Structure-based design and optimization of 2-aminothiazole-4-carboxamide as a new class of CHK1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.02.108

Drug design efforts in the emerging 2-aminothiazole-4-carboxamide class of CHK1 inhibitors have uncovered specific combinations of key substructures within the molecule; resulting in significant improvements in cell-based activity while retaining a greater than one hundred-fold selectivity against CDK2. The X-ray crystal structure of a complex between compound 39 and the CHK1 protein detailing a 'U-shaped' topology and key interactions with the protein surface at the ATP site is also reported.

Full text of DOI:10.1016/j.bmcl.2013.02.108

Bioorganic & Medicinal Chemistry Letters 23 (2013) 2590–2594
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Structure-based design and optimization of 2-aminothiazole-4-car-
boxamide as a new class of CHK1 inhibitors
a
b
b
a
Xiaohua Huang a, , Cliff C. Cheng , Thierry O. Fischmann , José S. Duca , Matthew Richards ,
⇑
Praveen K. Tadikonda a, Panduranga Adulla Reddy a, Lianyun Zhao a, M. Arshad Siddiqui a,
David Parry c, Nicole Davis c, Wolfgang Seghezzi c, Derek Wiswell c, Gerald W. Shipps Jr. a,
⇑
a Merck Research Laboratories, 320 Bent Street, Cambridge, MA 02141, United States
b Merck Research Laboratories, 2000 Galloping Hill Road, Kenilworth, NJ 07033, United States
c Merck Research Laboratories, 901 South California Avenue, Palo Alto, CA 94304, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Drug design efforts in the emerging 2-aminothiazole-4-carboxamide class of CHK1 inhibitors have
uncovered specific combinations of key substructures within the molecule; resulting in significant
improvements in cell-based activity while retaining a greater than one hundred-fold selectivity against
CDK2. The X-ray crystal structure of a complex between compound 39 and the CHK1 protein detailing
a ‘U-shaped’ topology and key interactions with the protein surface at the ATP site is also reported.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 9 January 2013
Revised 15 February 2013
Accepted 25 February 2013
Available online 7 March 2013
Keywords:
CHK1 protein kinase
2-Aminothiazole-4-carboxamide
Hybrid compounds
Structure-based drug design
CHK1 is a serine/threonine protein kinase that arrests the cell
cycle progression in response to the DNA damage induced by par-
ticular anticancer therapeutics; enabling the deployment of cellu-
lar repair mechanisms.1 CHK1 mediates the S and G2/M arrest
through the proteolysis of CDC25A2–4 that leads to the inactivation
of CDK2, whose activity is required in the cell cycle transition.5
Coupled with the loss of the G1 checkpoint in most tumor cells
due to the mutation of the tumor suppressor gene p53,6 CHK1
downregulation abrogates additional checkpoints in the cell cycle
causing tumor cells with damaged DNA to progress into cata-
strophic mitosis. Therefore, CHK1 inhibition should sensitize
p53-defective tumor cells to the cytotoxic effects of anticancer
DNA-damaging agents and lead to an expanded therapeutic win-
dow; a hypothesis that has prompted widespread attention to
the discovery and development of CHK1 inhibitors over the past
few years.7–9
the desired therapeutic effects of CHK1 inhibition the series was
also optimized for its selectivity against CDK2 and some of the
most promising compounds exhibited well above a hundred-fold
selectivity. However the progression of this series was limited by
weak cell-based activity, with the highest potency in low micro-
molar ranges assessed by the c-H2AX cellular biomarker for DNA
double-strand breaks.12 When cancerous cells are treated with a
ribonucleotide reductase inhibitor hydroxyurea (HU),13 the check-
point kinase CHK1 is activated to allow the cell cycle arrest and
DNA repair. The addition of a CHK1 inhibitor would abrogate this
mechanism and accelerate the progression of the damaged cancer-
ous cells into mitosis, where they undergo the replication fork col-
lapse and apoptosis, marked by the phosphorylation of histone
H2AX known as c-H2AX.
