8-substituted O 6-cyclohexylmethylguanine CDK2 inhibitors: Using structure-based inhibitor design to optimize an alternative binding mode

Article abstract of DOI:10.1021/jm401555v

Evaluation of the effects of purine C-8 substitution within a series of CDK1/2-selective O⁶-cyclohexylmethylguanine derivatives revealed that potency decreases initially with increasing size of the alkyl substituent. Structural analysis showed

Full text of DOI:10.1021/jm401555v

Article
pubs.acs.org/jmc
8‑Substituted O6‑Cyclohexylmethylguanine CDK2 Inhibitors: Using
Structure-Based Inhibitor Design to Optimize an Alternative Binding
Mode
Benoit Carbain,⊥,‡ David J. Paterson,⊥,† Elizabeth Anscombe,† Allyson J. Campbell,‡ Celine Cano,‡
Aude Echalier,†,▽ Jane A. Endicott,*,†,# Bernard T. Golding,‡ Karen Haggerty,‡ Ian R. Hardcastle,‡
Philip J. Jewsbury,§ David R. Newell,∥ Martin E. M. Noble,†,# Celine Roche,‡ Lan Z. Wang,∥
and Roger J. Griffin*,‡
†Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, U.K.
‡Newcastle Cancer Centre, Northern Institute for Cancer Research, School of Chemistry, Newcastle University, Bedson Building,
Newcastle upon Tyne NE1 7RU, U.K.
§AstraZeneca Pharmaceuticals, Alderley Park, Cheshire SK10 4TG, U.K.
∥Newcastle Cancer Centre, Northern Institute for Cancer Research, Newcastle University, Paul O’Gorman Building, Medical School,
Framlington Place, Newcastle upon Tyne NE2 4HH, U.K.
S
* Supporting Information
ABSTRACT: Evaluation of the effects of purine C-8 substitution within a
series of CDK1/2-selective O6-cyclohexylmethylguanine derivatives revealed
that potency decreases initially with increasing size of the alkyl substituent.
Structural analysis showed that C-8 substitution is poorly tolerated, and to
avoid unacceptable steric interactions, these compounds adopt novel binding
modes. Thus, 2-amino-6-cyclohexylmethoxy-8-isopropyl-9H-purine adopts a
“reverse” binding mode where the purine backbone has flipped 180°. This
provided a novel lead chemotype from which we have designed more potent
CDK2 inhibitors using, in the first instance, quantum mechanical energy
calculations. Introduction of an ortho-tolyl or ortho-chlorophenyl group at the
purine C-8 position restored the potency of these “reverse” binding mode inhibitors to that of the parent 2-amino-6-
cyclohexylmethoxy-9H-purine. By contrast, the corresponding 8-(2-methyl-3-sulfamoylphenyl)-purine derivative exhibited
submicromolar CDK2-inhibitory activity by virtue of engineered additional interactions with Asp86 and Lys89 in the reversed
binding mode, as confirmed by X-ray crystallography.
CDK inhibitor has yet to be approved by the FDA for clinical
use.7,10,11
INTRODUCTION
■
Deregulation of the cell cycle is a vital motif in most, if not all,
human malignancies.1,2 Phosphorylation of pRb, the product of
the retinoblastoma gene, by cyclin-dependent kinase (CDK)
family members, is required for G1 progression, and mutation of
the pRb gene or abnormalities in the pRb signaling pathway are a
common feature of cancer. For example, overexpression and/or
mutation of CDK4 and cyclin D has been reported in various
tumor types, and mutation of the p16INK4A gene that results in its
deletion or silencing is one of the most frequent genetic
alterations found in human tumors.3,4 Overexpression of cyclin
E and reduced levels of p27Kip1 have also been observed to
correlate with disease progression,5,6 suggesting that deregulation
at various checkpoints throughout the cell cycle is a key
contributor to the process of tumorigenesis. As a result of these
studies, the CDK family is considered an important target for
therapeutic intervention in cancer and other diseases, and a wide
range of ATP-competitive inhibitors have been developed.7−9
However, although a number of small-molecule CDK inhibitors
have progressed to clinical trials (e.g., 1−4), a kinase-specific
Initial enthusiasm for CDK2 as a cancer therapeutic target was
tempered by subsequent studies demonstrating that CDK2
siRNA or antisense constructs failed to induce cell cycle arrest in a
number of human tumor cell lines12 and that CDK2 knockout
mice are viable.13,14 However, more recent evidence suggests that
CDK2 inhibition may elicit antitumor activity in a subset of
tumors with defined genetic characteristics. For example, CDK2
knockdown or treatment with roscovitine (1) is selectively toxic
to MYCN-amplified neuroblastoma cells compared with their
nonamplified counterparts.15 Compelling evidence to support a
therapeutic role for pharmacological CDK2 inhibition in cancer
treatment has also been adduced very recently by employing a
chemical genetic approach.16 Thus, a comparison of selective
small-molecule CDK2 inhibition with that imposed by siRNA
knockdown in normal and tumor cells highlighted important
Received: August 20, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
complex with monomeric unphosphorylated CDK2 revealed that
the purine heterocycle binds within the ATP-binding pocket in a
different orientation from that previously reported for ATP,
isopentenyladenine, and the prototypic purine CDK inhibitor
olomoucine (8).23 The 9-NH group of 5 forms a hydrogen-bond
with the backbone carbonyl of Glu81, while the purine N-3 and 2-
amino group make hydrogen bonds with the backbone nitrogen
NH and carbonyl of Leu83, respectively. In this orientation, the
C-8 position of 5 is directed toward the side-chain of Phe80 in the
gatekeeper pocket (Figure 1). A number of purine-based CDK
inhibitors derive potency from interacting with this residue; for
example, purvalanol B (9)24 and H717 (10)25 both pack
hydrophobic aliphatic substituents against the Phe80 side chain.
In contrast, a series of bisanilinopyrimidines (e.g., 11) abut polar
and polarizable groups against the phenyl ring of Phe80, and the
observed gain in potency for this compound series is attributed to
induced dipole−dipole interactions.26,27
The close proximity of the imidazole ring of 5 to the Phe80 side
chain observed in the structure of the inhibitor in complex with
CDK2 implied that substitution at the purine C-8 position would
not be tolerated. To confirm this prediction, and as part of SAR
studies for this chemotype, three derivatives of 5 bearing alkyl
groups at the 8-position (24−26) were initially synthesized, and
the crystal structure of the 8-isopropylpurine (26) bound to
CDK2 was determined. Purine 26 was found to adopt a different
binding orientation from that observed for 5 within the ATP-
binding domain, offering the opportunity to exploit alternative
inhibitor−protein interactions through judicious structural
modifications to the parent purine 5. In this paper, we describe
the elaboration of this binding mode through iterative compound
synthesis, molecular modeling, and X-ray crystallographic
analysis of CDK2/cyclin A/inhibitor complexes. These studies
have achieved improved CDK2-inhibitory activity by virtue of an
engineered additional CDK2−inhibitor interaction.
differences between these approaches, with pharmacological
inhibition resulting in growth inhibition of a panel of human
cancer cells transformed with various oncogenes. A recent study
demonstrating that a combination of phosphatidylinositol-3-
kinase (PI3-K) and CDK2 inhibitors induced apoptosis in
malignant glioma xenografts via a synthetic−lethal interaction
also highlights the potential of CDK2-selective inhibitors.17
CHEMISTRY
■
The initial target 8-substituted O6-cyclohexylmethylguanine
derivatives (24−29, 32) were readily synthesized following the
proceduresillustratedinScheme1.Thus,acylationofthe6-amino
group of 2,6-diamino-4-cyclohexylmethoxy-5-nitrosopyrimidine
(17)18 afforded the 6-acylaminopyrimidines (18-23), which were
readily cyclized to the required purines (24−29) in good overall
yields on reduction of the 5-nitroso substituent by catalytic
hydrogenation. Efforts to prepare the 8-trifluoromethylpurine 32
by this approach were unsuccessful, with 17 proving refractory to
We have previously reported the structure-based design of a
series of potent CDK2-selective inhibitors derived from the lead
compound O6-cyclohexylmethylguanine (NU2058, 5; CDK2
IC50 = 16.0 μM), as exemplified by 6 and 7 (CDK2; IC50 = 1.0 μM
and 5.0 nM, respectively).18−22 The X-ray crystal structure of 5 in
Figure 1. (A) Purine 5 bound to the CDK2 ATP-binding domain. The kinase is presented in dark green in ribbon representation, and selected CDK2
residues and 5 are drawn in ball-and-stick mode. Carbon atoms of 5 are colored light green, and hydrogen bonds are denoted by dashed lines. (B)
Schematic representation of the bonds madebetween CDK2and 5. Hydrogen bondsare drawn as arrows. The double-headed arrow denotes the distance
(Å) between the purine C8 atom and the side-chain of Phe80.
B
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1
a
Reagents and conditions: (i) 18−21, (RCO)2O, DMF, 80 °C, 22 and 23, ArCOCl, pyridine, DMAP, 25 °C; (ii) H2, Pd/C, AcOH, THF, 20 °C;
(iii) TFA, reflux; (iv) DMAP, pyridine, reflux.
a
Scheme 2
a
Reagents and conditions: (i) (a) TsCl, pyridine, 0−25 °C, (b) PMB-Cl, DMAP, 25 °C; (ii) bis(pinacolato)diboron, PdCl2[[(C4H9)2PC5H4]2Fe],
[(C4H9)2PC5H4]2Fe, KOAc, dioxane, MW 150 °C; (iii) PMB-Cl, NaH, DMF, 80 °C (13) 25 °C (14); (iv) BnBr, K2CO3, DMF, 25 °C; (v)
Pd(OAc)2, CuI, CsCO3, DMF, MW 200 °C; (vi) H2, Pd/C, AcOH, aq. HCl, 20 °C; (vii) (tBuO)2CO, K2CO3, DMF, 20 °C; (viii) pyridinium
tribromide, DCM, 70 °C; (ix) NBS, MeCN, 20 °C; (x) appropriate arylboronate, PdCl2[[(C4H9)2PC5H4]2Fe], Cs2CO3, aq dioxane, 110 °C; (xi)
TFA, 20−110 °C.
acylation with TFA-anhydride or ethyl trifluoroacetate. However,
direct acylation of triaminopyrimidine 30 with TFA gave the
required trifluoroacetylaminopyrimidine 31, which slowly
cyclized to the requisite purine 32 on prolonged heating in
pyridine-DMAP.
benzoic acid 37 was achieved in moderate yield by catalytic
hydrogenation under acidic conditions. An alternative approach
utilized the 8-bromopurine 39, readily synthesized by treatment
of the t-BOC-protected purine derivative 38 with N-bromosucci-
nimide, which concomitantly removed the protecting group, or
via direct bromination of 5 with pyridinium tribromide. Suzuki−
Miyaura cross coupling of 39 with the appropriate boronate
derivatives(15,16), preparedfromthecorrespondingarylhalides
(12, 14) by standard methods, or generated in situ from 13, gave
the PMB-protected compounds (40−42), which were smoothly
deprotected to the required 8-substituted purines 43−45 on
treatment with TFA.
The preparation of O6-cyclohexylmethylguanine derivatives
bearing substituted aryl groups at the purine 8-position required
palladium-catalyzed cross-coupling methodology (Scheme 2).
Direct 8-arylation of the N9-benzylpurine 33 with the appropriate
aryl bromides, employing CuI−Pd(OAc)2 under microwave
irradiation, furnished benzyl-protected intermediates 34 and 35,
and subsequent conversion into the requisite phenol 36 and
C
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 1. Inhibition of CDK1 and CDK2 by C8-Substituted O6-Cyclohexylmethylguanines
a
IC50 (μM) or % inhibition
no.
structure
R
R′
CDK2/cyclin A CDK1/cyclin B
b
5
A
B
H
H
16 2.0
26 3.5
10.0 1.6
0.25 0.05
6
0.97 0.03
0.005 0.001
44.7 3.2
7
B
SO2NH2
Me
Et
c
24
25
26
27
28
29
32
36
37
43
44
45
A
A
A
C
C
C
A
C
C
C
C
C
7%, 11%
30.3 1.5
77.5
i-Pr
H
27.7 3.2
92.3 4.3
c
c
H
H
H
10%
35%
Cl
17 1.8
>100
d
Me
CF3
Me
Me
Me
H
18.2 4.3
93.6
e
49 2.6 %
>100
>100
>100
OH
23.5 4.8
>100
CO2H
d
CONH2
SO2NH2
SO2NH2
39.8 13.9
8.7 3.4
62%
>100
d
Me
0.51 0.05
2.8
a
b
c
IC50 values determined as described in ref 20. Standard deviations obtained from three separate experiments. Activity determined at 10 μM owing
d
e
to limiting solubility. Single determination. Activity measured at 100 μM.
Figure2. (A)Stereoviewof26incomplexwithCDK2. Thekinaseispresentedindarkgreeninribbonrepresentation, andselectedCDK2residuesand26
aredrawninball-and-stickmode. Carbonatomsof26arecoloredlightgreen,andhydrogenbondsaredenotedbydashedlines.Watermoleculesaredrawn
as red spheres. (B) Schematic representation of the bonds made between CDK2 and 26. Hydrogen bonds are drawn as arrows. The double-headed arrow
denotes the distance (Å) between the purine C2 substituent and the side chain of Phe80.
a highly selective CDK2 inhibitor that was developed by a
structure-based inhibitor design approach. The markedly
increased potency of 7 (CDK2; IC50 = 5 nM) compared with 5
(CDK2; IC50 = 16 μM) is attributed to additional favorable
interactions between the sulfonamide function of 7 and the
specificitysurfaceoftheATP-bindingdomainofthekinase.20 The
RESULTS AND DISCUSSION
■
Theinhibitoryactivity ofthepurinederivatives againstCDK1and
CDK2 is shown in Table 1. Our previous studies with O6-
alkylguanines were founded on the benchmark O6-alkylguanine
NU2058 (5) and culminated in the identification of NU6102 (7),
D
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. (A) Stereoview of 32 in complex with CDK2/cyclin A. The kinase is presented in dark green in ribbon representation, with selected CDK2
residues and 32 drawn in ball-and-stick mode. Carbon atoms of 32 are colored light green, and hydrogen bonds are denoted by dashed lines. Water
molecules are drawn as red spheres. (B) Schematic representation of the bonds made between CDK2 and 32. Hydrogen bonds and polar interactions are
drawn as arrows. The double-headed arrow denotes a potential hydrophobic interaction between the purine C8 substituent and the side chain of Phe80.
primary sulfonamide function of 7 also provided an opportunity
to develop a water-soluble prodrug, by virtue of N-acetylation and
formulation as the potassium salt, and in vivo studies are in
progress.28 With a view to further delineating SARs for this
interestingclass ofkinaseinhibitor, the effect ofsubstitution atthe
C-8 position of 5 was investigated.
