-
tion at 4 °C. The single-crystal X-ray diffraction analysis reveals
that 1 crystallizes in the monoclinic P21/n space group (Fig. 1).17
In the crystal, we observed two intermolecular hydrogen bonds:
N(4)-H(4A)Á Á ÁN(1) and N(5)-H(5A)Á Á ÁN(3) (Fig. S1 in Supplemen-
tary data) and the bond distances were 2.931 and 3.041, respec-
tively. And estimated interplane distance between the phenyl
rings of the two molecules is about 3.37 Å, thus the compound in
Abl-imatinib complex crystal, but the rest was deviated from the
original position (Figs. S3 and S4). The docking molecule bound
to Abl via three H-bond interactions: the pyridine-N with the back-
bone-NH of Met318, the anilino-NH with the side chain of Thr315,
the amide-NH to the side chain of Glu286, but there were lack of
three H-bonds observed in the X-ray crystal structure of Abl-imati-
nib complex (Figs. S3 and S4), indicating that the direct docking is
not effective. The reason may be that we did not consider the pro-
tonation of the nitrogen of N-methylpiperazine in the ligand. Tak-
ing the protonation into account, a significant improvement was
observed for an indirect docking. The position of the docking
molecule was completely overlapped with that of the ligand in
the Abl-imatinib complex crystal, and the number and location of
the H-bonds between the docking ligand and the receptor were
the same as those of the ligand in the complex crystal (Fig. 2).
The docking score and RMSD value were 13.08 and 0.39, respec-
tively, indicative of that the indirect docking results upon the pro-
tonation of the ligand was more reasonable and accurate than that
of the direct docking.
Then the six imatinib analogs, of which the structures were
built in SYBYL and minimized by Tripos force field with an energy
convergence value of 0.05 kcal molÀ1, were separately docked into
the binding pocket for docking-scoring calculations using the same
method validated. The docking scores are 11.20 (1), 12.10 (2), 8.72
(3), 9.41 (4), 6.51 (5) and 5.99 (6), respectively. These results are
consistent with the inhibitory potency order of the compounds
characterized by MS method, that is, the higher the score of the
compound, the higher its inhibitory efficiency.18
crystal shows extensive intermolecular
pÁ Á Á
p
interactions
(Fig. S2). The dihedral angels between the pyrimidine and the pyr-
idine rings, the pyrimidine and the neighboring benzene, and the
pyrimidine and the rightmost benzene were calculated to be
3.53°, 71.67° and 89.05°, respectively. Standard data relating to
the X-ray crystal structure of compound 1 have been deposited
in the Cambridge Crystallographic Data Centre with CCDC refer-
ence number 845049.
We have previously developed a mass spectrometric approach
to evaluate the inhibitory potency of imitanib derivative com-
pounds 1–6.18 Given that the inhibitory efficiency of 10
lM imati-
nib towards Abl was 100%, the relatively inhibitory efficiency of
compounds 1–6 at the same concentration was determined to be
34.3, 74.9, 20.0, 19.9, 16.3 and 0, respectively. And the IC50 values
of imitanib and the derivative compounds 1, 2 and 5 against Abl
are 0.234, 12.7, 1.1 and >25 lM, respectively.
The docking analysis was performed using the Surflex-Dock
module, a fully automatic docking tool available on SYBYL version
X 1.1 (Tripos Inc.), running on Dual-core Intel(R) E5300 CPU
2.60 GHz, RAM Memory 2 GB under the Windows XP system. The
crystal structure of Abl-imatinib complex was collected from PDB
under code 2hyy.19 All the hydrogen atoms were added to define
the correct configuration and tautomeric states. With standard
set parameters, then the modeled structure was energy-minimized
using AMBER7 F99 force field with the Powell energy minimization
algorithm, distance dependent dielectric function and current
charges. After extracted the bound ligand imatinib, the structure
of Abl kinase corresponding to the constringent energy gradient
(0.05 kcal molÀ1) was used for re-docking and scoring calculations
of imatinib so as to check the accuracy of the Surflex-Dock
program.
