Modulating the interaction between CDK2 and cyclin A with a quinoline-based inhibitor

Article abstract of DOI:10.1016/j.bmcl.2013.11.041

A new class of quinoline-based kinase inhibitors has been discovered that both disrupt cyclin dependent 2 (CDK2) interaction with its cyclin A subunit and act as ATP competitive inhibitors. The key strategy for discovering this class of protein-protein disrupter compounds was to screen the monomer CDK2 in an affinity-selection/mass spectrometry-based technique and to perform secondary assays that identified compounds that bound only to the inactive CDK2 monomer and not the active CDK2/cyclin A heterodimer. Through a series of chemical modifications the affinity (K_d) of the original hit improved from 1 to 0.005 μM.

Full text of DOI:10.1016/j.bmcl.2013.11.041

Bioorganic & Medicinal Chemistry Letters 24 (2014) 199–203
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Modulating the interaction between CDK2 and cyclin A with
a quinoline-based inhibitor
a,
Yongqi Deng a, , Gerald W. Shipps Jr. , Lianyun Zhao a, M. Arshad Siddiqui a,à, Janeta Popovici-Muller a,§
,
⇑
Patrick J. Curran a, Jose S. Duca a,–, Alan W. Hruza b, Thierry O. Fischmann b, Vincent S. Madison b,
Rumin Zhang b, Charles W. McNemar b, Todd W. Mayhood b, Rosalinda Syto b, Allen Annis a,k
,
Paul Kirschmeier b, , Emma M. Lees c,àà, David A. Parry c, William T. Windsor b,
⇑
a Merck Research Laboratories, 320 Bent Street, Cambridge, MA 02141, USA
b Merck Research Laboratories, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
c Merck Research Laboratories, 901 California Avenue, Palo Alto, CA 94304, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A new class of quinoline-based kinase inhibitors has been discovered that both disrupt cyclin dependent
2 (CDK2) interaction with its cyclin A subunit and act as ATP competitive inhibitors. The key strategy for
discovering this class of protein–protein disrupter compounds was to screen the monomer CDK2 in an
affinity-selection/mass spectrometry-based technique and to perform secondary assays that identified
compounds that bound only to the inactive CDK2 monomer and not the active CDK2/cyclin A heterodi-
mer. Through a series of chemical modifications the affinity (Kd) of the original hit improved from 1 to
Received 21 October 2013
Revised 13 November 2013
Accepted 15 November 2013
Available online 23 November 2013
Keywords:
Kinases
Tumor
0.005 lM.
Ó 2013 Elsevier Ltd. All rights reserved.
CDK2
Modulator
High throughput screening
Multiple cellular signals can stimulate growth, differentiation,
and apoptosis. A key mechanism for regulating these processes in-
volves the cell cycle, which controls cell division by regulating pas-
sage through the G1, S, and G2/M phases of DNA synthesis and
mitosis.1,2 Progression through the eukaryotic cell cycle is con-
trolled in part by cyclin dependent kinases CDK1 (cdc2), CDK2,
CDK4 and CDK6.3 These CDK proteins are serine/threonine kinases
that regulate different phases of the cell cycle by binding to distinct
regulatory subunits called cyclins and phosphorylation of a wide
range of proteins. The temporal expression of cyclin A, cyclin B, cy-
clin D, and cyclin E, is tightly controlled to ensure successful
progression through the cell cycle.4 Loss of normal cell cycle regu-
lation is thought to contribute to human proliferative disorders;
therefore inhibition of unregulated CDK activity by small molecule
inhibitors would be beneficial in the treatment of cancer. Although
a large number of drug discovery programs have been directed to-
ward developing CDK specific ATP-competitive inhibitors, only a
few molecules have progressed into human clinical trials.5–11
Therefore, there remains a need to develop additional CDK inhibi-
tors for the treatment of human cancers.
CDK2 activation requires association with a cyclin (cyclin E at
the G1–S phase transition and cyclin A during the S phase) as well
as phosphorylation of Thr160 in the kinase activation segment. A
structural comparison between the CDK2 monomer and the
CDK2/cyclin A complex shows the activation loop moves ꢀ25 Å fol-
lowing binding of the cyclin, stabilizing the complex and providing
a structural realignment of the C-Helix to form the fully active con-
formation.12 Thus far, most inhibitor screening research efforts
have focused on screening the active (phosphorylated) form of
CDK2 complexed with either cyclin A or cyclin E. The compounds
developed using these strategies have led to ATP-competitive
inhibitors. With the intent of finding new modes of inhibition we
decided to screen unphosphorylated (inactive) CDK2 in the ab-
sence of a bound cyclin. It was hypothesized that the uncomplexed
protein might be less rigid and able to undergo conformational
⇑
Corresponding authors.