1), the new orientation adopted by 26 can be described by a 180°
rotation of the purine ring about a vector coincident with the C-6
to N-3 bond. In this orientation, C-8 of 26 is directed toward the
“selectivity surface” of the kinase,20,29 with the purine 2-amino
group being projected toward Phe80. Thus, both 5 and 26 form a
conserved triplet of hydrogen bonds with the CDK2 hinge region
and the O6-cyclohexylmethyl substituent is retained in the ATP
ribose-binding pocket. In both cases, the conformation and
location of the O6-cyclohexylmethyl group is analogous to that
previously observed for the structure of 7 bound to CDK2/cyclin
The crystal structure of 5 bound to CDK2/cyclin A indicated
that 8-substitution would not be well tolerated, as a result of a
steric clash between the C-8 moiety and the side-chain of Phe80
(rC8−F80Cβ = 3.7 Å in the CDK2/5 structure).20 This observation
appeared to be confirmed by the CDK2-inhibitory activity of the
three initial derivatives, which showed a modest decrease in
potency compared with 5 on addition of a methyl (24), ethyl,
(25), or isopropyl (26) group at the purine 8-position (IC50
values of 45, 30, and 28 μM, respectively), albeit that 25 and 26
were essentially equipotent with 24 despite steric differences
(Table 1). Where measurable, activity against CDK1 compared
with CDK2 also generally tracked that observed for 5, with 25 and
26 proving some 2−3 times less potent against CDK1. Notably, a
dramatic reduction in potency against both kinases was observed
upon introduction of the bulky C-8 phenyl group, although a
precise determination of the activity of 27 was limited by very
poor aqueous solubility.
Crystal Structure Analysis. To establish the binding mode
of this inhibitor series within the ATP pocket, X-ray crystallo-
graphic analysis of the 8-isopropyl- and 8-trifluoromethyl-purines
(26 and 32) in complex with CDK2 was undertaken by following
crystal soaking with each inhibitor. Interestingly, 26 was found to
adopt a novel orientation within the ATP-binding site distinct
fromthatobservedfortheparentCDK2inhibitor5(Figure2). By
comparison with the “conventional” binding mode of 5 (Figure
A
20 (Supporting Information Figure S1).
Additional interactions of 26 with the glycine-rich loop of
CDK2 contribute to stabilization of the activation segment, as
previously observed for the structure of 5 bound to CDK2.18
These interactions are predominantly lipophilic in nature, arising
between the O6-cyclohexyl ring of 5 and 26 and a hydrophobic
patch on the glycine loop contributed by the side chain of residue
Val18 and the peptide backbone around residue Gly11. The
binding conformation adopted by 26 has not been reported
previously despite extensive structural analyses of various
substituted purine-based CDK ATP-competitive li-
gands.18−20,22−24,30,31
ToassesswhethersmallerC-8substituentswouldalsopromote
this “reverse” binding mode, the 8-trifluoromethylpurine 32 was
cocrystallized with CDK2/cyclin A phosphorylated on CDK2
Thr160(CDK2/cyclin A). Previous comparative crystallographic
studies of inhibitor binding to CDK2 and to CDK2/cyclin A have
shown that the interactions made between inhibitors within this
series, and the two alternative forms of CDK2 within the adenine
binding pocket, are conserved. Surprisingly, 32 did not share the
same binding mode as 26, but rather adopted another that is
E
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
reminiscent of, but distinct from, that of 5 (Figure 3). In contrast
to the 8-isopropyl group of 26, the 8-trifluoromethyl substituent
of 32 was better accommodated, although this group is regarded
as having a similar size to isopropyl.32 The purine of 32 was found
tobindinanearidentical orientationtothatof 5, albeitlessdeeply
within the CDK2 ATP-binding pocket. As a consequence, both
the structure of the glycine-rich loop and the nature of the
hydrogen bond interactions between 32 and the residues within
the CDK2 hinge are modified. Most notably, the characteristic
hydrogen bond interaction between the purine 9-NH and the
backbone carbonyl moiety of Glu81 observed for 5 and 7 was
absent with 32. N-2 is positioned to enable hydrogen bonding
with the carbonyl group of Leu83 or, for this inhibitor series, to
form a novel interaction with the backbone carbonyl of His84.
The 9-NH of 32 forms two indirect hydrogen bonds with CDK2
via a water molecule, which is also engaged in a polar contact with
one of the fluorine atoms of the C-8 trifluoromethyl group that is
packed against the Phe80 aromatic ring (r(CF3-aromF80) = 4.2
Å). A second fluorine atom mediates a network of polar contacts
with two water molecules that in turn interact with the side chains
of Glu51 and Lys33 and the amide group of Asp145.
The CDK-inhibitory activity (Table 1) of the C-8 substituted
purines (24−27, 32) was reassessed in light of the structural
biologyresults described above. Thechangesinpotencyobserved
against both CDK1 and CDK2 that accompany C-8 alkyl
substitution, compared with the parent purine 5, can be modeled
as occurring in two phases as the size of the alkyl substituent
increases. First, the initial decrease in potency is consistent with
the formation of suboptimal interactions imposed to accom-
modate small alkyl groups such as methyl (24) at the purine C8
position. As illustrated by the structure of 32 in complex with
CDK2/cyclin A (Figure 3), the inhibitor is forced away from
Phe80 and, consequently, the requisite triplet of hydrogen bonds
with the hinge cannot be maintained. Further small adjustments
of the ATP site residues are also necessary to accommodate the
inhibitor, and taken together, these changes lead to a decrease in
binding affinity compared with 5 as reflected by CDK1/CDK2
inhibitory activity in the 100 μM range for 32. Increasing the size
of the C-8 substituent further results in an unacceptable steric
clash with Phe80, and the inhibitor is forced to adopt an
alternativebindingmodewheretheC-8groupfacestheselectivity
surface, as observed with 26 (Figure 2).
Replicating the N-2 Aryl Group Interactions of 6 and 7
with CDK2 for C-8 Substituted Purines. The unique binding
orientation adopted by the 8-isopropylpurine (26) offered an
opportunity to design novel CDK2 inhibitors based on the 6-
alkoxypurine template. Accordingly, we sought to exploit this
finding by applying the structure-aided design principles
employed previously for the optimization of the 2-arylaminopur-
ines 6 and 720 for the identification of more potent derivatives of
26. A key feature of the binding of 6 and 7 to CDK2 is the packing
of the N-2 aryl ring against the CDK2 surface to form a π−π
stacking interaction with the peptide backbone between Gln85
andAsp86.20 Notably, thearylringappearstohavealowbarrierto
rotation about the N-2 to C-1′ bond, as demonstrated by the
different purine−aryl interplane torsion angles of these and other
derivatives.20,22
A structural superposition of 7 and 26 in complex with CDK2
and CDK2/cyclin A,20 respectively (Supporting Information
Figure S1), indicated that to replicate the aryl−backbone
interactions made between 7 and CDK2, the aryl ring would
need to be attached directly to the C-8 position, a compound that
is most closely modeled in this series by the 8-phenylpurine 27.
However, the weak CDK2-inhibitory activity of 27 suggested that
if, as predicted, the compound binds to CDK2 in the reverse
binding mode, the requisite additional interactions are not arising
between the 8-phenyl ring and the kinase. One possible
explanation for this observation is that the purine and phenyl
rings of 27 are essentially coplanar as a consequence of a favorable
interaction of their π-systems, which would preclude the
conformation necessary to facilitate an optimal interaction
between the 8-phenyl ring and CDK2. Indeed, evidence in
support of this is provided by the crystal structure of the aloisine
derivative RP90 (6-phenyl[5H]pyrrolo[2,3-b]pyrazine) in com-
plex with CDK2, where the pyrrole and 6-phenyl rings are found
to adopt a coplanar conformation.33
Withaviewtooptimizingtheinteractionofthe8-phenylringof
27 with the backbone of the ATP-binding domain of CDK2, the
possibility of imposing a favorable conformational twist between
the imidazole and phenyl rings, by virtue of ortho-substitution on
the 8-phenyl ring, was considered. To this end, quantum
mechanical calculations were employed to elucidate the
conformation-energy profiles for nine derivatives of 27 to
establish whether a noncoplanar conformation was energetically
favorable, and to identify substituents with the potential to confer
a nonplanar arrangement between the tworing systems. Theideal
ortho-substituent on the 8-phenyl ring of 27 would induce an
interplane twist of approximately 40°, with relatively free rotation
arising about the optimum conformation of 20°. This latter
aspect would allow for optimization of aryl ring packing against
the kinase. Such a twist was predicted to be tolerated based on
literature evidence, where relatively large variations of torsion
angles between nonconjugated ring systems have been observed
for a number of purine-based inhibitors.20,22,24
Quantum Mechanical Calculations on Model Com-
pounds Derived from 27. A simplified model of 27 was used to
investigate the possibility of inducing a phenyl−purine twist with
C-8 phenyl ortho-substituents. As a first approximation, it was
assumed that the O6-cyclohexylmethoxy substituent of 27 would
have little effect on the phenyl−purine twist angle, and this group
wasnotincludedinthecalculations. Ninemodelcompounds(A−
I) encompassing combinations of fluoro, chloro, and methyl
substituentswereinvestigatedincomparisonwith27.Eachmodel
was constructed and energy minimized, with the purine and aryl
rings constrained to a coplanar orientation. A rigid potential
energy scan was then performed at a range of interplanar angles
between the two coplanar conformations (a full description of the
results of this analysis for each compound can be found in the
Supporting Information results and Figure S2).
The ortho-fluorophenylpurine (A) (Figure 4) was found to
have both the desired characteristics of inducing an interplane
twist of approximately 40° and allowing relatively free rotation
about this optimum conformation. However, A possesses a
significantlymorestableconformationat180°,presumablydueto
an intramolecular hydrogen bond. The 2-chlorophenyl derivative
B has minima at approximately 40° and 140° that would,
potentially, allow the formation of favorable aryl−backbone
packing interactions. The 40° conformation is in a relatively
shallow energy well, indicating some degree of rotation between
the two ring systems. However, the more stable, and therefore
preferred, conformation at 140° has a relatively steep energy well,
implying a more restricted rotational freedom. By contrast, the
ortho-methylphenylpurine C has a global minimum at around 40°
in a relatively flat energy well. The ortho-disubstituted C8-
phenylpurines E, F, H, and I all show potential to form the
required conformation, especially the 8-(2-chloro-6-
F
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 4. Conformation-energy profiles of 27 and three representative model compounds (A−C).
methylphenyl)purine (I), which has a minimum in a relatively flat
energy well at ∼50°.
the C-8 atoms of compounds 26 and 29 almost coincide with the
C-17 atom of 7 (first carbon atom of the 2-arylamino group,
Figure 5B).
Importantly, the structure of 29 bound to CDK2/cyclin A
supported the results of the quantum mechanical calculations.
The o-tolyl group on the 8-aryl ring of the purine induces an inter-
ring twist of approximately 38°, which is close to the predicted
value of∼40°. Thisvalue isalsosimilartotheinhibitor interplanar
ringtiltsobservedfortheCDK2/cyclin A/6andCDK2/cyclinA/
7 structures (50° and ∼34°, respectively),20 suggesting that an
equivalent disposition of two-ring systems can be accommodated
in either binding mode. The introduction of an ortho-methyl (or
ortho-chloro) substituent onto the C-8 phenyl ring of 27 restores
the potency of the compounds that adopt this altered binding
mode to that of 5, the benchmark 6-alkoxypurine CDK2 inhibitor
in the series (Table 1). However, there remained a 20-fold
difference in potency between 29 (CDK2; IC50 = 18.2 μM) and
the 2-phenylaminopurine derivative 6 (CDK2; IC50 = 1.0 μM)
that does not adopt the alternative binding mode. The lower
potency of 29 likely results from misalignment of the purine,
resulting in weakened hydrogen bond interactions with the
CDK2 backbone in the hinge region. A concomitant shift of the
purine ring out of the cleft will also result in suboptimal packing of
the C-8 phenyl ring against the peptide backbone at residues
Gln85 and Asp86.
To assess the validity of the modeling studies, the ortho-
chlorophenyl and ortho-tolyl derivatives B and C were
synthesized (28 and 29) and evaluated for CDK-inhibitory
activity, with both compounds exhibiting potency comparable
with the parent 2-amino-6-cyclohexylmethylpurine 5 (Table 1).
