Synthesis and evaluation of alkenyl indazoles as selective Aurora kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2010.03.018

A series of alkenyl indazoles were synthesized and evaluated in Aurora kinase enzyme assays. Several promising leads were optimized for selectivity towards Aurora B. Excellent binding affinity and good selectivity were achieved with optimized compounds in isolated Aurora subfamily assays.

Full text of DOI:10.1016/j.bmcl.2010.03.018

Bioorganic & Medicinal Chemistry Letters 20 (2010) 2443–2447
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and evaluation of alkenyl indazoles as selective Aurora
kinase inhibitors
*
Stéphanie Blanchard , Anthony D. William, Angeline C.-H. Lee, Anders Poulsen, Ee Ling Teo, Weiping Deng,
Noah Tu, Evelyn Tan, Kay Lin Goh, Wai Chung Ong, Chee Pang Ng, Kee Chuan Goh, Zahid Bonday, Eric T. Sun
S*BIO Pte. Ltd, 1 Science Park Road, #05-09 The Capricorn, Singapore Science Park II, Singapore 117528, Singapore
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of alkenyl indazoles were synthesized and evaluated in Aurora kinase enzyme assays. Several
promising leads were optimized for selectivity towards Aurora B. Excellent binding affinity and good
selectivity were achieved with optimized compounds in isolated Aurora subfamily assays.
Ó 2010 Elsevier Ltd. All rights reserved.
Received 21 December 2009
Revised 1 March 2010
Accepted 4 March 2010
Available online 7 March 2010
Keywords:
Aurora kinase
Indazole
Selectivity
Cancer cells typically contain mutations in a number of genes,
which ultimately result in uncontrolled cell growth and tumor
metastasis. As enzymes specific for and essential to cell growth
and division, Aurora kinases hold the potential to be important
control points for slowing the growth and spread of tumors. Aurora
family kinases regulate important events during mitosis including
centrosome maturation and separation, mitotic spindle assembly,
and chromosome segregation. Misregulation of Aurora kinases
due to genetic amplification and protein overexpression results
in aneuploidy and may contribute to tumorigenesis.
Three Aurora kinase isoforms A, B and C are a family of serine–
threonine kinases that are believed to play multiple roles in the
development and progression of cancer, by acting as regulators
of cell proliferation, by transforming normal cells into cancer cells
and by down-regulating the tumor suppressor p53. The Aurora en-
zymes are expressed in mammals, each of which is thought to play
important roles in regulating mitosis. Overexpression of Aurora A
leads to centrosome amplification and polyploidy. Aurora B is
known to phosphorylate histone H3 during mitosis, which may
play a role in chromosome condensation. Both Aurora A and B
are over-expressed in many tumor types, including colon cancer,
breast cancer and leukemia. Aurora B has been described as the
more suitable anticancer target because its inhibition results in
catastrophic mitosis that leads to cell death.1–3 Validation in mouse
xenograft studies4,5 supports the concept of targeting Aurora
kinases with small molecule inhibitors as a potentially powerful
approach to new antitumor therapies. Moreover advanced pro-
grams in several companies have led to new small molecule Aurora
inhibitors recently entering the clinic.2
As part of our program towards the development of anticancer
kinase inhibitors, we have developed indazoles with Aurora kinase
subfamily selectivity.
Our in-house compound library was screened for inhibitors of
Aurora A and B resulting in the discovery of compound 1 (Fig. 1)
as a potent Aurora A inhibitor with an IC50 of 0.21 lM. The subse-
quent hit to lead process was heavily influenced by prior art in the
area of kinase inhibitors and led to the design of new molecules
based on indazole moiety 2.
Herein we disclose the optimization strategy applied to the
indazole scaffold leading to potent and selective Aurora
inhibitors.
B
Introduction of various benzimidazole replacements at the
indazole 5-position (R2) were initially investigated. A styryl group
(double bond spacer, see compounds 5, 6 and 7) was identified as
the preferred linker over alkyne or cyclopropyl linking moieties for
4
R
Cl
N
NH
2
R
O
N
N
N
N
1
H
1
R
2
H
* Corresponding author. Tel.: +65 68275050.
E-mail address: stephanie_blanchard@sbio.com (S. Blanchard).
Figure 1.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.03.018