E-mail addresses: yongqi.deng@ merck.com (Y. Deng), william.windsor@merck
.com (W.T. Windsor).
Current address: The New Venture Fund, Stoneham, MA 02180, USA.
Current address: Paraza Pharma, Inc., 500 Boulevard Cartier (West), Laval, Quebec
à
H7V 5B7, Canada.
§
Current address: 38 Sidney Street, Cambridge, MA 02139-4169, USA.
Current address: Novartis Institutes for BioMedical Research, Inc., 100 Technology
–
Square, Cambridge, MA 02139, USA.
k
Current address: Aileron Therapeutics, Inc., 281 Albany St., Cambridge, MA 02139,
USA.
Current address: 106 Culberson Rd, Basking Ridge, NJ 07920, USA.
Current address: Novartis Institutes for BioMedical Research, Emeryville, CA, USA.
àà
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.11.041

200
Y. Deng et al. / Bioorg. Med. Chem. Lett. 24 (2014) 199–203
changes that would reveal new states capable of binding small
molecules and that these inhibited states might have cellular func-
tion. Multiple inhibitory mechanisms were thereby enabled;
including modulation of important protein–protein interactions
(i.e., cyclin binding), decreasing phosphorylation (activation) of
CDK2, or direct inhibition of kinase activity in ATP competitive,
noncompetitive, and uncompetitive modalities.
To identify compounds with the above inhibition mechanisms an
unbiased high-throughput screening campaign was performed
using unphosphorylated CDK2 in the affinity-selection/mass spec-
trometry-based technique called Automated Ligand Identification
System (ALIS) versus a library of 5 million compounds.13 The major-
ity of compounds that were identified from the ALIS screen were
well known ATP competitive inhibitors (i.e., purine-based) or pre-
dicted to be ATP competitive. These compounds competed with
an ATP-competitive fluorescent probe (B-Alexa-Fluor647) in a fluo-
rescence polarization binding assay and also inhibited CDK2/cyclin
A in the enzyme assay. One class of quinoline-based compounds
had the desired binding/inhibition profile which demonstrated
displacement of the fluorescent probe in the fluorescence polariza-
tion CDK2 assay but did not inhibit the CDK2/cyclin A enzyme assay
and did not bind in the CDK2/cyclin A FP assay. Although these
inhibitors did not inhibit the kinase activity of CDK2/cyclin A, it
was anticipated that due to the temporal ‘cyclic expression’ of cyclin
subunits in the cell that free CDK2 would exist as a monomer during
the course of the cell cycle and thus monomer selective inhibitors
would have the potential to be biologically active.
Compound 1 (Fig. 1) had a Ki of 0.9
l
M in the CDK2 fluorescence
polarization (FP) binding assays14 and an IC50 >10
lM in the CDK2/
cyclin A enzyme assay. Two other binding assays were used to as-
sess the binding affinity of ligands that bound preferentially to the
inactive state of CDK2. The temperature-dependent circular
dichroism assay (TdCD)15 was used to determine the affinity of li-
gands bound to proteins by measuring the degree to which a
bound ligand stabilized the thermal denaturation (change in sec-
ondary structure) of CDK2 as measured by following changes in
the circular dichroism ellipticity at 220 nm. Isothermal titration
calorimetry (ITC)16 was also used for several compounds to obtain
HO
HO
O
O
S
Cl
O
O
N
N
HN
HN
O
O
O
O
OH
OH
1
2
Figure 1. Quinoline-based compounds 1 and 2.