Interestingly, 28 and 29 were essentially inactive against CDK1
(IC50 values of >100 and 94 μM, respectively), although the poor
aqueoussolubilityof27militatedagainstadirectcomparisonwith
28 and 29. The crystal structure of 29 in complex with CDK2/
cyclin A was determined. As anticipated from the analysis of 26,
the ortho-tolylpurine 29 binds in the same reverse binding mode
and makes an identical triplet of hydrogen bonds with Glu81 and
Leu83 (Figure 5A). However, the interaction with Lys33 is not
maintained, as the residue adopts a conformation in which the
side chain is directed away from the inhibitor. The o-tolyl group of
29 is in the same orientation as the isopropyl group of 26 and
projectstoward theselectivitysurface oftheATPpocketsuchthat
Our previous studies with other purine- and pyrimidine-based
CDK inhibitor series suggested the possibility of exploiting
sequence differences between kinases on the “selectivity surface”,
G
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
donor and acceptor properties, with a view to recapitulating the
interactions with Asp86 observed for 7. Although the potency of
the 3-carboxamido derivative 43 (CDK2; IC50 = 41/53 μM)
suggested suboptimal binding, the introduction of a 3-
sulfonamide group (45) conferred submicromolar CDK2-
inhibitory activity (IC50 = 0.51 0.05 μM). Thus, 45 proved
approximately 50-fold more potent than 26 and exhibited activity
comparablewiththe2-phenylaminopurine6. Interestingly, the8-
arylpurine 44, which differs from 45 only in lacking the ortho-
methyl group, was some 17-fold less potent (CDK2; IC50 = 8.7
3.4 μM), supportive of the role of the ortho-methyl function in
imposing the desired conformational twist of the 8-aryl ring of the
purine.
The predicted binding modes of 44 and 45 were confirmed by
determination of the structures of these purines bound to CDK2/
cyclin A. Both inhibitors adopt the reverse binding mode, and, as
predicted, the sulfonamide substituent makes hydrogen bonds to
Asp86 and Lys89 (Figure 6). The presence of the 2-methyl group
Figure5. (A)Bindingof29toCDK2/cyclinA.(B)Comparativebinding
of7 and29. (A)CDK2is shown indark green. Selectedresidues from the
CDK2 ATP binding pocket and 29 are shown as ball-and-stick
representation and colored dark green and light green, respectively.
Water molecules in the vicinity of the inhibitor are included as red
spheres. (B) The inhibitors and selected residues within the CDK2 ATP
binding site are shown in ball-and-stick representation. Carbon atoms of
CDK2 residues from 7 and 29 complex structures are colored gray and
dark green, respectively. Purine 7 is shown in coral and 29 in light green.
Figure 6. Crystal structures of (A) 44 and (B) 45 bound to CDK2/cyclin
A. CDK2 is shown in green, with selected residues from the ATP-binding
pocket shown in ball-and-stick mode. 44 and 45 are also drawn in ball-
and-stick representation with carbon atoms colored light green.
on the C-8 aryl ring of 45 imposes the requisite rotation of the
entire group and also precludes a steric clash between the methyl
moiety and the N7 position. This rotation also results in a slight
decrease in the distance between the sulfonamide substituent and
Asp86 and Lys89 that may contribute to the observed increases in
potency. The hydrogen bonds between the purine ring of 45 and
the CDK2 backbone within the hinge sequence are maintained,
but again a water molecule mediates an indirect network of
interactions between N9 and the side chains of Lys33 and Glu81.
through further elaboration of the 8-aryl ring of 29. For example, a
number of inhibitor series probe both the conserved aspartate
residue (Asp86 in CDK2) and the sequence differences that exist
betweenCDKsattheresidueequivalenttoCDK2Lys89(Thr102
and Val100 in CDK4 and CDK7, respectively) to derive
considerable gains in potency and selectivity.20,24,26,34,35 The
potent CDK2 inhibitor 7 evolved from 6 by the rational
introduction of a sulfonamide group at the 4-position of the 2-
phenylamino ring.20 The resultant 200-fold increase in potency is
attributable, atleastinpart,tooptimalH-bondinteractionsarising
between the sulfonamide moiety and the backbone amide and
side chain of Asp86. In an effort to emulate these interactions in
the context of the reversed binding mode, compound 29 was
initially further substituted with hydroxyl (36) and carboxylate
(37) groups at the meta-position on the C-8 aryl ring. However,
these substitutions did not lead to a significant increase in
potency, and while 36 (CDK2; IC50 = 28 μM) retained activity
approaching that of 29 (CDK2; IC50 = 18.2 μM), the carboxylic
acid derivative 37 was essentially devoid of activity against both
CDK1 and CDK2 (IC50 > 100 μM). Determination of the
structures of compounds 36 and 37 bound to CDK2/cyclin A
revealed the molecular details of the inhibitor binding modes
(Supporting Information Figure S3).
CONCLUSION
■
In the current study, the effect of substitutions at the C-8 position
within a series of O6-cyclohexylmethylguanines on their ability to
inhibit CDK2 has been evaluated. The introduction of alkyl
groups was poorly tolerated, leading to decreases in potency, the
origins of which were elucidated by crystallographic analysis of
selected inhibitors bound to CDK2/cyclin A. The crystal
structures of compounds 26 and 32 bound to CDK2/cyclin A
revealed that they adopt novel binding modes within the CDK2
ATP-binding site. An increase in the size of the C-8 alkyl group
from methyl to isopropyl induced a rotation of the binding mode
so that the C-8 substituent was positioned to face the CDK2
specificity surface. Quantum mechanical calculations were
employed to identify 29, a compound that adopts the novel,
reverse binding mode and is equipotent with the parent purine 5
against CDK2.
The disappointing activity of 36 and 37 prompted the
introduction of meta-substituents on the 8-aryl ring having both
H
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
formic acid in methanol (10−90%). All compounds gave ≥95% purity by
1H NMR, HPLC, or LCMS unless stated otherwise. Elemental analyses
were performed by The School of Pharmacy, Analytical Facility,
University of London, WC1N 1AX. Accurate masses were measured
using a Finnigan MAR 95 XP or a Finnigan MAR 900 XLT at the EPSRC
National Mass Spectrometry Service Centre (Chemistry Department,
University of Wales, Swansea SA2 8PP, Wales).
Initial further elaboration of 29 by meta-substitution to exploit
potential interactions with Asp86 failed to increase potency.
Structure determination of compounds 36 and 37 bound to
CDK2/cyclin A revealed that these compounds occupy a similar
position to compound 29 within the CDK2 ATP-binding cleft.
However, despite interacting with the same region as the
sulfonamidegroupof 7, thehydroxyl andcarboxylate substituents
of 36 and 37, respectively, are not oriented to reproduce the
favorable interactions observed in the structure of 7. However, in
derivative 45, a sulfonamide substituent at the meta-position of
the C-8 arylpurine, in tandem with an ortho-methyl group,
maintained the reversed binding mode and also emulated the
interactions made by 7, leading to the predicted increases in
potency. Unfortunately, the relatively modest potency of 45
compared with the parent 2-arylamino-6-alkoxypurine CDK2
inhibitor 7, combined with only limited opportunities for further
elaboration, militate against further optimization of this series.
Small-molecule CDK inhibitors have enjoyed a resurgence of
interest, as evidenced by a very recent review of the patent
literature,36 and clinical studies with the CDK4/6 inhibitor
palbociclib (PD-0332991) are at an advanced stage.37 These
clinical studies are underpinned by a greater appreciation of the
importance of CDK activity to normal cells, and the more careful
selection of patient populations to ensure that effective
therapeutic windows can be achieved. Previous experience has
highlighted the importance of achieving selectivity for the target
CDK over other family members, and the availability of crystal
structures for the majority of the CDKs to guide inhibitor design,
has proven invaluable. In this paper, we have demonstrated the
application of molecular modeling and structure-based design for
the exploitation of an unorthodox binding orientation of a purine-
based CDK2 inhibitor. The results of the study contribute to a
further understanding of the possible interactions of this inhibitor
class with the ATP-binding domain of CDK2, and will be of value
both in the design of more potent CDK2 inhibitors and for
abrogating CDK2-inhibitory activity where this is required.
3-Bromo-N-(4-methoxybenzyl)-2-methylbenzamide (12). p-Tol-
uenesulfonyl chloride (1.06 g, 5.58 mmol) was added dropwise to a
stirred solution of 3-bromo-2-methylbenzoic acid (1.0 g, 4.65 mmol) in
anhydrous pyridine (5 mL) at 0 °C, and the mixture was allowed to warm
to room temperature with stirring over 72 h. The solvent was removed in
vacuo, and the residue was redissolved in EtOAc (20 mL), washed
sequentiallywithaqueousHClsolution(0.1M,20mL)andbrine(2 ×20
mL), dried(MgSO4), andconcentratedunderreducedpressuretogive3-
bromo-2-methylbenzoic 4-methylbenzenesulfonic anhydride as a white
solid (1.72 g, 100%), which was used directly for the next reaction. To a
solution of the tosylate ester (0.46 g, 1.24 mmol) in DCM (5 mL) was
addedDMAP(15mg, 0.12mmol), andthemixturewasstirredfor15min
at room temperature prior to addition of 4-methoxybenzylamine (0.8
mL, 6.3 mmol). After stirring for a further 30 min, DCM (10 mL) was
added and the reaction mixture was washed sequentially with aqueous
HCl solution (0.1 M, 20 mL) and brine (2 × 20 mL) and dried (MgSO4),
and the organic fraction was concentrated under reduced pressure.
Purification by chromatography on silica, employing a gradient elution of
petrol:EtOAc (9:1) to petrol:EtOAc (1:1), gave the title compound as a
white powder (0.41 g, 100%); mp 121−122 °C. IR 3248, 1612, 1513
1
cm−1. H NMR (CDCl3, 500 MHz) δ 2.44 (0.74H, s, Ar-CH3), 2.47
(2.26H, s, Ar-CH3), 3.78 (0.74H, s, OCH3), 3.80 (2.26, s, OCH3), 4.05
(0.49H, d, J = 6.1 Hz, NCH2), 4.50 (0.25H, t, J = 5.9 Hz, NH), 4.55
(1.51H, d, J = 5.7 Hz, NCH2), 5.95 (0.75H, brs, NH), 6.78−6.82(0.48H,
m, H-3′), 6.86−6.91 (1.51H, m, H-3′), 7.02−7.07 (0.75H, m, H-5),
7.08−7.13 (0.5H, m, H-2′), 7.25−7.29 (2.1H, m, H-4, H-2′), 7.31 (0.5H,
d, J=8.3Hz),7.58(0.74H, dd,J=8.0,1.1Hz, H-6), 7.73−7.77(0.5H,m).
13C NMR (CDCl3, 125 MHz) δ 20.2, 21.7, 43.7, 46.9, 55.4, 55.5, 114.4,
114.2, 125.8, 126.8, 127.2, 127.3, 128.4, 129.4, 129.9, 130.0, 134.0, 135.7,
137.0, 138.8, 143.6, 159.5, 159.3, 169.2. HRMS [M + H]+
(C16H17BrNO2) calcd 334.0437, obsd 334.0438.
3-Bromo-N,N-bis(4-methoxybenzyl)benzenesulfonamide (13).
NaH (60% in mineral oil, 75 mg; 1.86 mmol) was added to a solution
of 3-bromobenzenesulfonamide (0.2 g, 0.85 mmol) in DMF (2 mL), and
the solution was stirred at room temperature for 5 min. 4-Methoxybenzyl
chloride (0.25 mL, 1.86 mmol) was added dropwise, and the mixture was
stirred at 80 °C for 2 h and saturated aqueous ammonium chloride
solution (5 mL). The resulting solution was extracted with DCM (3 × 20
mL), and the combined organic phases were dried (MgSO4) and
concentrated in vacuo. Purification by chromatography on silica, using a
gradient elution of 100% EtOAc to petrol:EtOAc (5:1), gave the title
compound as a crystalline white solid (0.34 g; 84%); mp 136−137 °C. IR
3009, 2837, 1608, 1509 cm−1. 1HNMR(CDCl3, 500MHz)δ3.79(6H, s,
OCH3), 4.27 (4H, s, NCH2), 6.78 (4H, d, J = 8.6 Hz, H-3′), 7.01 (4H, d, J
= 8.6 Hz, H-2′), 7.35 (1H, dd, J = 7.9, 8.0 Hz, H-5), 7.67 (1H, br d, J = 7.9
Hz, H-4), 7.72 (1H, br d, J = 7.9 Hz, H-6), 7.83 (1H, m, H-2). 13C NMR
(CDCl3, 125 MHz) δ 49.8, 55.4, 114.0, 125.7, 127.3, 130.0, 130.1, 130.6,
135.4, 159.3. HRMS [M + H]+ (C22H23BrNO4S) calcd 476.0526, obsd
476.0522.
3-Chloro-N,N-bis(4-methoxybenzyl)-2-methylbenzenesulfona-
mide (14). To a solution of 3-chloro-2-methylbenzenesulfonamide (0.8
g, 3.89 mmol) in DMF (5 mL) was added NaH (60% in mineral oil, 0.2 g,
8.56 mmol), followed by 4-methoxybenzyl chloride (1.16 mL, 8.56
mmol). After stirring at room temperature for 1 h, saturated aqueous
NH4Cl solution (5 mL) was added to the reaction mixture, followed by
H2O (10 mL). The resulting solution was extracted with EtOAc (3 × 20
mL), and the combined organic fractions were washed with brine and
dried (MgSO4), and the solvent was removed in vacuo. Purification by
chromatography on silica using gradient elution of 100% petrol to 30%
petrol:EtOAc (3:1) afforded the benzenesulfonamide as a colorless
crystalline solid (1.52 g, 88%); mp 66−68 °C. IR 2838, 1611, 1511 cm−1.
1H NMR (CDCl3, 500 MHz) δ 2.63 (3H, s, Ar-CH3), 3.80 (6H, s,
EXPERIMENTAL SECTION
■
GeneralSyntheticProcedures. Reagentswerepurchasedfromfine
chemicals vendors and used as received unless otherwise stated. Solvents
were purified and stored according to standard procedures. Petrol refers
to that fraction in the boiling range 40−60 °C. THF refers to anhydrous
tetrahydrofuran, either by distillation from sodium benzophenone or
from commercial sources. Melting points were obtained on a Stuart
Scientific SMP3 apparatus and are uncorrected. Thin layer chromatog-
raphy was performed using silica gel plates (Kieselgel 60F254; 0.2 mm)
and visualized with UV light or potassium permanganate. Chromatog-
raphy was conducted under medium pressure in glass columns or using a
Biotage SP4 instrument in prepacked columns (FLASH+ Silica columns
[40−63μm, 60Å]). SemiprepHPLCwasconductedwithVarianProStar
HPLC system equipped with 2 ProStar 210 solvent delivery modules, a
ProStar 320 UV−vis detector, a ProStar 701 fraction collector, and
controlled by Varian Star chromatography workstation. Proton (1H) and
carbon (13C) nuclear magnetic resonance (NMR) spectra were recorded
on a Bruker Spectrospin AC 300E (1H at 300 MHz, 13C at 75 MHz), a
Jeol JNM-LA500 spectrometer (1H at 500 MHz, 13C at 125 MHz), or a
Bruker Avance II 500 (1H at 500 MHz, 13C at 125 MHz). Chemical shifts
(δ) are reported in parts per million (ppm) downfield from
tetramethylsilane (TMS), using known solvent peaks as internal
standards. Coupling constants (J) are reported in hertz (Hz).
