Structure-based drug design of a highly potent CDK1,2,4,6 inhibitor with novel macrocyclic quinoxalin-2-one structure

Article abstract of DOI:10.1016/j.bmcl.2006.07.026

The design of a novel series of cyclin-dependent kinase (CDK) inhibitors containing a macrocyclic quinoxaline-2-one is reported. Structure-based drug design and optimization from the starting point of diarylurea 2, which we previously reported as a modera

Full text of DOI:10.1016/j.bmcl.2006.07.026

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5122–5126
Structure-based drug design of a highly potent CDK1,2,4,6 inhibitor
with novel macrocyclic quinoxalin-2-one structure
Nobuhiko Kawanishi,* Tetsuya Sugimoto, Jun Shibata, Kaori Nakamura,
Kouta Masutani, Mari Ikuta and Hiroshi Hirai
Department of Medicinal Chemistry, Banyu Tsukuba Research Institute in collaboration with Merck Research Laboratories,
Okubo-3, Tsukuba 300-2611, Ibaraki, Japan
Received 3 April 2006; revised 6 July 2006; accepted 11 July 2006
Available online 28 July 2006
Abstract—The design of a novel series of cyclin-dependent kinase (CDK) inhibitors containing a macrocyclic quinoxaline-2-one is
reported. Structure-based drug design and optimization from the starting point of diarylurea 2, which we previously reported as a
moderate CDK1,2,4,6 inhibitor [J. Biol.Chem. 2001, 276, 27548], led to the discovery of potent CDK1,2,4,6 inhibitor that were
suitable for iv administration for in vivo study.
Ó 2006 Elsevier Ltd. All rights reserved.
The cyclin-dependent kinase (CDK) protein family
plays key roles in cell-cycle regulation in eukaryotic
cells.2 Orderly cell-cycle progression requires CDK
activation, which is mainly controlled by expression of
their activator subunit, cyclin. CDK1 (CDC2) com-
plexed with cyclin B is key in G2/M phase progression.
CDK4 and CDK6 complexed with cyclin D, and CDK2
complexed with cyclin E or A, sequentially phosphory-
late retinoblastoma protein to facilitate G1/S progres-
sion. Retinoblastoma protein is a negative regulator of
transcription factor E2F; when hyperphosphorylated,
it releases E2F, which activates the transcription of
genes whose products are critical for cell-cycle progres-
sion. Deregulation of the cell cycle is a hallmark of
cancer;3 indeed, genetic or epigenetic changes that lead
to cyclin overexpression, or the absence or reduction
of CDK-inhibitor proteins are commonly observed in
human cancers. Consequently, CDK1,2,4, and 6 are
attractive targets for new anticancer drugs. Recently
the clinical progress of CDK1,2,4, and 6 inhibitors has
been reviewed.4
cal profile and value in cancer therapy of a
CDK1,2,4,6 inhibitor are expected to be different from
those of the CDK4-selective inhibitors; a potent
CDK1,2,4,6 inhibitor with minimum off-target activity
is desired. In the course of our structure–activity rela-
tionship study of the diarylurea class of compounds,
we found that compound 2, which lacks side chains on
the pyridine ring, inhibited CDK1,2,4, and 6 to a similar
degree (Table 1).5,6 We took compound 2 as the starting
point for the structure-based drug design of a potent
CDK1,2,4,6 inhibitor.7,8
The X-ray crystal structure of the CDK4 mimic CDK2
bound with compound 2 (Fig. 1) revealed that the urea
hydrogen and carbonyl group formed hydrogen bonds
with Val83 in the ATP-binding hinge region of the pro-
tein; compound 2 was coplanar with an intramolecular
hydrogen bond between pyridine and the urea hydro-
gen, as shown in Table 1.1,9 Compound 2 was insuffi-
ciently potent so we hypothesized that conversion of
the linear system to a ring to force coplanarity would
improve potency.