Table 1
Structure–activity relationships of CDK2 inhibitors based on the quinoline scaffolda,b
R3
5
8
4
R1
6
3
2
H
N
7
N
Ar
1
O
R2
Compd
R1
R2
R3
Ar
FP Ki (lM)
TdCD Kd
(l
M)
ITC Kd (lM)
1
2
7
8
9
2-Thiophen
3-Cl-phenyl
2-Cl-phenyl
4-Cl-phenyl
3,5-di-Cl-phenyl
Phenyl
Morpholine
3-CF3-phenyl
4-Cl-phenethyl
4-Cl-styryl
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
CO2Me
H
CO2H
Me
Propyl
CONHMe
CON(Me)2
5-Methyloxazole
CH2OMe
CH2OMe
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
CH2OH
CONH2
CONHSO2Me
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
COOH
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-F-phenyl
4-OMe-phenyl
3-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
4-OH-phenyl
0.90
0.14
0.7
0.9
0.2
10
2.8
<0.10
<0.10
<0.10
<0.10
1.1
0.30
0.8
0.9
0.32
12
1.00
0.30
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
2.4
0.04
0.048
0.005
0.076
>40
>40
>40
4.0
>40
>40
>40
>10
4
0.4
0.3
0.35
0.15
0.24
0.025
0.050
0.055
n-Hexyl
3-CF3-phenyl
2-Thiophen
2-Thiophen
3-CF3-phenyl
3-CF3-phenyl
3-CF3-phenyl
2-Thiophen
3-CF3-phenyl
3-CF3-phenyl
3-CF3-phenyl
4-Cl-styryl
>40
>40
>40
2.5
>40
>40
>40
12
2.5
0.3
0.10
0.16
0.20
0.1
4-Cl-styryl
3-CF3-phenyl
3-CF3-phenyl
4-Cl-styryl
<0.10
a
The CDK2/cyclin A enzyme IC50 is greater than 10
Binding activity reported for the inactive CDK2 monomer.
l
M for all of these compounds.
b

Y. Deng et al. / Bioorg. Med. Chem. Lett. 24 (2014) 199–203
201
Figure 2. X-ray crystal structure of 2/CDK2 complex superimposed on the CDK2/cyclin A complex (ATP removed). The 2/CDK2 complex is shown with the kinase in brown
and the compound in yellow. The CDK2/cyclin A complex is in blue. The yellow arrow identifies the shift in conformation for the C-Helix in the 2/CDK2 complex versus the
location of the C-Helix in the CDK2/cyclin A complex. Compound 2 binding to the kinase domain induces a conformational change that would cause steric interference in
binding cyclin A. PDB code is 4NJ3.
TdCD-Kd and ITC-Kd values determined for 1 and the quinoline de-
rived series of compounds are listed in Table 1. Interestingly, the
corresponding R isomers of 1 and its analogs show no binding
activities (data not shown).
A series of new synthesis led to 2 which contains a more elec-
tron-deficient 3-choloro phenyl appendage at R1 and has a signifi-
cant improvement in the FP-Ki to 0.14 lM, (consistent with lower
Kd values in TdCD and ITC studies). The binding mode of 2 bound to
the monomeric state of CDK2 determined by X-ray crystallography
shows 2 binds not only in the ATP active site but also extends into a
back pocket behind the gatekeeper and induces several conforma-
tional changes (Fig. 2). The Tyrosine phenol of 2 interacts with the
hinge region (Fig. 3A) and shares a hydrogen bond with the car-
bonyl of Glu81 and the NH of Leu83 in the CDK2 hinge region,
respectively. The carbonyl group of the amide bond at the 2-posi-
tion of 2 forms a hydrogen bond with the conserved Lys33 residue.
The carboxylic acid at the 4-position of the quinoline group forms a
hydrogen bond with Ala149 adjacent to the DFG motif. The DFG
motif is in the ‘In’ orientation with Phe146 buried in the core of
the protein as found for active kinase conformations. The 3-chloro-
phenyl group at the 6-position of 2 binds deeply into a hydropho-
bic pocket formed by the side chains of Leu55, Leu58, Val123 and,
Phe146. Comparison of the 2/CDK2 structure with that of the active
pThr160-CDK2/cyclin A/ATP complex (Figs. 2 and 3B, CDK2 green,
CDK2/cyclin
A blue) reveals several protein conformational
changes between the two structures. First, the activation loop is
displaced starting from the Gly147 in the DFG motif. More signifi-
cantly the distal chlorophenyl pushes against the C-terminal cap of
the C-Helix. The helix rotates such that the N-terminal end moves
by 2.0 Å. The rotation occurs in a direction opposed to the C-Helix
Figure 3. (A and B) Binding mode of 2 in CDK2. (3A) Compound 2 (yellow) is bound
to the hinge region of CDK2 through the sharing of 2 hydrogen bonds with the
carbonyl of Glu81 and the NH of Leu83. The DFG147 group is in the ‘In’ position and
the quinoline-R3 group extends over Phe146 forming an array of hydrophobic
interactions toward the C-Helix. (3B) Overlay of the structure of 2 (yellow)/CDK2
(green) with the structure of CDK2/cyclin A (blue) (Ref. 15 pdb 1FIN). An 11.7 Å shift
and rotation in Helix C is observed between the CDK2/cyclin A structure and 2/CDK2
complex. This conformational change is required to bind the R3-quinoline group
otherwise there would be steric hindrance in binding 2 to the kinase domain.
