Structure-based design of a new class of highly selective aminoimidazo[1,2-a]pyridine-based inhibitors of cyclin dependent kinases

Article abstract of DOI:10.1016/j.bmcl.2005.01.052

Structure-based design approach was successfully used to guide the evolution of imidazopyridine scaffold yielding new structural class of highly selective inhibitors of cyclin dependent kinases that were able to form a new interaction with an identified residue of the protein, Lys89. Compounds from this series have shown no detectable effect when tested against a representative set of other serine/threonine kinases such as GSK3β, CAMKII, PKA, PKC-α,β,ε,γ. Compound 2i inhibits proliferation in HCT 116 cells in tissue culture. Synthesis, co-crystal structure of CDK2 in complex with compound 2i, and preliminary SAR study are disclosed.

Full text of DOI:10.1016/j.bmcl.2005.01.052

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1943–1947
Structure-based design of a new class of highly selective
aminoimidazo[1,2-a]pyridine-based inhibitors of cyclin
dependent kinases
Chafiq Hamdouchi,a,* Boyu Zhong,c Jose Mendoza,c Elizabeth Collins,a Carlos Jaramillo,b
a
a
Jose Eugenio De Diego,b Daniel Robertson,a Charles D.Spencer, Bryan D.Anderson,
Scott A.Watkins, Faming Zhanga and Harold B.Brooks
a
a
aLilly Research Laboratories, A Division of Eli Lilly and Company, Lilly Corporate Center, DC: 0540,
Indianapolis, IN 46285, United States
b
´
Centro de Investigacion Lilly, Avenida de la Industria, 30, 28108 Alcobendas (Madrid), Spain
cEli Lilly and Company—RTP Labs, 20 T.W. Alexander Drive, Research Triangle Park, NC 27709, United States
Received 2 December 2004; revised 19 January 2005; accepted 21 January 2005
Abstract—Structure-based design approach was successfully used to guide the evolution of imidazopyridine scaffold yielding new
structural class of highly selective inhibitors of cyclin dependent kinases that were able to form a new interaction with an identified
residue of the protein, Lys89.Compounds from this series have shown no detectable effect when tested against a representative set of
other serine/threonine kinases such as GSK3b, CAMKII, PKA, PKC-a,b,e,c.Compound 2i inhibits proliferation in HCT 116 cells
in tissue culture.Synthesis, co-crystal structure of CDK2 in complex with compound 2i, and preliminary SAR study are disclosed.
Ó 2005 Elsevier Ltd. All rights reserved.
Cyclin-dependent kinases (CDK) and their cyclin part-
ners are key players in regulating the entry into, passage
through, and exit from the cell cycle.1 Because of their
critical role in the regulation of cell cycle and the ob-
served expression/activity pattern in most human can-
cers, considerable effort has been focused on the
development of small molecule inhibitors.2 However,
the number of structural classes that act as CDK inhibi-
tors is limited, and most of them derive from relatively
nonspecific protein kinase inhibitor scaffolds that also
inhibit CDKs, such as staurosporins,3 flavonoids,4
indigoids,5 paulones,6 and purines.7 Considerable efforts
are still devoted to the search for new structural classes
of cyclin-dependent kinase inhibitors8 with potentially
new selectivity profiles and new physiochemical
properties.
Compounds represented by structure 1 were demon-
strated to inhibit cyclin dependent kinases by competing
with ATP for binding to a catalytic subunit of the pro-
tein.In view of these results, we sought to utilize the
structural information from the crystal structure of
compound 1 bound to monomeric human CDK2 to
identify new potential binding sites with greater impact
on selectivity (Fig.1 ).
Our initial hypothesis was: could the internal hydrogen
bonds in the aminoimidazo[1,2-a]pyridine 1, be replaced
by a newly identified external interaction with Lys89?