We previously reported the diarylurea class of com-
pounds, represented by compound 1 (Table 1), to be
novel selective CDK4 inhibitors.5,6 The pharmacologi-
The cyclic compounds that we synthesized were
designed to form hydrogen bonds with Val83; their
structures are shown in Table 2. Compound 3, which
had a cyclic urea structure, had no activity against
CDK4; steric hindrance might make coplanarity difficult
(Fig. 2). In contrast, the quinoxaline-2-one compound
(4) could take a coplanar conformation and fit, like
Keywords: CDK inhibitor; CDK1; CDK2; CDK4; CDK6; X-ray;
CDK4 mimic CDK4; E2F.
*
Corresponding author. Tel.: +81 29 877 2222; fax: +81 29 877
2029; e-mail: nobuhiko_kawanishi@merck.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.07.026

N. Kawanishi et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5122–5126
5123
Table 1. IC50 values of compounds 1 and 2 against members of the CDK family (enzyme assay)
Compound
IC50 (nM)
CDK1
1800
CDK2
440
CDK4
2.3
CDK6
O
O
N
Cl
H
1
NDa
H
N
HN
N
N
N
H
O
N
2
122
80
42
71
H
N
N
N
H
O
a ND, not determined.
O
O
O
N
N
N
H
H
N
H
N
N
N
N
H
O
N
O
O
N
H
H
3
4
2
Figure 2. Structures of compounds 2–4 showing 3’s steric hindrance.
2-one template to provide a more favorable coplanar
conformation in compounds 5 and 6. As expected, they
were better inhibitors of CDK4 (Table 3); compound 6
in particular showed high activity. The activities of com-
pounds 4-6 correlate with their dihedral angles, which
were optimized by a Gaussian method (HF/6-31G).10
Figure 1. X-ray crystal structure of 2 in the ATP-binding site of CDK4
mimic CDK2.1,9
Although the CDK-inhibitory activity of compound 6
was potent in the enzyme assay, its cellular potency
was less good (IC50 = 310 nM; E2F assay).11 We used
Table 2. Structures and IC50 values of compounds 2–4 against CDK4
O
N
Table 3. IC50 values of compounds 4–6 against CDK4 and the
dihedral angles between quinoxaline-2-one and benzoisothiazol-1-one
R1
R2
N
Compound
R1
IC50 (nM)
42
H
N
N
N
H
O
2
N
H
O
Compound
R2
IC50 (nM)
480
Dihedral anglea
38.4
N
O
3
4
>10,000
480
N
N
N
O
H
4
N
N
O
H
O
5 (X = CH2)
6 (X = S)
170
34.7
1.5
X
compound 2, in the ATP-binding pocket of CDK4; its
CDK4-inhibitory activity was moderate.
6.0
We tried to improve the potency further by optimizing
the isoindol-1-one moiety within the new quinoxalin-
a Data optimized by a Gaussian method (HF/6-31G).10

5124
N. Kawanishi et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5122–5126
Table 6. IC50 values of compounds 11 and 12 against CDK4 and the
solubility of their hydrochloride salts in saline
chains of 4-6 carbon atoms to link the quinoxaline-2-one
and the benzoisothiazol-1-one as shown in Table 4;
these linkers fixed the dihedral angle and filled the void
space around the ribose site of the ATP-binding pocket.
CDK4-inhibitory activities were measured for com-
pounds 7-9 (Table 4); compound 8, which has a 5-car-
bon linker, showed the best activity in both assays.
O
HCl
N
N
X
O
N
O
N
H
We further optimized compound 8 by inserting amines
into the linker. Table 5 summarizes the results: com-
pound 11, which has N-methylamino functionality,
showed little discrepancy between the two assays.
