Design and synthesis of a novel series of N,4-diphenylpyrimidin-2-amine derivatives as potent and selective PI3Kγ inhibitors

Article abstract of DOI:10.1039/c3md00301a

Due to the increasing evidence linking the PI3Kγ pathway to various disease states, PI3Kγ is becoming an important target for cancer treatment. Herein we designed and synthesized a novel series of N,4-diphenylpyrimidin-2-amine derivatives with low CDOCKER-INTERACTION-ENERGY and then evaluated their PI3Kγ in vitro inhibitory activities and in vitro antiproliferation assays against four human cancer cells. Among the compounds we synthesized, compound C8 (IC₅₀ = 65 nM) demonstrated the most potent inhibitory activity against PI3Kγ kinase as well as at the cellular level, compared to the control drug TG100713 (IC₅₀ = 127 nM). Moreover, molecular docking analysis was also performed to determine possible binding modes between PI3Kγ and the target compounds.

Full text of DOI:10.1039/c3md00301a

MedChemComm
View Article Online
View Journal | View Issue
CONCISE ARTICLE
Design and synthesis of a novel series of
N,4-diphenylpyrimidin-2-amine derivatives as
potent and selective PI3Kg inhibitors
Cite this: Med. Chem. Commun., 2014,
5, 219
Hua-Lin Yang,† Fei Fang,† Chang-Po Zhao, Dong-Dong Li, Jing-Ran Li, Jian Sun,
*
Qian-Ru Du and Hai-Liang Zhu
Due to the increasing evidence linking the PI3Kg pathway to various disease states, PI3Kg is becoming an
important target for cancer treatment. Herein we designed and synthesized a novel series of N,4-
diphenylpyrimidin-2-amine derivatives with low CDOCKER_INTERACTION_ENERGY and then evaluated
their PI3Kg in vitro inhibitory activities and in vitro antiproliferation assays against four human cancer
cells. Among the compounds we synthesized, compound C8 (IC50 ¼ 65 nM) demonstrated the most
potent inhibitory activity against PI3Kg kinase as well as at the cellular level, compared to the control
drug TG100713 (IC50 ¼ 127 nM). Moreover, molecular docking analysis was also performed to determine
possible binding modes between PI3Kg and the target compounds.
Received 8th October 2013
Accepted 29th November 2013
DOI: 10.1039/c3md00301a
www.rsc.org/medchemcomm
Phosphoinositide 3-kinases (PI3Ks), the family of evolutionarily viable approach for the treatment of a variety of inammatory
conserved lipid kinases, are categorized into class I, II, and III and cancer diseases.13,14 As different subtypes of PI3Ks, the
based on their subunit structure, regulation and substrate efforts to develop PI3K-targeted therapeutics have led to two
selectivity.1,2 They all could catalyze phosphorylation of the 3- paradigms: namely simultaneous inhibition of all PI3K iso-
hydroxyl position of phosphatidylinositol 4,5-diphosphate forms (pan inhibition) and only one (isozyme selective).9
(PIP2) to generate phosphatidylinositol 3,4,5-triphosphate Thanks to these efforts, there are a considerable number of
(PIP3). PIP3 is known as a signicant second intracellular compounds as anti-tumor agents now in clinic practice (Fig. 1).
messenger in the control of a diverse set of cellular processes, LY294002,15 as the rst PI3K inhibitor, is a pan inhibitor of PI3K
including cell growth, proliferation, differentiation, survival, kinase that binds reversibly to the ATP site and is also a
intracellular trafficking and membrane ruffling.2–6 The class IA commonly referenced compound in PI3K inhibition studies.
(PI3Ka, PI3Kb and PI3Kd), as the most extensively studied of the PIK-75, TGX221 (ref. 7) and TG10071331 are isoform selective
three PI3K classes, is predominately activated by receptor tyro- inhibitors, which exhibit potent, efficacious, antiangiogenic
sine kinases and class IB (PI3Kg) is activated by G-protein activity of PI3Ka, PI3Kb and PI3Kg, respectively.16 Compared
coupled receptors.6,7 There is a regulatory subunit (p85) and an with isoform selective inhibitors, the pan inhibitors may cause
isoform-specic catalytic subunit (p110a, p110b and p110d) to some unexpected toxicities,9 such as platelet aggregation and
build up class IA, whereas PI3Kg, the sole representative of class inhibition of immune responses. Based on these consider-
IB, consists of only one member: a p110g catalytic subunit and a ations, selective inhibition of PI3Ks has been considered as a
p101 regulatory subunit.7–10 Three different subtypes of class IA comparative fashion for cancer treatment.