motion induced by cyclin binding such that the Glu51 Ca in the 2/
0
CDK2 complex is about 12 ÅA away from its position in the cyclin
bound kinase structure (Fig. 3B). The C-Helix partly overlaps with
cyclin in this new orientation (Fig 2). Hence the locking of the C-
Helix in this orientation and steric hindrance of binding cyclin A
provides a structural rationale as to why cyclin A cannot bind to
the pre-formed 2/CDK2 complex. At the beginning of the T-Loop,
Phe146 (of the DFG motif) overlaps in the two structures and thus
the 2/CDK2 complex has a classical DFG-In conformation.
a full thermodynamic characterization of the binding affinity to
verify the affinity measured from FP and TdCD assays. The FP-Ki,

202
Y. Deng et al. / Bioorg. Med. Chem. Lett. 24 (2014) 199–203
shows when CDK2/cyclin A is incubated at 40 °C without 2, fol-
lowed by centrifugation that both CDK2 and cyclin A remain in
the supernatant fraction because the heterodimer complex is sta-
ble at 40 °C. However, a similar incubation with increasing concen-
trations of 2 results in the loss of cyclin A (which is insoluble as a
monomer at 40 °C) from the supernatant fraction as evident by the
decrease in the corresponding Coomassie band for the supernatant
fractions and an increasing amount of cyclin A in the pellet frac-
tion. Thus, CDK2-bound 2 inhibits cyclin A binding.
The synthesis of 2 and its derivatives with variations at the 2, 4
and 6-position is summarized in Scheme 1. 5-Bromoisatin 3, was
converted to the dicarboxylic acid 4 using pyruvic acid and potas-
sium hydroxide.19 The aryl group in 5 was introduced using stan-
dard coupling conditions with the corresponding boronic acid.
The dicarboxylic acid 5 was then activated by converting to di-pen-
tafluorophenol (PFP) ester 6 in the presence of DCC. It was found
that the PFP ester at the 2-position of the quinoline was more reac-
tive than at the 4-position and reacting di-PFP ester 6 with one
equivalent of tyrosine methyl ester, followed by hydrolysis, gener-
ated 2 with ꢀ60% yield. A similar approach yielded 7–30.
The SAR for the analogs of 2 proved to be very consistent with
these structural observations. Substitutions at the 6-position (R1)
of the quinoline showed that a meta substituent on the phenyl ring
was preferred and that limited space around the ortho and para-
positions was available due to steric hindrance from Phe146 and
Leu 58 (7, 8). The un-substituted phenyl group reduces the potency
drastically (10). The R1 group binds in a hydrophobic environment,
and introduction of polar groups such as morpholine resulted in
less potent compounds (11). Hydrophobic substitutions at R1 (12,
3-CF3-phenyl) or extended hydrophobic substitutions (13 and 14,
4-Cl-phenylethyl and 4-Cl-styryl, respectively) were favorable
and resulted in single digit nM binding as determined by TdCD
(Kd = 5 nM for 14). These substitutions are keys for differentiating
the binding of the quinoline compounds to CDK2 and not CDK2/cy-
clin A. In the CDK2/cyclin A binary complex, no binding pocket ex-
ists to accommodate the quinoline scaffold and substitutions at the
6-position (R1). In the structure of CDK2 bound with 2 the C-Helix
is pushed 11.7 Å outward which provides room for the compound
binding but sterically inhibits the binding of cyclin A to CDK2.
The interaction of the quinoline 4-position carboxylic acid with
the backbone of the protein proved to be very important. Reduc-
tion of the carboxylic acid to the hydroxymethyl group (16), con-
version to an amide (17) or to an acylsulfonamide (18)
completely abolished binding. In 2, the hydroxyl group of the phe-
nol (AR substitution) functions as both a hydrogen bond donor and
an acceptor and shares two key hydrogen bond interactions with
the hinge amide backbone (Glu81, Leu83). Replacement of the hy-
droxyl group with a fluorine group was 100 fold less potent (19).