This might also hold the amino hydrogen in a syn posi-
tion with the nitrogen at the 1-position, favoring the key
interactions with Leu83.Herein we report our efforts to
test this hypothesis.The synthesis of our target mole-
cules is outlined in Scheme 1.In path A, reaction of 2-
amino-5-iodo-pyridine 3 with p-tolylsulfonyl chloride
in pyridine and subsequent treatment with iodoacet-
We recently described the discovery of a new structural
class of protein serine/threonine kinase inhibitors com-
prising of an aminoimidazo[1,2-a]pyridine nucleus.9
amide in the presence of HunigÕs base in DMF provided
¨
the corresponding carbamide10 4 in good yields.Conver-
sion of 4 to the desired 2-(N-trifluoroacetylamino)imid-
azopyridine 5 was accomplished by treatment with
trifluoroacetic anhydride in refluxing methylene
*
Corresponding author.Te:l. +1 317 276 9582; fax: +1 317 276
5431; e-mail: hamdouchi_chafiq@lilly.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd.All rights reserved.
doi:10.1016/j.bmcl.2005.01.052

1944
C. Hamdouchi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1943–1947
H
Lys89
N
External
HB
H
Internal
HB
O
External
HB
NH2
NH2
Leu83
External
HB
S
NH2
Leu83
O
O
H
H
O
N
H
O
F
N
H
N
H
N
N
O
N
O
N
N
O
NH
Glu81
NH
Glu81
Cl
O
O
O
F
H
Cl
H
HO
N
HO
N
O
2
1
O
Asp145
O
Asp145
O
Figure 1. Two-dimensional representation of the binding mode of imidazopyridines 1 and 2 with hCDK2.
Path A
N
Tos
NH2
NH2
N
c
a,b
H
N
O
N
I
N
N
O
I
I
F3C
3
4
5
d,e
Cl
N
H
f
Cl
N
N
N
H
N
O
N
F3C
R
Cl
O
Cl
O
6
Cl
O
g
2
N
Br/Cl
N
Cl
10
d,e
Path B
NH HBr
O
i
NH2
h
N
Br/Cl
N
Br
N
N
Br
Br
O
7
8
9
Scheme 1. Synthesis of phenylaminoimidazo[1,2-a]pyridines: (a) TsCl, py, 89; (b) ICH2CONH2, DIPEA, DMF, 86%; (c) TFAA, CH2Cl2, 63%; (d) i-
PrMgCl, THF 2,6-Cl2–C6H3CHO, 81%; (e) MnO2, CH2Cl2, 79%; (f) ArNH2, >250 °C; (g) BrCH2CO2CH3, 97%; (h) POCl3, 105 °C, 98%; (i) ArNH2,
CoCl2, Ph2O, 150–200 °C, 20–90%.
chloride.Metalation of compound 5 with i-propyl mag-
nesium chloride and subsequent reaction with 2,6-
dichlorobenzaldehyde followed by oxidation with
manganese dioxide furnished the diaryl ketone 6.The
trifluoroacetamide group was then substituted by reac-
tion with a selection of anilines at high temperature to
give the corresponding phenylaminoimidazo[1,2-a]pyr-
idines 2.Alternatively, compounds 2 were also prepared
using a cobalt-mediated amination as illustrated in path
B.Reaction of the commercially available 2-amino-5-
bromopyridine 7 with methyl bromoacetate and subse-
quent treatment with phosphorus oxychloride gave a
1:3 mixture of 2-bromo and 2-chloro imidazo[1,2-a]pyr-
idines 9 in good yield.A halogen–metal exchange study
on building block 9 showed that use of i-propyl magne-
sium chloride is most effective for chemoselective

C. Hamdouchi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1943–1947
Table 1. Enzyme inhibition data for compounds 2a–n
1945
Cl
N
H
N
N
R
Cl
O
Compds
R
E/CDK2 Inhibition IC50 (lM)a D1/CDK4 Inhibition IC50 (lM)a B/CDK1 Inhibition IC50 (lM)a
O
I
S
2a
1.21
1.22
0.52
O
O
S
S
2b
2.86
3.14
1.7
O
I
I
2c
2d
13.85
17.73
7.40
3.04
4.8
S
11.90
I
I
O
2e
2f
1.63
2.72
1.61
3.95
0.58
1.92
NH2
O
NH2
I
I
O
S
2g
2h
2i
1.54
12.09
0.56
2.16
0.97
9.26
0.39
N
O
O
S
N
O
>20
I
I
O
S
1.34
1.29
NH2
O
O
I
I
S
NH
2j
0.25
0.16
N
H
O
NH2
O
S
O
N
H
O
2k
0.67
0.74
3.84
4.99
0.25
0.45
O
2l
I
I
I
OH
2m
>20
>20
>20
N
O
2n
>20
>20
>20
Flavopiridol
0.96
0.18
0.20
a Values are means of three experiments, see Ref. 9 for assays protocol.