PI3Ks (PI3Ka, PI3Kb and PI3Kd) are encoded by PIK3CA,
PIK3CB and PIK3CD, respectively.11
Given the linking of the PI3K pathway to various disease
states, the class IA PI3Ks has been studied extensively as targets
In recent years, it has been widely reported that the dereg- for the treatment of cancers and metabolic disorders.13,14
ulation of class I PI3Ks is responsible for cancer, inammation, However, there are only a handful of reports of PI3Kg as the
immune disorders, and cardiovascular diseases.12 Moreover, target for the treatment of inammatory disorders and cancers
PI3Ka and PI3Kb, ubiquitously expressed in mammalian in the previous literature, especially tumors. Recently, some
tissues, are involved in regulating tumor growth and prolifera- papers linking the transforming functions of PI3Kg to disrupt
tion, whereas PI3Kg (mainly present in leukocytes) could be a intercellular adhesion and promote cancer cell invasion have
been published.17–19 For example, PI3Kg mediates Kaposi's
sarcoma-associated herpesviruses GPCR-induced sarcoma-
genesis20 and PI3Kg activated by CXCL12 regulates tumor cell
State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University, Nanjing,
210093, P. R. China. E-mail: zhuhl@nju.edu.cn; Fax: +86-25-8359-2672; Tel: +86-
adhesion and invasion.21 Therefore, based on these convincing
25-8359-2572
pieces of evidence delineating a strong rationale in this
† These authors contributed equally to this work.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 219–225 | 219

View Article Online
MedChemComm
Concise Article
Fig. 1 Some known PI3K kinase inhibitors and design solution of novel N,4-diphenylpyrimidin-2-amine derivatives.
promising therapeutic area, it was decided to pursue PI3Kg as a
Aer virtual screening, twenty-one N,4-diphenylpyrimidin-2-
drug discovery target for cancer.
amine derivatives were synthesized for the study of anti-tumor
Pyrimidines were extensively explored and oen used as activity. A general synthetic approach was developed to prepare
central core structures22 for the inhibitors of a wide range of these 2-aniline-pyrimidines that could be easily obtained by three
kinases, such as Axl kinase,23 phosphoinositide-3-kinase,24 steps in Scheme 1. These candidate compounds with para
broblast growth factor receptor25,26 and Janus tyrosine substituents on the E ring displayed low binding energy compared
kinase.27 Therefore, in order to develop more potent selective with the reference compound (TG100713), indicating that these
PI3Kg inhibitors as anti-tumor agents, we attempted to use the compounds would be potential PI3Kg inhibitors. To further
pyrimidine core as the basic scaffold for designing the novel
PI3Kg inhibitors. As reported in previous studies, the exocyclic
NH along with the pyrimidine core could form the classical bi-
Table 1 Enzyme activities (IC50, nM) of compounds C1–C21 against
dentate hydrogen bond donor–acceptor interaction with the
hinge region of the kinase,28,29 and the pendant phenyl ring
was identied as a suitable region for initial modication of
the molecules.16 Thus, the replacement of the aniline fragment
at the 2-position of the pyrimidine ring would lead to alter-
ation of these targets (Fig. 1). Moreover, pyrimidine, as a bio-
isostere of the quinoxaline ring moiety, shows some similarity C4
in the structure to our referenced compound TG100713 (from
Selleck). Based on this, a series of novel N,4-diphenylpyr-
imidin-2-amine derivatives had been designed as the selective
PI3Kg inhibitors (Table 1) and the medicinal chemistry
program was initiated to explore the SAR. Subsequently, in C10
order to validate whether these small molecules could work on
PI3Kg and to verify the authenticity of the selected study only
in PI3Kg inhibitory activity at the same time, we have per-
formed two groups of molecular dockings between these
designed compounds and three protein–ligand complexes (as C16
for PI3Ka, PDB code: 3HHM; as for PI3Kb, PDB code: 4G11; as
human PI3Kg, PI3Ka, PI3Kb kinases, respectively
Compound
PI3Kg IC50 (nM)
PI3Ka IC50 (nM)
PI3Kb IC50 (nM)
C1
C2
C3
195
167
143
235
371
334
247
65
342
357
313
302
265
231
217
154
250
274
278
386
268
127
615
552
527
738
>1000
880
679
341
894
>1000
792
914
946
591
625
503
836
813
851
953
772
165
729
603
537
839
>1000
>1000
863
373
>1000
977
874
946
961
673
741
573
926
851
893
938
799
215
C5
C6
C7
C8
C9
C11
C12
C13
C14
C15
C17
C18
C19
C20
C21
TG100713
for PI3Kg, PDB code: 1E8W) retrieved from the RCSB Protein
Data Bank (http://www.rcsb.org/pdb/home/home.do), by the
means of the CDOCKER protocol30 in the Accelrys Discovery
Studio 3.5 suite.