2444
S. Blanchard et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2443–2447
Table 1
2
R
N
1
N
H
R
R2
R1
H
Aur A IC50
1.7
(l
M)
Aur B IC50
(lM)
9
10
Compound
H
N
3
4
5
6
7
8
—
O
O
O
O
O
O
H
N
H
0.64
1.64
—
H
N
H
0.022
0.019
0.017
0.24
H
N
Me
Me
Me
—
H
N
0.047
0.11
S
S
H
N
Aurora A activity, with a 10-fold improvement in potency over 1
(Table 1). Substitution of the indazole 6-position (R1) was well tol-
erated and conferred novelty. To guide further optimization stud-
ies, models of the Aurora A and B ATP binding sites were utilized
from published crystal structures7,8 enabling new compounds to
be docked yielding much information from the subsequent analy-
sis. These computational studies showed that the linker should be
flat to avoid a clash with the protein (clash seen with alkyl and
cyclopropyl linkers). Although the alkyne linker is flat, it showed
less hydrophobic contacts than the double bond.6
tro-indazole (24 prepared according to the literature11
) and
protection of the N1 position, Heck reaction followed by reductive
amination were carried out to give nitro compound 28. Then 28
was reduced to the corresponding amine which was reacted with
thiophen-2-yl-acetyl chloride to yield, after THP deprotection,
alkenyl indazole 17. Compounds 9–15 were prepared using a cor-
responding chloro-nitroindazole and the cyclic amines were intro-
duced by chloride-displacement under basic conditions.11
Our first goal was to optimize general potency by exploring
three key substitution positions: R1, R2 and R3 (Table 2). Target
compounds were chosen based on docking and in silico analysis
of drug-like properties such as cLog P and polar surface area. Potent
compounds in the mono-substituted series, R1 = R3 = H (18 and 21)
Alkene substituted indazoles were easily accessible using a high
yielding procedure; the synthesis of representative Aurora inhibi-
tor 17 is outlined in Scheme 1. After iodination of 6-methyl-5-ni-
I
I
O N
O N
2
O N
2
2
a
b
N
N
N
N
N
N
H
H
THP
26
24
25
CHO
N
N
c
d, e
O N
O N
2
2
N
N
N
N
THP
THP
28
27
N
N
N
N
H
H
N
f, g
h
N
N
N
S
O
S
O
N
N
H
THP
29
17
Scheme 1. Synthesis of Aurora inhibitor 17. Reagents and conditions: (a) I2, KOH, DMF (98%); (b) DHP, pTSA, CAN (98%); (c) 3-vinyl-benzaldehyde, Pd(OAc)2, P(o-tolyl)3, Et3N,
DMF, 100 °C (70%); (d) 1-methyl-piperazine, DCM; (e) Na(OAc)3BH (71%); (f) SnCl2ÁH2O, DCM/MeOH; (g) thiophen-2-yl-acetyl chloride, pyridine, DCM (55%); (h) 4 N HCl in
dioxane, MeOH (69%).

S. Blanchard et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2443–2447
2445
Table 2
Activity and selectivity of indazole Aurora inhibitors
R4
R3
R2
N
R1
N
H
9
10
Entry
R2
R1
R3
H
R4
H
Aur A IC50
0.16
(lM)
Aur B IC50
0.046
(lM)
Selectivity A/B
3.5
>12
N
9
N
N
H
N
N
10
11
H
H
H
H
>10
0.81
0.13
N
N
N
N
H
0.62
5
12
13
NO2
H
H
H
H
0.24
0.20
0.11
2.2
2.5
N
N
N
NH2
0.081
H
N
H
N
O2
N
N
S
14
H
0.14
0.062
2.3
O
H
N
N
15
7
H
H
H
H
0.091
0.017
0.11
0.8
S
S
S
O
O
O
H
N
Me
Me
0.047
0.36
H
N
N
H
N
N
16
H
0.013
0.046
0.28
O
H
N
N
N
N
17
18
19
Me
H
H
0.079
0.035
1.9
0.13
0.6
1.3
N
S
S
O
O
O
H
N
H
0.026
0.099
N
N
H
N
H
Me
19
S
OMe
H
N
20
21
Me
H
H
H
H
H
0.042
0.035
0.18
0.23
O
OMe
H
N
0.031
1.1
4.7
S
O2
N
N
22
23
CN
H
H
Me
Me
0.28
0.35
0.059
0.031
N
N
11.3
H2N
gave 30 nM IC50s for both Aurora A and B. Although these com-
pounds were potent they were neither novel nor selective. To ad-
dress these issues di-substitution on the indazole ring was
investigated firstly with the assessment of the effect of the size
of the 6-position group, R1. Various groups such as alkyl or cyclic
amines were introduced demonstrating that small to medium
sized lipophilic groups were preferred for potency (examples 7,
11 and 15) whereas larger groups such as methyl-piperazine (10)
showed a decrease in activity due to a clash with the Val147 resi-
due in the active site.
A and B but, encouragingly, exhibited some selectivity (2–5-fold)
towards Aurora B. Installation of an amide or sulfonamide moiety
boosted potency to below 100 nM against one or both isozymes
(7, 14, 20 and 21).
In the course of our SAR studies, exploration of R3 provided a
breakthrough in the quest for selectivity towards Aurora B. Switch-
ing the methyl group from R1 (17) to R3 (19) dramatically improves
selectivity (24-fold swing) towards Aurora B. Similar results were
seen with compounds 22 and 23. A docking study (19) showed that
the methyl group at this position changes the conformation of the
amide linker creating a clash with the Aurora A protein explaining
the selectivity towards Aurora B and the dramatic potency loss to-
Examining the role of R2, analogues containing a nitro or amine
function suffered a decrease in potency (11, 12 and 13) for Aurora