However, it was insufficiently soluble in water for iv dos-
ing so we went on to design more hydrophilic com-
pounds such as 12. Although its cellular activity was a
little decreased compared with compound 11, its solubil-
ity in aqueous saline solution was much improved
(Table 6). This result encouraged us to further optimize
the linker, and cellular potency was improved by the
introduction of a pyrrolidine ring (compound 13, Table
7). 3-D molecular modeling of compound 13 in the
Compound
X
IC50 (nM)
Solubility (lg/ml)
CDK4
E2F
in saline
11
12
S
NH
11
13
13
77
20
990
Table 7. IC50 values of compounds 12–14 as measured by the enzyme
and the cellular assay11
O
Linker
N
NH
O
N
Table 4. IC50 values of compounds 6–9 against CDK4 as measured by
the enzyme and the cellular assay11
O
N
H
Compound
IC50 (nM)
CDK4
Compound
Linker
IC50 (nM)
CDK4
E2F
E2F
77
N
12
13
14
13
O
N
S
6
6.0
310
(S)
(S)
29
27
N
N
O
N
H
(R)
6.4
3.6
N
7 (n = 4)
8 (n = 5)
160
NDa
30
O
(CH2)n
N
S
ATP-binding pocket of the CDK4 mimic CDK2 sug-
gested that incorporation of a methyl group on the pyr-
rolidine ring would improve binding affinity to CDK4
by means of additional lipophilic interactions with
Ala144, Leu134, and Asn132. Indeed, compound 14
was highly inhibitory against CDK4 in both the enzyme
and the cell-based assay. As shown in Figure 3, the
X-ray structure of compound 14 co-crystallized with
O
3.6
N
O
N
H
9 (n = 6)
11
>100
a ND, not determined.
Table 5. IC50 values of compounds 8, 10 and 11 against CDK4 as
measured by the enzyme and the cellular assay11
Lys33
O
N
X
S
O
Asn132
Ala144
N
O
N
H
Compound
X
IC50 (nM)
Leu134
Val83
CDK4
E2F
8
10
11
CH2
NH
3.6
6.0
11
30
600
13
Figure 3. X-ray crystal structure of the CDK inhibitor 14 in the ATP-
binding site of CDK2.12
NMe

N. Kawanishi et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5122–5126
5125
the CDK4 mimic CDK2 shows its coplanar structure; it
forms hydrogen bonds with Val83 like those of the lead
compound 2; there are additional hydrogen bonds with
Lys33 in the salt-bridge region and hydrophobic interac-
tions with Ala144, Leu134, and Asn132.12
pyrrolidin-3-ol8 to yield compound 22. The mesylation
of compound 22 followed by deprotection of the MOM
group with TFA at room temperature led to compound
23. Macrocyclization of compound 23 under basic con-
ditions followed by deprotection of the methyl group
with TFA under reflux afforded compound 14.15
Compound 14 was synthesized as shown in Scheme 1.
Deprotonation of the fluorine-adjacent position of 1-flu-
oro-2-iodobenzene, 15, with LDA, followed by carbon
dioxide trapping and then esterification of the carboxylic
acid, led to methyl 2-fluoro-3-iodobenzoate. Iodine-
magnesium exchange of methyl 2-fluoro-3-iodobenzoate
according to Knochel’s procedure13 followed by
reaction with chloroglyoxylic acid ethyl ester gave
ketoester 16. Compound 16 was coupled with 3-[(tert-
butyldimethylsilyl)oxy]benzene-1,2-diamine14 to yield
quinoxaline-2-one, 17. Activation of the 2-position of
quinoxaline-2-one 17 via the corresponding 2-chloroqui-
noxaline using SOCl2 was followed by replacement of the
TBS group by a MOM group with TBAF to yield
compound 18. The addition of sodium methoxide at
the 2-position of 2-chloroquinoxaline followed by
hydrolysis with NaOH gave compound 19. Allyl 2-[2-
(tert-butyldimethylsilyloxy)ethyl]hydrazinecarboxlate8
was condensed with compound 19 to yield compound
20. Deprotection of the Alloc group of compound 20
in the presence of a palladium catalyst followed by
cyclization to indazol-3-one under basic conditions
gave compound 21. Alcohol of compound 21 was
mesylated and then aminated with (3R,5R)-5-methyl-
The kinase-selectivity profile of compound 14 indicates
that it has potent inhibitory activity against
CDK1,2,4,6 as shown in Table 8 and is highly selective
for the CDK family (IC50 > 1000 nM against 62 off-
target kinases;16) it showed potent inhibitory activity
against GSK3b (IC50 value = 13 nM). Compound 14 is
available for iv administration and showed potent
CDK-inhibitory activity in a preclinical animal model.