Methylation of the same hydroxyl group or relocating to a meta-
position on the tyrosine ring resulted in complete loss of binding
(20 and 21). The methyl ester at R2 was found to be essential for
binding since 22 (R2 = H) and 23 (R2 = COOH) bound poorly to
CDK2. In anticipation of potential metabolic liabilities with the
methyl ester (R2), we decided to investigate its replacement with
more stable functional groups. The crystal structure of 2 with
CDK2 revealed a hydrophobic interaction between the methyl es-
ter moiety and the side chain of Leu 134. This position tolerated
a wide range of substitutions including alkyl (25), amide (26 and
27), heterocyclic (28) and methoxy–methyl groups (29 and 30).
The chemical and structural SAR described above clearly defines
the observed affinity SAR and why the quinoline series bind to the
CDK2 monomer and not the CDK2/cyclin A heterodimer. Despite
the improvement of binding affinity, none of the compounds de-
scribed above show significant activity in a cell based proliferation
assay or inhibition of pRb phosphorylation while CDK2 control
compounds were active in these assays. Several reasons may exist
Figure 4. CDK2/cyclin A complex dissociation by 2 at 40 °C. CDK2/cyclin A at 1.5 lM
was incubated with increasing concentrations of 2 at rt for 30 min and then heated at
40 °C for 30 min. The samples were centrifuged and the supernatant fractions run on
an SDS–PAGE and stained with Coomassie stain (Fig. 4-top). The loss of cyclin A
fractions at the higher concentrations of 2 is an indication of disruption of the CDK2/
cyclin
A complex. Figure 4-bottom is the centrifugation pellet fraction and
demonstrates that cyclin A becomes dissociated in the presence of increasing 2
and is retained in the pellet fraction due to poor thermal stability at 40 °C. Sample C is
a control sample with only 1% DMSO run at 40 °C under the same conditions as the
compound treated samples. The final DMSO in all samples was 1%.
Compound 2 has been shown to bind to CDK2 by direct compe-
tition of the ATP-competitive fluorescently labeled compound (B-
Alexa-Fluor647), binding in the TdF assay where bound 2 increases
the thermal stability of CDK2, and binds in the isothermal titration
calorimetry assay with a Kd = 0.3 lM. The unique binding mode of
2 with CDK2 suggests that 2 and its analogues may be selective
against other kinases. Indeed counter-screening 2 against a panel
of 310 kinases at 1 lM resulted in only one kinase with greater
than 20% inhibition (PDK1, 30% inhibition) while control kinase
inhibitors demonstrated the expected inhibition profile. No inhibi-
tion was observed for the active state of CDK1, CDK2, CDK5, CDK7,
CDK8, and CDK9. We cannot rule out that compound 2 binds to the
inactive (noncyclin bound) states of the CDK family. Binding stud-
ies would be needed to assess if compound 2 could bind to the
inactive states of these CDKs. The bound conformation of 2 to
CDK2 is similar to that observed for Lapatinib bound to EGFR (Type
I½) where these compounds bind into the core of the protein to-
ward the C-Helix but without inducing a DFG-out conformation.17
It is interesting to note that Betzi et al.18 have recently observed
that two 8-anilino-1-naphthalene sulfonate (ANS) hydrophobic
fluorescent probes (Kd was 37 lM) can also bind into the same re-
gion as the quinoline core and the R3 group and induces a similar
C-Helix conformational change. Selected compounds were also
evaluated in a cell based proliferation assay but none showed cell
proliferation inhibition possibly due to permeability issues due to
the R3 carboxylic acid group. Substitution of the ionic R3 carboxylic
acid group with bioisosteres that retain enzyme activity may im-
prove permeability and cellular activity.