IR spectra were recorded on a Bio-Rad FTS 3000MX diamond ATR.
Liquidchromatography−massspectrometry (LCMS) was carriedout on
a Micromass Platform instrument operating in positive and negative ion
electrospray mode, employing a 50 mm × 4.6 mm C18 column (Waters
Symmetry or Waters Atlantis) 5 or 12 min gradient elution with 0.05%
OCH3), 4.25 (4H, s, NCH2), 6.81 (4H, d, J = 8.6 Hz, H-3′), 6.96 (4H, d, J
I
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
= 8.6 Hz, H-2′), 7.23 (1H, dd, J = 7.9, 8.0 Hz, H-5), 7.59 (1H, dd, J = 8.0,
1.0 Hz, H-4), 7.89 (1H, dd, J = 7.9, 1.1 Hz, H-6). 13C NMR (500
MHz,CDCl3) 17.0, 48.8, 55.4, 114.1, 126.6, 127.2, 128.8, 130.2, 133.6,
135.9, 137.2, 140.9, 159.4. HRMS [M]+ (C23H24ClNO4S) calcd
445.1108, obsd 445.1109.
2-Amino-4-cyclohexylmethoxy-5-nitroso-6-propionylaminopyri-
midine (19). Synthesized from 17 (0.50 g, 2.0 mmol) and propionic
anhydride (0.29 g, 2.2 mmol) in DMF (6.0 mL) following method I,
purified by chromatography on silica employing petrol:EtOAc (3:2) as
eluent, togivethetitlecompoundasanaquamarinecrystallinesolid(0.28
g, 46%); mp 160−161 °C. 1H NMR (DMSO-d6, 300 MHz) δ 1.11−1.18
(3H, t, J = 7.3 Hz, CH2CH3), 1.23−1.30 (5H, m, cyclohexyl), 1.78−1.94
(6H, m, cyclohexyl), 2.95 (2H, q, J = 7.3 Hz, CH2CH3), 4.44 (2H, d, J =
6.3 Hz, CH2O), 8.72 (2H, br s, NH2), 12.4 (1 H, br s, NH). HRMS [M]+
(C14H21N5O3) calcd 307.1644, obsd 307.1632. Anal. (C14H21N5O3 + 0.3
EtOAc) C, H, N.
2-Amino-4-cyclohexylmethoxy-6-isobutylamino-5-nitrosopyrimi-
dine (20). From 17 (0.82 g, 3.25 mmol) and isobutyric anhydride (0.57 g
3.57 mmol) in DMF (12 mL) employing method I, purification by
chromatography on silica with petrol:EtOAc (3:2) as eluent, followed by
recrystallization from EtOAc:petrol to furnish 20 as a pale-blue
crystalline solid (0.63 g, 60%); mp 180−181 °C. 1H NMR (DMSO-d6,
300 MHz) δ 0.92−1.39 (5H, m, cyclohexyl), 1.22 (6H, d, J = 6.8 Hz,
CH(CH3)2), 1.78−1.93 (6 H, m, cyclohexyl), 3.26 (1H, quin, J = 6.8 Hz,
CH(CH3)2),4.44(2H, d,J=6.8Hz,CH2CH),8.74(2H,brs, NH2),12.6
(1 H, br s, NH). Anal. (C15H23N5O3) C, H, N.
2-Amino-6-benzoylamino-4-cyclohexylmethoxy-5-nitrosopyrimi-
dine (21). Synthesized following method I from 17 (0.50 g, 2.0 mmol)
and benzoic anhydride (0.50 g; 2.2 mmol) in DMF (8 mL) and isolated
by chromatography on silica using DCM:MeOH (95:5) as eluent.
Recrystallization from EtOAc:petrol yielded 21 as a blue crystalline solid
(0.24 g; 33.8%); mp 210−211 °C. 1H NMR (DMSO-d6, 300 MHz) δ
1.16−1.43 (5H, m, cyclohexyl), 1.81−1.98 (6 H, m, cyclohexyl), 4.50
(2H, d, J = 6.3 Hz, OCH2), 7.75−7.86 (3H, m, Ar-H), 8.08−8.12 (2H, m,
Ar-H), 8.88 (1 H, br s, NH), 8.99 (1H, br s, NH), 13.8 (1H, br s, NH).
HRMS [M]+ (C18H21N5O3) calcd 355.1644, obsd 355.1646. Anal.
(C18H21N5O3) C, H, N.
N-(4-Methoxybenzyl)-2-methyl-3-(4,4,5,5-tetramethyl-1,3,2-diox-
aborolan-2-yl)benzamide (15). A mixture of 12 (0.16 mg, 0.49 mmol),
bis(pinacolato)diboron (0.19 g, 0.74 mmol), [1,1′-bis-
(diphenylphosphino)ferrocene]dichloropalladium(II) (17 mg, 0.024
mmol), 1,1′-bis(diphenylphosphino)ferrocene (26 mg, 0.048 mmol),
and KOAc (0.144 g, 1.47 mmol) in dioxane (2 mL) was purged with N2
and heated under microwave irradiation at 150 °C for 90 min. After
cooling, the reactionmixture was filtered through Celite, EtOAc (10mL)
was added, and the mixture was washed with brine (2 × 10 mL) and dried
(MgSO4), and the solvents were evaporated under reduced pressure.
Purification by chromatography on silica, employing a gradient eluent of
petrol:EtOAc (10:1) to petrol:EtOAc (1:1), yielded the title compound
as a colorless syrup (0.144 g, 77%). IR 3484, 3181, 2980, 2933, 1644,
1612, 1580, 1512 cm−1. 1H NMR (CDCl3, 500 MHz) δ 1.34 (12H, s,
CH3), 2.44 (0.6H, s, Ar-CH3), 2.60 (2.4H, s, Ar-CH3), 3.78 (0.6H, s,
OCH3), 3.80 (2.4H, s, OCH3), 4.05 (0.4H, d, J = 6.1 Hz, NCH2), 4.51
(0.2H,t,J=5.9Hz,NH), 4.56(1.6H,d,J=5.6Hz, NCH2), 5.88(0.8H,t,J
= 4.9 Hz, NH), 6.78−6.82 (0.4H, m, H-3′), 6.85−6.90 (1.6H, m, H-3′),
7.08−7.12 (0.4H, m, H-2′), 7.17 (0.8H, dd, J = 7.5, 7.6 Hz, H-5), 7.26−
7.29(1.6H,m, H-2′), 7.31(0.4H,d, J=8.3Hz),7.38(0.8H,dd,J=7.6,1.4
Hz, H-4), 7.74−7.80 (1H, m, H-6). 13C NMR (CDCl3, 125 MHz) δ 19.1,
21.6, 24.9, 43.5, 46.9, 55.4, 55.4, 83.8, 114.2, 114.3, 124.9, 127.3, 128.3,
129.1, 129.3, 129.4, 129.8, 130.3, 137.0, 137.3, 137.4, 141.8, 143.6, 159.2,
159.5, 170.6. HRMS [M]+ (C22H28BNO4) calcd 381.2220, obsd
381.2222.
N,N-Bis(4-methoxybenzyl)-2-methyl-3-(4,4,5,5-tetramethyl-1,3,2-
dioxaborolan-2-yl)benzenesulfonamide (16). A mixture of 14 (0.18
mg, 0.41 mmol), bis(pinacolato)diboron (0.16 g, 0.62 mmol), [1,1′-
bis(diphenylphosphino)ferrocene]dichloropalladium(II) (14 mg, 0.02
mmol), 1,1′-bis(diphenylphosphino)ferrocene (22 mg, 0.04 mmol), and
KOAc (0.119 g, 1.21 mmol) in dioxane (2 mL) was purged under N2 for
10 min and heated under microwave conditions 150 °C for 120 min. The
cooled reaction mixture was filtered through Celite, diluted with EtOAc
(10 mL), and washed with brine (3 × 10 mL). The combined organic
fractions were dried (MgSO4), and the solvent was remove under
reduced pressure. Purification by chromatography on silica) using a
gradient elution of petrol:EtOAc (10:1) from petrol:EtOAc (1:1)
afforded the title compound as a white solid (87 mg, 40%); mp 126−128
°C. IR 2981, 2158, 1610, 1510 cm−1. 1H NMR (CDCl3, 500 MHz) δ 1.38
(12H, s, CH3), 2.79 (3H, s, Ar-CH3), 3.79 (6H, s, OCH3), 4.23 (4H, s,
NCH2), 6.76−6.81 (4H, m, H-3′), 6.92−6.98 (4H, m, H-2′), 7.28 (1H,
dd,J=7.8, 7.4Hz, H-5), 7.95(1H, dd, J=7.4,1.3Hz, H-4), 8.07(1H, dd, J
= 7.8, 1.4 Hz, H-6). 13C NMR (CDCl3, 125 MHz): 19.1, 25.0, 48.6, 55.4,
84.1, 114.0, 125.2, 127.7, 130.2, 132.6, 139.0, 140.2, 144.6, 159.2. HRMS
[M]+ (C29H36NO6BS) calcd 537.2465, obsd 537.2465.
6-Acylamino-2-amino-4-cyclohexylmethoxy-5-nitrosopyrimi-
dines: Method I. General Procedure. To a solution of 2,6-diamino-4-
cyclohexylmethoxy-5-nitrosopyrimidine18 (17) (1.0 mol equiv) in DMF
(5−10 mL) was added the appropriate anhydride (1.1 mol equiv), and
the mixture was heated at 80 C° for the specified time. After cooling,
volatiles were removed under reduced pressure, water (10−20 mL) was
added, and the reaction mixture was extracted with EtOAc (3 × 20 mL).
The combined organic fractions were dried (Na2SO4) and concentrated
in vacuo, and the required acylaminopyrimidines (18−23) were isolated
by chromatography on silica as described.
6-Acetylamino-2-amino-4-cyclohexylmethoxy-5-nitrosopyrimi-
dine (18). Prepared from 17 (0.50 g, 2.0 mmol) and acetic anhydride (4
mL) in accordance with method I and purified by chromatography on
silica (eluent DCM:MeOH, 10:1) to furnish 18 as an aquamarine
crystalline solid (0.35 g; 60%); mp 149−150 °C. 1H NMR (DMSO-d6,
300 MHz) δ 1.13−1.29 (5H, m, cyclohexyl), 1.79−1.94 (6H, m,
cyclohexyl), 2.57 (3H, s, Me), 4.45(2H, d, J=6.3, CH2O), 8.73(2H, brs,
NH2), 12.3 (1 H, br s, NH). HRMS [M]+ (C13H19N5O3) calcd 293.1488,
obsd 293.1484. Anal. (C13H19N5O3 + 0.05 EtOAc) C, H, N.
2-Amino-6-(2-chlorobenzoyl)amino-4-cyclohexylmethoxy-5-ni-
trosopyrimidine(22).Toastirredsolutionof17(0.50g, 2.00mmol)and
N,N-dimethyl-4-aminopyridine (24mg, 0.20 mmol)in pyridine(12mL)
was added 2-chlorobenzoyl chloride (280 μL, 2.20 mmol). The solution
was stirred at 25 °C under N2 for 18 h and concentrated under reduced
pressure, and the residual solid was dissolved in DCM (40 mL) and
washed with aqueous hydrochloric acid solution (5% v/v, 3 × 20 mL).
Theorganicfractionwasdried(Na2SO4), andthesolventwasremovedin
vacuo to give a dark-green solid, which was recrystallized from
EtOAc:petrol to furnish the title compound as a blue−green crystalline
solid (0.6 g; 78%); mp 184−185 °C. IR 3467, 3202, 2925, 1728, 1594,
1
1536 cm−1. H NMR (DMSO-d6, 300 MHz) δ 0.94−1.28 (5H, m,
cyclohexyl), 1.60−1.97 (6H, m, cyclohexyl), 4.35 (2H, d, J = 6.6 Hz,
OCH2), 6.15 (2H, s, NH2), 7.32−7.53 (3H, m, Ar-H), 7.54 (1H, d, Ar-
H). 13C NMR, (DMSO-d6, 75 MHz) δ 26.0, 26.7, 30.0, 37.5, 74.0, 127.7,
129.9, 131.3, 132.0, 132.8, 135.1, 139.6, 164.9, 166.6. MS (ES+) m/z 390
[M(35Cl) + H]+, 392 [M(37Cl) + H]+. Anal. (C18H20ClN5O3 + 0.1
EtOAc) C, H, N.
2-Amino-4-cyclohexylmethoxy-6-(2-methylbenzoyl)amino-5-ni-
trosopyrimidine(23). Prepared analogouslyto22abovefrom17(0.51g,
2.00 mmol) and 2-methylbenzoyl chloride (300 μL, 2.20 mmol) in
pyridine (12 mL). Recrystallization from EtOAc:petrol gave a dark-blue
crystalline solid (0.57 g; 77%); mp 177−179 °C. IR 3418, 3233, 2916,
1713, 1598, 1520 cm−1. 1H NMR (DMSO-d6, 300 MHz) δ 0.94−1.28
(5H, m, cyclohexyl), 1.60−1.95 (6H, m, cyclohexyl), 2.47 (3H, s,
ArCH3), 4.35(2H, d, J=6.7Hz, OCH2), 6.17(1H, s), 7.22−7.57(3H, m,
Ar−H), 7.57 (1H, s), 7.59 (1H, d, J = 1.2 Hz). 13C NMR, (DMSO-d6, 75
MHz)δ21.0, 26.0, 26.7, 30.1, 37.5, 73.9, 126.8, 128.3, 132.2, 132.3, 134.5,
138.8, 139.8, 165.1, 168.9. MS (ES+) m/z 370 [M + H]+. Anal.
