Development of highly potent and selective diaminothiazole inhibitors of cyclin-dependent kinases

Article abstract of DOI:10.1021/jm301234k

Cyclin-dependent kinases (CDKs) are serine/threonine protein kinases that act as key regulatory elements in cell cycle progression. We describe the development of highly potent diaminothiazole inhibitors of CDK2 (IC₅₀ = 0.0009-0.0015 μM) from a

Full text of DOI:10.1021/jm301234k

Article
pubs.acs.org/jmc
Development of Highly Potent and Selective Diaminothiazole
Inhibitors of Cyclin-Dependent Kinases
Ernst Schonbrunn,*,† Stephane Betzi,† Riazul Alam,† Mathew P. Martin,† Andreas Becker,† Huijong Han,†
Rawle Francis,‡ Ramappa Chakrasali,‡ Sudhakar Jakkaraj,‡ Aslamuzzaman Kazi,† Said M. Sebti,†
Christopher L. Cubitt,§ Anthony W. Gebhard,∥ Lori A. Hazlehurst,∥ Joseph S. Tash,⊥
and Gunda I. Georg*,‡
†Drug Discovery Department, H. Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive, Tampa, Florida 33612,
United States
‡Department of Medicinal Chemistry, Institute for Therapeutics Discovery and Development, University of Minnesota, 717 Delaware
Street SE, Minneapolis, Minnesota 55414, United States
§Translational Research Laboratory, H. Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive, Tampa, Florida
33612, United States
∥Molecular Oncology Department, H. Lee Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive, Tampa, Florida
33612, United States
⊥Department of Molecular and Integrative Physiology, University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City,
Kansas 66160, United States
S
* Supporting Information
ABSTRACT: Cyclin-dependent kinases (CDKs) are serine/
threonine protein kinases that act as key regulatory elements in
cell cycle progression. We describe the development of highly
potent diaminothiazole inhibitors of CDK2 (IC50 = 0.0009−0.0015
μM) from a single hit compound with weak inhibitory activity
(IC50 = 15 μM), discovered by high-throughput screening.
Structure-based design was performed using 35 cocrystal structures
of CDK2 liganded with distinct analogues of the parent compound.
The profiling of compound 51 against a panel of 339 kinases
revealed high selectivity for CDKs, with preference for CDK2 and
CDK5 over CDK9, CDK1, CDK4, and CDK6. Compound 51
inhibited the proliferation of 13 out of 15 cancer cell lines with IC50
values between 0.27 and 6.9 μM, which correlated with the
complete suppression of retinoblastoma phosphorylation and the onset of apoptosis. Combined, the results demonstrate the
potential of this new inhibitors series for further development into CDK-specific chemical probes or therapeutics.
responsible for mitosis.6,7 CDK-specific inhibitors induce
INTRODUCTION
■
apoptosis by repressing transcription and/or by perturbing
Protein kinases have been implicated in a myriad of human
the cell cycle.8 As alterations in checkpoint regulation can lead
diseases, including various cancers and neurodegenerative
to aberrant cell division, CDK2 represents an attractive target
disorders. Cyclin-dependent kinases (CDKs) are a family of
for therapeutics designed to arrest or recover control of the cell
serine/threonine kinases that are involved in cell cycle
cycle.9,10 Additionally, CDK2 is essential for completion of
progression and transcription. Deregulation of CDKs has
prophase I during meiotic cell division in male and female germ
been associated with a number of medical conditions, and they
cells, and CDK2−/− knockout mice are viable but sterile.11−13
have therefore become validated and important targets in drug
discovery.1,2 The functionality of CDKs is dependent on
Therefore, CDK2 also provides promise as a target for the
development of nonhormonal male contraceptive agents that
do not exhibit the significant side effects associated with
hormone-based agents currently available for female contra-
ception and under development for male contraception.14−16
specific interactions with regulatory partner proteins, the
cyclins.3−5 Cell-cycle progression depends on the activity of
CDK1, CDK2, CDK4, and CDK6. CDK4/6 in complex with
cyclin D1, D2, or D3 and CDK2 in complex with cyclin E
promote S-phase entry by phosphorylating and inactivating the
retinoblastoma (Rb) protein. CDK2-cyclin A and CDK1-cyclin
A propel cells through the S-phase, and CDK1-cyclin B is
Received: August 27, 2012
Published: April 19, 2013
© 2013 American Chemical Society
3768
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Several CDK inhibitors have been in clinical development
since the 1990s. These first-generation inhibitors, namely
flavopiridol, (R)-roscovitine, SNS-032,17 and PHA-793887,18
were discontinued during phase II or phase III trials due to
unfavorable pharmacological properties and low specificity
resulting in undesirable off-target interactions. Recently, two
new-generation CDK inhibitors, dinaciclib (Merck,
SCH727965) and palbociclib (Pfizer, PD-0332991), advanced
to phase III trials for refractory chronic lymphocytic leukemia
and advanced estrogen receptor (ER)-positive breast cancer,
respectively.19,20 Dinaciclib is a selective inhibitor of CDK1,
CDK2, CDK5, and CDK9,21,22 while palbociclib is a selective
inhibitor of CDK4 and CDK6.23
In an attempt to discover new CDK2 inhibitors with
promising scaffolds for hit-to-lead development, we screened
103000 compounds from commercial libraries against the
human CDK2-cyclin A2 complex. Hit compound 2-(allylami-
no)-4-aminothiazol-5-yl)(phenyl)methanone (1) was selected
for thorough structure−activity relationship (SAR) analysis
using medicinal chemistry and protein crystallography, yielding
compounds with high potency and selectivity against CDK2
and CDK5 over CDK9, CDK1, CDK4, and CDK6. Compound
51 inhibited the proliferation/survival of five cancer cell lines
with IC50 values ≤0.5 μM, suppressed CDK-specific phosphor-
ylation of the Rb protein, and induced activation of caspase-3.
The comprehensiveness of this experimental structure−activity
data, including 35 cocrystal structures of CDK2 with analogues
of the parent compound, provides valuable guidance for the
refinement of algorithms used in molecular docking simulations
and structure-based drug design.
Figure 1. Interaction of HTS hit compound 1 with CDK2. (a) The
cocrystal structure of CDK2 with hit 1 revealed a typical type I
mechanism of action by binding through the hinge-region of the ATP
site. The enzyme is shown in gray, with the hinge region (residues 81−
83) highlighted in orange, gatekeeper residue Phe80 in red, and DFG
motif in cyan. The exploded view shows the hydrogen bonding
interactions between the inhibitor (yellow) and enzyme residues
including Asp145 of the DFG motif. Water molecules are shown as
cyan spheres. The 2Fo − Fc electron density map at 1.85 Å resolution
is shown as blue mesh, contoured at 1σ around the inhibitor. The Fo −
Fc omit map is shown in the Supporting Information (Figure S5). (b)
Schematic representation of hydrogen bonding (dotted lines) and
hydrophobic interactions (green). PDB: 3QQK.
based assay was employed for inhibitors with submicromolar
activities, which allowed accurate IC50 determinations up to
∼0.02 μM (Figure S1 of Supporting Information). Activities of
the most potent compounds were determined using a highly
sensitive P33-radiolabeled activity assay.30 The pan-kinase
inhibitor staurosporine (IC50 < 1 nM against CDK2−cyclin
A) served as a control. Crystal structures of 35 CDK2−
inhibitor complexes were determined between 1.4 and 2.0 Å
resolution (Table S1 of Supporting Information). Scheme 1
shows the general method for synthesis of the targeted
compounds. An isothiocyanate is reacted with cyanamide in
the presence of tert-butoxide in THF for 15 min to form the
intermediate addition product, which is directly reacted with a
bromoacetyl derivative to form the desired (2,4-diaminothiazol-
5-yl)methanone derivative.
Modifications to R2 resulted in analogues 2−15 (Table 1a),
of which the meta-pyridine analogue (3) was 5-fold more active
(IC50 = 3.1 μM) than the parent compound. This is likely due
to the weak hydrogen bonding potential of the pyridine
nitrogen with the ε-amino group of Lys33 (Figure 3a), which
establishes a salt bridge with Glu51 of the PSTAIRE helix in the
activated CDK2−cyclin A2 complex.31 Remarkably, substitu-
tion with ortho- or para-pyridine (6, 10) resulted in loss of
activity, and cocrystal structures suggest that Lys33 is no longer
attracted to the pyridine ring (Figure 3b, 3c). Moreover,
binding of 10 induced a large conformational change in the P-
loop (residues Gly11−Val18), allowing a water molecule to
enter and to interact with the ring nitrogen, while 6 adopts an
unfavorable conformation upon binding to the active site.
Modifications to R1 yielded analogues 16−25 (Table 1b), of
which the phenyl (16) and cyclohexyl (17) substituents yielded
more than a 15-fold improvement in activity over the parent
compound. Combination of a meta-pyridine ring at R2 (as in
compound 3) with a phenyl group at R1 (as in compound 16)
further improved inhibitory activity (30, IC50 = 0.65 μM).
However, with a phenyl group at R1, substituents at R2 other
than meta-pyridine were not tolerated as well or were
detrimental to compound activity (Table 1c). Bulky R1
RESULTS AND DISCUSSION
■
HTS Hit Identification and Characterization. We
discovered compound 1 (2-(allylamino)-4-aminothiazol-5-yl-
(phenyl)methanone) by high-throughput screening (HTS) as
an inhibitor of the human CDK2-cyclin A2 complex with an
IC50 value of 15 μM. The cocrystal structure with CDK2 at 1.85
Å resolution (Figure 1, Table S1 in Supporting Information)
confirmed that 1 belongs to the type I class of kinase inhibitors,
binding to the ATP site through hydrogen bonding interactions
between the thiazolamine moiety and the hinge region
(Glu81−Leu83). A water molecule bridges the carbonyl oxygen
of 1 and the backbone amide nitrogen of Asp145, while the
phenyl and allyl moieties establish van der Waals carbon−
carbon interactions (3.2 Å < d < 3.8 Å) with neighboring
residues. Compounds with aminothiazole hinge-binding
scaffolds have been previously described as inhibitors of
checkpoint kinase I,24 p38α/MAPK,25 CDK2,26 and
CDK5.27−29 However, the diaminothiazole CDK-inhibitors
developed herein are unique with respect to structure, potency,
and selectivity.
Structure−Activity Relationship (SAR) Studies. Ana-
logues of 1 were designed such that the hinge-binding
functionality of the aminothiazole core remained unchanged,
while the flanking allyl (R1) and phenyl (R2) moieties were
systematically modified (Figure 2, Table 1). A total of 95
analogues were synthesized and evaluated using three different
enzymatic assays based on their inhibitory activities (see
Experimental Section). The pyruvate kinase/lactate dehydro-
genase (PK−LDH) spectrophotometric assay was employed
for HTS and initial SAR studies. As the CDK2−cyclin A
concentration in the PK−LDH assay limits the determination
of IC50 values up to ∼0.1 μM, a more sensitive fluorescence-
3769
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Figure 2. SAR milestones in the optimization of hit compound 1.
functional groups extend into a region commonly referred to as
the front specificity pocket, consisting of residues Gln85,
Asp86, and Lys89. To interrogate this pocket, we next
synthesized a series of analogues with meta-pyridine at R2
and diverse substituents at R1 (Table 1d). Introduction of a
sulfonamide moiety at the para position of the phenyl ring
yielded significant improvement over the parent compound
(42, IC50= 0.02 μM). Using the fluorescence-based assay, the
CDK2−cyclin A concentration of 9 nM became limiting for
highly potent inhibitors such as 42. Even the pan-kinase
inhibitor staurosporine, which is a known subnanomolar
inhibitor of CDK2−cyclin A, displayed an IC50 value of 20
nM in the fluorescence-based assay. Therefore, a highly
sensitive P33-radiolabel assay was employed, in which 42
inhibited the CDK2−cyclin A complex with an IC50 value of 0.9
nM (Figure S2 of Supporting Information), rendering this
compound among the most potent CDK2 inhibitors reported
to date.17 This 15000-fold increase in activity over the parent
compound is primarily attributed to an elaborate hydrogen
bonding network between the sulfonamide group and residues
His84, Gln85, Asp86, and Lys89 (Figure 4a). Of the other R1
substituents tested, the para-benzamide also conferred a
considerable increase in activity (43, IC50= 0.22 μM).