220 | Med. Chem. Commun., 2014, 5, 219–225
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
To get more direct insight into the binding mode of
N,4-diphenylpyrimidin-2-amine derivatives and PI3Kg, the
binding model of 4-(4-chlorophenyl)-N-(3,4,5-trimethoxyphenyl)-
pyrimidin-2-amine (compound C8) with the target protein
structure (1E8W) is shown in Fig. 3. Visual inspection of the pose
of compound C8 in the PI3Kg binding site revealed that this
candidate PI3Kg inhibitor was tightly embedded into the ATP
binding pocket. The model also depicted that extensive hydro-
phobic interactions were formed between compound C8 and
residues Ser 806, Lys 808, Pro 810, Ile 831, Asp 836, Ile 879, Glu
880, Asp 950 and Asp 964 of the ATP-binding pocket of PI3Kg
Scheme 1 Synthesis of compounds C1–C21. Reagents and condi-
tions: (i) DMF-DMA/120 ꢀC/16 h. (ii) Ar2/EtOH/70% HNO3/50% cyan-
amide/100 ꢀC/16–18 h. (iii) Ar2/NaOH/2-methoxyethanol/125 ꢀC/22 h.
evaluate the selectivity for PI3Kg, these candidates were also
docked with PI3Ka (PDB ID: 3HHM) and PI3Kb (PDB ID: 4G11),
respectively. For better display and comparison with each other,
these obtained results are plotted as a line–scatter graph presented
in Fig. 2. It was composed of three trend lines: the black solid line
that represented PI3Kg; the blue solid line that represented PI3Ka;
and the red solid line that represented PI3Kb. The general trends
of the corresponding CDOCKER_INTERACTION_ENERGY were
similar against three enzymes, as shown in Fig. 2. For instance,
the top compounds (C9 and C12) and the bottom ones (C3 and
C8) were the same. However, the corresponding CDOCK-
ER_INTERACTION_ENERGY of these compounds binding into
PI3Kgwas obviously lower than that binding to PI3Kaand PI3Kb. It
was concluded that these candidates were promising to synthesize
and evaluate as selective PI3Kg inhibitors.
Fig. 3 The two kinds of binding modes between the active confor-
mation of compound C8, TG100713 (the reference drug) and the
target protein PI3Kg (PDB code: 1E8W) provided by the CDOCKER
protocol (Ligplus and Discovery Studio 3.5, Accelrys, Co. Ltd). (A) 2D
molecular docking model of compound C8 with 1E8W. (B) 2D
Fig. 2 The CDOCKER_INTERACTION_ENERGY (kcal molꢁ1) obtained molecular docking model of TG100713 with 1E8W. (C) The 3D inter-
from the docking study of the targeted compounds by the CDOCKER action map between compound C8 and the PI3Kg protein. (D) The 3D
protocol (Discovery Studio 3.5, Accelrys, Co. Ltd).
interaction map between TG100713 and the PI3Kg protein.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 219–225 | 221

View Article Online
MedChemComm
Concise Article
kinase (Fig. 3A). From Fig. 3C and D, there were only two H- comparison of the para substitution on the F ring demonstrated
bonds formed between the reference compound TG100713 and that a para electron-withdrawing group (compounds C2, 3, 7,
1E8W, whereas ve H-bonds were formed between compound 12, and 13) may have more slightly improved inhibitory activity
C8 and 1E8W, inferring that the candidate was more tightly than a methoxy or methyl group (compounds C1, 6, 9, and 10),
embedded into the ATP binding pocket than TG100713. Three and it showed the most potent inhibitory activity when the para
H-bonds were formed between three methoxyl groups of the F position was substituted by –NO2 (compound C3, IC50 ¼ 143 nM
˚
˚
˚
ring and Lys 807, Lys 808, with their lengths 4 A, 4.3 A and 5.1 A, and compound C13, IC50 ¼ 265 nM). Besides, the nitrogen at
respectively, as shown in Fig. 3C; another one, with its length 4.1 positions 1 and 3 of the pyrimidine ring could enhance the
˚
A, was formed between p-Cl of the E ring and Val 882; the lengths binding potency as PI3Kg inhibitors. Based on the above
˚
4 A of H-bond was formed between N1 of the pyrimidine core results, these compounds in this series deserved further
and Lys 883, which was signicant to bind the ATP-binding investigation.