Significant improvements in cell-based activity and additional
robust structure–activity relationship (SAR) development were
necessary for further progression of the series. In light of the break-
through in the biochemical potency and selectivity against CDK2
that was achieved through the modification of the 2-aryl group
connected to the thiazole on the left hand side (LHS) of the mole-
cule, continuing modification at this region with a focus on the
nitrogen-linked appendage was prioritized (Table 1). These analogs
were prepared following Scheme 1. The amide intermediate was
prepared from 2-bromothiazole-4-carboxylic acid and the aniline
using HATU as the coupling reagent. Palladium catalyzed
amination of the 2-bromo intermediate using the phosphine ligand
In our previous account of a novel series of thiazole-4-carbox-
amide-based CHK1 inhibitors,10 which were developed from the
Automated Ligand Identification System (ALIS) screening hit 1,11
significant improvements were achieved in a biochemical CHK1 ki-
nase inhibition assay resulting in compounds with single digit
nanomolar potency. As CDK2 activity was anticipated to oppose
⇑
Corresponding authors. Tel.: +1 617 992 3174.
E-mail addresses: xiaohua.huang@merck.com (X. Huang), gshipps@comcast.net
(G.W. Shipps).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.02.108

X. Huang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2590–2594
2591
Table 1
SAR at the 2-aminothiazole positiona,b
O
HN
HN
Me
N
Me
N
N
N
N
N
N
N
S
H
N
N
F
F
HN
N
O
1
N
R
2
R
N
H
ALIS Hit
MeO
OMe
Compound #
CHK1 IC50 (nm)
CDK2/CHK1
1
75
2.8
10000
2
92
9
3
4
5
6
7
8
1300
3
9
4000
nd
10
29
380
nd
11,000
nd
nd
240
19
nd
>50,000
nd
nd
38
90
nd
38
90
nd
H2AX EC50 (nm)
nd
nd
nd
O
N
N
HN
N
N
N
N
N
N
N
N
N
N
N
N
S
N
H
N
N
N
N
HN
S
N
S
1
R
N
2
R
N
H
Compound #
CHK1 IC50 (nm)
CDK2/CHK1
11
39
120
nd
12
8
290
1400
13
6900
nd
14
190
nd
15
5700
nd
16
62
nd
nd
17
280
nd
18
10000
nd
19
120
nd
20
2
500
750
H2AX EC50 (nm)
nd
nd
nd
nd
nd
nd
NH
NH
N
Me
N
O
N
N
N
N
N
N
N
NH
N
O
N
H
N
H
HN
N
NH
N
2
N
H
S
H
N
N
MeO
1
N
MeO
R
2
R
N
H
MeO
Compound #
CHK1 IC50 (nm)
CDK2/CHK1
21
110
nd
22
5
4800
250
23
2
720
350
24
14
880
7500
25
3
4300
3750
26
4500
nd
27
980
nd
H2AX EC50 (nm)
nd
nd
nd
a
Biochemical data represent average values of duplicates or triplicates with a standard deviation of 10%.
nd, not determined.
b
O
O
O
Ar
N
Ar
N
Ar
Ar
N
O
H N
N
H
N
H
2
N
S
S
N
S
C
OH
B
H
A
N
N
N
S
N
N
1
Br
1
R
N
R
2
2
Br
R
N
Boc
N
Boc
N
Boc
R
N
H
Scheme 1. Synthesis of substituted 2-aminothiazole-4-carboxamides. Reagents and conditions: (A) HATU (N,N,N0,N0-tetramethyl-O-(7-azabenzotriazol-1-yl)uronium
hexafluorophosphate), DIEA (N,N-diisopropylethylamine), DMF, rt; (B) Pd2(DBA)3, Xantphos (4,5-bis(diphenylphosphino)-9,9-dimethylxanthene) or X-Phos (2-dicyclohex-
ylphosphino-20,40,60-triisopropylbiphenyl), HNR1R2, Cs2CO3, dioxane, 100 °C; (C) TFA (trifluoroacetic acid).
Xantphos14 or X-Phos15 afforded the substituted 2-aminothiazole
derivatives. The removal of the Boc protecting group on the right
hand side of the compounds using TFA gave the final products.