Introduction of a sulfonamide group in the meta position
resulted in a loss of activity (49, IC50 = 7.8 μM), and
modifications to 49 did not restore inhibitory activity (78−83,
Table S2 of Supporting Information).
Several analogues of 42 were prepared with modifications to
R2, of which compounds 51−55 were the most potent
inhibitors, with IC50 values between 0.02 and 0.07 μM
(Table 1e). Using the P33-radiolabeled assay, 51 inhibited
CDK2−cyclin A2 with an IC50 value of 1.5 nM (Figure S2 of
Supporting Information). Intriguingly, the ortho-nitro group of
51 is nested in a hydrophobic pocket composed of residues
Val18, Ala31, Lys33, and gatekeeper Phe80 (Figure 4b). The
unusual proximity and orientation of the nitro group toward the
Phe80 side chain indicate attractive interactions through van
der Waals forces or the establishment of “nonclassical”
hydrogen bonds between the phenyl ring and the nitro
group, a phenomenon previously described for the interaction
of nitro-substituted inhibitors of aldolase reductase with an
active site tyrosine ring.32 This notion is supported by the
observation that other substitutions in the ortho position, such
as methyl (62) or methylester (68), were detrimental to
inhibitory activity. Although 52−55 displayed similar activities,
the binding interactions of the respective R2 substituents are
markedly different (Figure 5). The meta-substituents of 52
(fluorine), 54 (nitro), and 56 (methoxy) share the same
binding pocket, composed of residues Gly13, Val18, Lys33, and
Asp145, whereas the meta-amino group of 55 causes the ring to
rotate by ∼180° to form hydrogen bonds with Gln131 and
Asn132. Other meta-substituents, such as trifluoromethyl (58),
hydroxy (59), or chlorine (65), significantly decreased
inhibitory activity. The para-sulfonamide moiety of 53
establishes hydrogen bonding interactions with Asn 132 and
Asp145, but other para substituents, such as nitro (63), amino
(64), methoxy (65), or bulky functional groups, resulted in loss
of activity. Combined, these findings indicate that anchoring
the inhibitor through a sulfonamide moiety in R1 allows diverse
R2 substituents to efficiently interact with residues around
Asp145 and the P-loop. Notably, introduction of aniline
substituents in R1 to mimic the triazole carbothioamide
structure of the highly potent pan-kinase inhibitor CDK1/2
inhibitor III resulted in complete loss of activity (Table S6 of
Supporting Information).
Kinase Profiling. To assess kinase selectivity, a representa-
tive set of inhibitors was initially profiled against a panel of 48
different kinases (Figure S3 of Supporting Information).
GSK3β, a well-known cross-reactive target of CDK2 inhib-
itors,33,34 was the only other kinase significantly affected by
these compounds. Next, compound 51 was thoroughly
characterized using the P33-radiolabeled assay against a panel
of different CDK−cyclin complexes together with GSK3β
(Table 2, Figure S4 of Supporting Information). The highest
inhibitory activity was seen against CDK2 and CDK5 (IC50
values between 1.1 and 1.8 nM), followed by CDK1, CDK4,
and CDK6 (IC50 values between 4.0 and 7.6 nM), CDK9 and
CDK3 (IC50 values of 13 and 38 nM), and CDK7 (IC50 > 1
3770
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
b
Table 1. CDK2−Cyclin A2 Inhibitory Activities of Analogues of Hit Compound 1
3771
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Table 1. continued
a
b
Value determined by P33-radiolabeled assay. Values represent the mean of duplicate or triplicate dose−response measurements with a standard
deviation of <10%.
μM). Notably, 51 was 90-fold less active against GSK3β.
Comparison of 51 and staurosporine showed similar activities
against most of the enzymes tested, with the exception of
CDK1, CDK3, and GSK3β, which were significantly more
tolerant toward 51. Finally, compound 51 was tested against a
panel of 339 kinases by Reaction Biology Corporation at a
3772
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
lines, with IC50 values between 4 and 20 μM (Table 3).
Compounds 51, 52, and 57 showed higher activity against
spermatocyte-derived cells (GC2) over spermatogonial-derived
cells (GC1) cells and a Sertoli cell line (TM-4).
Scheme 1
Next we determined the cellular activity of compounds 51,
52, 53, 55, and 69 against a panel of 14 different cancer cell
lines using a resazurin/resorufin-based fluorescence assay
(Table 4). The potent cell active CDK2 inhibitor dinaciclib
and the antineoplastic drug doxorubicin served as positive
controls. Compound 51 inhibited the proliferation/survival of
four cell lines (A-673, SK-Br-3, SK-ES-1, and T-47D) with IC50
values between 0.27 and 0.56 μM, eight cell lines (SK-UT-1,
MDA-MB-231, H929, U266, MDA-MB-436, MNNG HOS,
MCF-7, and SW-872) were moderately inhibited with IC50
values between 1.4 and 6.9 μM, and two cell lines (RPMI-8226
and BT-549) were largely tolerant with IC50 values >10 μM.
Compound 52 displayed a similar activity profile with IC50
values between 0.45 and 6.8 μM for 13 cell lines but was
significantly less active in those cell lines that were prominently
inhibited by 51. The weak CDK2 inhibitor 69 was largely
ineffective, and the potent CDK2 inhibitors 53 and 55 were
also ineffective, presumably due to poor cell permeability. As
expected, dinaciclib was extremely effective against most cell
lines with IC50 values between <0.005 and 0.01 μM. Notably,
the BT-549 cell line was tolerant to all CDK2 inhibitors
single concentration of 0.1 μM (≈ 100 × IC50) (Figure 6, Table
S7 of Supporting Information). The data confirmed that 51 is
highly selective against CDKs within the human kinome
(Figure 6b). Compound 51 displayed highest inhibitory activity
against CDK5 and CDK2 (95−96%), followed by CDK9 and
CDK1 (90−93%) and CDK4 and CDK6 (87 and 85%).
GSK3α and ERK7 were inhibited to 73−75%, followed by
TAO1, GSK3β, and CLK2 (60−66%). The vast majority of
kinases (282) were not affected by compound 51.
Cellular Activity. Compounds with the highest activity
against the CDK2−cyclin A complex were evaluated for
inhibitory potential against a total of 15 cancer cell lines and
three testis cell lines. Initial studies were performed on the
proliferation/survival of MDA-MB-468 breast cancer cells
treated with increasing concentrations of inhibitors for 72 h
and determined by MTT assay. Compound 51 significantly
inhibited cell viability in a concentration-dependent manner
with an IC50 value of 0.3 μM (Table 3, Figure S6 of Supporting
Information). In contrast, 42 was much less potent (IC50 = 2.8
μM), suggesting that the R2 pyridine moiety negatively impacts
cellular uptake of the inhibitor. This is supported by the
observation that the pyridine-containing 45 was equally less
effective (IC50 = 3.1 μM) as compared to 51, whereas the
fluorophenyl moiety of 52 conferred higher cellular activity
(IC50 = 0.85 μM). As expected from its poor in vitro activity
against CDK2, compound 69 displayed the lowest ability to
inhibit MDA-MB-468 cells (IC50 = 5.5 μM). To assess the
ability of the compounds to selectively inhibit the division of
testis cells important for spermatogenesis, MTT assays were
carried out using cell lines that represent important
spermatogenic stages where cell division is critical for
maintenance of spermatogenesis: spermatogonia (GC-1)35,36
and spermatocytes (GC-2).36,37 The results were compared to
TM-4 cells,38,39 which represent Sertoli cells that normally stop
dividing by the time spermatogenesis has commenced. Most
compounds showed moderate to poor activity against these cell
including dinaciclib but was sensitive to doxorubicin (IC50
=
0.11 μM).
Mechanism of Action of Compound 51. To evaluate if
the antiproliferative activity of 51 is caused by inhibition of
cellular CDK2, we selected inhibition of Rb phosphorylation at
the CDK-specific sites Ser807/811 and activation of caspase-3
at the onset of apoptosis as mechanism-based markers for
CDK-specific inhibitors.8,22 The most sensitive cell line (A-
673) and a moderately inhibited H929 cell line were treated for
24 h with compound 51 or dinaciclib and were evaluated by
immunoblotting and caspase-3 activation (Figure 7). Com-
pound 51 suppressed phosphorylation of the Rb protein above
0.25 μM in A-673 cells (Figure 7a), which is in accord with the
IC50 value of 0.27 μM in cell viability assays. The activity of
caspase-3 was increased by 196% with a half-maximal effective
concentration (EC50) of 0.4 μM. Dinaciclib completely
suppressed phosphorylation of Rb Ser807/811 at 0.05 μM
Figure 3. Influence of R2 pyridine substituents on binding interactions with CDK2. Crystal structures of CDK2 liganded with the pyridine analogues
3 (a), 6 (b), and 10 (c). meta-Pyridine (3) attracts the ε-amino group of Lys33, whereas para-pyridine (10) induces a large conformational change in
the P-loop, allowing a water molecule to interact with the ring nitrogen. The ortho-pyridine (6) adopts a conformation unfavorable for interaction
with CDK2. Figure 3d shows the differences in binding interactions upon superposition of the crystal structures. The enzymes are shown in gray
with the hinge region (residues 81−83) colored in orange, the gatekeeper Phe80 residue in red, and Asp145 of the DFG motif in cyan. The
hydrogen bonding interactions between the inhibitors (yellow) and enzyme residues are shown as black dotted lines. The key water molecules are
shown as cyan spheres, with the 2Fo − Fc electron density map shown as a blue mesh around the inhibitor, and a green mesh around the water both
contoured at 1σ. PDB entries 3QTQ, 3S0O, and 3R8Z.
3773
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Figure 4. Interaction of sulfonamide analogues 42 and 51 with CDK2. Crystal structures of the most potent inhibitors 42 (a) and 51 (b). Addition
of a para-phenyl sulfonamide moiety at R1 substantially increased the inhibitory activity through establishment of a hydrogen bonding network with
residues of the front specificity pocket (top exploded view). Addition of an ortho-nitrophenyl group at R1 of 51 retained the high in vitro activity and
substantially increased cellular activity. The exploded view (bottom) shows the unusual proximity of the nitro group to neighboring hydrophobic
residues. The 2Fo − Fc electron density is displayed as blue mesh, contoured at 1σ around the compound. Hydrogen bonding and potential π-
bonding interactions are shown as black and orange dotted lines, respectively. Fo − Fc electron density maps from refinements omitting the inhibitors
are shown in the Supporting Information (Figure S5). Schematic representations of the respective CDK2−inhibitor interactions are shown in the
right panel. Black dotted lines indicate hydrogen bonding interactions, while van der Waals (hydrophobic) interactions are shown in green. PDB
entries 3QU0 and 3QXP.