1946
C. Hamdouchi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1943–1947
functionalization at position-6.Reaction of compound 5
with i-propyl magnesium chloride and subsequent reac-
tion with 2,6-dichlorobenzaldehyde followed by oxida-
tion with manganese dioxide furnished the diaryl
ketone 10 in 65% yield from 9.The Br/Cl group was
then substituted by reaction with a selection of anilines
in the presence of cobalt chloride in diphenyl ether at
150–200 °C.11
Our assumption was tested through both SAR studies
and X-ray crystallographic analysis.Initial compounds
that were designed to satisfy the new interaction with
Lys89 were the methyl sulfones 2a and 2b.As listed in
Table 1, these two compounds showed moderate activity
against cyclin dependent kinases CDK2, CDK4, and
CDK1, 2a (IC50 = 1.21, 1.22, and 0.52 lM) and 2b
(IC50 = 2.86, 3.14, and 1.7 lM).
Interestingly when the oxygens were removed (sulfide
analogues 2c and 2d) the activity dropped significantly.
Figure 3. Binding of imidazopyridine 2i to CDK2 highlighting the
surrounding residues and hydrogen bond interactions (red lines).
Phenyl sulfonamide and carboxylic amides were also
found to be potent against cyclin dependent kinases.
Our initial data suggests the substitution at the para-po-
sition to be more favorable for CDKs inhibition.In
addition, incorporation of proton prone to forming
stronger hydrogen bond such as in 2i, 2j, 2k, and 2l
increases potency.
bonyl oxygen of Leu83 (heavy atom distance of
˚
2.61 A).In addition, there is an indication of hydrogen
bond interactions between the sulfonamide group and
˚
both the Lys89 (O–N distance of 4.13 A) and Asp86
˚
(N–O distance of 3.44 A) as shown in Figure 3.The
Lys89 bridges the inhibitor sulfonamide and the surface
˚
Arg297 guanidinium group (N–N distance of 4.54 A).
The crystal structure of compound 2i bound to mono-
meric human CDK2 was solved to reveal the basis for
the binding behavior.12 Inhibitor 2i was found to occupy
the ATP binding site of CDK2 as shown in Figure 2.In
particular, the N-1 of the imidazopyridines acts as a
hydrogen bond acceptor with the backbone NH of
˚
The side chain of Asp145 is 3.32 A from one of the aro-
matic chlorine, which creates speculation of the poten-
tial protonation of the Asp145 side chain.
X-ray crystallographic analysis, preliminary SAR study,
conformational analysis and docking studies furnished a
number of guidelines for analogue design.A complete
loss of affinity was found when the 2-position was either
4-aminopyridine (compound 2m) or phenyl ether (2n).
˚
Leu83 (heavy atom distance of 2.95 A), and the amino
group donates a hydrogen bond to the backbone car-
This is not unexpected as the pyridine of the compound
2m is too short to reach the Lys89 and adequately form
a hydrogen bond, while compound 2n has the bulkier
hydrophobic phenyl group, which does not have enough
room to be accommodated in this area.
The original lead compound 19 was used to validate our
CDK2 docking model using the CDocker methodol-
ogy.13 Using this CDK2 docking model, new proposed
molecules were docked to find their optimal binding
geometry.Then their binding affinity was predicted
using the DoMCoSAR (Docking mode that correlates
with SAR) methodology.11 DoMCoSAR uses a statisti-
cal model based on the energetics of the ligand/protein
residue interactions.The comparison of the binding
geometries of the CDocker methodology, however, has
shown good ability to predict a mode that can be con-
firmed by crystallography.The results of the prediction
versus the original lead compound 110 and the crystal-
lized compound 2i are shown in Figure 4.The prediction
agrees with the crystallographic results with 1.06 RMS,
with the main deviation being a slight distortion of the
sulfonamide.