Elaboration of the 2-thiazole position was guided by both ligand
and structure-based design considerations. An array of amines
with steric and electronic variations were linked to the 2-thiazole.
The substitution with aniline provided compound 2 with the IC50
of 92 nM and the selectivity against CDK2 of ninefold. Elongation
to benzylamino (3) resulted in the significant potency decrease
to 11,000 nM. Capping the aniline or benzyl amine with a methyl
group (4 or 5) also proved to weaker relative to 1. In light of the
preferred substructure at the 2-thiazole position, the fused bicyclic
aromatics that were reported in the previous account,10 we were
particularly interested in their N-containing isosteres, tetrahydro-
quinolines and indoline analogs. Interestingly, substitution with
tetrahydroquinolinyl (6) and tetrahydroisoquinolinyl (7) demon-
strated improvements in potency to less than 50 nM and selectiv-
ity against CDK2 to 90-fold. Modification of the phenyl ring in the
tetrahydroisoquinolinyl proved to be quite sensitive in affecting
the potency and selectivity against CDK2. While substitution with
electron-withdrawing atom fluorine at the 4- and 7-positions (8)
or changing the phenyl ring to imidazole (9) led to a decrease in
potency to IC50’s in the micromolar range, substitution with the
electron donating group methoxy at the 5- and 6-positions (10)
or changing the phenyl ring to a thiophene (11) retained a potency
to an IC50 value below 50 nM and selectivity against CDK2 over
100-fold. Changing the tetrahydroisoquinoline to an isoindoline
(12) was rewarded with an improvement in the biochemical activ-
ity to single-digit nanomolar and selectivity against CDK2 to ꢀ300-
fold. Moving the endocyclic isoindoline nitrogen out of the ring
into an exocyclic position (13) resulted in a much weaker CHK1
inhibitor. The change of the phenyl ring in the isoindoline to other
heterocycles such as pyridine (14), 2-methylpyrimidine (15),
pyrazole (16) and 2-methylthiazole (17) was accompanied by a
decrease of potency ranging from double-digit nanomolar to
micromolar.
One of the design principles we wanted to explore was the addi-
tion of new H-bond interactions with CHK1. An area that we tar-
geted early on was to compliment the backbone H-bond donor of
Cys87. As the 2-position of the thiazole was modified it became
evident that notable differences in potency could be obtained by
changing the heteroaryl groups used in this LHS position. While
the imidazole analogue (18) was a weak inhibitor with an IC50 of
10 lM, pyrazole (19) was 100-fold more potent with an IC50 of

2592
X. Huang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2590–2594
120 nM. Fused heterocycles also behaved differently in their inter-
action with CHK1 protein. The use of indazole (20) on the LHS
markedly improved the biochemical potency to 2 nM and the
cell-based activity to 750 nM with a 500-fold selectivity against
CDK2; consistent with a possible new interaction between the lone
pair of the indazole nitrogen and the Cys87 amide. The benzotria-
zole analog (21) was less active with an IC50 of 110 nM; possibly
due its weaker ability to accept an H-bond from the Cys87 and/
or repulsive interactions with the additional triazole nitrogen
and a nearby protein carbonyl group. Substitution with 6-meth-
oxyisoindolinone (22) or 3-aminoindazole via the primary amino
group (23) provided compounds with the biochemical activity in
the same range as compound 20, but was about threefold more
active in the cell assay with better selectivity against CDK2. 3-
Aryl-5-aminopyrazole derivatives were also examined. The 3-
phenyl-5-aminopyrazole analog (24) suffered from poor cellular
potency despite its active inhibition of CHK1 kinase activity.