Further analysis on the interactions between Abl and com-
pounds 1 and 2 shows that these two compounds make three
hydrogen bonds with the receptor: the pyridine-N with the back-
bone-NH of Met318, the anilino-NH with the side chain of
Thr315, the amide-NH to the side chain of Glu286, but both lack
of two H-bonds which form between the N-methylpiperazine of
imatinib and the carboxyl of Ile360 and His361 residues, due to
the substitution of piperazine by benzene (for 1) or sulfurmethyl-
benzene (for 2). The leftmost part of compounds 1 and 2 over-
lapped with imatinib, and the amide-carbonyl of 1 and 2 shifted
away from Asp381, leading to absence of the H-bond between
the amide-carbonyl group and amide-NH of Asp381 (Fig. 3). The
presence of Thr315 appears to be a key requirement for the ability
of this class of compounds to inhibit wild-type Abl. This threonine
residue is replaced by a methionine residue in many protein
kinases, for example, insulin receptor tyrosine kinase (IRK). Methi-
onine residues cannot form this kind of hydrogen bond with imati-
nib ligand, and its side chain would also interfere with the binding
of the phenyl-moiety of imatinib, but not affect the binding of ATP.
For a direct docking, for which no any configuration restraints
are set up, the re-docking score and RMSD value for the Abl-imati-
nib complex were 10.66 and 2.29, respectively. It can be seen that
the pyrimidine ring and pyridine ring of the docking ligand
molecule were superposed with those of the ligand in the
A
B
GLU286
THR315
ILE360
HIS361
ASP381
MET318
Figure 2. (A) The superposition of the bound imatinib (green) to Abl in the crystal
structure3,5 and docked imatinib (colored by atoms) to Abl in the molecular model
constructed by indirect docking. (B) Details of potential hydrogen bonds (red dotted
lines) between imatinib and Abl generated by indirect docking analysis.
Figure 1. A view of compound 1 with displacement ellipsoids drawn at the 30%
probability level.

S. Lü et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6964–6968
6967
HIS361
ILE360
ASP381
MET318
GLU286
Figure 4. The docked conformers/poses of compound 1 (color by atoms) at the
active site of Abl kinase as generated via Surflex docking-scoring combinations
compared with the X-ray structure of unbound 1 (red).
THR315
Figure 3. Detail of the binding of compounds 1 and 2 to Abl showing potential
hydrogen bonds in red dotted lines. The ligands colored by atoms, in purple and in
green represent imatinib in Abl-imatinib complex crystal and docking compounds 1
and 2, respectively.
inhibitory potency of the compounds characterized by MS method.
Furthermore, the molecular models constructed by docking analy-
sis indicate that the formation of ligand-receptor hydrogen bonds
between the py-N and Met318, and the anilino N–H and Thr315
are crucial to achieve accurate poses and high binding affinities
for the ATP-competitive kinase inhibitors. In particular, the pres-
ence of Thr315 in the ATP-binding pocket of the kinases is a key
structural requirement for this class of imatinib analogs binding
to the receptors. The absence of the H-bond between the docking
ligands and the Thr315 in the ATP-binding pocket leads to dramat-
ically reduction of inhibitory activities of the imatinib derivatives.
Our findings indicate that the virtual screening method based on
Surflex-Docking can provide rational insights to design novel ki-
nase inhibitors for development of next generation of molecular
targeting anticancer agents.
This may be the reason why the kinases with Met residue in the
place of Thr315 in the ATP-binding pocket are resistant to imatinib.