(C19H23N5O3) C, H, N.
8-Substituted 2-Amino-6-cyclohexylmethoxypurines: Method II.
General Procedure. A stirred solution of the appropriate 6-acylamino-2-
amino-4-cyclohexylmethyloxy-5-nitrosopyrimidine (0.5−2.0 mmol) in
THF (20−30 mL) containing AcOH (glacial, 1−2 mL) was cooled to
−78 °C and purged with N2 prior to addition of Pd (10% on carbon, 15%
w/w). ThereactionmixturewasstirredunderH2 atatmosphericpressure
for 24 h at ambient temperature and filtered through Celite, and the
filtrate was neutralized with saturated NaHCO3 solution (30−50 mL)
J
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
and extracted with EtOAc (3 × 30 mL). The combined organic fractions
were dried (Na2SO4) and evaporated under reduced pressure to give the
required 8-substituted purine. Compound purification was achieved by
chromatography on silica and recrystallization as described.
2-Amino-6-cyclohexylmethoxy-8-methylpurine (24). Prepared in
accordance with method II from 18 (0.35 g, 1.20 mmol) and isolated by
chromatography on silica employing DCM:MeOH (10:1) as eluent.
Recrystallization from a mixture of EtOAc:MeOH gave 24 as a cream
crystallinesolid(0.19 g, 60%);mp99−100°C. 1HNMR(DMSO-d6, 300
MHz)δ1.08−1.45 (5H, m, cyclohexyl), 1.78−1.92 (6H, m, cyclohexyl),
2.42 (3H, s, Me), 4.26 (2 H, d, J = 6.2 Hz, OCH2), 6.20 (2 H, br s, NH2),
12.2 (1 H, br s, NH). HRMS [M]+ (C13H19N5O) calcd 261.1590, obsd
261.1587. Anal. (C13H19N5O + 1.15 MeOH) C, H, N.
2-Amino-6-cyclohexylmethoxy-8-ethylpurine (25). Synthesized by
method II from pyrimidine 19 (0.28 g, 0.91 mmol) and isolated by
chromatography on silica employing petrol:EtOAc (3:2) as eluent.
Recrystallization from a mixture of EtOAc:Et2O gave 25 as a cream solid
(0.10 g, 39.9%); mp 137−138 °C. 1H NMR (DMSO-d6, 300 MHz) δ
1.06−1.18 (5 H, m, cyclohexyl), 1.31(3 H, t, J= 7.5 Hz, CH2CH3), 1.76−
1.90(6H, m,cyclohexyl), 2.74(2H, q, J=7.5Hz, CH2CH3), 4.23(2H, d,
J = 6.1 Hz, OCH2), 6.18 (2 H, br s, NH2), 12.2 (1 H, br s, NH). HRMS
[M]+ (C14H21N5O)calcd275.1746, obsd275.1743. Anal. (C14H21N5O+
0.1 EtOAc) C, H, N.
palladium(II) (8 mg; 0.01 mmol) in toluene (15 mL) was degassed with
nitrogen and heated under at 100 °C with stirring for 3 h. After cooling,
the reaction mixture was filtered through Celite and the filtrate was
concentrated under reduced pressure to give a yellow solid. The crude
material was purified by chromatography on silica, employing a gradient
of petrol:EtOAc (1:1 to 0:1) as eluent to afford 29 as a beige−white solid
(60 mg; 70%). All analytical data (mp, NMR, MS) were identical with 29
prepared previously.
2,4,6-Triamino-5-cyclohexylmethoxypyrimidine (30). Zinc powder
(0.26 g, 3.9 mmol) was added to a solution of 2,6-diamino-4-
cyclohexylmethoxy-5-nitrosopyrimidine (17, 0.20 g, 0.79 mmol) in
AcOH (5 mL), and the solution was stirred for 12 h at room temperature.
The reaction mixture was filtered through Celite, and the filtrate was
evaporated under reduced pressure to furnish a yellow solid, which was
redissolved in water (20 mL), and the solution was adjusted to pH 7.0
with saturated aqueous sodium bicarbonate solution. The mixture was
extracted with EtOAc (3 × 30 mL), the combined organic layers were
dried (Na2SO4), and the solvent was removed in vacuo to give a pale-
yellow solid. Recrystallization from MeOH afforded the title compound
(0.14 g, 74%); mp 151−153 °C. IR 3446, 3390, 3309, 3178, 2918, 2847,
1613, 1577 cm−1. 1H NMR (DMSO-d6, 300 MHz) δ 0.92−1.29 (5H, m,
cyclohexyl), 1.67−1.79 (6H, m, cyclohexyl), 3.39 (2H, br, NH2), 3.93
(2H, d, J = 6.0 Hz, OCH2), 5.27 (2H, br, NH2), 5.66 (2H, br, NH2). 13
C
2-Amino-6-cyclohexylmethoxy-8-isopropylpurine (26). In accord-
ance with method II from 20 (0.63 g; 2.0 mmol). Isolated by
chromatography on silica with DCM:MeOH (10:1) as eluent, followed
by recrystallization from EtOAc:Et2O to yield the title purine as a cream
crystalline solid (0.39 g, 70%); mp 136−138 °C. 1H NMR (DMSO-d6,
300 MHz) δ 1.08−1.19 (5 H, m, cyclohexyl), 1.36 (6 H, d, J = 6.9 Hz,
CH(CH3)2) 1.79−1.92 (6 H, m, cyclohexyl), 3.07 (1 H, quin, J = 6.9 Hz,
(CH(CH3)2), 4.25 (2 H, d, J=6.1 Hz, OCH2), 6.20 (2H, br s, NH2), 12.6
(1 H, s, NH). (C15H23N5O + 0.25 H2O) C, H, N.
2-Amino-6-cyclohexylmethoxy-8-phenylpurine (27). Prepared fol-
lowing method II from 21 (0.24 g, 0.67 mmol) and isolated by
chromatography on silica employing DCM:MeOH (10:1) as eluent.
Recrystallization from MeOH:EtOAc gave 27 as pale-pink crystals (0.16
g, 76%); mp 271−272 °C. 1H NMR (DMSO-d6, 300 MHz) δ 1.12−1.31
(5H, m, cyclohexyl), 1.81−1.97 (6H, m, cyclohexyl), 4.33 (2H, d, J = 6.2
Hz, OCH2), 6.43 (2H, br s, NH2), 7.56−7.60 (3H, m, Ar-H), 8.16 (2 H,
m, Ar-H), 13.0 (1H, br s, NH). HRMS [M]+ (C18H21N5O) calcd
323.1746, obsd 323.1746. (C18H21N5O + 0.4 MeOH) C, H, N.
2-Amino-8-(2-chlorophenyl)-6-cyclohexylmethoxypurine (28).
Prepared from 22 (375 mg; 0.971 mmol) following method II and
purified by chromatography on silica employing EtOAc:petrol (2:1), to
afford the title compound as an off-white powder (241 mg, 70%); mp
190−191 °C. IR 3487, 3289, 3171, 2021, 1578 cm−1. 1H NMR (DMSO-
d6, 300 MHz) δ 1.02−1.33 (5H, m, cyclohexyl), 1.64−1.83 (6H, m,
cyclohexyl), 4.23 (2H, d, J = 6.24 Hz, OCH2), 6.34 (2H, br s, NH2),
7.45−7.54 (2H, m, Ar-H), 7.61−7.64 (1H, m, Ar-H), 7.73−7.75 (1H, m,
Ar-H), 12.69 (1 H, br s, NH). 13C NMR, (DMSO-d6, 75 MHz) δ 25.6,
26.4, 29.6, 37.2, 70.9, 114.7, 127.6, 130.4, 130.5, 131.3, 132.2, 145.1,
156.2, 160.2, 160.8. MS (ES+) m/z 358 [M(35Cl) + H]+, 360 [M(37Cl) +
H]+. Anal. (C18H20ClN5O + 0.1 EtOAc) C, H, N.
2-Amino-6-cyclohexylmethoxy-8-(2-methylphenyl)purine (29).
Synthesized in accordance with method II from 23 (327 mg; 0.88
mmol) and purified by chromatography on silica with EtOAc:petrol
(2:1); pale-yellow powder (302 mg, 100%); mp 187−188 °C. IR 3480,
3310, 3179, 3085, 2045, 1582 cm−1. 1H NMR (DMSO-d6, 300 MHz) δ
1.01−1.29 (5H, m, cyclohexyl), 1.64−1.84 (6H, m, cyclohexyl), 2.54
(3H, s, Ar-CH3), 4.25 (2H, d, J = 6.4 Hz, OCH2), 6.28 (2H, br s, NH2),
7.28−7.38 (3H, m, Ar-H), 7.64 (1H, m, Ar-H), 12.60 (1H, br s, NH). 13C
NMR (DMSO-d6, 75 MHz) δ 21.3, 25.5, 26.3, 29.6, 37.2, 70.9, 114.9,
126.2, 129.3, 129.6, 130.3, 131.4, 137.6, 147.8, 156.3, 160.0, 160.6. MS
(ES+) m/z 338 [M + H]+. HRMS [M + H]+ (C19H24N5O) calcd
338.1975, found 338.1973; [M − H]− (C19H22N5O) calcd 336.1830,
found 336.1826. Anal. (C19H23N5O + 0.2 EtOAc) C, H, N.
NMR, (DMSO-d6, 75 MHz) δ 24.9, 25.7, 28.9, 36.7, 69.6, 99.9, 155.2,
155.4, 157.8. HRMS[M]+ (C11H19N5O) calcd237.1587, obsd 237.1590.
Anal. (C11H19N5O + 0.1 H2O) C, H, N.
2,5-Diamino-4-cyclohexylmethoxy-6-(2,2,2-trifluoroacetyl)-
aminopyrimidine (31). A solution of 30(171 mg; 0.72 mmol) in TFA (5
mL) was heated under reflux for 6 h, and the solvent was evaporated in
vacuo. The title compound was isolated as a yellow solid (213 mg; 91%)
by chromatography on silica using DCM:MeOH, (9:1) as eluent and
used directly for the next reaction. 1H NMR (DMSO-d6, 300 MHz) δ
0.95−1.17 (5H, m, cyclohexyl), 1.64−1.67 (6H, m, cyclohexyl), 3.98
(2H, d, J = 3.9, OCH2), 6.98 (2H, br s, NH2), 7.24 (2H, br s, NH2), 9.97
(1H, s, NHCO). MS (ES+) m/z 334 [M + H]+.
2-Amino-6-cyclohexylmethoxy-8-trifluoromethylpurine (32). To a
solution of 31 (213 mg; 0.64 mmol) in pyridine (6 mL) was added N,N-
dimethyl-4-aminopyridine (21 mg; 0.17 mmol), and the mixture was
heated under reflux for 5 days. The residual red oil remaining, following
removal of the solvent under reduced pressure, was purified by
chromatography on silica utilizing using DCM:MeOH (95:5) furnished
the title compound as a pale-orange crystalline solid (106 mg; 53%); mp
1
252−253 °C. IR 3496, 1632, 1585 cm−1. H NMR (DMSO-d6, 300
MHz) δ0.97−1.32(5 H, m, cyclohexyl), 1.63−1.81(6 H, m, cyclohexyl),
4.23 (2H, d, J = 6.2 Hz, OCH2), 6.76 (2H, br s, NH2), 13.71 (1H, s, NH).
13C NMR (DMSO-d6, 75 MHz) δ 25.5, 26.3, 29.5, 37.1, 71.3, 114.2,
117.7, 121.3, 124.8, 155.3, 160.5, 161.9. MS (ES+) m/z 316 [M + H]+.
Anal. (C13H16F3N5O + 0.1 H2O) C, H, N.
2-Amino-9-benzyl-6-cyclohexylmethoxypurine (33). To a solution
of 5 (2.60 g, 10.5 mmol) in DMF (20 mL) was added K2CO3 (7.3 g, 52.6
mmol) and benzyl bromide (1.5 mL, 12.6 mmol), and the mixture was
stirred at room temperature for 12 h. Volatiles were removed in vacuo,
and the residual oil was redissolved in EtOAc (20 mL) and washed with
water (3 × 25 mL), and the organic fraction was dried (Na2SO4) and
evaporated under reduced pressure. Purification by chromatography on
silica, using EtOAc:petrol (1:1) as eluent, gave the 9-benzylpurine (33)
as a white solid (1.04 g, 29%). IR 3394, 3341, 3225, 3123, 2923, 2851,
1733, 1643, 1518 cm−1. 1H NMR (DMSO-d6, 300 MHz) δ 0.95−1.27
(5H, m, cyclohexyl), 1.61−1.95 (6H, m, cyclohexyl), 4.21 (2H, d, J = 6.3
Hz, OCH2), 5.25 (2H, s, CH2Ph), 6.42 (2H, s, NH2), 7.15−7.40 (5H, m,
Ar−H), 7.95 (1H, s, purine C8−H). 13C NMR (DMSO-d6, 75 MHz) δ
26.1, 26.8, 30.2, 37.8, 47.2, 72.4, 116.0, 128.0, 128.5, 129.3, 136.4, 139.3,
154.7, 159.9, 162.3.
9-Benzyl-8-(3-(benzyloxy)-2-methylphenyl)-6-(cyclohexylme-
thoxy)-9H-purin-2-amine (34). 1-Benzyloxy-3-bromo-2-methylben-
zene (616 mg, 2.22 mmol) and cesium carbonate (604 mg, 1.85
mmol) were added to a solution of 33 (250 mg, 0.74 mmol) in DMF (15
mL). The mixture was purged with N2 for 15 min prior to addition of
copper(I) iodide (423 mg, 2.22 mmol) and Pd(OAc)2 (0.1 M in DMF,
0.37 mL) and heated for 40 min under microwave irradiation at 200 °C.