para-Methoxyl substitution to 3-phenyl group (25) that fit well
in the solvent exposed region improved the potency by several
stitution (30) was also obtained; the other 3 isomeric pyridines
(28, 29, and 31) suffered from potency losses ranging from a few
fold to over tenfold relative to compound 1 as they may not place
the pyridyl nitrogen in an optimal position to H-bond to Lys38. A
loss of selectivity against CDK2 in compound 29 was also evident,
presumably because the side-chain homologous CDK2 Lys33
shows extended flexibility and more readily accommodates a less
favorable orientation of the pyridinyl H-bond acceptor compared
to CHK1.17 Consequently, additional compounds with the same
RHS as 30 were prepared incorporating other favorable groups
on the LHS identified from both the previous account, such as 5-
benzofuran (32), 5-indole (33), 5- and 2-benzothiophene (34 and
35), and others detailed in Table 1 of this article (see Table 3). In
the class of 2-aryl thiazoles cell-based activity still remained in
the micromolar range, and in the case of para-phenoxy phenyl sub-
stitution (36), a significant drop in the biochemical activity was
also observed. Interestingly, the 2-amino substituent appended
to the thiazole on the LHS and the pyridine on the RHS demon-
strated a combined favorable influence on the cell-based activity,
and pronounced improvement to an EC50 of 30 nM in the best case
(39) was observed (see the Supplementary data for the activity
assessment in the presence and absence of hydroxyurea). The
des-OMe analog (38) is threefold less potent in H2AX cell-based as-
say. While the combination of isoindoline on the LHS and pyridine
on the RHS (37) did not gain much in terms of cell-based activity,
the hybrids with indazole (40) or 3-aminoindazole (41) on the LHS
gave a 10-fold potency boost in the cell-based assay to 75 and
40 nM respectively, relative to the analogs with the same LHS
group in Table 1 (20 and 23).
To elucidate the binding mode of this new class of CHK1 inhib-
itors with the protein, the X-ray crystal structure of the complex
between compound 39 and the CHK1 protein at a resolution of
1.70 Å was successfully obtained (Fig. 1).18 Many of key interac-
tions observed between compound 1 and CHK1 were retained in
the binding of compound 39.10 Intramolecular H-bonds amongst
the secondary amide, tertiary anilinic N and the thiazolyl N stabi-
lize a ‘U-shaped’ topology to 39. The salt bridge between the distal
piperazinyl NH and Glu91 and Glu134 and H-bond between the
carbonyl O of the amide and conserved water are identical to the
previously reported X-ray.10 A slight difference in the hinge inter-
action is observed including the attraction between carbonyl O in
Glu85 and thiazolyl S.19 In agreement with the initial hypothesis
a new H-bond between Cys87 NH and the carbonyl O of the isoind-
olinone group is evident. In addition, the design efforts on the RHS
of the compound yielded another hydrogen bond between the
pyridinyl N and the Lys38. To further probe this H-bond interac-
tion, modifications in the vicinity of the pyridinyl N were made
(Table 4). The replacement of pyridine with fluorobenzene (42) re-
tained the biochemical potency in the single-digit nanomolar
range and excellent selectivity against CDK2, suggesting the
fold,16 but remained above 1
lM in cells. Adding a methyl group
(26) or moving the connection (27) to the pyrazolyl nitrogen
brought down the enzymatic potency drastically to the micromolar
range, possibly due to the disruption of interactions with the
H-bond donor on the Cys87 residue by the methyl group (26) or
exocyclic NH functionality (27), and/or an unfavorable orientation
of the distal phenyl ring.
To establish additional SAR trends, a scan of different isomeric
pyridines was performed on the phenyl ring of the right hand side
(RHS) of the molecule (Table 2). With the possibility of gaining an
additional H-bond interaction between a pyridine ring nitrogen
and the catalytic Lys38, we prioritized the synthesis of compound
30. A slight improvement in the biochemical potency and selectiv-
ity against CDK2 in the pyridine analog with a para-piperazine sub-
Table 2
Replacement of phenyl ring with pyridines on the RHS of the moleculea,b
N
N
N
O
N
N
H
S
N
N
N
N
N
N
N
N
H
N
H
N
H
N
H
N
H
N
H
O
Compound #
CHK1 IC50 (nm)
CDK2/CHK1
1
75
2.8
28
870
nd
29
230
0.11
30
55
8.2
31
4700
3
a
Biochemical data represent average values of duplicates or triplicates with a
standard deviation of 10%.
b
nd, not determined.