However, the formation of H-bond between imatinib derivatives
and Met318 of Abl kinase strengthens the binding of the ligands
to Abl. The adenine group of ATP generally makes two H-bonds
with backbone atoms of the peptide chain connecting the N- and
C-lobes of kinase domains. The extracyclic amino group of ATP do-
nates a hydrogen bond to the carbonyl oxygen of the residue cor-
responding to Glu316 in Abl, and the nitrogen (N1) of the purine
ring accepts a hydrogen bond from the amide nitrogen of residue
Met318.5,20 Many small molecular inhibitors of protein kinases
are anchored to the kinase domain by the pair of hydrogen bonds
that mimic those formed by adenine.
Acknowledgments
The pattern of hydrogen bonds formed by compound 3 with the
Abl receptor is similar to that of compounds 1 and 2 with Abl
kinase. The carbonyl oxygen deviated from the Asp-Phe-Gly
(DGF) motif in Abl and accepted a hydrogen bond from the amide
nitrogen of residue Lys271 which replaced the residue Asp381
(Fig. S5). Compounds 4, 5 and 6 had many similarities to each an-
other: deviation from the original position of imatinib in the C-lobe
of Abl, and accepting hydrogen bond from residue Tyr253. Notably,
compound 6 lost the two important H-bonds formed with Thr315
and Met318 residues (Figs. S6–S8), leading it to be the weakest
inhibitor to Abl as indicated by both docking analysis and MS-
based assay.18
We thank NSFC (Grant No. 90713020, 20975103, 21005081 (for
X.L.) and 21020102039), the 973 Program of MOST (Grant No.
2007CB935601) and the Chinese Academy of Sciences (Hundred
Talent Program) for support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.09.127.
References and notes
We also compared the conformational differences of compound
1 between the docking molecule in the Abl using Surflex-Docking
and the molecule in the crystal structure (Fig. 1). We first kept
the possibility of H-bonding between the N4Á Á ÁH of compound 1
and the side chain of Thr315 because of the importance of this
hydrogen bond. The molecular model (Fig. 4) showed that when
we overlapped the pyrimidine ring of the two molecules, the rest
of the docking molecule deviated from the crystal conformation
to meet the requirements for docking into the ATP-binding pocket.
The dihedral angel between the pyrimidine and the pyridine rings
changed from 3.53° to 14.93°, while the other two dihedral angels
changed little. And the position of the two benzene rings in the
docking molecule deviated from the original position towards the
wall of the active pocket. These imply that the conformation of
the kinase inhibitors can be induced to fit the binding to the
ATP-pocket in the Abl kinase.
1. Nagar, B.; Hantschel, O.; Young, M. A.; Scheffzek, K.; Veach, D.; Bornmann, V.;
Clarkson, B.; Superti-Furga, G.; Kuriyan, J. Cell 2003, 112, 859.
2. Zimmermann, J.; Buchdunger, E.; Mett, H.; Meyer, T.; Lydon, N. B. Bioorg. Med.
Chem. Lett. 1997, 7, 187.
3. Schindler, T.; Bornmann, W.; Pellicena, P.; Miller, W. T.; Clarkson, B.; Kuriyan, J.
Science 2000, 289, 1938.
4. Manley, P. W.; Cowan-Jacob, S. W.; Buchdunger, E.; Fabbro, D.; Fendrich, G.;
Furet, P.; Meyer, T.; Zimmermann, J. Eur. J. Cancer 2002, 38, S19.
5. Nagar, B.; Bornmann, W. G.; Pellicena, P.; Schindler, T.; Veach, D. R.; Miller, W.
T.; Clarkson, B.; Kuriyan, J. Cancer Res. 2002, 62, 4236.
6. Blagosklonny, M. V. Leukemia 2002, 16, 570.
7. Weisberg, E.; Manley, P.; Mestan, J.; Cowan-Jacob, S.; Ray, A.; Griffin, J. D. Br. J.
Cancer 2006, 94, 1765.