Table 2. Inhibitory Activity of Compound 51 against Various
CDK−Cyclin Complexes and GSK3β Using the P33-
Radiolabeled Assay
kinase
compound 51 IC50 (nM) staurosporine IC50 (nM)
CDK1/cyclin A
CDK2/cyclin A
CDK2/Cyclin A1
CDK3/cyclin E
CDK4/cyclin D1
CDK5/p25
7.6 0.7
1.1 0.1
1.8 0.2
38 5.3
4.0 0.2
1.5 0.2
6.6 0.3
>1000
1.7 0.1
0.4 0.03
0.5 0.04
4.0 0.6
6.6 0.2
1.2 0.1
2.9 0.1
910 115
13 0.5
15 1.1
CDK6/cyclin D1
CDK7/cyclin H
CDK9/cyclin K
GSK3b
13 1.0
107 20
and increased caspase-3 activity by 280% with an EC50 value of
0.011 μM. Similar results were obtained for H929 cells (Figure
7b), which displayed inhibition of Rb phosphorylation above
1.5 μM (IC50 = 1.8 μM in viability assays). Upon treatment
with 1.5 μM of compound 51, 26−59% of the cells were
apoptotic. Dinaciclib suppressed the phosphorylation of Rb at
0.005 μM with 63−84% apoptotic cells. The correlation
between the suppression of Rb-phosphorylation and onset of
apoptosis suggests that the mechanism of action of compound
51 in these two cancer cell lines may proceed through
inhibition of CDKs critical for the phosphorylation of Rb
protein at Ser807/811.
Figure 5. Sulfonamide inhibitors tolerate diverse R2 substituents.
Crystal structures of CDK2 liganded with 52 (a), 53 (b), 54 (c), and
55 (d), other highly potent inhibitors of the para-sulfonamide series of
Table 1e. The respective 2Fo − Fc density maps, contoured at 1σ, are
shown as blue mesh. Black, green, and orange dotted lines indicate
potential hydrogen bonding, hydrophobic, and π-interactions,
respectively. PDB entries 3QTZ, 3QTU, 3QTX, and 3RPV.
3774
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Figure 6. Profiling of compound 51 against a panel of 339 kinases. (a) Quantitative distribution of kinases binned according to the inhibition by
compound 51. (b) Distribution of inhibited kinases within the human kinome. The color code for inhibition is indicated. Residual enzymatic activity
was determined in single dose duplicate at a compound concentration of 0.1 μM. The ATP concentration was 10 μM. Staurosporine served as a
positive control. The experiment was performed by Reaction Biology Corp. using a P33-radiolabel assay. The data sets are shown in Table S7 of
Supporting Information.
refinement of algorithms used in molecular docking simulations
and structure-based drug design.
Table 3. Cellular Activity of Inhibitors against MDA-MB-468
Breast Cancer and Testis Cell Lines As Determined by MTT
Assay
Extensive kinase profiling indicates that compound 51 is
highly selective for CDKs, with preference for CDK2 and
CDK5 over CDK9, CDK1, CDK4, and CDK6. Among the
inhibitors tested, only compound 51 significantly inhibited the
proliferation/survival of cancer cells with IC50 values ≤0.5 μM
in five of the 15 cancer cell lines tested. Mechanism-based
marker analysis established that the cellular activity of 51
correlates with the suppression of Rb-phosphorylation and the
onset of apoptosis. The observation that the reportedly Rb-
deficient cell line BT-54940 is insensitive to compound 51 and
dinaciclib (Table 4) seems to corroborate the notion that
suppression of Rb-phosphorylation is the primary mode of
action of 51. However, the MDA-MB-468 cell line is also Rb-
deficient but sensitive to 51 (Table 3), indicating the
involvement of other essential CDK substrates or possible
off-target interactions. A more thorough cell-based study is
planned to further address the mechanism of action of this
inhibitor series using multiple marker analysis.8 Although
compound 51 and dinaciclib share similar CDK inhibitory and
selectivity profiles in enzymatic activity assays, dinaciclib is an
extremely potent antiproliferative compound in cancer cell
lines.8,22 The modest activity of our inhibitors in cell-based
assays suggests poor cell permeability, and a new series of
compounds is planned with suitable solubilizing groups in
solvent exposed regions to facilitate cellular uptake. Combined,
these results demonstrate the potential of these inhibitors for
further development into CDK-specific chemical probes and
therapeutics.
cell line
MDA-MB-468
a
TM-4
GC-2
GC-1
a
a
a
compd
IC50 (μM)
IC50 (μM)
IC50 (μM)
IC50 (μM)
16
17
30
31
32
41
42
45
51
52
54
56
57
69
nd
nd
12 1.1
9.3 1.0
5.5 1.1
14 1.1
5.1 1.0
>60
11 1.0
8.2 1.0
9.5 1.1
17 1.1
4.1 1.1
>60
11 1.0
9.2 1.1
5.9 1.0
16 1.0
3.6 1.0
>60
nd
nd
nd
nd
2.8 0.3
3.1 0.2
0.3 0.1
0.8 0.1
nd
13 4.9
nd
>30
20 1.7
nd
nd
16 1.1
4.8 1.0
4.0 0.6
6.4 1.5
9.1 1.1
nd
9.0 1.2
3.9 1.0
20 5.4
11 1.2
6.1 1.0
nd
20 1.7
6.4 1.1
22 6.7
9.1 1.1
17 1.1
nd
nd
nd
5.5 0.5
a
Average of dose−response experiments performed in triplicate using
9 inhibitor concentrations between 0.02 and 100 μM (nd = not
determined).
CONCLUSIONS
■
Starting from a single HTS hit compound with weak inhibitory
activity, we employed protein crystallography as the main tool
to design novel, highly potent, and selective inhibitors of CDK2
as potential leads for further development as chemical probes
or therapeutics. Throughout the course of this project, 35 high-
resolution cocrystal structures of CDK2 were determined for
distinct analogues of the hit compound, with inhibitory
activities ranging from excellent (IC50 = 0.001 μM) to poor
(IC50 > 25 μM). The comprehensiveness of crystallographic
structure−activity relationship data for a single chemical
scaffold against a single target protein is unprecedented to
our knowledge and may provide valuable guidance for the
EXPERIMENTAL SECTION
■
Reagents and Compounds. Reagents and compounds for
biochemical and crystallization experiments were purchased from
Sigma-Aldrich (St. Louis, MO) and Hampton Research (Aliso Viejo,
CA) unless otherwise indicated. Compounds 1, 5, 8, 11−14, and 19
were purchased from Sigma-Aldrich, Ryan Scientific (Mt. Pleasant,
SC), ChemDiv (San Diego, CA), and Labotest OHG (Niederschoena,
Germany). Staurosporine was purchased from Tocris Bioscience
3775
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
Table 4. Cellular Activity of Inhibitors against a Panel of 14 Cancer Cell Lines As Determined by a Resazurin/Resorufin-Based
Fluorescent Viability Assay
compd
51
52
53
55
69
dinaciclib
a
doxorubicin
a
a
a
a
a
a
cell line
IC50 (μM)
IC50 (μM)
IC50 (μM)
IC50 (μM)
IC50 (μM)
IC50 (μM)
IC50 (μM)
A-673 rhabdomyosarcoma
SK-Br-3 adenocarcinoma
SK-ES-1 osteosarcoma
T-47D ductal carcinoma
SK-UT-1 leiomyosarcoma
MDA-MB-231 adenocarcinoma
H929 myeloma
0.27
0.49
0.56
0.5
0.45
1.2
>10
>10
>10
8.7
8.8
4.1
3.7
6.1
<0.005
<0.005
<0.005
0.01
0.17
0.09
0.042
0.1
1.1
9.8
4.8
0.98
1.6
5.8
4.5
1.4
>10
>10
>10
7.2
>10
>10
6.8
10
0.006
0.076
0.17
0.065
0.67
0.1
1.5
1.2
>10
4.3
<0.005
<0.005
0.006
1.6
1
U266 myeloma
1.8
1.5
5.5
7.9
MDA-MB-436 adenocarcinoma
MNNG HOS osteosarcoma
MCF-7 adenocarcinoma
SW-872 liposarcoma
2
1.3
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
5.3
<0.005
0.0055
0.02
3
2.2
9.7
0.59
0.24
0.14
0.27
0.11
3.1
1.3
5.9
6.9
3.1
>10
>10
>10
0.0095
0.009
RPMI-8226 myeloma
>10
>10
6.8
BT-549 ductal carcinoma
>10
>10
a
Average of three data sets, each consisting of dose−response experiments performed in quadruplicate using nine inhibitor concentrations between
0.004 μM and 12.5 μM.
Figure 7. Mechanism of action of compound 51. (a) Asynchronously growing A-673 sarcoma cells were treated with increasing concentrations of
compound 51 for 24 h, and cell lysates were analyzed for phosphorylation of Rb protein at Ser807/811 by immunoblotting (left). Dinaciclib (0.05
μM) served as a positive inhibitor control and actin as a loading control. Onset of apoptosis was assessed by measuring the activity of caspase 3 using
■
▲
●
a bioluminescent assay upon 24 h treatment of cells with increasing concentrations of compound 51 ( ), dinaciclib ( ), or doxorubicin ( ) (right).
Dinaciclib and doxorubicin served as positive controls. The EC50 values for caspase-3 activation were 0.41 μM for compound 51, 0.011 μM for
dinaciclib, and 0.39 μM for doxorubicin, which served as a positive control for the onset of apotosis. (b) Asynchronously growing H929 myeloma
cells were treated for 24 h with compound 51 or dinaciclib and phosphorylation of Rb protein was analyzed by immunoblotting (left). Onset of
apoptosis was assessed by measuring the level of cleaved, activated caspase 3 using flow cytometry (right). Shown are the percentages of cells with
increased caspase-3 levels upon treatment with 1.5 μM compound 51 or 0.005 μM dinaciclib from three experiments.
(Minneapolis, MN). The peptide substrate for activity assays was
synthesized by Biomatik (Wilmington, DE). Protein concentrations
were determined by the Bradford method using Bio-Rad protein assay
reagent (Bio-Rad, Hercules, CA), with bovine serum albumin as a
standard. The concentration of crystallization-grade proteins was
determined by A280 molar absorbance using a nanodrop ND-1000
spectrophotometer (Nanodrop Technologies, Wilmington, DE).
Nonlinear regression analysis was performed using SigmaPlot (Systat
Software, Chicago, IL). Cloning, expression, and purification of CDK2,
cyclin A2, and CAK1 were performed as described previously.41
3776
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
DMEM, penicillin, and streptomycin were purchased from Invitrogen
(Carlsbad, CA). Fetal bovine serum was obtained from Atlanta
Biologicals (Atlanta, GA), and MTT reagent was purchased from
Sigma-Aldrich.
2H). 13C NMR (100 MHz, DMSO-d6) δ 175.9, 168.9, 154.9, 147.4,
137.4, 134.1, 125.2, 121.4, 116.0, 45.7.
(2-(Allylamino)-4-aminothiazol-5-yl)(4-nitrophenyl)methanone
1
(9). Brown solid: mp (EtOAc/MeOH) 183−185 °C. H NMR (400
MHz, DMSO-d6) δ 8.80 (s, 1H), 8.20−7.98 (m, 4H), 7.64−7.51 (m,
2H), 5.88 (ddd, J = 22.2, 10.3, 5.2 Hz, 1H), 5.33−5.06 (m, 2H), 3.93
(s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 178.9, 167.1, 148.0, 147.7,
133.8, 133.7, 128.0, 123.7, 116.4, 46.2. HRMS calcd for C13H13N4O3 S
(M + H) 305.0708, found 305.0719.
Chemical Synthesis. Commercially available chemicals were used
as purchased without further purification. THF was dried over an
1
activated alumina column before use. H NMR spectra were obtained
on a 400 MHz spectrometer, and chemical shifts (ppm) were
referenced to the residual solvent peak. 13C NMR spectra were
recorded at 100 MHz, and chemical shifts (ppm) were referenced to
the appropriate residual solvent peak. High-resolution mass spectra
(HRMS) were recorded with electrospray ionization. NMR spectra are
shown in the Supporting Information. Compounds were purified by
HPLC to >95% purity as reflected by their ability to readily
cocrystallize with CDK2. The synthesis of 42 is given as a
representative procedure that was used for the synthesis of the
targeted library (* denotes that cocrystal structures with CDK2 were
obtained).