Table 3
Hybrid compoundsa,b
N
N
O
NH
N
N
N
N
O
O
HN
S
N
N
S
H
N
N
O
O
HN
S
R
O
N
H
Compound #
32
28
nd
nd
33
10
430
1000
34
7
210
1250
35
3
65
1000
36
430
nd
37
8
130
900
38
2
2580
100
39
1
11000
30
40
3
100
75
41
3
550
40
CHK1 IC
(nm)
50
CDK2/CHK1
H2AX EC
(nm)
nd
50
a
Biochemical data represent average values of duplicates or triplicates with a standard deviation of 10%.
nd, not determined.
b

X. Huang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2590–2594
2593
Figure 1. X-ray of crystal structure of compound 39 bound to CHK1 protein.
Table 4
a,b
Optimization to Probe Interaction with Lys38
N
N
F
N
N
Me
N
N
N
N
N
N
N
O
R
N
N
N
S
H
N
N
N
O
N
N
H
N
H
N
H
N
H
N
H
N
H
N
H
O
Compound #
CHK1 IC50 (nm)
CDK2/CHK1
39
1
11,000
30
42
4
10,000
6000
43
10
5000
1000
44
46
nd
nd
45
281
nd
46
5200
nd
H2AX EC50 (nm)
nd
nd
a
Biochemical data represent average values of duplicates or triplicates with a standard deviation of 10%.
nd, not determined.
b
charge-dipole or H-bond interaction between C–F bond and
Lys38.20 The large loss in cellular potency to 6
M might be due
enzymes up to 30
using human ether-a-go-go-related gene (hERG) rubidium efflux
(65% inhibition @ 5 g/mL), IonWorks (42% inhibition @ 10 M),
and voltage clamp (82% inhibition @ 1 M), and hERG blockade
was seen in compound 39, most likely caused by the basic pipera-
zine. The bioavailability of compound 39 was determined in an oral
lM. The cardiovascular liability was assessed
l
to the confounding off-target effect or efflux from the cell. Adding
a methyl group next to the pyridinyl N (43) or changing the pyri-
dine to pyrimidine (44), pyridazine (45) or quinoline (46) imparted
a reduction in biochemical activity from tenfold to several thou-
sand folds due to unfavorable steric or electronic interactions.
Compound 39 was further profiled for the in-house kinase
selectivity, GPCR counter screening, drug interaction and hERG
inhibition (see the Supplementary data). Good selectivity against
an in-house kinase panel and a clean profile in GPCR counter
screening were demonstrated. The good kinase selectivity might
be partially explained by the salt bridge interaction with Glu91
which shows some variability among kinases. Compound 39 dem-
onstrated no inhibition against a panel of cytochrome P450 (CYP)
l
l
l
rat PK experiment, where an AUC of 7.2
a 10 mg/kg single dose.
lM h was determined after
In summary, drug design efforts have led to the discovery of 2-
aminothiazole-4-carboxamide as a new class of CHK1 inhibitors
that are potent in both enzymatic and cell-based assays with over
a hundred fold selectivity against CDK2. The overall profile of this
novel series of CHK1 inhibitors represented by compound 39 pro-
vided a solid base for further optimization efforts towards a poten-
tial clinical candidate as will be reported at a later date.

2594
X. Huang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 2590–2594
8. Garrett, M. D.; Collins, I. Trends Pharmacol. Sci. 2011, 32, 308.
Acknowledgments
9. (a) Oza, V.; Ashwell, S.; Almeida, L.; Brassil, P.; Breed, J.; Deng, C.; Gero, T.;
Grondine, M.; Horn, C.; Ioannidis, S.; Liu, D.; Lyne, P.; Newcombe, N.; Pass, M.;
Read, J.; Ready, S.; Rowsell, S.; Su, M.; Toader, D.; Vasbinder, M.; Yu, D.; Yu, Y.;
Xue, Y.; Zabludoff, S.; Janetka, J. J. Med. Chem. 2012, 55, 5130; (b) Montano, R.;
Chung, I.; Garner, K. M.; Parry, D.; Eastman, A. Mol. Cancer Ther. 2012, 11, 427.