2446
S. Blanchard et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2443–2447
Figure 2A. Compound 16 docked into the ATP binding site of the Aurora A X-ray structure 1MQ4. Compound 16 is displayed in tube representation with green carbons.
Aurora A is displayed in wire representation with gray carbons. The alpha carbons are labeled with three letter residue name and number. Hydrogen bonds between 16 and
Aurora A are displayed with yellow dashed lines. Glu211 and Ala213 are the same hinge region residues which hydrogen bond to the purine group in ATP.
Figure 2B. Compound 19 docked into the ATP binding site of the Aurora B X-ray structure 2BFX. Due to the 4-methyl substituent the amide of the 5-position is rotated 180°
compared to 16 docked into Aurora A. This orientation of the amide substituent can only be docked into Aurora B in a low energy conformation.
wards Aurora A6 (Fig. 2B). A comparison of Aurora A and B X-ray
structures revealed that the binding sites only differ by two resi-
dues. Tyr212 and Thr217 in Aurora A correspond to Phe172 and
Glu177, respectively in Aurora B6.
In parallel we investigated the right hand side of the molecule
with an assessment of R4. Introduction of hydrophilic groups (14,
16, 17 and 18) lowers cLog P while maintaining good potency.
Docking of 16 showed additional hydrogen bonds to the Aurora
A protein besides the two important interactions from the indazole
moiety itself to the hinge region (Fig. 2A). The compounds were
equipotent against both enzymes.
In conclusion we have designed and synthesized a series of
alkenyl indazoles with potent in vitro inhibition of both Aurora A
and B in the low nanomolar range. Furthermore, it has been shown
that small differences in active site architecture can be exploited to
achieve a significant degree of selectivity of one isozyme over the
other. Our studies culminated in the discovery of 19 which had
potent Aurora
Aurora A.
B activity and nearly 20-fold selectivity over
Although these compounds were potent and drug like with
good cell penetration properties we were unable to demonstrate
good cellular inhibition. Further work is needed to progress this
series into more advanced anticancer studies.
References and notes
1. Keen, N.; Taylor, S. Nat. Rev. 2004, 4, 927.
2. Pollard, J. R.; Mortimore, M. J. Med. Chem. 2009, 52, 2629.
3. Yeung, S. C.; Gully, C.; Lee, M. H. Mini-Rev. Med. Chem. 2008, 8, 1514.
4. Yang, J.; Ikezoe, T.; Nishioka, C.; Tasaka, T.; Taniguchi, A.; Kuwayama, Y.;
Komatsu, N.; Bandobashi, K.; Togitani, K.; Koeffler, H. P.; Taguchi, H.;
Yokoyama, A. Blood 2007, 110, 2034.
5. Manfredi, M. G.; Ecsedy, J. A.; Meetze, K. A.; Balani, S. K.; Burenkova, O.; Chen,
W.; Galvin, K. M.; Hoar, K. M.; Huck, J. J.; Leroy, P. J.; Ray, E. T.; Sells, T. B.;
Stringer, B.; Stroud, S. G.; Vos, T. J.; Weatherhead, G. S.; Wysong, D. R.; Zhang,
M.; Bolen, J. B.; Claiborne, C. F. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 4106.