Figure 2. Crystal structure of imidazopyridine 2i bound to CDK2.The
compound occupies the space of the ATR-pocket.The inhibitor
carbon atoms are in light blue, the nitrogen atoms are in blue, the
chlorine atoms are in grey, the oxygen atoms are in red and the sulfur
atoms are in yellow.The protein carbon atoms are in grey.

C. Hamdouchi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1943–1947
1947
Pines, J. Trends Biochem. Sci. 1993, 18, 195; (g) Gillet, C.
E.; Barnes, D. M. J. Clin. Pathol. 1998, 51, 310.
2. (a) Gray, N.; Detivaud, L.; Doerig, C.; Meijer, L. Curr.
Med. Chem. 1999, 6, 859; (b) Knockaert, M.; Greengard,
P.; Meijer, L. Trends Pharmacol. Sci. 2002, 23, 417; (c)
Lees, J.A;. Weinberg, R.A. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 4221; (d) Kong, N.; Fotouhi, N.; Wovkulich, P.
M.; Roberts, J. Drugs Future 2003, 9, 881; (e) Shapiro, G.
I. Mol. Biol. Intel. Unit 2002, 15, 208.
3. (a) Gadbois, D.; Hamaguchi, J. R.; Swank, R. A.;
Bradbury, E.M. Biochem. Biophys. Res. Commun. 1992,
184, 80–85; (b) Meijer, L.; Thunnissen, A. M.; White, A.
W.; Garnier, M.; Nikolic, M.; Tsai, L. H.; Walter, J.;
Cleverley, K. E.; Salinas, P. C.; Wu, Y. Z.; Biernat, J.;
Mandelkow, E. M.; Kim, S.-H.; Pettit, G. R. Chem. Biol.
2000, 7, 51.
4. (a) Sedlacek, H. H.; Czech, J.; Naik, R.; Kaur, G.;
Worland, P.; Losiewicz, M.; Parker, B.; Carlson, B.;
Smith, A.; Senderowicz, A.; Sausville, E. Int. J. Oncol.
1996, 9, 1143; (b) Shapiro, G.I. Clin. Cancer Res. 2004,
10, 4270; (c) Newcomb, E.W. Anti-Cancer Drugs 2004, 15,
411.
5. Hoessel, R.; Leclerc, S.; Endicott, J.; Nobel, M.; Lawrie,
A.; Tunnah, P.; Leost, M. E. D.; Marie, D.; Marko, D.;
Niederberger, E.; Tang, W.; Eisenbrand, G.; Meijer, L.
Nat. Cell Biol. 1999, 1, 60.
Figure 4. Comparison of the prospective predicted binding geometry
of the difluoroanalog of 2i (carbons in yellow) and the observed crystal
structures of the original lead compound 1 (carbons in green) and the
2i compound (carbons in light blue) bound to CDK2.
6. Zaharevitz, D.; Gussio, R.; Leost, M.; Senderowicz, A.;
Lahusen, T.; Kunick, C.; Meijer, L.; Sausville, E. A.
Cancer Res. 1999, 59, 2566.
7. Vesely, J.; Havlicek, L.; Strnad, M.; Blow, J.; Donella-
Deana, A.; Pinna, L.; Letham, D.; Kato, J.; Detivaud, L.;
Leclerc, S. Eur. J. Biochem. 1994, 224, 771.
Of particular interest is the observed specificity of this
class of compounds for CDKs over other kinases.Com-
pound 2i has shown no detectable effect when tested
against a representative set of other serine/threonine ki-
nases such as GSK3b, CAMKII, PKA, PKC-a,b,e,c.
This compound inhibits proliferation in HCT 116 cells
in tissue culture as determined by an MTT assay
(IC50 = 5. 5lM).