Purine 29 was also synthesized by the following alternative method: a
mixture of 8-bromo-6-(cyclohexylmethoxy)-9H-purin-2-amine (83 mg;
0.25 mmol), o-tolyl boronic acid (69 mg; 0.51 mmol), Cs2CO3 (166 mg;
0.51 mmol), and dichloro [1,1′ bis(di-tert-butylphosphino)]ferrocene
K
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
The solvent was removed under reduced pressure, and the residual oil
was redissolved in EtOAc (25 mL) and washed with water (3 × 25 mL)
and dried (Na2SO4), and the organic fraction was evaporated in vacuo.
Theproductwaspurifiedbychromatographyonsilicawithpetrol:AcOEt
(2:1) as eluent to furnish the title compound as a pale-yellow solid (190
mg, 48%). IR: 3376, 3346, 3223, 2923, 2848, 1648, 1605, 1578 cm−1. 1H
NMR (CDCl3, 300 MHz) δ 0.95−1.15 (5H, m, cyclohexyl), 1.50−1.95
(6H, m, cyclohexyl), 1.85 (3H, s, Ar-CH3), 4.23 (2H, d, J = 6.0 Hz,
OCH2), 4.84 (2H, br s, NH2), 4.99 (2H, s, N-CH2-Ar), 5.04 (2H, s, O-
CH2-Ar), 6.71−7.40 (13H, m, Ar-H). 13C NMR (CDCl3, 75 MHz) δ
13.5, 26.2, 26.9, 30.3, 37.9, 46.7, 70.9, 72.5, 113.8, 123.3, 126.6, 127.5,
127.9, 128.0, 128.2, 128.7, 128.9, 129.3, 131.9, 136.9, 137.7, 157.5, 159.7.
Benzyl 3-(2-Amino-9-benzyl-6-(cyclohexylmethoxy)-9H-purin-8-
yl)-2-methyl benzoate (35). Synthesized in a manner analogous to
that for 34 above from 33 (250 mg, 0.74 mmol) and benzyl 3-bromo-2-
methylbenzoate (678 mg, 2.22 mmol). Purification by chromatography
on silica with petrol:AcOEt (2:1) as eluent gave the title compound as a
yellow powder (131 mg, 32%). IR 3349, 3225, 2923, 2851, 1721, 1653,
cm−1. 1H NMR (CDCl3, 300 MHz) δ 1.04−1.31 (5H, m, cyclohexyl),
1.70−1.91 (6H, m, cyclohexyl), 4.28 (2H, d, J = 5.8 Hz, OCH2), 5.10 (2
H, br s, NH2). 13C NMR (CDCl3, 75 MHz): 26.1, 26.8, 30.1, 37.6, 72.8,
155.6, 158.8, 161.0. HRMS [M]+ (C12H16BrN5O) calcd 325.0536, obsd
325.0538.
Bromopurine 39 was also prepared directly from 5 as follows: A
mixture of 5 (1.0 g; 4.0 mmol) and pyridinium tribromide (2.58 g; 8.1
mmol) in DCM (10 mL) was stirred at 70 °C for 96 h. Saturated aqueous
sodium metabisulfite solution (10 mL) was added, followed by water (10
mL), and the mixture was extracted with EtOAc (3 × 20 mL). The
organic phase was washed with brine (2 × 10 mL), dried (MgSO4), and
concentrated in vacuo. The residual solid was purified by chromatog-
raphy on silica, employing EtOAc: MeOH (10:1) as eluent, to give 39 as
an off-white solid (0.39 g; 29%), identical (mp, NMR, MS) to that
prepared above.
PMB-Protected 2-Amino-8-aryl-6-cyclohexylmethoxypurines
(40−42): Method III. General Procedure. A mixture of 39 (1.0 mol
equiv), the appropriate benzamide (15) or sulfonamide (16) (1.0 mol
equiv), dichloro [1,1′-bis(di-tert-butylphosphino)]ferrocene palladium-
(II) (‘Pd-118’) (0.5 mol equiv), and Cs2CO3 (2.0 mol equiv) in
dioxane:H2O (4:1, 2.5 mL) was purged under N2 for 15 min and stirred
for4hat110°C.ThereactionmixturewasfilteredthroughCelite, EtOAc
(10 mL) was added, and the organic fraction was washed sequentially
with saturated aqueousNH4Clandbrine, dried overMgSO4, filtered, and
concentrated in vacuo. The target compound was isolated by
chromatography on silica as described.
1
1603, 1580 cm−1. H NMR (CDCl3, 300 MHz) δ 0.92−1.17 (5H, m,
cyclohexyl), 1.55−1.97 (6H, m, cyclohexyl), 2.18 (3H, s, ArCH3), 4.32
(2H, d, J = 6.0 Hz, OCH2), 4.93 (2H, s, NH2), 5.05 (2H, s, N-CH2-Ar),
5.37 (2H, s, O-CH2-Ar), 6.82−6.87 (2H, m, Ar-H), 7.10−7.50 (10H, m,
Ar-H), 8.01 (1H, d, J = 7.5 Hz, Ar-H). 13C NMR (CDCl3, 75 MHz) δ
18.2, 26.1, 26.9, 30.3, 37.8, 46.7, 67.2, 72.5, 115.7, 125.6, 128.0, 128.5,
128.6, 128.8, 128.9, 131.8, 132.2, 132.6, 134.3, 136.4, 136.6, 140.6, 149.2,
155.6, 159.8, 162.1, 167.5.
3-(2-Amino-6-cyclohexylmethoxy-9H-purin-8-yl)-2-methylphenol
(36). To a solution of 34 (125 mg, 0.23 mmol) in acetic acid (10 mL),
containing hydrochloric acid (12M, 0.25 mL), was added Pd (10% on
carbon, 125 mg), and the reaction mixture was stirred under H2 at
atmospheric pressure for 12 h. Water (10 mL) was added, and the
mixture was heated under reflux for 1 h and filtered through Celite, and
the solvents were evaporated under reduced pressure. Purification by
semipreparative HPLC afforded the title compound as white crystals (11
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-N-(4-methoxy-
benzyl)-2-methylbenzamide (40). Prepared inaccordance withmethod
III from 39 (98 mg, 0.26 mmol) and 15 (84 mg, 0.26 mmol). Purification
by chromatography on silica using a gradient eluent of 100% DCM to
DCM:MeOH (5:1) afforded the title compound as a beige solid (54 mg;
42%); mp 252−255 °C. IR 3485, 3185, 2926, 1580, 1512 cm−1. 1H NMR
(DMSO-d6,500MHz)δ0.98−1.09(2H, m,cyclohexyl), 1.13−1.32(3H,
m, cyclohexyl), 1.62−1.75 (3H, m, cyclohexyl), 1.77−1.87 (3H, m,
cyclohexyl), 2.43 (3H, s, Ar-CH3), 3.74 (3H, s, OCH3), 4.24 (2H, d, J =
6.4 Hz, OCH2), 4.39 (2H, d, J = 6.0 Hz, NCH2), 6.28 (2H, s, NH2), 6.91
(2H, d, J = 8.7 Hz, H-3′), 7.28 (2H, d, J = 8.7 Hz, H-2′), 7.32−7.41 (2H,
m, H-4, H-5), 7.61 (1H, br d, J = 7.0 Hz, H-6), 8.87 (1H, t, J = 6.0 Hz,
NH), 12.63 (1H, s, N9-H). 13C NMR (DMSO-d6, 125 MHz) δ 17.3, 25.2,
26.0, 29.3, 36.8, 41.8, 55.0, 70.6, 113.7, 125.5, 127.6, 128.5, 130.4, 131.5,
139.2, 158.2, 159.7, 168.9. HRMS [M + H]+ (C28H33N6O3) calcd
501.2609, obsd 501.2598.
1
mg, 13%). H NMR (DMSO-d6, 300 MHz) δ 0.99−1.84 (m, 11H,
cyclohexyl), 2.17 (3H, s, Ar-CH3), 4.21 (2H, d, J = 6.3 Hz, OCH2), 6.80
(1H, d, J = 7.8 Hz, Ar-H), 6.90 (1H, d, J = 7.5 Hz, Ar-H), 6.99−7.05 (1H,
dd, J = 7.8, 7.5 Hz, Ar-H), 8.43 (2H, br s, NH2). HRMS [M + H]+
(C19H24N5O2) calcd 354.1925, obsd 354.1925.
3-(2-Amino-6-cyclohexylmethoxy-9H-purin-8-yl)-2-methylben-
zoic Acid (37). Prepared as described for 36 from 35 (105 mg, 0.19
mmol)andPd(10% oncarbon, 105 mg);purificationbysemipreparative
1
HPLC gave the title compound as a white powder (17 mg, 24%). H
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-N,N-bis(4-
methoxybenzyl)benzenesulfonamide (41). Prepared following meth-
od III from 8-bromo-6-(cyclohexylmethoxy)-9H-purin-2-amine 39 (32
mg, 0.098 mmol), except that the diisopropyl (3-(N,N-bis(4-
methoxybenzyl)sulfamoyl)phenyl)boronate was generated in situ as
follows: To a solution of 3-bromo-N,N-bis(4-methoxybenzyl)-
benzenesulfonamide 13 (62 mg, 0.13 mmol) in THF (2 mL) at −78
°Cwas addedn-BuLi(2.5Minhexane, 55 μL, 0.14 mmol)dropwise. The
mixture was stirred for 10 min, triisopropylborate (35 μL, 0.15 mmol)
wasadded, andstirringwascontinuedforafurther15minat −78°C. The
reaction mixture was allowed to warm to room temperature overnight,
and the solvent was evaporated under reduced pressure to give
diisopropyl (3-(N,N-bis(4-methoxybenzyl)sulfamoyl)phenyl)boronate
(72 mg, 0.137 mmol), which was used directly. Purification by
chromatography on silica, using a gradient elution from EtOAc:petrol
(65:15) to 100% EtOAc, yielded the title compound as a white solid (18
mg; 29%). 1H NMR (CDCl3, 500 MHz) δ 0.91−1.05 (2H, m,
cyclohexyl), 1.14−1.29 (3H, m, cyclohexyl), 1.63−1.87 (6H, m,
cyclohexyl), 3.69 (6H, s, OCH3), 4.03−4.23 (6H, m, NCH2, OCH2),
4.94 (2H, s, NH2), 6.66 (4H, d, J = 8.1 Hz, H-3′), 6.85 (4H, d, J = 7.9 Hz,
H-2′), 7.48 (1H, dd, J = 7.6, 7.9 Hz, H-3), 7.72 (1H, d, J = 7.6 Hz, H-4),
8.29 (1H, br s, H-6), 8.40 (1H, s, H-2). 13C NMR (CDCl3, 125 MHz) δ
25.8, 26.6, 29.9, 37.2, 49.8, 55.3, 72.3, 113. 9, 116.6, 124.6, 127.5, 127.8,
129.8, 129.9, 130.4, 131.1, 141.9, 146.6, 155.5, 159.2, 159.4, 161.7.
HRMS [M + H]+ (C34H39N6O5S) calcd 643.2697, obsd 643.2693.
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-N,N-bis(4-me-
thoxybenzyl)-2-methylbenzenesulfonamide (42). Prepared in accord-
ance with methodIII from 39(104 mg, 0.32 mmol) and 16(171mg, 0.32
NMR DMSO-d6, 300 MHz) δ 0.99−1.71 (m, 11H, cyclohexyl), 2.36
(3H, s, ArCH3), 4.21 (2H, d, J = 6.3 Hz, CH2), 7.15−7.20 (1H, m, Ar-H),
7.34 (1H, dd, J = 1.2, 7.5 Hz, Ar-H), 7.41 (1H, dd, J = 1.2, 7.5 Hz, Ar-H).
13C NMR (DMSO-d6, 75 MHz) 18.3, 27.3, 27.9, 31.0, 31.3, 39.3, 41.1,
73.3, 126.7, 129.9, 130.7, 132.6, 135.2, 142.8, 145, 150.0, 161.8. HRMS
[M + H]+ (C20H24N5O3) calcd 382.1874, obsd 382.1872.
2-Amino-6-cyclohexylmethoxy-purine-9-carboxylic Acid t-Butyl
Ester (38). To a stirred suspension of 5 (98.4 mg; 4.0 mmol) and
potassium carbonate (67.1 mg; 4.8 mmol) in DMF (12 mL) was added
di-t-butyl dicarbonate (1.2 mL; 5.0 mmol), and the mixture was stirred at
room temperature for 18 h. Water (20 mL) was added, and the reaction
mixture was stirred a further 1 h, whereupon the resulting white
precipitate was collected and triturated with Et2O to give the product as a
white solid (1.16 g; 84%); Rf = 0.61 (F). 1H NMR (CDCl3, 300 MHz) δ
1.01−1.32 (5 H, m, cyclohexyl), 1.68 (9H, s, C(CH3)3), 1.74−1.92 (6H,
m,cyclohexyl), 4.28(2H,d, J=6.3Hz, OCH2), 5.14(2H, brs,NH2), 8.00
(1H, s, NH). 13C NMR (CDCl3, 75 MHz) δ 26.1, 26.8, 28.4, 30.1, 37.6,
72.5, 86.5, 137.7, 146.6, 153.9, 161.1, 162.3. HRMS [M]+ (C17H25N5O3)
calcd 347.1948, obsd 347.1957.