pocket of PI3Kg kinase for all designed compounds. Based on
These target compounds were also chosen to evaluate in vitro
these considerations, these results could provide a molecular antiproliferation assays against four different human cancer
level foundation for compound C8 as the most potent PI3Kg cells (A549, MCF-7, HepG-2 and HeLa) as control for non-
inhibitor.
specic cytotoxicity. Interestingly, from the obtained results as
Based on the docking results, herein we only focused on the listed in Table 2, compounds C1–C21 almost showed moderate
replacement of substituents at 2-aniline and benzene ring at the potent inhibitory activities against four cancer cells. And the
4-position of the pyrimidine skeleton. The synthesis of series A representative compound C8 exhibited the best inhibition
((E)-3-(dimethylamino)-1-phenylprop-2-en-1-one) started from activity with the IC50 values of 0.09 mM, 0.29 mM, 0.36 mM, and
different substituted acetophenones by dissolving in DMF– 3.15 mM, much better than those of the positive drug TG100713
DMA. Under Ar2, substituted anilines were reacted with 50% in some cell lines (HEPG2 and A549). In accordance with the
cyanamide in EtOH to prepare the corresponding phenyl- PI3Kg kinase assay, it was concluded that compounds C8, C3,
guanidine derivatives B. Subsequently, a mixture of series A, C16, and C4, with the level of IC50 values in the low micromolar
series B, and NaOH in 2-methoxyethanol was heated at reux to range, might be good candidates for further optimization. In
yield the target products C (Scheme 1).
Subsequently, all the synthesized compounds were evaluated
for their inhibitory activities toward PI3Kg, PI3Ka, and PI3Kb,
Table 2 In vitro anticancer activities (IC50, mM) against four human
tumor cell lines
and the encouraging results are depicted as concentrations of
IC50 in Table 1. As expected, most of the synthesized N,4-
diphenylpyrimidin-2-amine derivatives showed moderate to
potent inhibitory activities for these three target protein
IC50 (mM)a
kinases, and the tendency of inhibition was roughly similar to Compound
our energy prediction. Moreover, most of these molecules also
HEPG2b
HeLab
A549b
MCF-7b
C1
13.88
4.50
1.04
0.55
0.31
0.98
6.53
20.89
2.73
3.45
0.81
0.77
showed high selectivity over PI3Ka and PI3Kb, compared with
their IC50 in three target kinases from Table 1. Based on the
C2
C3
obtained result, it was concluded that these synthesized
compounds may have potential for further development as
selective PI3Kg inhibitors. In particular, compound C8 of which
the IC50 value could reach up to 65 nM, was a more potent PI3Kg
inhibitor than the referenced compound TG100713 (IC50 ¼ 127
nM), indicating that compound C8 had a great opportunity as
the new PI3Kg selective inhibitor for cancer therapy.
According to the data shown in Table 1, it could be
concluded that molecules with bulky substituents of the
pyrimidine core could more easily bind to the active domain of
PI3Kg kinase. The inhibitory activity of compounds (C1–C14)
with different para substituents on the E ring increased in the
following order: 4-F < 4-Br < 4-methyl < 4-methoxy, inferring that
those compounds with strong electron-donating groups per-
formed better than those with electron-withdrawing substi-
tutes. In addition, comparing all these potential inhibitors
(especially among C10, C15, and C19), the disubstituents on the
E ring were likely to be superior to monosubstitution inhibitors.