(2-(Allylamino)-4-aminothiazol-5-yl)(4-chlorophenyl)methanone
1
(15). Yellow solid: mp (EtOAc/hexane) 198−200 °C. H NMR (400
MHz, DMSO-d6) δ 8.84 (br s, 1H), 8.41 (br s, 1H), 7.95 (br s, 1H),
7.64 (ddd, J = 19.8, 10.6, 5.0 Hz, 1H), 7.51 (dd, J = 8.5, 1.7 Hz, 1H),
7.92−7.83 (m, 2H), 5.17 (ddd, J = 22.2, 10.3, 5.2 Hz, 1H), 5.33−5.06
(m, 2H), 3.91 (br s, 2H).
*(4-Amino-2-(phenylamino)thiazol-5-yl)(phenyl)methanone
1
(16). Yellow solid: mp (EtOAc/MeOH) 194−196 °C. H NMR (400
MHz, DMSO-d6) δ 10.78 (s, 1H), 8.21 (s, 2H), 7.76−7.55 (m, 4H),
7.54−7.41 (m, 3H), 7.36 (t, J = 7.8 Hz, 2H), 7.08 (t, J = 7.3 Hz, 1H).
13C NMR (100 MHz, DMSO-d6) δ 182.7, 167.2, 165.6, 141.9, 139.6,
130.4, 129.1, 128.4, 126.8, 123.4, 119.1.
*4-(4-Amino-5-nicotinoylthiazol-2-ylamino)benzenesulfonamide
(42). At 20 °C, a mixture of 4-isothiocyanatobenzenesulfonamide
(2.14 g, 10 mmol) and cyanamide (0.42 g, 10 mmol) in
tetrahydrofuran (35 mL) was treated with potassium tert-butoxide
(2.46 g, 22 mmol) for 15 min. 3-(2-Bromoacetyl)pyridinium bromide
(2.66 g, 9.5 mmol) was added, and the suspension was stirred for 6 h.
The solvent was removed, and the semisolid residue was triturated
with water to give the crude product, which was filtered and dried in a
vacuum oven at 50 °C for 2 h. The dried crude product was
recrystallized from EtOAc:DMF (8:2), which provided 2.2 g (60%) of
4-(4-amino-5-nicotinoylthiazol-2-ylamino)benzenesulfonamide as a
(4-Amino-2-(cyclohexylamino)thiazol-5-yl)(phenyl)methanone
1
(17). Yellow solid: mp (EtOAc/MeOH) 181−182 °C. H NMR (400
MHz, DMSO-d6) δ 8.56 (s, 2H), 7.66−7.56 (m, 2H), 7.48−7.38 (m,
3H), 3.59 (s, 1H), 2.16−0.88 (m, 11H). 13C NMR (100 MHz,
DMSO-d6) δ 181.4, 172.3, 166.6, 142.2, 130.0, 128.2, 126.6, 32.1, 25.0,
54.0, 24.4.
(4-Amino-2-(isopropylamino)thiazol-5-yl)(phenyl)methanone
1
(18). Yellow solid: mp (EtOAc/MeOH) 161−163 °C. H NMR (400
MHz, DMSO-d6) δ 9.07−7.66 (m, 3H), 7.67−7.29 (m, 5H), 3.81 (brs,
1H), 1.17 (d, J = 6.5 Hz, 6H). 13C NMR (100 MHz, DMSO-d6) δ
181.4, 166.6, 142.2, 130.0, 128.2, 126.6, 46.4, 22.1. HRMS calcd for
C13H15N3O S (M + H) 262.1014, found 262.1014.
1
yellow solid: mp (EtOAc/DMF) 218−220 °C. H NMR (400 MHz,
DMSO-d6) δ 11.18 (s, 1H), 8.86 (d, J = 1.7 Hz, 1H), 8.69 (dd, J = 4.8,
1.6 Hz, 1H), 8.33 (s, 2H), 8.13−7.97 (m, 1H), 7.85−7.72 (m, 4H),
7.52 (ddd, J = 7.9, 4.8, 0.6 Hz, 1H), 7.29 (s, 2H). 13C NMR (100
MHz, DMSO-d6) δ 180.6, 166.6, 165.5, 151.3, 147.6, 142.3, 138.2,
137.1, 134.5, 127.0, 123.7, 118.3; HRMS calcd for C15H14N5O3S2 (M
+ H) 376.0538, found 376.0549.
(4-Amino-2-(propylamino)thiazol-5-yl)(phenyl)methanone (20).
Yellow solid: mp (EtOAc/hexane) 159−161 °C. 1H NMR (400
MHz, DMSO-d6) δ 8.63 (s, 1H), 8.56−7.47 (m, 4H), 7.47−7.39 (m,
3H), 3.19 (s, 2H), 1.65−1.45 (m, 2H), 0.88 (t, J = 7.4 Hz, 3H). 13C
NMR (100 MHz, DMSO-d6) δ 181.5, 171.3, 166.6, 142.2, 130.0,
128.2, 126.7, 91.9, 45.8, 21.8, 11.3. HRMS calcd for C13H16N3O S (M
+ H) 262.1002, found 262.1014.
*(4-Amino-2-(2-methylallylamino)thiazol-5-yl)(phenyl)-
methanone (21). Yellow solid: mp (EtOAc/hexane) 174−176 °C. 1H
NMR (400 MHz, DMSO-d6) δ 8.81 (br s, 2H), 7.63 (d, J = 8.2, 3.5
Hz, 2H), 7.48−7.25 (m, 3H), 4.85 (d, J = 6.4 Hz, 2H), 3.84 (s, 2H),
1.70 (s, 3H). 13C NMR (100 MHz, DMSO-d6) δ181.7, 166.4, 142.2,
141.1, 130.0, 128.3, 126.7, 111.0, 92.2, 60.0, 49.3, 20.0.
Physical Chemical Properties. Solubility: 1.4 μM. Plasma protein
binding: 84%. In vitro absorption (Caco-2, ph 6.5/7.4): A−B = 1.0 ×
10−6; B−A = 24.5 × 10−6.
(2-(Allylamino)-4-aminothiazol-5-yl)(3-nitrophenyl)methanone
1
(2). Brown solid: mp (EtOAc/MeOH) 167−169 °C. H NMR (400
MHz, DMSO-d6) δ 8.89 (s, 1H), 8.80 (d, J = 1.5 Hz, 1H), 8.63 (dd, J
= 4.7, 1.3 Hz, 1H), 8.55−7.61 (m, 3H), 7.46 (ddd, J = 7.9, 4.8, 0.6 Hz,
1H), 5.86 (ddd, J = 22.2, 10.3, 5.1 Hz, 1H), 5.24 −5.04 (m, 2H), 3.91
(s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 178.2, 167.2, 147.6, 143.2,
133.7, 133.1, 130.2, 124.7, 121.4, 116.4, 46.1. HRMS calcd for
C13H13N4O3 S (M + H) 305.0708, found 305.0694.
(4-Amino-2-(4-(trifluoromethylphenyl)amino)thiazol-5-yl)-
(phenyl)methanone (25). Yellow solid: mp (EtOAc/MeOH/DMF)
1
210−212 °C. H NMR (400 MHz, DMSO-d6) δ 11.11 (s, 1H), 8.24
*(2-(Allylamino)-4-aminothiazol-5-yl)(pyridin-3-yl)methanone
(s, 2H), 7.85 (d, J = 8.6 Hz, 2H), 7.75−7.58 (m, 4H), 7.56−7.42 (m,
3H). 13C NMR (100 MHz, DMSO-d6) δ 183.2, 166.4, 165.1, 143.0,
141.7, 130.6, 128.4, 126.8, 126.4, 126.3, 125.7, 123.0, 123.0, 122.7,
118.4.
1
(3). Yellow solid: mp (EtOAc/hexane) 184−186 °C. H NMR (400
MHz, DMSO-d6) δ 8.94 (s, 1H), 8.35−8.65 (m, 2H), 8.30 (dd, J = 8.1,
2.0 Hz, 1H), 8.07 (d, J = 7.7 Hz, 2H), 7.75 (t, J = 7.9 Hz, 1H), 5.87
(ddd, J = 22.0, 10.2, 5.1 Hz, 1H), 5.18 (ddd, J = 13.7, 11.6, 1.3 Hz,
2H), 3.94 (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 179.0, 171.6,
171.3, 166.7, 150.8, 147.5, 137.5, 134.3, 133.8, 123.5, 116.4, 46.1.
*(2-(Allylamino)-4-aminothiazol-5-yl)(naphthalen-2-yl)-
(4-Amino-2-(cyclohexylamino)thiazol-5-yl)(3-nitrophenyl)-
methanone (26). Yellow solid: mp (EtOAc/MeOH) 185−187 °C. 1H
NMR (400 MHz, DMSO-d6) δ 9.13−7.85 (m, 6H), 7.75 (t, J = 8.0
Hz, 1H), 3.72 (s, 1H), 2.12−1.44 (m, 5H), 1.39−1.03 (m, 5H). 13C
NMR (100 MHz, DMSO-d6) δ 177.8, 167.5, 147.6, 143.3, 133.1,
130.2, 124.6, 121.3, 53.0, 32.0, 25.0, 24.3.
*(4-Amino-2-(cyclohexylamino)thiazol-5-yl)(3-methoxyphenyl)-
methanone (27). Yellow solid: mp (EtOAc/MeOH) 184−186 °C. 1H
NMR (400 MHz, DMSO-d6) δ 8.56 (s, 3H), 7.34 (t, J = 7.9 Hz, 1H),
7.24−7.06 (m, 2H), 7.01 (dd, J = 8.1, 2.1 Hz, 1H), 3.78 (s, 3H), 1.89
(d, J = 9.5 Hz, 2H), 1.76−1.49 (m, 3H), 1.44−0.92 (m, 6H). 13C
NMR (100 MHz, DMSO-d6) δ 181.0, 166.7, 159.0, 143.6, 129.4,
118.8, 115.8, 111.9, 55.2, 45.7, 32.1, 25.0, 24.4.
1
methanone (4). Yellow solid: mp (EtOAc/MeOH) 178−180 °C. H
NMR (400 MHz, DMSO-d6) δ 8.80 (s, 1H), 8.20 (s, 1H), 7.99 (ddd, J
= 4.96, 10.61, 19.80 Hz, 3H), 7.73 (dd, J = 1.68, 8.48 Hz, 1H), 7.64−
7.53 (m, 2H), 5.88 (ddd, J = 5.18, 10.35, 22.17 Hz, 1H), 5.49 (D, J =
40.52 Hz, 1H), 5.18 (ddd, J = 1.48, 11.66, 13.74 Hz, 2H), 3.93 (s, 2H).
13C NMR (100 MHz, DMSO-d6) δ 181.5, 166.3, 155.1, 139.5, 133.8,
133.5, 132.1, 128.6, 127.9, 127.6, 127.2, 126.7, 126.1, 124.3, 116.2,
46.1. HRMS calcd for C17H16N3O S (M + H) 310.1014, found
310.1001.
*(2-(Allylamino)-4-aminothiazol-5-yl)(pyridin-2-yl)methanone
*(4-Amino-2-(phenylamino)thiazol-5-yl)(pyridin-3-yl)methanone
(30). Yellow solid: mp (EtOAc/MeOH/DMF) 202−204 °C. 1H NMR
(400 MHz, DMSO-d6) δ 10.87 (s, 1H), 8.84 (d, J = 5.7 Hz, 1H), 8.68
(dd, J = 4.8, 1.6 Hz, 1H), 8.30 (s, 2H), 8.09−7.99 (m, 1H), 7.62 (d, J =
7.8 Hz, 2H), 7.51 (ddd, J = 7.8, 4.8, 0.6 Hz, 1H), 7.36 (d, J = 8.0 Hz,
1
(6). Yellow solid: mp (EtOAc/hexane) 178−180 °C. H NMR (400
MHz, DMSO-d6) δ 8.99 (s, 1H), 8.73 (s, 1H), 8.64 (d, J = 4.7 Hz,
1H), 8.10 (d, J = 7.8 Hz, 1H), 7.97−7.95 (m, 2H), 7.55−7.45 (m,
1H), 5.99−5.83 (m, 1H), 5.27−5.12 (m, 2H), 3.99 (d, J = 25.4 Hz,
3777
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
2H), 7.10 (t, J = 7.4 Hz, 1H). 13C NMR (100 MHz, DMSO-d6) δ
180.1, 167.4, 165.9, 151.1, 147.6, 139.4, 137.3, 134.4, 129.2, 123.6,
123.6, 119.2.