Its biological profile will be reported separately.
In conclusion, structure-based drug design starting from
the diarylurea scaffold of lead compound 2 led to the
discovery of compound 14, a potent CDK1,2,4,6
inhibitor that has a novel macrocyclic structure and
low off-target activity.
Table 8. IC50 values of compound 14 against members of the CDK
family
Compound
IC50 (nM)
CDK1
1.0
CDK2
CDK4
6.4
CDK6
12
14
3.4
CO2Me
F
CO2Me
CO2Me
F
OMOM
OTBS
F
I
a,b,c
47%
e,f
d
F
N
N
48%
88%
COCO2Et
Cl
N
O
N
H
15
16
17
18
OH
NHAlloc
N
O
CO2H
O
N
NH
OTBS
OMOM
F
OMOM
g,h
F
i
j,k
OMOM
N
N
N
N
77%
69%
78%
MeO
MeO
N
MeO
N
19
20
21
OMs
OH
N
N
N
O
O
O
N
N
N
NH
NH
NH
l
m,n
o,p
OH
OMOM
O
N
N
N
83%
85%
74%
MeO
N
MeO
N
O
N
H
23
22
14
Scheme 1. Reagents: (a) LDA, CO2 gas, THF; (b) concd H2SO4, MeOH; (c) i-PrMgCl, chloroglyoxylic acid ethyl ester, THF; (d) 3-[(tert-
butyldimethylsilyl)oxy]benzene-1,2-diamine,14 AcOH, toluene; (e) SOCl2, DMF; (f) MOMCl, TBAF, THF; (g) NaH, MeOH; (h) NaOH, THF,
MeOH, H2O; (i) allyl 2-[2-(tert-butyldimethylsilyloxy)ethyl]hydrazinecarboxylate,8 DMC, Et3N, CHCl3; (j) Pd(PPh3)4, HCO2H, Et2NH, THF; (k)
N,N-diisopropylethylamine, DMF; (l) MsCl, N,N-diisopropylethylamine, CHCl3, (3R,5R)-5-methylpyrrolidin-3-ol;8 (m) MsCl, N,N-diisopropyl-
ethylamine, CHCl3; (n) TFA, H2O; (o) K2CO3, DMF; (p) TFA, H2O.

5126
N. Kawanishi et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5122–5126
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.;
Acknowledgments
Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz,
J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-
Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challa-
combe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong,
M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision
C.02, Gaussian Inc.: Wallingford CT, 2004.
We would like to acknowledge the excellent contributions
of the following scientists to this work: for biology, Ikuko
Takahashi; for structural chemistry, Toshiharu Iwama.
References and notes
1. Ikuta, M.; Kamata, K.; Fukasawa, K.; Honma, T.;
Machida, T.; Hirai, H.; Suzuki-Takahashi, I.; Hayama,
T.; Nishimura, S. J. Biol. Chem. 2001, 276, 27548.
2. (a) Sherr, C. J. Science 1996, 274, 1672; (b) Sherr, C. J.
Cancer Res. 2000, 60, 3689; (c) Pines, J. Adv. Cancer Res.
1995, 66, 181; (d) Sherr, C. J.; Roberts, J. M. Genes Dev.
1999, 13, 1501; (e) Dyson, N. Genes Dev. 1998, 12, 2245.