Evidence that 2-induced conformational change prevents cyclin
binding to CDK2 was obtained by incubating the compound with
CDK2/cyclin A and measuring the amount of soluble cyclin A after
incubating at 40 °C. Figure 4 is a Coomassie stained SDS gel that

Y. Deng et al. / Bioorg. Med. Chem. Lett. 24 (2014) 199–203
203
CO2H
CO2H
O
R1
Br
(ii)
Br
(i)
O
N
N
CO2H
N
CO2H
H
5
3
4
CO2H
R1
CO2PFP
R1
H
N
(iv) (v)
(iii)
N
O
Ar
N
CO2PFP
R2
6
Scheme 1. Reagents and conditions : (i) Pyruvic acid, KOH, 40 °C, 12 h, 95%; (ii) 3-chlorophenyl boronic acid, Pd(AcO)2, K3PO4, dioxane, water, 80 °C, 12 h, 85%; (iii)
pentafluoro-phenol, DCC, rt, 4 h, 70%; (iv) tyrosine methyl ester, DIEA, 0 °C, 12 h; (v) water, DIEA, rt, 60%.
Fischmann, T.; Wang, Y. L.; Oft, M.; Chen, T. Y.; Kirschmeier, P.; Lees, E. M. Mol.
Cancer Ther. 2010, 9, 2344.
from the carboxylic acid functionality, difficulty of the current
11. Lapenna, S.; Giordano, A. Nat. Rev. Drug Discovery 2009, 8, 547.
for the lack of cellular activity such as poor permeability resulting
inhibitors to compete with cyclin binding to the monomeric
CDK2 state (thus cyclin A would shift the equilibrium from free
CDK2 to a CDK2/cyclin A complex incapable of binding compound
2) or a combination of both. More experiments are needed to fur-
ther understand how cyclin A regulates the activity of CDK2 and to
improve the cell activity of this class of compounds.
In conclusion, we have identified and developed a series of com-
pounds based on a quinoline scaffold which bind selectively to the
CDK2 monomer and not the CDK2/cyclin A complex. The binding of
the quinoline series has the potential functional consequence of
not only inhibiting ATP binding to the active site but also the dis-
ruption of binding cyclins due to the conformational change in the
critical C-Helix.
12. Heitz, F.; Morris, M. C.; Fesquet, D.; Cavadore, J. C.; Doree, M.; Divita, G.
Biochemistry 1997, 36, 4995.
13. Annis, D. A.; Nickbarg, E.; Yang, X.; Ziebell, M. R.; Whitehurst, C. E. Curr. Opin.
Chem. Biol. 2007, 11, 518.
14. Zhang, R. M.; Mayhood, T.; Lipari, P.; Wang, Y. L.; Durkin, J.; Syto, R.; Gesell, J.;
McNemar, C.; Windsor, W. Anal. Biochem. 2004, 331, 138. The reference
compound BMS-250595 (BMS) is an ATP competitive compound that binds to
CDK2 with a Kd = 9 nM. The deacetylated form of compound BMS was reacted
with Alexa Fluor 647 to generate a compound (B-Alexa-Fluor647) that bound
to both proteins and provided a fluorescence polarization signal that could be
used to measure the affinity of compounds from the high throughput screening
study. The binding assay was performed in 10 mM HEPES pH 7.5, 150 mM
NaCl, 10% Glycerol, 1 mM DTT, 0.01% BOG with 200 nM CDK2 or 450 nM CDK2/
cyclinA. The affinity of the fluorescence probe B-Alexa-Fluor647 was ꢀ200 nM
to both proteins. Ki values of test compounds were determined by adding
increasing concentrations of unlabeled test compound to solutions containing
50 nM B-Alexa-Fluor647 and CDK2 or CDK2/cyclin
A solutions. This
displacement assay provided accurate affinities for inhibitors greater than
ꢀ100 nM. For compounds with a Ki of <100 nM other complimentary assays
were utilized.
Acknowledgments
15. Mayhood, T. W.; Windsor, W. T. Anal. Biochem. 2005, 345, 187. Briefly, CDK2,
CDK2/cycin A or cyclin A were at ꢀ0.2 mg/ml in 10 mM Hepes, 150 mM NaCl,
10% glycerol, and 1 mM DTT (pH 7.5) in a 0.1-cm cell. Secondary structure
unfolding was measured using a six-cell Peltier temperature-controlled Jasco
810 spectropolarimeter at a wavelength of 220 nM. A typical run time was
approximately 3 h using six samples for temperature scans between 20 and
80 °C (ꢀ0.33 °C/min). Experiments performed at 1 °C/min gave similar results
We would like to thank Drs. Huw Nash, and Hung Le for their
helpful discussions.