3H), 1.41−0.99 (m, 5H). 13C NMR (100 MHz, DMSO-d6) δ 178.7,
166.9, 150.7, 147.5, 137.6, 134.3, 123.5, 53.08, 32.1, 24.9, 24.4.
3-(4-Amino-5-nicotinoylthiazol-2-ylamino)benzenesulfonamide
(49). Yellow solid: mp (EtOAc/MeOH/DMF) 194−196 °C. 1H NMR
(400 MHz, DMSO-d6) δ 11.63 (s, 1H), 8.99 (d, J = 1.5 Hz, 1H), 8.83
(dd, J = 5.1, 1.1 Hz, 1H), 8.39 (d, J = 8.0 Hz, 2H), 8.15 (s, 1H), 8.02−
7.95 (m, 1H), 7.81 (dd, J = 7.8, 5.3 Hz, 1H), 7.66−7.36 (m, 5H). 13C
NMR (100 MHz, DMSO-d6) δ 178.3, 177.9, 167.1, 149.4, 145.0,
144.5, 143.9, 139.9, 138.3, 129.8, 129.3, 125.3, 121.8, 120.3, 115.9.
HRMS calcd for C15H14N5O3 S2 (M + H) 376.0538, found 376.0549.
*4-(4-Amino-5-(2-nitrobenzoyl)thiazol-2-ylamino)-
(4-Amino-2-(phenylamino)thiazol-5-yl)(3-fluorophenyl)-
methanone (32). Yellow solid: mp (EtOAc/MeOH/DMF) 189−191
1
°C. H NMR (400 MHz, DMSO-d6) δ 10.83 (s, 1H), 8.27 (s, 2H),
7.62 (d, J = 8.0 Hz, 2H), 7.54 (dd, J = 12.4, 5.4 Hz, 2H), 7.47−7.28
(m, 4H), 7.10 (t, J = 7.3 Hz, 1H).
(4-Amino-2-(phenylamino)thiazol-5-yl)(3-nitrophenyl)-
methanone (33). Yellow solid: mp (EtOAc/MeOH) 209−211 °C. 1H
1
NMR (400 MHz, DMSO-d6) H NMR δ 10.92 (s, 1H), 8.97−7.95
1
(m, 5H), 7.79 (t, J = 8.0 Hz, 1H), 7.64 (d, J = 7.9 Hz, 2H), 7.37 (t, J =
7.9 Hz, 2H), 7.10 (t, J = 7.4 Hz, 1H). 13C NMR (100 MHz, DMSO-
d6) δ 179.2, 167.4, 166.5, 147.7, 142.9, 139.4, 133.2, 130.3, 129.6,
125.0, 123.6, 121.4, 119.2.
benzenesulfonamide (51). Orange solid: mp 250 °C. H NMR (400
MHz, DMSO-d6): δ 11.16 (s, 1H disappear on D2O shake), 8.17 (brs
2H, overlapping, disappear on D2O shake), 8.08 (d, J = 8.4 Hz, 1H,
overlapping), 7.82 (dt, J = 7.6, 1.2 Hz, 1H), 7.79−7.76 (m, 4H), 7.69
(dt, J = 8.4, 1.6 Hz, 1H D2O shake), 7.64 (d, J = 7.6 Hz, 1H), 7.30 (s,
2H, disappear on D2O shake). 13C NMR (100 MHz, DMSO-d6): δ
180.24, 167.06, 164.89, 146.76, 142.50, 138.35, 137.23, 134.64, 131.29,
128.74, 127.47, 124.99, 118.74, 94.09. LCMS (ESI+) m/z 420.05 (M +
H)+. HRMS (ESI+) m/z calcd for C16H14N5O5S2 (M + H1)+
420.0431, found 420.0428.
*(4-Amino-2-(phenylamino)thiazol-5-yl)(3-methoxyphenyl)-
methanone (34). Yellow solid: mp (EtOAc/MeOH/DMF) 194−196
1
°C. H NMR (400 MHz, DMSO-d6) δ 10.76 (s, 1H), 8.21 (s, 2H),
7.62 (d, J = 7.8 Hz, 2H), 7.44−7.30 (m, 3H), 7.28−7.15 (m, 2H),
7.13−7.01 (m, 2H), 3.80 (s, 3H). 13C NMR (100 MHz, DMSO-d6) δ
182.3, 167.2, 165.7, 159.1, 143.3, 139.6, 129.6, 129.1, 123.4, 119.1,
119.0, 116.2, 112.1, 55.2. HRMS calcd for C17H16N3O2 S (M + H)
326.0958, found 326.0957.
*4-(4-Amino-5-(3-fluorobenzoyl)thiazol-2-ylamino)-
benzenesulfonamide (52). Yellow solid: mp (EtOAc/MeOH/DMF)
1
(4-Amino-2-(phenylamino)thiazol-5-yl)(3-chlorophenyl)-
202−204 °C. H NMR (400 MHz, DMSO-d6) δ 11.14 (s, 1H), 8.30
(s, 2H), 7.86−7.75 (m, 4H), 7.59−7.24 (m, 6H). 13C NMR (100
MHz, DMSO-d6) δ 181.3, 166.6, 165.5, 163.2, 160.7, 144.0, 142.3,
138.1, 130.7, 127.0, 122.9, 118.3, 117.4, 113.8, 113.6.
methanone (35). Yellow solid: mp (EtOAc/MeOH/DMF) 205−207
1
°C. H NMR (400 MHz, DMSO-d6) δ 10.83 (s, 1H), 8.28 (s, 2H),
7.75−7.41 (m, 6H), 7.36 (d, J = 7.8 Hz, 2H), 7.09 (t, J = 7.3 Hz, 1H).
13C NMR (100 MHz, DMSO-d6) δ 180.5, 167.3, 166.0, 143.8, 139.5,
*4-(4-Amino-5-4-(4-Amino-5-(3-aminobenzoyl)thiazol-2-
ylamino)benzenesulfonamide (55). Yellow solid: mp (EtOAc/
MeOH/DMF) 201−203 °C. 1H NMR (400 MHz, DMSO-d6) δ
11.04 (s, 1H), 8.12 (s, 2H), 7.79 (s, 4H), 7.42−6.47 (m, 6H), 5.31 (s,
2H). 13C NMR (100 MHz, DMSO-d6) δ 184.2, 166.3, 165.9, 164.7,
148.8, 142.6, 137.8, 128.8, 126.9, 118.1, 115.9, 114.1, 112.0.
*4-(4-Amino-5-(3-methoxybenzoyl)thiazol-2-ylamino)-
benzenesulfonamide (56). Yellow solid: mp (EtOAc/MeOH/DMF)
216−218 °C. 1H NMR (400 MHz, DMSO-d6) δ 11.10 (s, 1H), 8.65−
7.88 (m, 2H), 7.87−7.75 (m, 4H), 7.44−7.14 (m, 5H), 7.08 (dd, J =
8.2, 2.0 Hz, 1H), 3.81 (s, 3H)). 13C NMR (100 MHz, DMSO-d6) δ
182.8, 166.5, 165.2, 159.2, 143.1, 142.4, 138.0, 129.7, 127.0, 119.0,
118.2, 116.5, 112.1, 55.3.
133.2, 130.4, 130.2, 129.1, 126.6, 125.4, 123.5, 119.1.
(4-Amino-2-(phenylamino)thiazol-5-yl)(3,5-bis(trifluoromethyl)-
phenyl)methanone (36). Yellow solid: mp (EtOAc/MeOH) 192−
1
194 °C. H NMR (400 MHz, DMSO-d6) δ 10.93 (br s, 1H), 9.03−
7.90 (m, 5H), 7.66 (d, J = 7.9 Hz, 2H), 7.35 (t, J = 7.9 Hz, 2H), 7.08
(t, J = 7.4 Hz, 1H). 13C NMR (100 MHz, DMSO-d6) δ 178.1, 167.9,
166.8, 143.8, 139.4, 131.0, 130.6, 130.3, 130.0, 129.1, 127.5, 127.2,
124.5, 123.9, 123.6, 121.7, 119.9.
*(4-Amino-2-(phenylamino)thiazol-5-yl)(3-(trifluoromethyl)-
phenyl)methanone (37). Yellow solid: mp (EtOAc/MeOH) 179−
1
181 °C. H NMR (400 MHz, DMSO-d6) δ 10.86 (s, 1H), 8.34 (s,
2H), 7.93 (dd, J = 41.4, 9.3 Hz, 3H), 7.77−7.53 (m, 3H), 7.36 (t, J =
7.6 Hz, 2H), 7.09 (t, J = 7.2 Hz, 1H). 13C NMR (100 MHz, DMSO-
d6) δ 180.3, 167.2, 166.2, 142.6, 139.4, 130.8, 129.8, 129.2, 129.1,
128.9, 126.9, 125.3, 123.5, 123.3, 122.6, 119.1.
*4-(4-Amino-5-benzoylthiazol-2-ylamino)benzenesulfonamide
(57). Yellow solid: mp (EtOAc/MeOH/DMF) 218−220 °C. 1H NMR
(400 MHz, DMSO-d6) δ 11.09 (s, 1H), 8.23 (s, 2H), 7.84−7.76 (m,
4H), 7.69 (dd, J = 7.4, 1.5 Hz, 2H), 7.54−7.44 (m, 3H), 7.28 (s, 2H).
13C NMR (100 MHz, DMSO-d6) δ 183.2, 166.4, 165.1, 142.4, 141.7,
(4-Amino-2-(phenylamino)thiazol-5-yl)(naphthalen-2-yl)-
methanone (40). Yellow solid: mp (EtOAc/MeOH/DMF) 191−196
1
°C. H NMR (400 MHz, DMSO-d6) δ 10.81 (s, 1H), 8.28 (s, 3H),
138.0, 130.6, 128.5, 127.0, 126.8, 118.2.
8.06−7.95 (m, 3H), 7.78 (dd, J = 8.5, 1.7 Hz, 1H), 7.70−7.49 (m,
4H), 7.36 (dd, J = 8.2, 7.7 Hz, 2H), 7.08 (t, J = 7.4 Hz, 1H). 13C NMR
(100 MHz, DMSO-d6) δ 182.6, 167.2, 165.6, 139.6, 139.3, 133.7,
132.1, 129.1, 128.7, 128.1, 127.7, 127.4, 126.8, 126.5, 124.3, 123.4,
119.0.
*4-(4-Amino-5-(3-(trifluoromethyl)benzoyl)thiazol-2-ylamino)-
benzenesulfonamide (58). Yellow solid: mp (EtOAc/MeOH/DMF)
1
220−222 °C. H NMR (400 MHz, DMSO-d6) δ 11.18 (s, 1H), 8.36
(s, 2H), 7.99 (d, J = 8.9 Hz, 2H), 7.89 (d, J = 7.8 Hz, 1H), 7.85−7.68
(m, 5H), 7.28 (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 180.9,
166.6, 165.7, 142.4, 142.3, 138.2, 130.9, 129.9, 129.3, 129.0, 127.1,
127.0, 123.4, 123.3, 118.2.
*4-Amino-2-(phenylamino)thiazole-5-carboxamide (41). Yellow
solid: mp (EtOAc/MeOH) 145−147 °C. 1H NMR (400 MHz,
DMSO-d6) δ 11.54 (s, 1H), 8.03 (s, 1H), 7.72 (s, 1H), 7.44 (ddd, J =
32.2, 11.2, 4.8 Hz, 4H), 7.22 (t, J = 7.3 Hz, 1H), 3.98 (s, 2H).