In the light of steric and electronic factors, it could be noted that
compounds with the same substituent on the phenyl E ring
exhibited distinct kinase inhibitory activity due to different
substituents which were introduced into the phenyl F ring. A
C4
C5
C6
C7
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
TG100713
0.19
23.03
91.52
145.01
3.51
7.33
4.17
344.10
127.64
7.95
164.37
35.91
83.95
0.36
63.19
82.66
3.71
6.11
75.99
3.78
>1000
412.45
0.40
0.09
0.29
3.15
74.57
148.05
7.03
47.48
42.98
0.72
71.04
152.71
11.53
36.68
25.70
1.08
209.65
278.17
111.06
4.18
680.73
2.49
3.37
0.5
4.97
1.84
6
0.72
21.01
3.51
28.45
13.15
96.9
99.33
8.08
23.03
7.66
76.39
167.93
1.56
25.96
18.45
32.16
119.38
11.59
0.49
45.64
248.63
192.03
192.45
3.18
0.82
0.15
1.03
a
Antiproliferation activity was measured using the MTT assay. Values
were determined from replicates of 6 wells from three independent
experiments.
Laboratory of Pharmaceutical Biotechnology, Nanjing University;
HepG-2 (hepatocellular liver carcinoma cell line), MCF-7 (breast
cancer), Hela (immortal cell line), and A549 (carcinomic human
alveolar basal epithelial cell).
b
Cancer cells kindly supplied by the State Key
222 | Med. Chem. Commun., 2014, 5, 219–225
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
Table 3 IC50 data for compound C8 against several-related kinases
screened
compounds, 3D-QSAR modeling was performed based on the
docking results. The 3D-QSAR model was built by using the
corresponding pIC50 values which were converted from
the obtained IC50 (nM) values of PI3Kg kinase inhibition and
performed by built-in QSAR soware of DS 3.5 (Discovery Studio
3.5, Accelrys, Co. Ltd). The way of this transformation was
derived from an online calculator developed from an Indian
medicinal chemistry lab (http://www.sanjeevslab.org/tools-
IC50.html). The training and test sets were divided by the
random diverse molecules method of DS 3.5, in which the test
set accounts for 80% while the training set was set to 20%. 17
compounds were selected into training sets and the rest were in
relative text sets. The results are presented in Table 4. During
the development of 3D-QSAR modeling, one of the important
steps was the determination of active conformation and align-
ment of molecules. The efficient solution of this model
depended on a docking study and the reliability of a previous
study about activity data. The alignment conformation of each
molecule was the one with the lowest energy in the docked
results of CDOCKER.
Kinase
IC50 (mM)
Fold selectivity
PI3Kg
EGFR
CDK2
PTK2
JAK2
0.065
2.145
18.782
10.143
21.513
—
33
289
156
331
short, these synthesized compounds played a role in inhibiting
many tumor cell lines.
The kinase selectivity prole of this series was assessed,
using compound C8 as a representative. The obtained results
(Table 3) presented their activity data depicted as IC50 values.
Interestingly, compound C8 showed moderate inhibitory activ-
ities for the majority of these target protein kinases, which
belong to tyrosine and serine/threonine kinase. In addition,
compound C8 performed better as a PI3Kg inhibitor than
against other kinases. These results could provide sufficient
evidence that compound C8 could be selected as a potential
PI3Kg inhibitor and anti-tumor agent for further optimization.
For the sake of further assessing the inhibition of the target
kinase by the synthesized compounds, an apoptosis experiment
of compound C8 was performed in the A549 cell line using
Annexin-V and propium iodide (PI) double staining by ow
cytometry. The uptake of Annexin V-PE notably increased, and
the uptake of normal cells signicantly decreased in a time-
dependent manner. Finally the percentage of apoptotic cells
from 5.47 ꢂ 0.43% to 13.50 ꢂ 0.79% was markedly elevated in a
density-dependent manner at 48 h (Fig. 4). This is consistent
with its good binding affinity to PI3Kg and its potent activity in
the inhibition of cell growth.