References and notes
1. Malumbres, M.; Barbacid, M. Nat. Rev. Cancer 2001, 1, 222.
2. Pavletich, N. P. J. Mol. Biol. 1999, 287, 821.
3. Murray, A. W. Cell 2004, 116, 221.
for
DTm values with and without ligand.
16. ITC was used to obtain a full thermodynamic characterization of the binding
affinity for several lead compounds. The procedure is similar to that reported
by Mayhood and co-workers in Ref. 14. Briefly, human CDK2 was dialyzed in to
10 mM Hepes, 150 mM NaCl, 10% glycerol, 1 mM DTT, pH 7.5 overnight with 2
changes of buffer. Both protein and ligand samples were prepared in 2 ml
4. Morgan, D. O.; Roberts, J. M. Nature 2002, 418, 495.
5. Bible, K. C.; Lensing, J. L.; Nelson, S. A.; Lee, Y. K.; Reid, J. M.; Ames, M. M.; Isham,
C. R.; Piens, J.; Rubin, S. L.; Rubin, J.; Kaufmann, S. H.; Atherton, P. J.; Sloan, J. A.;
Daiss, M. K.; Adjei, A. A.; Erlichman, C. Clin. Cancer Res. 2005, 11, 5935.
6. McClue, S. J.; Blake, D.; Clarke, R.; Cowan, A.; Cummings, L.; Fischer, P. M.;
MacKenzie, M.; Melville, J.; Stewart, K.; Wang, S.; Zhelev, N.; Zheleva, D.; Lane,
D. P. Int. J. Cancer 2002, 102, 463.
volumes. To prepare the ligand solution 20
DMSO) was added to 2 ml of buffer. The ITC runs were performed at 20 °C
using approximately 10–15 M CDK2 and 100–150 M compound and mixed
at a rate of 350 rpm. Injection volumes varied from 15 to 18 l and 360 s of
ll of a 10 mM compound stock (in
l
l
l
7. Misra, R. N.; Xiao, H. Y.; Kim, K. S.; Lu, S.; Han, W. C.; Barbosa, S. A.; Hunt, J. T.;
Rawlins, D. B.; Shan, W.; Ahmed, S. Z.; Qian, L.; Chen, B. C.; Zhao, R.; Bednarz, M.
S.; Kellar, K. A.; Mulheron, J. G.; Batorsky, R.; Roongta, U.; Kamath, A.; Marathe,
P.; Ranadive, S. A.; Sack, J. S.; Tokarski, J. S.; Pavletich, N. P.; Lee, F. Y.; Webster,
K. R.; Kimball, S. D. J. Med. Chem. 2004, 47, 1719.
8. Chu, X. J.; DePinto, W.; Bartkovitz, D.; So, S. S.; Vu, B. T.; Packman, K.; Lukacs, C.;
Ding, Q.; Jiang, N.; Wang, K.; Goelzer, P.; Yin, X.; Smith, M. A.; Higgins, B. X.;
Chen, Y.; Xiang, Q.; Moliterni, J.; Kaplan, G.; Graves, B.; Lovey, A.; Fotouhi, N. J.
Med. Chem. 2006, 49, 6549.
data was recorded for each titration. When a clear baseline was not observed a
blank titration was performed to subtract it from the raw data. The MicroCal
MCS ITC and included software were used for performing the titrations and
analysis.
17. Wood, E. R.; Truesdale, A. T.; McDonald, O. B.; Yuan, D.; Hassell, A.; Dickerson,
S. H.; Ellis, B.; Pennisi, C.; Horne, E.; Lackey, K.; Alligood, K. J.; Rusnak, D. W.;
Gilmer, T. M.; Shewchuk, L. Cancer Res. 2004, 64, 6652.
18. Betzi, S.; Alam, R.; Martin, M.; Lubbers, D. J.; Han, H. J.; Jakkaraj, S. R.; Georg, G.
I.; Schonbrunn, E. ACS Chem. Biol. 2011, 6, 492.
9. 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. J. Med. Chem. 2005, 48, 2388.
10. 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.;
19. Deng, Y.; Curran, P. J.; Shipps, G. W., Jr.; Zhao, L.; Siddiqui, M. A.; Popovici-
Muller, J.; Duca, J. S.; Hruza, A. W.; Fischmann, T. O.; Madison, V. S.; Zhang, R.;
McNemar, C. W.; Mayhood, T. W.; Windsor, W. T.; Lees, E. M.; Parry, D. A. U.S.
Patent 7,511,063, 2009.