8. (a) Misra, R. N.; Xiao, H.-y.; Williams, D. K.; Kim, K. S.;
Lu, S.; Keller, K. A.; Mulheron, J. G.; Batorsky, R.;
Tokarski, J. S.; Sack, J. S.; Kimball, S. D.; Lee, F. Y.;
Webster, K.R. Bioorg. Med. Chem. Lett. 2004, 14, 2973;
(b) Anderson, M.; Beattie, J. F.; Breault, G. A.; Breed, J.;
Byth, K. F.; Culshaw, J. D.; Ellston, R. P. A.; Green, S.;
Minshull, C. A.; Norman, R. A.; Pauptit, R. A.; Stanway,
J.; Thomas, A. P.; Jewsbury, P. J. Bioorg. Med. Chem.
Lett. 2003, 13, 3021.
9. (a) Hamdouchi, C.; Keyser, H.; Collins, E.; Jaramillo, C.;
de Diego, E.; Spencer, C.; Dempsey, J.; Anderson, B.;
Leggett, T.; Stamm, N.; Schultz, R.; Watkins, S.; Cocke,
K.; Lemke, S.; Burke, T.; Beckmann, R.; Dixon, J.;
Gurganus, T.; Rankl, N.; Houck, K. A.; Zhang, F.; Vieth,
M.; Espinosa, J.; Timm, D.; Campbell, R.; Patel, B.;
Brooks, H.B. Mol. Cancer Ther. 2004, 3, 1; (b) Jaramillo,
C.; de Diego, E.; Hamdouchi, C.; Collins, E.; Keyser, H.;
Sanchez-Martinez, C.; delPrado, M.; Norman, B.; Brooks,
H.; Watkins, S.; Spencer, C.; Dempsey, J.; Anderson, B.;
Leggett, T.; Schultz, R.; Watkins, S.; Lemke, S.; Burke, T.;
Beckmann, R.; Dixon, J.; Gurganus, T.; Rankl, N.;
Houck, K. A.; Zhang, F.; Vieth, M.; Espinosa, J.; Timm,
D.; Campbell, R.; Patel, B. Bioorg. Med. Chem. Lett. 2004,
14, 6095.
In conclusion, we have identified a highly selective struc-
tural class of cyclin dependent kinase inhibitors by modif-
ying aminoimidazo[1,2-a]pyridines scaffold.Structural-
based design approach was successfully used to investi-
gate and design new compounds that were able to form
a new interaction with an identified residue of the pro-
tein, Lys89.This new interaction eliminated the need
for internal hydrogen bonds.These compounds were
found to compete with ATP for binding to a catalytic
subunit of the protein.Initial structure features required
for activity were defined.Further studies, including in
vivo activities and additional structure-activity relation-
ships are underway.
Acknowledgements
We are grateful to Dr.Chuan Shih and the Lilly Kinase
Platform for the advice and helpful discussions.
10. (a) Hamdouchi, C.; Sanchez, C.; Ezquerra, J. Synthesis
1998, 867; (b) Hamdouchi, C.; de Blas, J.; del Prado, M.;
Gruber, J.; Heinz, B. A.; Vance, L. J. Med. Chem. 1999,
42, 50.
11.Final compounds were characterized by 400/500 MHz 1H
NMR, HRMS, and LC/MS (97%).
References and notes
12.The crystal structure of compound 2i bound to monomeric
human CDK2 has been deposited in the Protein Data
Bank with deposition code 1YKR together with structure
factors and detailed experimental condition.
13. Wu, G.; Robertson, D. H.; Brooks, C. L., III; Vieth, M. J.
Comput. Chem. 2003, 24, 1549.
1.(a) Sherr, C.J.
Science 1996, 274, 1672; (b) Grana, X.;
Reddy, E.P. Oncogene 1995, 11, 211; (c) Hunter, T. Cell
1997, 88, 333; (d) Malumbres, M.; Barbacid, M. Nat. Rev.
Cancer 2001, 1, 222; (e) Stein, G.; Baserga, R.; Giordano,
A.; Denhardt, D. T. The Molecular Basic of Cell Cycle
Control; John Wiley & Sons: New York, 1998; p 389; (f)