10. Huang, X.; Cheng, C. C.; Fischmann, T. O.; Duca, J. S.; Yang, X.; Richards, M.;
Shipps, G. W., Jr. ACS Med. Chem. Lett. 2012, 3, 123.
11. Annis, A.; Chuang, C.-C.; Nazef, N. ALIS: An Affinity Selection-Mass
Spectrometry System for the Discovery and Characterization of Protein-
Ligand Interactions. In Methods and Principles in Medicinal Chemistry; Wanner,
K., Höfner, G., Eds.; Mass Spectrometry in Medicinal Chemistry: Applications in
Drug Discovery; Mannhold, R., Kubinyi, H., Folkers, G., Eds.; Wiley-VCH:
Weinheim, 2007; Vol. 36, Chapter 3.
We thank all legacy Schering-Plough colleagues who contrib-
uted in many meaningful ways to initiating and sustaining the
hit to lead and lead optimization of the thiazole-4-carboxamide-
based CHK1 inhibitor project and all supporting management,
including Timothy J. Guzi, Satish Jindal, Neng-Yang Shih and John
J. Piwinski.
Supplementary data
12. Rogakou, E. P.; Pilch, D. R.; Orr, A. H.; Ivanova, V. S.; Bonner, W. M. J. Biol. Chem.
1998, 273, 5858.
13. Elford, H. L. Biochem. Biophys. Res. Commun. 1968, 33, 129.
14. Yin, J.; Zhao, M. M.; Huffman, M. A.; McNamara, J. M. Org. Lett. 2002, 4, 3481.
15. Anderson, K. W.; Tundel, R. E.; Ikawa, T.; Altman, R. A.; Buchwald, S. L. Angew.
Chem., Int. Ed. 2006, 45, 6523.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.02.
108.
16. Caldwell, J. J.; Welsh, E. J.; Matijssen, C.; Anderson, V. E.; Antoni, L.; Boxall, K.;
Urban, F.; Hayes, A.; Raynaud, F. I.; Rigoreau, L. J. M.; Raynham, T.; Aherne, G.
W.; Pearl, L. H.; Oliver, A. W.; Garrett, M. D.; Collins, I. J. Med. Chem. 2011, 54,
580.
17. Fischmann, T. O.; Hruza, A.; Duca, J. S.; Ramanathan, L.; Mayhood, T.; Windsor,
W. T.; Le, H. V.; Guzi, T. J.; Dwyer, M. P.; Paruch, K.; Doll, R. J.; Lees, E.; Parry, D.;
Seghezzi, W.; Madison, V. Biopolymers 2008, 89, 372.
18. The atomic coordinates of compound 39 bound to CHK1 have been deposited
in the Protein Data Bank: PDB ID, 4HYH.
19. Lu, J.; Lu, Y.; Yang, S.; Zhu, W. Struct. Chem. 2011, 22, 757.
20. (a) Müller, K.; Faeh, C.; Diederich, F. Science 2007, 317, 1881; (b) Razgulin, A. V.;
Mecozzi, S. J. Med. Chem. 2006, 49, 7902.
References and notes
1. Tapia-Alveal, C.; Calonge, T. M.; O’Connell, M. J. Cell Div. 2009, 4.
2. Mailand, N.; Falck, J.; Lukas, C.; Syljuåsen, R. G.; Welcker, M.; Bartek, J.; Lukas, J.
Science 2000, 288, 1425.
3. Xiao, Z.; Chen, Z.; Gunasekera, A. H.; Sowin, T. J.; Rosenberg, S. H.; Fesik, S.;
Zhang, H. J. Biol. Chem. 2003, 278, 21767.
4. Xiao, Z.; Xue, J.; Sowin, T. J.; Rosenberg, S. H.; Zhang, H. Oncogene 2005, 24,
1403.
5. Guadagno, T. M.; Newport, J. W. Cell 1996, 84, 73.
6. Levine, A. J. Cell 1997, 88, 323.
7. Ma, C. X.; Janetka, J. W.; Piwnica-Worms, H. Trends Mol. Med. 2011, 17, 88.