8. Weisberg, E.; Manley, P. W.; Breitenstein, W.; Bruggen, J.; Cowan-Jacob, S. W.;
Ray, A.; Huntly, B.; Fabbro, D.; Fendrich, G.; Hall-Meyers, E.; Kung, A. L.;
Mestan, J.; Daley, G. Q.; Callahan, L.; Catley, L.; Cavazza, C.; Mohammed, A.;
Neuberg, D.; Wright, R. D.; Gilliland, D. G.; Griffin, J. D. Cancer Cell 2005, 7, 129.
9. Katsoulas, A.; Rachid, Z.; McNamee, J. P.; Williams, C.; Jean-Claude, B. J. Mol.
Cancer Ther. 2008, 7, 1033.
10. Huang, W. S.; Zhu, X. T.; Wang, Y. H.; Azam, M.; Wen, D.; Sundaramoorthi, R.;
Thomas, R. M.; Liu, S.; Banda, G.; Lentini, S. P.; Das, S.; Xu, Q. H.; Keats, J.; Wang,
F.; Wardwell, S.; Ning, Y. Y.; Snodgrass, J. T.; Broudy, M. I.; Russian, K.; Daley, G.
Q.; Iuliucci, J.; Dalgarno, D. C.; Clackson, T.; Sawyer, T. K.; Shakespeare, W. C. J.
Med. Chem. 2009, 52, 4743.
In summary, with the protonation of the nitrogen of N-methyl-
piperazine in Abl kinase inhibitor imatinib as a restraint, the pres-
ent work demonstrates that Surflex-Docking provides a robust and
accurate virtual screening method for rational design of PTK inhib-
itors as potential anticancer agents. The docking-scores for the
newly synthesized imatinib analogs are consistent with the
11. Keseru, G. M. J. Comput. Aided Mol. Des. 2001, 15, 649.
12. Aparna, V.; Rambabu, G.; Panigrahi, S. K.; Sarma, J.; Desiraju, G. R. J. Chem. Inf.
Model. 2005, 45, 725.
13. Wang, R. X.; Lu, Y. P.; Wang, S. M. J. Med. Chem. 2003, 46, 2287.

6968
S. Lü et al. / Bioorg. Med. Chem. Lett. 21 (2011) 6964–6968
14. Juan, A. A. S. J. Mol. Graphics Modell. 2007, 26, 482.
18. Lü, S.; Luo, Q.; Li, X. C.; Wu, J. H.; Liu, J. A.; Xiong, S. X.; Feng, Y. Q.; Wang, F. Y.
Analyst 2010, 135, 2858.
19. Cowan-Jacob, S. W.; Fendrich, G.; Floersheimer, A.; Furet, P.; Liebetanz, J.;
Rummel, G.; Rheinberger, P.; Centeleghe, M.; Fabbro, D.; Manley, P. W. Acta
Crystallogr., Sect. D 2007, 63, 80.
15. Bissantz, C.; Folkers, G.; Rognan, D. J. Med. Chem. 2000, 43, 4759.
16. Liu, Y. F.; Wang, C. L.; Bai, Y. J.; Han, N.; Jiao, J. P.; Qi, X. L. Org Process Res. Dev.
2008, 12, 490.
17. Crystal data for 1: C23H19N5O, M = 381.43, monoclinic, space group P21/n,
a = 8.0283(16) Å, b = 15.129(3) Å, c = 16.088(3) Å, V = 1948.3(7) Å3, Z = 4,
Dc = 1.300 mg/m3, F(000) = 800, crystal dimensions 0.46 Â 0.42 Â 0.36 mm,
20. Zheng, J. H.; Knighton, D. R.; Teneyck, L. F.; Karlsson, R.; Xuong, N. H.; Taylor, S.
S.; Sowadski, J. M. Biochemistry (Mosc) 1993, 32, 2154.
R1 = 0.0528, wR2 = 0.1030 for 3428 reflections with I > 2
largest peak/hole 0.233/À0.238.
r(I), GOF = 1.153,