*4-(4-Amino-5-nicotinoylthiazol-2-ylamino)benzamide (43). Yel-
*4-(4-Amino-5-(3-chlorobenzoyl)thiazol-2-ylamino)-
benzenesulfonamide (65). Yellow solid: mp (EtOAc/MeOH/DMF)
1
210−212 °C. H NMR (400 MHz, DMSO-d6) δ 11.15 (s, 1H), 8.31
1
(s, 2H), 7.93−7.74 (m, 4H), 7.59 (ddd, J = 36.4, 18.0, 10.4 Hz, 4H),
7.29 (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 181.1, 166.6, 165.6,
143.6, 142.3, 138.1, 133.3, 130.5, 130.4, 127.0, 126.7, 125.5, 118.3.
*4-(4-Amino-5-(4-methoxybenzoyl)thiazol-2-ylamino)-
benzenesulfonamide (66). Yellow solid: mp (EtOAc/MeOH/DMF)
low solid: mp (EtOAc/MeOH/DMF) 229−231 °C. H NMR (400
MHz, DMSO-d6) δ 11.08 (s, 1H), 8.87 (s, 1H), 8.69 (d, J = 3.9 Hz,
1H), 8.35 (s, 2H), 8.13−7.62 (m, 6H), 7.51 (dd, J = 7.4, 5.0 Hz, 1H),
7.31 (s, 1H). 13C NMR (100 MHz, DMSO-d6) δ 180.5, 167.3, 166.8,
165.7, 151.2, 147.6, 142.0, 137.2, 134.5, 128.7, 123.6, 117.9.
1
217−219 °C. H NMR (400 MHz, DMSO-d6) δ 11.07 (s, 1H), 8.17
*(4-Amino-2-(isopropylamino)thiazol-5-yl)(pyridin-3-yl)-
methanone (44). Yellow solid: mp (EtOAc/MeOH) 174−176 °C. 1H
NMR (400 MHz, DMSO-d6) δ 8.79 (d, J = 1.8 Hz, 1H), 8.64 (dd, J =
4.8, 1.6 Hz, 2H), 8.37−7.70 (m, 2H), 7.47 (dd, J = 7.8, 4.8 Hz, 1H),
3.85 (s, 1H), 1.17 (d, J = 6.5 Hz, 6H). 13C NMR (100 MHz, DMSO-
d6) δ 178.7, 166.9, 150.7, 147.5, 137.5, 134.3, 123.5, 46.4, 22.1.
*(4-Amino-2-(cyclohexylamino)thiazol-5-yl)(pyridin-3-yl)-
methanone (45). Yellow solid: mp (EtOAc/MeOH) 174−176 °C. 1H
NMR (400 MHz, DMSO-d6) δ 9.23−7.69 (m, 6H), 7.47 (dd, J = 7.8,
4.8 Hz, 1H), 3.68 (s, 1H), 1.93 (t, J = 19.1 Hz, 2H), 1.77−1.42 (m,
(s, 2H), 7.88−7.74 (m, 4H), 7.70 (d, J = 8.7 Hz, 2H), 7.28 (s, 2H),
7.03 (d, J = 8.7 Hz, 2H), 3.82 (s, 3H). 13C NMR (100 MHz, DMSO-
d6) δ 182.5, 166.0, 164.9, 161.1, 142.5, 137.9, 134.0, 128.9, 127.0,
118.1, 113.7, 55.3.
4-(4-Amino-5-(3,5-bis(trifluoromethyl)benzoyl)thiazol-2-
ylamino)benzenesulfonamide (67). Yellow solid: mp (EtOAc/
MeOH/DMF) 224−226 °C. 1H NMR (400 MHz, DMSO-d6) δ
11.25 (s, 1H), 8.46 (s, 2H), 8.30 (d, J = 9.2 Hz, 3H), 7.82 (q, J = 9.0
Hz, 4H), 7.30 (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ 178.8,
3778
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
166.7, 166.2, 143.6, 142.2, 138.3, 131.0, 130.7, 130.4, 130.0, 127.6,
127.2, 127.0, 124.5, 124.1, 121.7, 118.3. HRMS calcd for
C18H12F6N4O3 S2 (M + H) 511.0335, found 511.0336.
The profiling of compound 51 against a panel of 339 kinases and the
dose−response kinetics against different CDK−cyclin complexes were
performed by Reaction Biology Corp. using a P33-radiolabeled assay.
Residual enzymatic activity (in % of DMSO controls) was determined
in duplicate at a compound concentration of 0.1 μM for the profiling
and between 0.05 nM and 1000 nM for the dose−response analysis;
the ATP concentration was 10 μM and staurosporine served as a
positive control.
X-ray Crystallography. Crystallization was performed at 19 °C
using the hanging drop vapor diffusion method. Crystals of CDK2−
inhibitor complexes were grown from a solution of 10 mg mL−1 CDK2
in the presence of 0.5−3 mM compound with 0.1 M HEPES/NaOH
(pH 7.5) and 10% (v/v) polyethylene glycol 3350. Crystals were
harvested in cryoprotectant (50 mM HEPES pH 7.5, 50 mM
phosphate (Na/K) pH 7.5, 15% (v/v) polyethylene glycol 3350, 23%
(v/v) ethylene glycol, and 0.5 mM inhibitor) prior to data collection.
X-ray diffraction data were recorded at −180 °C in the Moffitt Cancer
Center Chemical Biology Core using the rotation method on single
crystals (Cu Kα radiation generated by Rigaku Micro-Max 007-HF,
equipped with an HTC image plate and a Rigaku CCD Saturn 944
detector). The data set for compound 51 was recorded at the APS
(Argonne National Laboratories) beamline 22-ID (SER-CAT) at a
wavelength of 1.0 Å. Data were reduced with XDS46 or HKL-2000.47
CNS48 was employed for phasing and refinement (minimization and
simulated annealing), and model building was performed using O49 or
Coot.50 Initial models for the inhibitors were generated using
Chem3DPro (CambridgeSoft), and parameter and topology files
were generated using xplo2d.51 Figures were drawn with PyMol
*4-(5-(2-Naphthoyl)-4-aminothiazol-2-ylamino)-
benzenesulfonamide (69). Yellow solid: mp (EtOAc/MeOH/DMF)
1
220−222 °C. H NMR (400 MHz, DMSO-d6) δ 11.13 (s, 1H), 8.31
(s, 3H), 8.01 (ddd, J = 13.1, 9.0, 6.7 Hz, 3H), 7.86−7.75 (m, 5H),
7.65−7.56 (m, 2H), 7.29 (s, 2H). 13C NMR (100 MHz, DMSO-d6) δ
183.3, 166.7, 165.4, 142.7, 139.3, 138.2, 134.0, 132.3, 129.0, 128.4,
127.9, 127.7, 127.2, 127.1, 126.9, 124.5, 118.4.
3-(4-Amino-5-(3-fluorobenzoyl)thiazol-2-ylamino)-
benzenesulfonamide (83). Yellow solid: mp (EtOAc/MeOH/DMF)
1
182−184 °C. H NMR (400 MHz, DMSO-d6) δ 11.08 (s, 1H), 8.27
(s, 2H), 8.07 (s, 1H), 7.96 (d, J = 7.2 Hz, 1H), 7.69−7.22 (m, 8H).
4-Amino-N-(4-nitrophenyl)-2-(4-sulfamoylphenylamino)-
thiazole-5-carboxamide (93). Yellow solid: mp (EtOAc/MeOH)
1
228−230 °C. H NMR (400 MHz, DMSO-d6) δ 10.99 (s, 1H), 9.62
(s, 1H), 8.17 (d, J = 8.9 Hz, 2H), 7.93 (d, J = 9.0 Hz, 2H), 7.87−7.64
(m, 4H), 7.33 (d, J = 47.4 Hz, 4H). 13C NMR (100 MHz, DMSO-d6)
δ 163.7, 163.6, 162.7, 146.6, 142.9, 141.3, 137.5, 127.0, 124.6, 119.2,
117.7.
Enzyme Assays. The synthetic peptide PKTPKKAKKL42 served
as a substrate for the activated CDK2−cyclin A2 complex, and the
formation of ADP from ATP was coupled to the oxidation of NADH
using pyruvate kinase (PK) and lactate dehydrogenase (LDH),
monitored at 340 nm.43,44 Assays for the HTS campaign (Z-factor =
0.74) and dose−response measurements for SAR were carried out in
384-well plates at room temperature in 50 mM Tris-HCl buffer (pH
7.5) containing 10 mM MgCl2, 0.24 mM NADH, 5 mM DTT, 6 U
mL−1 LDH, 10 U mL−1 PK, 1 mM phosphoenolpyruvate, 5% (v/v)
DMSO, and 0.15 μM (0.01 mg mL−1) activated CDK−cyclin A2
complex. Inhibitor was added to the mixture, and the reaction was
initiated by the addition of 0.2 mM ATP and 0.125 mM peptide
substrate. The more sensitive ADP Quest fluorescence assay
(DiscoveRx, Fremont, CA)45 was employed for compounds with
IC50 values below 1.0 μM. In this assay, the formation of ADP is
coupled to the generation of a fluorescent resorufin dye (excitation
and emission wavelengths of 540 and 590 nm, respectively). Reactions
were carried out at room temperature in 15 mM HEPES buffer (pH
7.4) containing 20 mM NaCl, 1 mM EGTA, 0.02% Tween 20, 10 mM
MgCl2, 5% (v/v) DMSO, and 8.8 nM (0.56 μg mL−1) activated
CDK2−cyclin A2. Inhibitor was added to the mixture, and the reaction
was initiated by the addition of 150 μM ATP and 100 μM peptide
substrate. All kinetic assays were performed in 384-well plates using a
Wallac Envision 2102 plate reader (Perkin-Elmer). IC50 values were
obtained by fitting data to eq 1, where A is the fraction of activity
remaining, [I] is the concentration of inhibitor, and n is the Hill slope
coefficient.
(Schrodinger, Inc.). Data and refinement statistics are shown in the
̈
Supporting Information (Table S1), along with the Fo − Fc electron
density maps from refinement cycles omitting the respective ligands
(Figure S5 of Supporting Information).
Cell Viability Assays. Cell lines were obtained from the American
Type Culture Collection (Manassas, VA) and maintained in RPMI or
DMEM with 10% fetal bovine serum, 37 °C, and 5% CO2. All cell lines
tested free of mycoplasma with MycoAlert tests (Lonza Rockland, Inc.,
Rockland, ME), and the cell line identities were confirmed using
StemElite ID system (Promega Corp., Madison, WI).
For viability assays using MTT (3-(4,5-dimethythiazol-2-yl)-2,5-
diphenyltetrazolium bromide), cells were plated in 96-well plates in
100 μL of medium and were allowed to attach overnight. Cells were
then incubated for 72 h with increasing concentrations of inhibitor.