The 3D-QSAR model, generated from DS 3.5, dened the
critical regions (steric or electrostatic) affecting the binding
affinity, which was a PLS model built on 400 independent
variables (conventional r2 ¼ 0.94124). The observed and pre-
dicted values and their residual values for the training set and
test set molecules in the 3D-QSAR model are given in Table 4
and their graphical relationship is illustrated in Fig. 5, respec-
tively. The plot of the observed IC50 vs. the predicted results
Table 4 Experimental, predicted inhibitory activity of compounds C1–
C21
PI3Kg kinase inhibition
For further validation of the binding mode and the structural
framework of the structure–activity relationship of these
Predicted
pIC50
Residual
error
b
Compound
Actual pIC50
C1
C2
C3
6.7100
6.7773
6.8447
6.6289
6.4306
6.4763
6.6073
7.1871
6.4660
6.4473
6.5045
6.5200
6.5768
6.6364
6.6635
6.8125
6.6021
6.5622
6.5560
6.4134
6.5719
6.7163
6.7808
6.8345
6.6376
6.6673
6.4712
6.5979
7.1629
6.4677
6.5926
6.4923
6.5557
6.6008
6.6233
6.5826
6.8140
6.5951
6.5371
6.5531
6.4014
6.5675
ꢁ0.0063
ꢁ0.0035
0.0102
ꢁ0.0087
ꢁ0.2367
0.0051
C4
C5a
C6
C7
0.0094
0.0242
C8a
C9
ꢁ0.0217
ꢁ0.1453
0.0122
ꢁ0.0357
ꢁ0.0240
0.0131
C10a
C11
C12
C13
C14
C15a
C16
C17
C18
C19
C20
C21
0.0809
ꢁ0.0015
0.0070
0.0251
0.0029
0.0120
0.0044
a
Compounds were selected as the test sets while the rest were in the
b
Fig. 4 Analysis of apoptosis induced by compound C8 in the A549 cell
line. Data represent the percentage of apoptotic cells. Cells treated
with 1.0, 4.0, and 16.0 mM with compound C8.
training sets. The IC50 values of the compounds against PI3Kg
(Table 1) were converted into pIC50 values by using the online
calculator (http://www.sanjeevslab.org/tools-IC50.html).
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 219–225 | 223

View Article Online
MedChemComm
Concise Article
showed that this model has a good predictive power which polar group in the blue electrostatic potential areas (which
could be used in the prediction of activity for new N,4-diphe- were dominant close to the skeleton). Several key features of
nylpyrimidin-2-amine derivatives as PI3Kg inhibitors.
the 3D-QSAR contour map are predicted to increase PI3Kg
Also the molecules aligned with the iso-surfaces of the 3D- inhibitor affinity: (1) a more positive environment all around
QSAR model coefficients on the electrostatic eld region the para position of ring E and the meta position of ring F
favorable (in blue) or unfavorable (red) and on the energy (electronic study); (2) a more negative environment around
grids corresponding to the favorable (in green) or unfavorable the para position of ring F (electronic study); (3) less bulky p-
(in yellow) steric effects for the PI3Kg affinity are shown in substituent of the group of ring E and ring F (steric study);
Fig. 5B and C, respectively. It was widely acceptable that a (4) more bulky group substituted in the meta position of ring
better inhibitor based on the 3D-QSAR model should have F (steric study).
strong van der Waals attraction in the green areas and a
In summary, a series of N,4-diphenylpyrimidin-2-amine
analogues have been synthesized based on the results of the
virtual screening. These compounds showed modest potency
and selectivity against PI3Kg, with IC50 values up to nanomolar
levels. Among these small molecules, the docking results indi-
cated that compound C8 could bind well inside the active site of
PI3Kg with the lowest CDOCKER_INTERACTION_ENERGY.
Also, compound C8 displayed the most potent anticancer
activity against PI3Kg (IC50 ¼ 65 nM), being comparable with
that of the positive control TG100713 (IC50 ¼ 127 nM). More-
over, the apoptotic cells with agent treatment accounted for
13.50%, as compared to 5.47% of apoptotic cells in the
untreated control. This is consistent with its good binding
affinity to PI3Kg and its potent activity in the inhibition of cell
growth. These results, along with the SAR analysis, were
signicant evidence to demonstrate that compound C8 could be
optimized as a potential selective PI3Kg inhibitor in further
studies.
Acknowledgements
This work was supported by the Jiangsu National Science
Foundation (no. BK2009239) and the Fundamental Research
Fund for the Central Universities (no. 1092020804).
References
1 D. A. Sabbah, N. A. Simms, M. G. Brattain, J. L. Vennerstrom
and H. Z. Zhong, Bioorg. Med. Chem. Lett., 2012, 22, 876.