S. Blanchard et al. / Bioorg. Med. Chem. Lett. 20 (2010) 2443–2447
2447
6. Poulsen, A.; William, A. D.; Lee, C. H. A.; Blanchard, S.; Teo, E. L.; Deng, W.;
Tu, N. P.; Tan, E. Y. L.; Sun, E. T.; Goh, K. L.; Ong, W. C.; Ng, C. P.; Goh, K. C.;
Bonday, Z. J. Comput. Aided Mol. Des. 2008, 22, 897. doi:10.1007/s10822-008-
9224-5.
7. Cheetham, G. M.; Knegtel, R. M.; Coll, J. T.; Renwick, S. B.; Swenson, L.; Weber,
P.; Lippke, J. A.; Austen, D. A. J. Biol. Chem. 2002, 277, 42419.
using the Prism 4.0 software (GraphPad Software Pte Ltd). All IC50 values are
the mean from at least two independent experiments.
10. Aurora B assay: The recombinant Aurora B kinase is a GST-tagged full-length
protein expressed in Hi-5 cells using the expression plasmid pFB-GST
(Invitrogen, USA). The activity assay is based on the principles of time-
resolved fluorescence (TRF). The reaction mixture consisted of 20–40 nM of
8. Nowakowski, J.; Cronin, C. N.; McRee, D. E.; Knuth, M. W.; Nelson, C. G.;
Pavletich, N. P.; Rogers, J.; Sang, B. C.; Scheibe, D. N.; Swanson, R. V.; Thompson,
D. A. Structure 2002, 10, 1659.
9. Aurora A assay: The recombinant Aurora A kinase is a hexahistidine-tagged full-
length protein expressed in Hi-5 insect cells using the expression plasmid
pFBHTb (Invitrogen, USA). The activity assay uses the PKLight™ HTS Protein
Kinase Assay Kit from Cambrex Corporation (New Jersey, USA). This assay
measures the consumption of ATP and is based on the bioluminescent
detection of the ATP remaining in the wells after the kinase reaction. The
enzyme, 0.5
Cat#1300) and 40
Buffer Cat# 9802). The compounds were tested at eight concentrations
prepared from fourfold serial dilution starting at 10 M. The reaction was
incubated at room temperature for 1 h. Twenty-five microliters of STOP buffer
(60 mM Hepes pH 7.5, 60 mM EDTA) added and 30 L quenched mixture was
transferred to Neutravidin coated plates (Cat#15402B, Pierce). After incubation
at room temperature (rt) for 30 min, the plates were washed with 4 Â 80 L Tris-
buffered saline with 0.5% Tween-20 (TBST). About 50 L of antibody mixture
(CST antibody Cat# 5070 and Perkin–Elmer antibody Cat# AD0124) was added
and the reaction incubated for 30 min, followed by washing with 4 Â 80 L TBST.
About 50 L of Dissociation-enhanced lanthanide fluorescence immunoassay
(DELFIA) Enhancement Solution (Perkin–Elmer Cat#4001-0010) was added and
the reaction incubated for 5 min. TRF signals were detected on a multi-label plate
reader (Victor2 V 1420, Perkin–Elmer). IC50 values were generated from the raw
data using the Prism 4.0 software (GraphPad Software Pte Ltd). All IC50 values are
the mean from at least two independent experiments.
l
M of biotinylated substrate (Cell Signaling Technology, Inc. (CST)
l
M of ATP (its Km value) in 50 L assay buffer (CST Kinase
l
l
l
l
l
reaction mixture consisted of 10–15 nM of recombinant kinase and 5.0
lM of
ATP (its Km value) in 40 L assay buffer (25 mM Hepes pH 7.5, 25 mM KCl,
l
l
5 mM MgCl2, 5 mM b-glycerophosphate, 1 mM dithiothreitol). The affinity-
purified kinase undergoes robust autophosphorylation and no exogenous
substrate is added. The compounds were tested at eight concentrations
l
prepared from fourfold serial dilution starting at 10 lM. The reaction was
incubated at room temperature for 2 h. Twenty microliters of PKLight™ ATP
detection reagent was added and the reaction was incubated for another
10 min. Luminescence signals were detected on a multi-label plate reader
(Victor2 V 1420, Perkin–Elmer). IC50 values were generated from the raw data
11. Blanchard, S.; Deng, W.; Lee, C. H. A.; Poulsen, A.; Teo, E. L.; Tu, N. P.; William A.
D. WO2007/058626, 2007.