Media was aspirated and replaced with 100 μL of complete media
containing 1 mg mL−1 MTT ((3-(4,5-dimethythiazol-2-yl)-2,5-
diphenyltetrazolium bromide) and incubated for 3 h at 37 °C in 5%
CO2 humidified incubator. Media was then aspirated, and DMSO was
added. Cells were incubated for 10 min at room temperature while
shaking, and optical density was determined at 540 nm using a μQuant
spectrophotometric plate reader (Bio-TEK, Winooski, VT). NCI-
H929 and U266 cells were cultured in RPMI-1640 media (GIBCO,
Life Technologies Carlsbad, CA), supplemented with 10% fetal bovine
serum (GIBCO) including 0.05 mM 2-mercaptoethanol for NCI-
H929 cells. Cells were plated in 96-well plate at a density of 10 × 103
cells/well. MTT assays with GC-1, GC-2, and TM-4 testis cells were
carried out as above, with modifications in culture conditions as
appropriate for these cell lines.52−54
For viability assays using high-throughput CellTiter-Blue, 1200 cells
were plated in each well of 384-well plates using a Precision XS liquid
handling station (Bio-Tek Instruments, Inc., Winooski, VT) and
allowed to incubate overnight at 37 °C and 5% CO2. A liquid handling
station was used to serially dilute all drugs (2:3) in media, and 6 μL of
these dilutions were added to appropriate wells. Four replicate wells
were used for each drug concentration, and an additional four control
wells received a diluent control without drug. After a 72 h incubation
period with drugs, 5 μL of CellTiter-Blue reagent (Promega Corp.)
were added to each well. Cell viability was assessed by the ability of the
remaining viable cells to reduce resazurin to resorufin. The
fluorescence of resorufin (579 nm excitation/584 nm emission) was
measured with a Synergy 4 microplate reader (Bio-Tek Instruments,
Inc.). The fluorescence data were transferred to Microsoft Excel to
1
A =
n
[I]
1 +
(IC )
(1)
50
The most potent compounds, 42 and 51, were further analyzed by
Reaction Biology Corporation (Malvern, PA) using a highly sensitive
P33-radiolabeled assay against CDK2−cyclin A2.
Kinase Profiling. Selected compounds were tested against a panel
of 48 kinases provided by Caliper Life Sciences (Hopkinton, MA),
which is representative of the entire human kinome. This kinase panel,
available as ProfilerPro Kits 1 and 2, was tested using a Caliper
LC3000 instrument in an electrophoretic mobility-shift assay at the
High Throughput Screening Laboratory in the Institute for
Therapeutics Discovery and Development at the University of
Minnesota, College of Pharmacy. Compounds were added to the
plates and preincubated with enzyme for 15 min, followed by addition
of peptide substrates and ATP for a 90 min reaction period at 28 °C.
Inhibitory activity was determined by measuring the conversion of
nonphosphorylated peptide substrate to phosphorylated product,
expressed relative to plate controls. The Spotfire visualization program
(Tibco Software, Palo Alto, CA) was used to generate a gradient heat
map of the inhibitory activities (Figure S3 of Supporting Information).
3779
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
calculate the percent viability relative to the four replicate cell wells
that did not receive drug. IC50 values were determined using a
sigmoidal equilibrium model regression using XLfit version 5.2 (ID
Business Solutions Ltd.).
ACKNOWLEDGMENTS
■
We thank the Moffitt Core facilities Chemical Biology,
Translational Research, and Flow Cytometry for assistance in
X-ray data collection and cell biological studies. We acknowl-
edge the technical support of Nestor Rodriguez, Anne
Hoffman, and Todd Hall (University of Kansas Medical
School) for performing MTT assays on testis cell lines. We
thank Drs. Bi-Cheng Wang and Lirong Chen (SER-CAT,
University of Georgia) for support during Synchrotron data
collection at Argonne National Laboratory. We thank Katinka
Baehr (University of Kansas) for assistance with the production
of CDK2-cyclin A, and Dr. Subhashree Francis (University of
Minnesota) for assistance with mass spectrometry analysis of
compounds. We thank Donna J. Ingles (Moffitt) for help with
the preparation of the manuscript and Dr. Jin-Yi Zhu for
assistance with X-ray data collection. This work was supported
by the Contraceptive Discovery and Development Branch of
the Eunice Kennedy Shriver National Institute of Child Health
a n d H u m a n D e v e l o p m e n t t h r o u g h c o n t r a c t
HHSN275200503400C (N01-HD-5-3400). We thank program
officers Drs. H. K. Kim, Diana Blithe, and Min Lee for their
support of this project.
Assessment of Activated Caspase. Caspase 3/7 activation was
measured using a plate-based Caspase-Glo 3/7 (Promega) lumines-
cent assay. Cells were plated to white 384-well plates as described for
cell viability assays. Cells were treated for 24 h with serial dilutions of
each compound. For the detection of cleaved caspase-3, H929 cells
were plated at a density of 4 × 105 cells/mL and treated with DMSO
control, 1.5 μM compound 51, or 0.005 μM dinaciclib for 24 h.
Cleaved caspase-3 levels were assessed using the mAb FITC Active
caspase-3 apoptosis kit (BD Pharmingen). Samples were analyzed
using a FACScan flow cytometer (BD Biosciences). The experiment
was performed in triplicate.
Immunoblotting. A-673 cells were grown to ∼70% confluence,
and compounds were added. After exposure to the compounds for 24
h, the cells were lysed in cell lysis buffer (Cell Signaling, Danvers,
MA). The soluble protein concentration was determined using the
Bio-Rad Protein Assay reagent, and 50 μg of total protein wad loaded
on each lane of a 10% SDS-polyacrylamide gel. The proteins were then
transferred to nitrocellulose membrane, washed with TBS/0.1%
Tween 20, and incubated in 1× TBS−Tween20/2% BSA overnight
at 4 °C with primary antibodies. Antibody for phospho-Rb (Ser807/
811, no. 9308) was obtained from Cell Signaling. β-Actin (clone AC-
74) antibody was obtained from Sigma-Aldrich (St. Louis, MO; no.
A2228). The membrane was then washed with TBS/0.1% Tween 20,
incubated for 1 h at room temperature with goat antimouse 800 nm
fluorescent and goat antirabbit 680 nm fluorescent labeled antibodies,
and scanned on a Li-Cor (Lincoln, NE) Odyssey infrared scanner.
H929 cells were plated at a density of 4 × 105 cells/mL. After
treatment for 24 h, cells were washed with cold PBS and lysed in 100
μL of RIPA lysis buffer (Upstate, Lake Placid, NY). Then 10 μg of
total protein was loaded on an 8% SDS-PAGE gel. Proteins were
resolved by electrophoresis, transferred to a polyvinylidene fluoride
(PVDF) membrane, and then blocked overnight in 5% nonfat milk in
TBS-0.05% Tween-20. Protein levels were examined with antisera
specific antibodies to pRb (Ser807/811) or GAPDH (Cell Signaling,
Beverly, MA) and visualized with chemoluminescence (Pierce,
Rockford, IL).
ABBREVIATIONS USED
■
CAK, CDK-activating kinase; DMEM, Dulbecco’s Modified
Eagle’s Medium; EGTA, ethylene glycol tetraacetic acid;
HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid;
HTS, high throughput screening; LDH, lactate dehydrogenase;
MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
bromide; PK, pyruvate kinase; SAR, structure−activity relation-
ship
REFERENCES
■
(1) Johnson, L. N. Protein kinase inhibitors: contributions from
structure to clinical compounds. Q. Rev. Biophys. 2009, 42, 1−40.
(2) Hall, M.; Peters, G. Genetic alterations of cyclins, cyclin-
dependent kinases, and Cdk inhibitors in human cancer. Adv. Cancer.
Res. 1996, 68, 67−108.
(3) Grana, X.; Reddy, E. P. Cell cycle control in mammalian cells:
role of cyclins, cyclin dependent kinases (CDKs), growth suppressor
genes and cyclin-dependent kinase inhibitors (CKIs). Oncogene 1995,
11, 211−219.
(4) Morgan, D. O. Principles of CDK regulation. Nature 1995, 374,
131−134.
ASSOCIATED CONTENT
■
S
* Supporting Information
X-ray crystallography data and refinement statistics, Fo − Fc
electron density omit maps, additional SAR data along with
dose−response graphs, kinase profiling results, and NMR
spectra. This material is available free of charge via the Internet
at http://pubs.acs.org.
(5) Sherr, C. J. Cancer cell cycles. Science 1996, 274, 1672−1777.
(6) Weinberg, R. A. The retinoblastoma protein and cell cycle
control. Cell 1995, 81, 323−330.
Accession Codes
(7) van den Heuvel, S.; Harlow, E. Distinct roles for cyclin-
dependent kinases in cell cycle control. Science 1993, 262, 2050−2054.
(8) Fu, W.; Ma, L.; Chu, B.; Wang, X.; Bui, M. M.; Gemmer, J.;
Altiok, S.; Pledger, W. J. The cyclin-dependent kinase inhibitor SCH
727965 (dinacliclib) induces the apoptosis of osteosarcoma cells. Mol.
Cancer Ther. 2011, 10, 1018−1027.
(9) Malumbres, M.; Barbacid, M. To cycle or not to cycle: a critical
decision in cancer. Nature Rev. Cancer 2001, 1, 222−231.
(10) Horiuchi, D.; Huskey, N. E.; Kusdra, L.; Wohlbold, L.; Merrick,
K. A.; Zhang, C.; Creasman, K. J.; Shokat, K. M.; Fisher, R. P.; Goga,
A. Chemical-genetic analysis of cyclin dependent kinase 2 function
reveals an important role in cellular transformation by multiple
oncogenic pathways. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 1019−
1027.
The atomic coordinates and structure factors for the crystal
structures determined as part of this work have been deposited
in the Protein Data Bank (entries 3QQK, 3R8V, 3QTQ, 3R8U,
3S00, 3S0O, 3R8Z, 3QTR, 3RJC, 3R9N, 3RAH, 3QTW,
3QTS, 3RPR, 3RZB, 3QU0, 3RK7, 3RK9, 3RKB, 3RK5, 3QXP,
3QTZ, 3QTU, 3QTX, 3RPV, 3RAL, 3RAK, 3RPY, 3SQQ,
3S1H, 3RMF, 3RNI, 3R9D, 3R9O, 3R9H, and 4GCJ).
AUTHOR INFORMATION
■
Corresponding Author
*For E.S: phone, +813-745-4503; e-mail, ernst.schonbrunn@
moffitt.org. For G.I.G.: phone, +612-626-6320; e-mail, georg@
umn.edu.
(11) Ortega, S.; Prieto, I.; Odajima, J.; Martin, A.; Dubus, P.; Sotillo,
R.; Barbero, J. L.; Malumbres, M.; Barbacid, M. Cyclin-dependent
kinase 2 is essential for meiosis but not for mitotic cell division in mice.
Nature Genet. 2003, 35, 25−31.
Notes
The authors declare no competing financial interest.
3780
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
(12) van der Meer, T.; Chan, W. Y.; Palazon, L. S.; Nieduszynski, C.;
Murphy, M.; Sobczak-Thepot, J.; Carrington, M.; Colledge, W. H.
Cyclin A1 protein shows haplo-insufficiency for normal fertility in
male mice. Reproduction 2004, 127, 503−511.
(13) Yang, R.; Muller, C.; Huynh, V.; Fung, Y. K.; Yee, A. S.;
Koeffler, H. P. Functions of cyclin A1 in the cell cycle and its
interactions with transcription factor E2F-1 and the Rb family of
proteins. Mol. Cell. Biol. 1999, 19, 2400−2407.
(14) Nya-Ngatchou, J. J.; Amory, J. K. New approaches to male non-
hormonal contraception. Contraception 2013, 87, 296−299.
(15) Page, S. T.; Amory, J. K.; Bremner, W. J. Advances in male
contraception. Endocr. Rev. 2008, 29, 465−493.
(16) Tash, J. S.; Attardi, B.; Hild, S. A.; Chakrasali, R.; Jakkaraj, S. R.;
Georg, G. I. A novel potent indazole carboxylic acid derivative blocks
spermatogenesis and is contraceptive in rats after a single oral dose.
Biol. Reprod. 2008, 78, 1127−1138.
potential treatment for Alzheimer’s disease. Bioorg. Med. Chem. Lett.
2004, 14, 5521−5525.
(29) Laha, J. K.; Zhang, X.; Qiao, L.; Liu, M.; Chatterjee, S.;
Robinson, S.; Kosik, K. S.; Cuny, G. D. Structure−activity relationship
study of 2,4-diaminothiazoles as Cdk5/p25 kinase inhibitors. Bioorg.
Med. Chem. Lett. 2011, 21, 2098−2101.
(30) Anastassiadis, T.; Deacon, S. W.; Devarajan, K.; Ma, H.;
Peterson, J. R. Comprehensive assay of kinase catalytic activity reveals
features of kinase inhibitor selectivity. Nature Biotechnol. 2011, 29,
1039−1045.