2 L. Bi, I. Okabe, D. Bernard, A. Wynshaw-Boris and
R. Nussbaum, J. Biol. Chem., 1999, 274, 10963.
3 L. C. Cantley, Science, 2002, 296, 1655–1657.
4 J. W. Leahy, C. A. Buhr, H. W. B. Johnson, B. Gyu, T. Baik,
J. Cannoy, T. P. Forsyth, J. W. Jeong, M. S. Lee, S. Ma,
K. Noson, L. Wang, M. Williams, J. M. Nuss, E. Brooks,
P. Foster, L. Goon, N. Heald, C. Holst, C. Jaeger, S. Lam,
J. Lougheed, L. Nguyen, A. Plonowski, J. Song, T. Stout,
X. Wu, F. Michael, P. Yakes; Yu, W. T. Zhang, P. Lamb and
O. Raeber, J. Med. Chem., 2012, 55, 5467.
5 D. Song, J. S. Yang, S. J. Kim, B. K. Kim, S. K. Park, M. Wonc,
K. Lee, H. M. Kim, K. Y. Choi, K. Lee and G. Han, Bioorg. Med.
Chem. Lett., 2013, 21, 788.
6 F. Giordanetto, A. Walberga, J. Cassel, S. Ghosal,
M. Kossenjans, Z. Q. Yuan, X. P. Wang and L. F. Liang,
Bioorg. Med. Chem. Lett., 2012, 22, 6665.
Fig. 5 (A) The predicted versus experimental pIC50 values for the
inhibition of PI3Kg (PDB: 1E8W). (B) Isosurface of the 3D-QSAR model
coefficients on electrostatic potential grids. The blue triangle mesh
represents positive electrostatic potential and the red area represents
negative electrostatic potential; (C) isosurface of the 3D-QSAR model
on the steric effects. The green triangle mesh representation indicates
that a larger volume is favorable; the yellow triangle mesh indicates the
smaller one.
7 J. Kim, S. Hong and S. Hong, Bioorg. Med. Chem. Lett., 2011,
21, 6977.
224 | Med. Chem. Commun., 2014, 5, 219–225
This journal is © The Royal Society of Chemistry 2014

View Article Online
Concise Article
MedChemComm
´ ´
8 J. D. Kendall, P. D. O'Connor, A. J. Marshall, R. Frederick, 16 K. Bell, M. Sunose, K. Ellard, A. Canseld, J. Taylor,
E. S. Marshall, C. L. Lill, W. J. Lee, S. Kolekar, M. Chao,
A. Malik, S. Yu, C. Chaussade, C. Buchanan,
W. Miller, N. Ramsden, G. Bergamini and G. Neubauer,
Bioorg. Med. Chem. Lett., 2012, 22, 5257.
G. W. Rewcastle, B. C. Baguley, J. U. Flanagan, 17 A. Ghigo, F. Damilano, L. Braccini and E. Hirsch, BioEssays,
S. M. F. Jamieson, W. A. Denny and P. R. Shepherd, Bioorg.
Med. Chem., 2012, 20, 69.
9 F. G. L. Turiso, Y. Shin, M. Brown, M. Cardozo, Y. Chen,
D. Fong, X. L. Hao, X. He, K. Henne, Y. L. Hu,
2010, 32, 185.
18 K. Reif, K. Okkenhaug, T. Sasaki, J. M. Penninger,
B. Vanhaesebroeck and J. G. Cyster, J. Immunol., 2004, 173,
2236.
M. G. Johnson, T. Kohn, J. Lohman, H. J. McBride, 19 M. Laffargue, R. Calvez, P. Finan, A. Trilieff, M. Barbier,
L. R. McGee, J. C. Medina, D. Metz, K. Miner, D. Mohn,
V. Pattaropong, J. Seganish, J. L. Simard, S. Wannberg,
F. Altruda, E. Hirsch and M. P. Wymann, Immunity, 2002,
16, 441.
D. A. Whittington, G. Yu and T. D. Cushing, J. Med. Chem., 20 D. Martin, R. Galisteo, A. A. Molinolo, R. Wetzker, E. Hirsch
2012, 55, 7667. and J. S. Gutkind, Cancer Cell, 2011, 19, 805.
10 C. Brock, M. Schaefer, H. P. Reusch, C. Czupalla, 21 M. Monterrubio, M. Mellado, A. C. Carrera and
M. Michalke, K. Spicher, G. Schultz and B. Nurnberg,
J. Cell Biol., 2003, 160, 89.
J. M. R.-F. Rodriguez-Frade, Biochem. Biophys. Res.
Commun., 2009, 388, 199.