3. Hall, M.; Peters, G. Adv. Cancer Res. 1996, 68, 67.
4. Misra, R. N. Drugs Future 2006, 31, 43.
5. Honma, T.; Hayashi, K.; Aoyama, T.; Hashimoto, N.;
Machida, T.; Fukasawa, K.; Iwama, T.; Ikeura, C.; Ikuta,
M.; Suzuki-Takahashi, I.; Iwasawa, Y.; Hayama, T.;
Nishimura, S.; Morishima, H. J. Med. Chem. 2001, 44,
4615.
11. Cellular potency of the CDK inhibitor was determined
by the E2F-dependent transcription assay described
elsewhere (manuscript in preparation). Briefly, human
glioma cell line T98G, stably containing
a CDC6
promoter gene which contains E2F binding sites, was
used: it was synchronously cultured from G1 to S
phase by relief from contact growth inhibition, and
induced E2F transcription activity was determined by
SEAP reporter gene under control of the CDC6
promoter. Cells were cultured in the presence of
CDK inhibitor, and inhibition of E2F reporter activity
was determined.
12. The X-ray coordinate has been deposited with the Protein
Data Bank, as entry 1DS1.
6. Honma, T.; Yoshizumi, T.; Hashimoto, N.; Hayashi, K.;
Kawanishi, N.; Fukasawa, K.; Takaki, T.; Ikeura, C.;
Ikuta, M.; Suzuki-Takahashi, I.; Hayama, T.; Nishimura,
S.; Morishima, H. J. Med. Chem. 2001, 44, 4628.
7. Hayama, T.; Kawanishi, N.; Takaki, T. WO 2002002550.
8. Hirai, H.; Kawanishi, N.; Hirose, M.; Sugimoto, T.;
Kamijyo, K.; Shibata, J.; Masutani, K. WO 2004039809.
9. The X-ray coordinate has been deposited with the Protein
Data Bank, as entry 1GII.
10. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G.
E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.,
Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.;
Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.;
Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda,
R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.;
Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
13. Knochel, P.; Dohle, W.; Gommermann, N.; Kneisel, F. F.;
Kopp, F.; Korn, T.; Sapountzis, I.; Vu, V. A. Angew.
Chem., Int. Ed. 2003, 42, 4302.
14. Ueno, Y.; Noguchi, T.; Hirota, K.; Sawada, N.;
Umezome, T. WO 2003106418.
15. 1H NMR (300 MHz, DMSO-d) d (ppm) 0.68 (m, 3H), 1.34
(m, 1H), 2.00–3.10 (m, 5H), 3.40–4.00 (m, 2H), 4.46 (m,
1H), 5.24 (br s, 1H), 6.94 (m, 2H), 7.14 (m, 1H), 7.48 (m,
1H), 7.83 (d, J = 8.4 Hz, 1H), 9.43 (m, 1H), 12.20 (br s,
1H), 12.60 (br s, 1H). MS (ESI+): m/z 404 (M+H)+.
16. Compound 14 was tested against a wider panel of kinases.
Notably, it had IC50 > 1 lM against Abl, Arg, Aurora-A,
Axl, Blk, Bmx, CaMKIV, Chk1, Chk2, c-RAF, CSK,
ERK1, ERK2, KDR, Flt-1, FGFR1, FGFR2, FGFR3,
IGF-1R, IKKa, IKKb, JNK1a1, JNK2a2, JNK3, Lyn,
MAPK1, MAPK2, MAPKAP-K2, MEK1, MKK4,
MKK6, MKK7b, p70S6K, PAK2, PDGFRa, PDGFRb,
PDK1, PKA, PKBa, PKBb, PKc, PKCa, PKCbII, PKCc,
PKCd, PKCe, PKCg, PKCi, PKCl, PKD2, PRAK,
PRK2, ROCK-II, Rsk2, SAPK2a, SAPK2b, SAPK3,
SAPK4, SGK, Syk, Tie-2 and ZAP-70.