(31) Jeffrey, P. D.; Russo, A. A.; Polyak, K.; Gibbs, E.; Hurwitz, J.;
Massague, J.; Pavletich, N. P. Mechanism of CDK activation revealed
by the structure of a cyclinA−CDK2 complex. Nature 1995, 376, 313−
320.
(32) Steuber, H.; Heine, A.; Klebe, G. Structural and thermodynamic
study on aldose reductase: nitro-substituted inhibitors with strong
enthalpic binding contribution. J. Mol. Biol. 2007, 368, 618−638.
(33) Lesuisse, D.; Dutruc-Rosset, G.; Tiraboschi, G.; Dreyer, M. K.;
Maignan, S.; Chevalier, A.; Halley, F.; Bertrand, P.; Burgevin, M. C.;
Quarteronet, D.; Rooney, T. Rational design of potent GSK3beta
inhibitors with selectivity for Cdk1 and Cdk2. Bioorg. Med. Chem. Lett.
2010, 20, 1985−1989.
(34) Echalier, A.; Endicott, J. A.; Noble, M. E. Recent developments
in cyclin-dependent kinase biochemical and structural studies. Biochim.
Biophys. Acta 2010, 1804, 511−519.
(35) Sirianni, R.; Chimento, A.; Ruggiero, C.; De Luca, A.; Lappano,
R.; Ando, S.; Maggiolini, M.; Pezzi, V. The novel estrogen receptor, G
protein-coupled receptor 30, mediates the proliferative effects induced
by 17beta-estradiol on mouse spermatogonial GC-1 cell line.
Endocrinology 2008, 149, 5043−5051.
(17) Cicenas, J.; Valius, M. The CDK inhibitors in cancer research
and therapy. J. Cancer. Res. Clin. Oncol. 2011, 137, 1409−1418.
(18) Massard, C.; Soria, J. C.; Anthoney, D. A.; Proctor, A.; Scaburri,
A.; Pacciarini, M. A.; Laffranchi, B.; Pellizzoni, C.; Kroemer, G.;
Armand, J. P.; Balheda, R.; Twelves, C. J. A first in man, phase I dose-
escalation study of PHA-793887, an inhibitor of multiple cyclin-
dependent kinases (CDK2, 1 and 4) reveals unexpected hepatotoxicity
in patients with solid tumors. Cell Cycle 2011, 10, 963−970.
(19) Guha, M. Cyclin-dependent kinase inhibitors move into Phase
III. Nature Rev. Drug Discovery 2012, 11, 892−894.
(20) Guha, M. Blockbuster dreams for Pfizer’s CDK inhibitor. Nature
Biotechnol. 2013, 31, 187.
(21) Paruch, K.; Dwyer, M. P.; Alvarez, C.; Brown, C.; Chan, T.-Y.;
Doll, R. J.; Keertikar, K.; Knutson, C.; McKittrick, B.; Rivera, J.;
Rossman, R.; Tucker, G.; Fischmann, T.; Hruza, A.; Madison, V.;
Nomeir, A. A.; Wang, Y.; Kirschmeier, P.; Lees, E.; Parry, D.;
Sgambellone, N.; Seghezzi, W.; Schultz, L.; Shanahan, F.; Wiswell, D.;
Xu, X.; Zhou, Q.; James, R. A.; Paradkar, V. M.; Park, H.; Rokosz, L.
R.; Stauffer, T. M.; Guzi, T. J. Discovery of Dinaciclib (SCH 727965):
A Potent and Selective Inhibitor of Cyclin-Dependent Kinases. ACS
Med. Chem. Lett. 2010, 1, 204−208.
(36) Chen, H.; Fok, K. L.; Jiang, X.; Jiang, J.; Chen, Z.; Gui, Y.; Chan,
H. C.; Cai, Z. CD147 regulates apoptosis in mouse spermatocytes but
not spermatogonia. Hum. Reprod. 2012, 27, 1568−1576.
(37) McKee, C. M.; Ye, Y.; Richburg, J. H. Testicular germ cell
sensitivity to TRAIL-induced apoptosis is dependent upon p53
expression and is synergistically enhanced by DR5 agonistic antibody
treatment. Apoptosis 2006, 11, 2237−2250.
(22) Parry, D.; Guzi, T.; Shanahan, F.; Davis, N.; Prabhavalkar, D.;
Wiswell, D.; Seghezzi, W.; Paruch, K.; Dwyer, M. P.; Doll, R.; Nomeir,
A.; Windsor, W.; Fischmann, T.; Wang, Y.; Oft, M.; Chen, T.;
Kirschmeier, P.; Lees, E. M. Dinaciclib (SCH 727965), a novel and
potent cyclin-dependent kinase inhibitor. Mol. Cancer Ther. 2010, 9,
2344−2353.
(23) Toogood, P. L.; Harvey, P. J.; Repine, J. T.; Sheehan, D. J.;
VanderWel, S. N.; Zhou, H.; Keller, P. R.; McNamara, D. J.; Sherry,
D.; Zhu, T.; Brodfuehrer, J.; Choi, C.; Barvian, M. R.; Fry, D. W.
Discovery of a potent and selective inhibitor of cyclin-dependent
kinase 4/6. J. Med. Chem. 2005, 48, 2388−2406.
(38) Robillard, K. R.; Hoque, T.; Bendayan, R. Expression of ATP-
binding cassette membrane transporters in rodent and human sertoli
cells: relevance to the permeability of antiretroviral therapy at the
blood−testis barrier. J. Pharmacol. Exp. Ther. 2012, 340, 96−108.
(39) Musa, F. R.; Tokuda, M.; Kuwata, Y.; Ogawa, T.; Tomizawa, K.;
Konishi, R.; Takenaka, I.; Hatase, O. Expression of cyclin-dependent
kinase 5 and associated cyclins in Leydig and Sertoli cells of the testis.
J. Androl. 1998, 19, 657−666.
(40) T’Ang, A.; Varley, J. M.; Chakraborty, S.; Murphree, A. L.; Fung,
Y. K. Structural rearrangement of the retinoblastoma gene in human
breast carcinoma. Science 1988, 242, 263−266.
(24) Matthews, T. P.; Klair, S.; Burns, S.; Boxall, K.; Cherry, M.;
Fisher, M.; Westwood, I. M.; Walton, M. I.; McHardy, T.; Cheung, K.
M.; Van Montfort, R.; Williams, D.; Aherne, G. W.; Garrett, M. D.;
Reader, J.; Collins, I. Identification of inhibitors of checkpoint kinase 1
through template screening. J. Med. Chem. 2009, 52, 4810−4819.
(25) Getlik, M.; Grutter, C.; Simard, J. R.; Nguyen, H. D.; Robubi,
A.; Aust, B.; van Otterlo, W. A.; Rauh, D. Structure-based design,
synthesis and biological evaluation of N-pyrazole, N′-thiazole urea
inhibitors of MAP kinase p38alpha. Eur. J. Med. Chem. 2012, 48, 1−15.
(26) Vulpetti, A.; Casale, E.; Roletto, F.; Amici, R.; Villa, M.;
Pevarello, P. Structure-based drug design to the discovery of new 2-
aminothiazole CDK2 inhibitors. J. Mol. Graphics Modell. 2006, 24,
341−348.
(27) Ahn, J. S.; Radhakrishnan, M. L.; Mapelli, M.; Choi, S.; Tidor,
B.; Cuny, G. D.; Musacchio, A.; Yeh, L. A.; Kosik, K. S. Defining Cdk5
ligand chemical space with small molecule inhibitors of tau
phosphorylation. Chem. Biol. 2005, 12, 811−823.
(28) Helal, C. J.; Sanner, M. A.; Cooper, C. B.; Gant, T.; Adam, M.;
Lucas, J. C.; Kang, Z.; Kupchinsky, S.; Ahlijanian, M. K.; Tate, B.;
Menniti, F. S.; Kelly, K.; Peterson, M. Discovery and SAR of 2-
aminothiazole inhibitors of cyclin-dependent kinase 5/p25 as a
(41) Betzi, S.; Alam, R.; Martin, M. P.; Lubbers, D. J.; Han, H.;
Jakkaraj, S. R.; Georg, G. I.; Schonbrunn, E. Discovery of a potential
allosteric ligand binding site in CDK2. ACS Chem. Biol. 2011, 6, 492−
501.
(42) Stevenson-Lindert, L. M.; Fowler, P.; Lew, J. Substrate
specificity of CDK2-cyclin A. What is optimal? J. Biol. Chem. 2003,
278, 50956−50960.
(43) Adams, J. A.; McGlone, M. L.; Gibson, R.; Taylor, S. S.
Phosphorylation modulates catalytic function and regulation in the
cAMP-dependent protein kinase. Biochemistry 1995, 34, 2447−2454.
(44) Cook, P. F.; Neville, M. E., Jr.; Vrana, K. E.; Hartl, F. T.;
Roskoski, R., Jr. Adenosine cyclic 3′,5′-monophosphate dependent
protein kinase: kinetic mechanism for the bovine skeletal muscle
catalytic subunit. Biochemistry 1982, 21, 5794−5799.
(45) Martin, M. P.; Zhu, J. Y.; Lawrence, H. R.; Pireddu, R.; Luo, Y.;
Alam, R.; Ozcan, S.; Sebti, S. M.; Lawrence, N. J.; Schonbrunn, E. A
novel mechanism by which small molecule inhibitors induce the DFG
flip in Aurora A. ACS Chem. Biol. 2012, 7, 698−706.
(46) Kabsch, W. Automatic processing of rotation diffraction data
from crystals of initially unknown symmetry and cell constants. J. Appl.
Crystallogr. 1993, 26, 795−800.
3781
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782

Journal of Medicinal Chemistry
Article
(47) Otwinowski, Z.; Minor, W. Processing of X-Ray Diffraction
Data Collected in Oscillation Mode. Methods Enzymol. 1997, 276,
307−326.
(48) Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.;
Gros, P.; Grosse-Kunstleve, R. W.; Jiang, J. S.; Kuszewski, J.; Nilges,
M.; Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: A new software suite for macro-
molecular structure determination. Acta Crystallogr., Sect. D: Biol.
Crystallogr. 1998, 54, 905−921.
(49) Jones, T. A.; Zou, J. Y.; Cowan, S. W.; Kjeldgaard, M. Improved
methods for building protein models in electron density maps and the
location of errors in these models. Acta. Crystallogr., Sect. A: Found.
Crystallogr. 1991, 47, 110−119.
(50) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta. Crystallogr., Sect. D: Biol. Crystallogr. 2004,
60, 2126−2132.
(51) Kleywegt, G. J.; Zou, J.-Y.; Kjeldgaard, M.; Jones, T. A. Around,
O Crystallography of biological macromolecules. In International
Tables for Crystallography; Rossmann, M. G.; Arnold, E., Eds.;
International Union of Crystallography: Chester, UK, 2001; Vol. F,
pp 353−356.
(52) Carson, D. D.; Rosenberg, L. I.; Blaner, W. S.; Kato, M.;
Lennarz, W. J. Synthesis and secretion of a novel binding protein for
retinol by a cell line derived from Sertoli cells. J. Biol. Chem. 1984, 259,
3117−3123.
(53) Lee, D. Y.; Lee, S. S.; Joo, W. A.; Lee, E. J.; Kim, C. W. Analysis
of differentially regulated proteins in TM4 cells treated with bisphenol
A. Biosci. Biotechnol. Biochem. 2004, 68, 1201−1208.
(54) Lizama, C.; Ludwig, A.; Moreno, R. D. Etoposide induces
apoptosis and upregulation of TACE/ADAM17 and ADAM10 in an in
vitro male germ cell line model. Biochim. Biophys. Acta 2011, 1813,
120−128.
3782
dx.doi.org/10.1021/jm301234k | J. Med. Chem. 2013, 56, 3768−3782