11 M. H. Norman, K. L. Andrews, Y. Y. Bo, S. K. Booker, 22 A. Stella, K. Van Belle, S. De Jonghe, T. Louat, J. Herman,
S. Caenepeel, V. J. Cee, N. D. D Angelo, D. J. Freeman,
B. J. Herberich, F. Hong, C. L. M. Jackson, J. Jiang,
J. Rozenski, M. Waer and P. Herdewijn, Bioorg. Med.
Chem., 2013, 21, 1209.
B. A. Lanman, L. Liu, J. D. McCarter, E. L. Mullady, 23 A. Mollard, S. L. Warner, L. T. Call, M. L. Wade, J. J. Bearss,
N. Nishimura, L. H. Pettus, A. B. Reed, T. S. Miguel,
A. L. Smith, M. M. Stec, S. Tadesse, A. Tasker, D. Aidasani,
A. Verma, S. Sharma, H. Vankayalapati and D. J. Bearss, ACS
Med. Chem. Lett., 2011, 2, 907.
X. Zhu, R. Subramanian, N. A. Tamayo, L. Wang, 24 M. T. Burger, S. Pecchi, A. Wagman, Z. Ni, M. Knapp,
D. A. Whittington, B. Wu, T. Wu, R. P. Wurz, K. Yang,
L. Zalameda, N. Zhang and P. E. Hughes, J. Med. Chem.,
2012, 55, 7796.
T. Hendrickson, G. Atallah, K. Pster, Y. Zhang and
S. Bartulis, ACS Med. Chem. Lett., 2011, 2, 774.
25 V. Guagnano, P. Furet, C. Spanka, V. Bordas, M. Le Douget,
C. Stamm, J. Brueggen, M. R. Jensen, C. Schnell and
H. Schmid, J. Med. Chem., 2011, 54, 7066.
12 M. A. Crackower, G. Y. Oudit, I. Kozieradzki, R. Sarao,
H. Sun, T. Sasaki, E. Hirsch, A. Suzuki, T. Shioi, J. Irie-
Sasaki, R. Sah, H. M. Cheng, V. O. Rybin, G. Lembo, 26 D. D. Li, F. Fang, J. R. Li, Q. R. Du, J. Sun, H. B. Hai-Bin Gong
L. Fratta, A. J. Oliveira-dos-Santos, J. L. Benovic, and H. L. Zhu, Bioorg. Med. Chem. Lett., 2012, 22, 5870.
C. R. Kahn, S. Izumo, S. F. Steinberg, M. P. Wymann, 27 J. J. Chen, K. D. Thakur, M. P. Clark, S. K. Laughlin,
P. H. Backx and J. M. Penninger, Cell, 2002, 110, 737.
13 T. B. Lanni Jr, K. L. Greene, C. N. Kolz, K. S. Para, M. Visnick,
J. L. Mobley, D. T. Dudley, T. J. Baginski and M. B. Liimatta,
Bioorg. Med. Chem. Lett., 2007, 17, 756.
K. M. George, R. G. Bookland, J. R. Davis, E. J. Cabrera,
V. Easwaran and B. De, Bioorg. Med. Chem. Lett., 2006, 16,
5633.
28 J.-L. L. Liao, J. Med. Chem., 2007, 50, 409.
¨
14 T. Sasaki, J. Irie-Sasaki, R. G. Jones, A. J. Oliveria-dos-Santos, 29 A. C. Backes, B. Zech, B. Felber, B. Klebl and G. Muller,
W. L. Stanford, B. Bolon, A. Wakeham, A. Itie, D. Bouchard, Expert Opin. Drug Discovery, 2008, 3, 1427.
I. Kozieradzki, N. Joza, T. W. Mak, P. S. Ohashi, A. Suzuki 30 G. Wu, D. H. Robertson, C. R. Brooks and M. Vieth,
and J. M. Penniger, Science, 2000, 287, 1040. J. Comput. Chem., 2003, 24, 1549.
15 C. J. Vlahos, W. F. Matter, K. Y. Hui and R. F. Brown, J. Biol. 31 J. Doukas, W. Wrasidlo and G. Noronha, Proc. Natl. Acad. Sci.
Chem., 1994, 269, 5241.
U. S. A., 2006, 52, 19866.
This journal is © The Royal Society of Chemistry 2014
Med. Chem. Commun., 2014, 5, 219–225 | 225