2-Amino-8-bromo-6-cyclohexylmethoxypurine (39). To a stirred
suspension of 38 (1.13 g; 3.26 mmol) in MeCN (36 mL) and water (8.6
mL) was added N-bromosuccinimide (87.4 mg; 4.89 mmol). The
mixture was stirred at room temperature for 1 h, filtered, and the filtrate
concentrated under reduced pressure to give a tan solid. Purification by
chromatography on silica using gradient elution with EtOAc:petrol (2:1)
as eluent afforded the title compound as an off-white solid (442 mg,
41.6%); mp 228−229 °C. IR 3482, 3303, 3163, 2921, 2850, 1623, 1563
L
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
mmol). Purification by chromatography on silica with a gradient elution
of 100% DCM to DCM:MeOH (10:1) gave the title compound as an-off
white solid (24 mg; 24%); mp 209−210 °C. IR 3366, 2926, 1582, 1512
cm−1. 1H NMR (DMSO-d6, 500 MHz) δ 0.99−1.11 (2H, m, cyclohexyl),
1.13−1.32 (3H, m, cyclohexyl), 1.63−1.76 (3H, m, cyclohexyl), 1.78−
1.88 (3H, m, cyclohexyl), 2.63 (3H, s, Ar-CH3), 3.72 (6H, s, OCH3),
4.21−4.30(6H, m, OCH2, NCH2), 6.35(2H, s, NH2), 6.84(4H, d, J=8.7
Hz, H-3′), 6.98 (4H, d, J = 8.7 Hz, H-2′), 7.50 (1H, dd, J = 7.3, 7.9 Hz, H-
5), 7.84 (1H, d, J = 7.3 Hz, H-4), 7.95 (1H, d, J = 7.9 Hz, H-6), 12.76 (1H,
s, N9-H). 13C NMR (DMSO-d6, 125 MHz) δ 16.9, 25.2, 25.9, 29.3, 36.8,
48.9, 55.0, 70.6, 113.8, 114.5, 126.0, 127.3, 129.7, 129.9, 133.7, 134.6,
136.3, 139.6, 146.0, 155.9, 158.7, 159.8, 160.4. HRMS [M + H]+
(C35H41N6O5S) calcd 657.2854, obsd 657.2840.
2-Amino-8-aryl-6-cyclohexylmethoxypurines (43−45): Method IV.
General Procedure. A solution of the appropriate PMB-protected 2-
amino-8-aryl-6-cyclohexymethyloxypurine (0.04−0.2 mmol) in TFA
(1−2 mL) was stirred at the specified temperature until the reaction was
complete. The reaction mixture was neutralized with saturated aqueous
Na2CO3 solution and extracted with EtOAc (3 × 10 mL), and the
combined organic fractions were dried (MgSO4), filtered, and
evaporated in vacuo. The target compounds were isolated by
chromatography on silica as described.
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-2-methylben-
zamide (43). Synthesized from 40 (90 mg, 0.18 mmol) and TFA (1 mL)
(48 h, 110 C) following method IV. Purification by chromatography on
silica, using a gradient elution of 100% EtOAc to EtOAc:MeOH (5:1),
gave the required purine as a white solid (4 mg; 6%); mp 271−272 °C. IR
3379, 3210, 2932, 2857, 1662, 1624, 1591, 1574, 1533 cm−1. 1H NMR
(DMSO-d6, 500 MHz) δ 1.02−0.14 (m, 2H, cyclohexyl), 1.16−1.36 (m,
3H, cyclohexyl), 1.65−1.79 (m, 3H, cyclohexyl), 1.80−1.92 (m, 3H,
cyclohexyl), 2.52 (s, 3H, Ar-CH3), 4.28 (d, 2H, J = 6.4 Hz, OCH2), 6.32
(s, 2H, NH2), 7.37 (dd, 1H, J = 7.4, 7.5 Hz, H-5), 7.43 (d, 1H, J = 7.4 Hz,
H-4), 7.52 (s, 1H, NH), 7.63 (d, 1H, J = 7.5 Hz, H-6), 7.88 (s, 1H, NH),
12.67 (s, 1H, NH). 13C NMR (DMSO-d6, 125 MHz) δ 17.5, 26.8, 27.5,
30.8, 38.7, 72.8, 126.9, 129.5, 132.4, 132.6, 135.7, 139.7, 161.6, 175.2.
HRMS [M + H]+ (C20H25N6O2) calcd 381.2034, obsd 381.2030.
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-
benzenesulfonamide (44). In accordance with method IV from 41 (29
mg; 0.045 mmol) and TFA (1 mL) (3h at 20 °C). Purification by
chromatography on silica, employing a gradient elution of EtOAc:petrol
(4:1) to 100% EtOAc, furnished the title compound as a white solid (11
mg;61%);mp159−162°C. IR3323, 2920, 2851, 1651, 1625, 1530cm−1.
1HNMR(MeOD-d4, 500MHz)δ1.09−1.20(2H, m, cyclohexyl), 1.24−
1.40 (3H, m, cyclohexyl), 1.68−1.84 (3H, m, cyclohexyl), 1.90−1.97
(3H, m, cyclohexyl), 4.33(2H, d, J=6.2Hz, OCH2), 7.68(1H, dd, J=7.8,
7.9 Hz, H-5), 7.99 (1H, d, J = 7.8 Hz, H-4), 8.22 (1H, d, J = 7.9 Hz, H-6),
8.58 (1H, s, H-2). 13C NMR (MeOD-d4, 125 MHz) δ 26.9, 27.57,30.81,
38.78, 73.0, 125.21, 128.4, 130.75, 130.88, 131.8, 146.22, 161.61. HRMS
[M + H]+ (C18H23N6O3S) calcd 403.1547, obsd 403.1546.
3-(2-Amino-6-(cyclohexylmethoxy)-9H-purin-8-yl)-2-methylben-
zenesulfonamide (45). Prepared in accordance with method IV from 42
(60 mg; 0.092 mmol) and TFA (1 mL) (90 min at 80 °C); isolated by
chromatography on silica, using a gradient elution of 100% EtOAc to
EtOAc:MeOH (5:1), to give the title compound as a white powder (30
mg;78%);mp258−259°C. IR3503, 3407, 2927, 2851, 1629, 1587cm−1.
1HNMR(DMSO-d6, 500MHz)δ0.94−1.10(2H, m,cyclohexyl), 1.15−
1.33 (3H, m, cyclohexyl), 1.61−1.75 (3H, m, cyclohexyl),1.77−1.86
(3H, m, cyclohexyl), 2.68 (3H, s, Ar-CH3), 4.24 (2H, d, J = 6.3 Hz,
OCH2), 6.32 (2H, s, NH2), 7.49 (1H, dd, J = 7.6, 7.8.Hz, H-5), 7.55 (2H,
s, NH2), 7.73 (1H, d, J = 7.6 Hz, H-4), 8.00 (1H, d, J = 7.8 Hz, H-6), 12.70
(1H, s, NH). 13C NMR (DMSO-d6, 125 MHz) δ 17.0, 25.2, 26.0, 29.2,
36.8, 70.6, 125.7, 143.4, 159.7, 160.3. HRMS [M + H]+ (C19H25N6O3S)
calcd 417.1703, obsd 417.1697.
Crystallography. Crystals of monomeric CDK2 (kindly supplied by
C.A. Minshull, AstraZeneca, Macclesfield, UK) were soaked for 24 h in a
5 mM solution of compound 26 in mother liquor (50 mM ammonium
acetate, 10% PEG-3350, 15 mM NaCl, 100 mM HEPES, pH = 7.4, 10%
DMSO). The crystals were cryoprotected in 50 mM ammonium acetate,
10% PEG-3350, 25% glycerol, 100 mM HEPES, pH 7.4, and mounted
into a 100 K nitrogen stream. Data collection was carried out at
AstraZeneca, Macclesfield, UK, using a Bruker-Nonius FR591 rotating
anode generator and an XrayResearch Mar345 image plate. Compounds
29, 32, 36, 37, 44, and 45 were cocrystallized with CDK2/cyclin A as
previously described.20 Before data collection, crystals were briefly
immersed in cryoprotectant (8 M sodium formate) before freezing.
Data processing was carried out using MOSFLM,38 SCALA,39 and
other programs of the CCP4 suite.40 The structures of CDK2/
compound 26 and of compounds 29, 32, 36, and 37 bound to CDK2/
cyclin A were solved by molecular replacement using MOLREP41 taking,
respectively, the protein atoms of an unpublished structure of CDK2
bound to a derivative of 5 or CDK2/cyclin A/7 (PDB accession code
1H1S22) as the search model. Structures of 44 and 45 bound to CDK2/
cyclin A were solved by Phaser,42 using as the search model a high-
resolution structure of a recruitment peptide bound to CDK2/cyclin A
(PDB accession code 2CCH). Each solution was then subjected to rigid
body refinement in REFMAC,43 revealing unambiguous electron density
intheCDK2ATP-bindingsite,consistentwiththeexpectedshapesofthe
inhibitors. Models of 26, 29, 32, 36, and 37 were created using the CCP4
Sketcher40 and manually docked into the density using either O43,44
(compounds 26 and 29) or Coot45 (compounds 32, 36, and 37). Models
of 44 and 45 were created using the Elbow Builder utility and built into
the electron density using Coot. The inhibitor atoms were kept in all
subsequent models during refinement. The structures were then refined
further by rounds of manual rebuilding in O (for compounds 26 and 29)
orinCoot(compounds32, 36, 37, 44,and45)andrestrained refinement
in REFMAC. Toward the end of refinement, waters were added using
ARP/wARP46 (compounds 26 and 29) or the Coot water picking utility
(32, 36, 37, 44, and 45) and manually verified. The statistics for the data
sets and for the crystallographic refinement are presented in Supporting
Information Table S1.
Modeling. The refined structure of mCDK2/compound 26 was
aligned with CDK2/inhibitor structures using LSQKAB.47 The kinases
were aligned based on the α carbon positions of the residues Glu81,
Phe82, Leu83, His84, Gln85, and Ala31. The hinge residues (Glu81−
Gln85) were used as all the inhibitors investigated form hydrogen bonds
with the backbone atoms in this region. Ala31 was used to allow the
reliable alignment of the monomeric and the cyclin-bound CDK2
structures.48
Conformation-energy profiles were calculated with Gaussian.47
Models were created using the CCP4 monomer sketcher40 and InsightII
(Accelrys, CA, USA) and then converted into Z matrices. The purine and
aryl rings were constrained to coplanarity, and the models energy
minimized with the 6-31G** basis set. Rigid potential energy scans were
carried out at 5° intervals between the two coplanar conformations using
the 3-21G basis set.
ASSOCIATED CONTENT
* Supporting Information
■
S
Quantum mechanical calculations, structural data for the CDK2/
cyclin A/17 and CDK2/cyclin A/18 complexes, and combustion
analyses. This material is available free of charge via the Internetat
http://pubs.acs.org.
Accession Codes
The structures of CDK2/compound 26, and compounds 32, 29,
36, 37, 44, and 45 bound to CDK2/cyclin A have been deposited
in the PDB with accession codes 1W8C, 4CFN, 4CFM, 4CFV,
4CFU, 4CFX, and 4CFW, respectively.
AUTHOR INFORMATION
Corresponding Authors
■
*For J.A.E. (structural biology): phone, (44) 191 246 4422; fax,
(44) 191 246 4301; E-mail, jane.endicott@ncl.ac.uk.
*For R.J.G. (medicinal chemistry): phone, (44) 191 222 8591;
fax, (44) 191 222 8591; E-mail, r.j.griffin@ncl.ac.uk.
Present Addresses
#J.A.E. and M.E.M.N.: Northern Institute for Cancer Research,
Newcastle University.
M
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
▽A.E.: Centre de Biochimie Structurale, Centre National de la
(14) Berthet, C.; Aleem, E.; Coppola, V.; Tessarollo, L.; Kaldis, P.
CDK2 knockout mice are viable. Curr. Biol. 2002, 13, 1175−1185.
(15) Molenaar, J. J.; Ebus, M. E.; Geerts, D.; Koster, J.; Lamers, F.;
Valentijn, L. J; Westerhout, E. M.; Versteeg, R.;Caron, H. N. Inactivation
of CDK2 is synthetically lethal to MYCN over-expressing cancer cells.
Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 12968−12973.
(16) 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, E1019−E1027.
(17) Cheng, C. K.; Gustafson, W. C.; Charron, E.; Houseman, B. T.;
Zunder, E.; Goga, A.; Gray, N. S.; Pollok, B.; Oakes, S. A.; James, C. D.;
Shokat, K. M.;Weiss, W. A.;Fan, Q.-W. Dualblockade oflipidandcyclin-
dependent kinases induces synthetic lethality in malignant glioma. Proc.
Natl. Acad. Sci. U. S. A. 2012, 109, 12722−12727.
(18)Arris, C. E.;Boyle, F. T.;Calvert, A. H.;Curtin, N. J.;Endicott, J. A.;
Garman, E. F.; Gibson, A. E.; Golding, B. T.; Grant, S.; Griffin, R. J.;
Jewsbury, P.; Johnson, L. N.; Lawrie, A. M.; Newell, D. R.; Noble, M. E.;
Sausville, E. A.; Schultz, R.; Yu, W. Identification of novel purine and
pyrimidine cyclin-dependent kinase inhibitors with distinct molecular
interactions and tumor cell growth inhibition profiles. J. Med. Chem.
2000, 43, 2797−2804.
(19) Gibson, A. E.; Arris, C. E.; Bentley, J.; Boyle, F. T.; Curtin, N. J.;
Davies, T. G.; Endicott, J. A.; Golding, B. T.; Grant, S.; Griffin, R. J.;
Jewsbury, P.; Johnson, L. N.; Mesguiche, V.; Newell, D. R.; Noble, M. E.;
Tucker, J. A.; Whitfield, H. J. Probing the ATP ribose-binding domain of
cyclin-dependent kinases 1 and 2 with O6-substituted guanine
derivatives. J. Med. Chem. 2002, 45, 3381−3393.
Recherche Scientifique (CNRS), Institut National de la Sante
de la Recherche Medicale-Unites Mixtes de Recherche 5048 and
1054, Universite Montpellier I, Montepellier, France.
Author Contributions
́
et
́
́
́
⊥These authors contributed equally to the work described in this
paper.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We would like to thank Julie Tucker and staff at AstraZeneca for
providing facilities and for their assistance with data collection.
We also thank beamline staff at the ESRF, Elettra, and The
Diamond Light Source who provided excellent facilities for data
collection. The authors would like to thank I. Taylor for technical
support and T. Davies for useful discussions. This research was
supported bygrantsfrom CancerResearch UK, Medical Research
Council, and AstraZeneca.
ABBREVIATIONS USED
■
CDK, cyclin-dependent kinase; pRb, retinoblastoma protein;
PI3-K, phosphatidylinositol-3-kinase; DMAP, 4-dimethylamino-
pyridine; PMB, 4-methoxybenzyl; HEPES, 4-(2-hydroxyethyl)-
1-piperazineethanesulfonic acid
(20) Davies, T. G.; Bentley, J.; Arris, C. E.; Boyle, F. T.; Curtin, N. J.;
Endicott, J.-A.; Gibson, A. E.; Golding, B. T.; Griffin, R. J.; Hardcastle, I.
R.; Jewsbury, P.; Johnson, L. N.; Mesguiche, V.; Newell, D. R.; Noble, M.
E. M.; Tucker, J. A.; Wang, L.; Whitfield, H.-J. Structure-based design of a
potent purine-based cyclin-dependent kinase inhibitor. Nature Struct.
Biol. 2002, 9, 745−749.
(21) Sayle, K. L.; Bentley, J.; Boyle, F. T.; Calvert, A. H.; Cheng, Y.;
Curtin, N.J.;Endicott, J.A.;Golding,B.T.;Hardcastle, I.R.;Jewsbury, P.;
Mesguiche, V.; Newell, D. R.; Noble, M. E.; Parsons, R. J.; Pratt, D. J.;
Wang, L. Z.; Griffin, R. J. Structure-based design of 2-arylamino-4-
cyclohexylmethyl-5-nitroso-6-aminopyrimidine inhibitors of cyclin-
dependent kinases 1 and 2. Bioorg. Med. Chem. Lett. 2003, 13, 3079−
3082.
(22) Hardcastle, I. R.; Arris, C. E.; Bentley, J.; Boyle, F. T.; Chen, Y.;
Curtin, N. J.; Endicott, J. A.; Gibson, A. E.; Golding, B. T.; Griffin, R. J.;
Jewsbury, P.; Menyerol, J.; Mesguiche, V.; Newell, D. R.; Noble, M. E.;
Pratt, D. J.; Wang, L. Z.; Whitfield, H. J. N2-Substituted O6-
cyclohexylmethylguanine derivatives: potent inhibitors of cyclin-
dependent kinases 1 and 2. J. Med. Chem. 2004, 47, 3710−3722.
(23) Schulze-Gahmen, U.; Brandsen, J.; Jones, H. D.; Morgan, D. O.;
Meijer, L.; Vesely, J.; Kim, S. H. Multiple modes of ligand recognition:
crystal structures of cyclin-dependent protein kinase 2 in complex with
ATP and two inhibitors, olomoucine and isopentenyladenine. Proteins
1995, 22, 378−391.
(24) Gray, N. S.; Wodicka, L.; Thunnissen, A. M.; Norman, T. C.;
Kwon, S.; Espinoza, F. H.; Morgan, D. O.; Barnes, G.;LeClerc, S.;Meijer,
L.; Kim, S. H.; Lockhart, D. J.; Schultz, P. G. Exploiting chemical libraries,
structure, and genomics in the search for kinase inhibitors. Science 1998,
281, 533−538.
REFERENCES
■
(1) Malumbres, M.; Barbacid, M. Cell cycle, CDKs and cancer: a
changing paradigm. Nature Rev. Cancer 2009, 9, 153−166.
(2) Lapenna, S.; Giordano, A. Cell cycle kinases as therapeutic targets
for cancer. Nature Rev. Drug Discovery 2009, 8, 547−566.
(3) Romagosa, C.; Simonetti, S.; Lopez-Vicente, L.; Mazo, A.; Lleonart,
M. E.; Castellvi, J.; Ramon, Y. C. S. p16(Ink4a) overexpression in cancer:
a tumor suppressor gene associated with senescence and high-grade
tumors. Oncogene 2011, 30, 2087−2097.
(4) Sherr, C. J.; Roberts, J. M. CDK inhibitors: positive and negative
regulators of G1-phase progression. Genes Dev. 1999, 13, 1501−1512.
(5) Hunt, K. K.; Keyomarsi, K. Cyclin E as a prognostic and predictive
marker in breast cancer. Semin. Cancer Biol. 2005, 15, 319−326.
(6) Wander, S. A.; Zhao, D.; Slingerland, J. M. p27: a barometer of
signaling deregulation and potential predictor of response to targeted
therapies. Clin. Cancer Res. 2011, 17, 12−18.
(7) Malumbres, M.; Pevarello, P.; Barbacid, M.; Bischoff, J. R. CDK
inhibitors in cancer therapy: what is next? Trends Pharmacol. Sci. 2008,
29, 16−21.
(8) Hardcastle, I. R.; Golding, B. T.; Griffin, R. J. Designing inhibitors of
cyclin-dependent kinases. Annu. Rev. Pharmacol. Toxicol. 2002, 42, 325−
348.
(9) Bettayeb, K.; Tirado, O. M.; Marionneau-Lambot, S.; Ferandin, Y.;
Lozach, O.; Morris, J. C.;Mateo-Lozano, S.; Drueckes, P.; Schachtele, C.;
Kubbutat, M. H.;Liger, F.;Marquet, B.;Joseph, B.;Echalier, A.;Endicott,
J. A.; Notario, V.; Meijer, L. Meriolins, a new class of cell death inducing
kinase inhibitors with enhanced selectivity for cyclin-dependent kinases.
Cancer Res. 2007, 67, 8325−8334.
(10) Sausville, E. Cell cycle regulatory kinase modulators: interim
progress and issues. Curr. Top. Med. Chem. 2005, 5, 1109−1117.
(11) Galons, H.; Oumata, N.; Meijer, L. Cyclin-dependent kinase
inhibitors: a survey of recent patent literature. Expert Opin. Ther. Pat.
2010, 20, 377−404.
(12) Tetsu, O.; McCormick, F. Proliferation of cancer cells despite
CDK2 inhibition. Cancer Cell 2003, 3, 233−245.
(13) Ortega, S.; Presto, I.; Odajima, J.; Martín, A.; Dubus, P.; Sotillo, R.;
Barbero, J. L.; Malumbres, M.; Barbacid, M. Cyclin-dependent kinase 2 is
essentialformeioticbutnotformitoticcelldivisioninmice. NatureGenet.
2003, 35, 25−31.
(25) Dreyer, M. K.;Borcherding, D. R.; Dumont, J. A.; Peet, N. P.;Tsay,
J. T.; Wright, P. S.; Bitonti, A. J.; Shen, J.; Kim, S. H. Crystal structure of
human cyclin-dependent kinase 2 in complex with the adenine-derived
inhibitor H717. J. Med. Chem. 2001, 44, 524−530.
(26) Beattie, J. F.; Breault, G. A.;Ellston, R. P.;Green, S.;Jewsbury, P. J.;
Midgley, C. J.; Naven, R. T.; Minshull, C. A.; Pauptit, R. A.; Tucker, J. A.;
Pease, J. E. Cyclin-dependent kinase 4 inhibitors as a treatment for
cancer. Part 1: Identification and optimization of substituted 4,6-bis
anilino pyrimidines. Bioorg. Med. Chem. Lett. 2003, 13, 2955−2960.
(27) Breault, G. A.; Ellston, R. P.; Green, S.; James, S. R.; Jewsbury, P. J.;
Midgley, C. J.; Pauptit, R. A.; Minshull, C. A.; Tucker, J. A.; Pease, J. E.
N
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Cyclin-dependent kinase 4 inhibitors as a treatment for cancer. Part 2:
Identification and optimization of substituted 2,4-bis anilino pyrimi-
dines. Bioorg. Med. Chem. Lett. 2003, 13, 2961−2966.
(28) Thomas, H. D.; Wang, L.-Z.; Roche, C.; Bentley, J.; Cheng, Y.;
Hardcastle, I. R.; Golding, B. T.; Griffin, R. J.; Curtin, N. J.; Newell, D. R.
Preclinical invitroandinvivo evaluationof thepotent andspecific cyclin-
dependent kinase 2 inhibitor NU6102 and a water soluble prodrug
NU6301. Eur. J. Cancer 2011, 47, 2052−2059.
(29) Fabbro, D.; Ruetz, S.; Buchdunger, E.; Cowan-Jacob, S.; Fendrich,
G.; Liebetanz, J.; Mestan, J.; O’Reilly, T.; Traxler, P.; Chaudhuri, B.;
Fretz, H.; Zimmermann, J.; Meyer, T.; Caravatti, G.; Furet, P.; Manley, P.
Protein kinases as targets for anticancer agents: from inhibitors to useful
drugs. Pharmacol. Ther. 2002, 93, 79.
(30) De-Azevedo, W. F.; Leclerc, S.;Meijer, L.; Havlicek, L.; Strnad, M.;
Kim, S. H. Inhibition of cyclin-dependent kinases by purine analogues:
crystal structure of human cdk2 complexed with roscovitine. Eur. J.
Biochem. 1997, 243, 518−526.
(46) Lamzin, V. S.; Wilson, K. S. Automated refinement of protein
models. Acta Crystallogr., Sect. D: Biol. Crystallogr. 1993, 49, 129−147.
(47) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.;
Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.;
Cammi, R.;Mennucci, B.;Pomelli, C.;Adamo, C.;Clifford, S.;Ochterski,
J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J.
V.; Baboul, A. G.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-
Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe,
M.; Gill, P. M. W.; Bohnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian
1998, version A.7; Gaussian, Inc.: Pittsburgh, PA, 1998.
(48) Davies, T. G.; Tunnah, P.; Meijer, L.; Marko, D.; Eisenbrand, G.;
Endicott, J. A.; Noble, M. E. Inhibitor binding to active and inactive
CDK2: the crystal structure of CDK2-cyclin A/indirubin-5-sulphonate.
Structure (Cambridge, MA, U. S.) 2001, 9, 389−397.
(31) Legraverend, M.; Tunnah, P.; Noble, M.; Ducrot, P.; Ludwig, O.;
Grierson, D. S.; Leost, M.; Meijer, L.; Endicott, J. Cyclin-dependent
kinase inhibition by new C-2 alkynylated purine derivatives and
molecular structure of a CDK2-inhibitor complex. J. Med. Chem. 2000,
43, 1282−1292.
(32) Schlosser, M.; Michel, D. About the “physiological size” of fluorine
substituents: comparison of sensorially active compounds with fluorine
and methyl substituted analogues. Tetrahedron 1996, 52, 99−108.
(33) Mettey, Y.; Gompel, M.; Thomas, V.; Garnier, M.; Leost, M.;
Ceballos-Picot, I.; Noble, M.; Endicott, J.; Vierfond, J. M.; Meijer, L.
Aloisines, a new family of CDK/GSK-3 inhibitors. SAR study, crystal
structure in complex with CDK2, enzyme selectivity, and cellular effects.
J. Med. Chem. 2003, 46, 222−236.
(34)Wang, S.;Meades, C.;Wood, G.;Osnowski, A.;Anderson, S.;Yuill,
R.; Thomas, M.; Mezna, M.; Jackson, W.; Midgley, C.; Griffiths, G.;
Fleming, I.; Green, S.; McNae, I.; Wu, S. Y.; McInnes, C.; Zheleva, D.;
Walkinshaw, M. D.; Fischer, P. M. 2-Anilino-4-(thiazol-5-yl)pyrimidine
CDK inhibitors: synthesis, SAR analysis, X-ray crystallography, and
biological activity. J. Med. Chem. 2004, 47, 1662−1675.
(35)Brana, M.F.;Cacho, M.;Garcia, M. L.;Mayoral, E. P.;Lopez, B.;de
Pascual-Teresa, B.; Ramos, A.; Acero, N.; Llinares, F.; Munoz-Mingarro,
D.; Lozach, O.; Meijer, L. Pyrazolo[3,4-c]pyridazines as novel and
selective inhibitors of cyclin-dependent kinases. J. Med. Chem. 2005, 48,
6843−6854.
(36) Galons, H.; Oumata, N.; Gloulou, O.; Meijer, L. Cyclin-dependent
kinase inhibitors closer to market launch? Expert Opin. Ther. Pat. 2013,
23, 945−963.
(37) Guha, M. Cyclin-dependent kinase inhibitors move into Phase III.
Nature Rev. Drug. Discovery 2012, 11, 892−894.
(38) Leslie, A. G. W. Recent Changes to the MOSFLM Package for
Processing Film and Image Plate Data; Science and Engineering Research
Council UK, Daresbury Laboratory, Warrington, UK, 1992, p 26.
(39) Evans, P. R. Data Reduction; Science and Engineering Research
Council UK, Daresbury Laboratory: Warrington, UK, 1993; pp 114−
122.
(40) CCP4. The CCP4 suite: programs for protein crystallography.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 1994, 50, 760−763.
(41) Vagin, A.; Teplyakov, A. MOLREP: an automated program for
molecular replacement. J. Appl. Crystallogr. 1997, 30, 1022−1025.
(42) McCoy, A. J.; Grosse-Kunstleve, R. W.; Storoni, L. C.; Read, R. J.
Likelihood-enhanced fast translation functions. Acta Crystallogr., Sect. D:
Biol. Crystallogr. 2005, 61, 458−64.
(43) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-liklihood method. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 1997, 53, 240−255.
(44) Jones, T. A.; Zou, J. Y.; Cowan, S. W. Kjeldgaard. Improved
methods for binding protein models in electron density maps and the
location of errors in these models. Acta Crystallogr., Sect. A: Found.
Crystallogr. 1991, 47, 110−119.
(45) Emsley, P.; Cowtan, K. Coot: model-building tools for molecular
graphics. ActaCrystallogr., Sect. D:Biol. Crystallogr. 2004, 60, 2126−2132.
O
dx.doi.org/10.1021/jm401555v | J. Med. Chem. XXXX, XXX, XXX−XXX