Design and structure-activity relationship of 3-benzimidazol-2-yl-1H-indazoles as inhibitors of receptor tyrosine kinases

Article abstract of DOI:10.1016/j.bmcl.2006.03.069

3-Benzimidazol-2-yl-1H-indazole analogs were developed as inhibitors of receptor tyrosine kinases (RTK). The synthesis and SAR of this series is reported.

Full text of DOI:10.1016/j.bmcl.2006.03.069

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3595–3599
Design and structure–activity relationship of 3-benzimidazol-2-yl-
1H-indazoles as inhibitors of receptor tyrosine kinases
Christopher M. McBride, Paul A. Renhowe, Carla Heise, Johanna M. Jansen,
Gena Lapointe, Sylvia Ma, Ramon Pineda, Jayesh Vora, Marion Wiesmann and
˜
Cynthia M. Shafer*
Small Molecule Drug Discovery, Biopharma Division, Chiron Corporation, 4560 Horton Street, Emeryville, CA 94608, USA
Received 22 February 2006; revised 20 March 2006; accepted 21 March 2006
Available online 5 April 2006
Abstract—3-Benzimidazol-2-yl-1H-indazole analogs were developed as inhibitors of receptor tyrosine kinases (RTK). The synthesis
and SAR of this series is reported.
Ó 2006 Elsevier Ltd. All rights reserved.
Tumor angiogenesis is the process that leads to the for-
mation of blood vessels in a tumor, which in turn sup-
ports cancer cell survival, local tumor growth, and the
development of distant metastasis.1–3 The key mediator
of this process is vascular endothelial growth factor
(VEGF) binding to its respective receptor tyrosine ki-
nase (RTK) to promote the proliferation, migration,
and survival of endothelial cells. Additional pro-angio-
genic RTKs and their growth factors are basic fibroblast
growth factor (bFGF) and platelet derived growth fac-
tor (PDGF) with additional functions in early angiogen-
esis and the stabilization of newly formed blood vessels,
respectively. Demonstration of several tumors to secrete
VEGF has made VEGFR-2 as well as PDGFR and
FGFR expression aid tumorigenesis directly with auto-
crine and/or paracrine signaling through these receptors,
resulting in deregulated autonomous growth.4,5 There-
fore blocking, in particular, VEGF signaling through
these receptors has become an attractive target for
anti-cancer therapy. The approaches to block signaling
include the neutralizing anti-VEGF antibody, soluble
VEGF receptors, antisense oligonucleotides targeting
VEGF, and small molecule inhibitors of the receptor
tyrosine kinases. There are multiple RTK inhibitors in
development targeting VEGFR-2 and in some cases
PDGFR. The most advanced small molecule inhibitors,
BAY43-9006 and SU11248, have both been approved
for advanced renal cell carcinoma,6,7 while PTK787 is
currently in Phase III trials.8
During the course of our RTK inhibitor program, we
desired an alternate scaffold to the 4-amino-3-ben-
zimidazol-2-ylhydroquinolin-2-one series (Fig. 1, 1).9–12
One idea entailed contracting the B ring from a 6-mem-
bered ring to a 5-membered ring which would change
the orientation of the benzimidazole (rings C and D) rel-
ative to the B ring, while retaining the donor–acceptor–
donor motif necessary for binding to the hinge region of
the RTKs. The scaffold that arose from this effort was
the 3-benzimidazol-2-yl-1H-indazole scaffold (2, hereaf-
ter referred to as indazole benzimidazole). While the un-
adorned indazole benzimidazole (2) exhibited
D
D
NH2
N
O
N C
C
NH
N
A
B
H
N
A
B
N
N
H
H
1
2
VEGFR-2 IC50 0.025 μM
HMVEC EC50 0.078 μM
VEGFR-2 IC50 3.3 μM
HMVEC EC50 5.9 μM
Keywords: Indazole benzimidazole; Receptor tyrosine kinase (RTK);
Vascular endothelial growth factor (VEGF); Basic fibroblast growth
factor (bFGF); Platelet derived growth factor (PDGF).
Figure 1. Potency of 4-Amino-3-benzimidazol-2-ylhydroquinolin-2-
one (1) and 3-Benzimidazol-2-yl-1H-indazole (2) in VEGFR-2 and
inhibition of VEGF-mediated proliferation of endothelial cells
(HMVEC).
*
Corresponding author. Tel.: +1 510 923 8086; e-mail:
cynthia_shafer@chiron.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.03.069

3596
C. M. McBride et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3595–3599
significantly less activity against VEGFR-2 compared to
the unsubstituted 4-amino-3-benzimidazol-2-ylhydro-
quinolin-2-one (1), we felt this series could be optimized
to yield potent RTK inhibitors. A dedicated effort en-
sued to improve the affinity of this scaffold toward
VEGFR-2 and the other RTKs of interest and to ex-
plore the SAR of this series.13
on ring D, including aromatic heterocycles (8), ethers
(9) and alkylamino heterocycles (10).
Because of the attractive in vitro profile of 6, the SAR of
the A ring was investigated using the 50-piperidinylpi-
peridine D ring substituent. Substitution at C-4 of
ring A was sensitive to the nature of the substituent
(Table 2, 11–12) with the 4-benzyloxy substituent result-
ing in lower VEGFR-2 affinity, while the 4-tert-butyl
urea gave equipotent affinity compared to 6.
Initially, a number of basic amines were surveyed on the
D ring to attempt to improve both in vitro potency and
solubility. Substitution of the 40 position of the benz-
imidazole with morpholine (Table 1, 3) led to loss of
activity, while moving the morpholine to the 50 position
(4) improved affinity 20-fold over the unsubstituted
indazole benzimidazole (2). Replacing the morpholine
with 1-methylpiperazine (5) yielded similar affinity but
somewhat improved the inhibition of VEGF-mediated
proliferation of endothelial cells (HMVEC). The piper-
idinylpiperidine substituent (6) conferred even greater
potency against VEGFR-2, while retaining the good cel-
lular activity achieved with 1-methylpiperazine. Substi-
tution with an acyclic amine (7) showed similar
potency to the cyclic amines 4 and 5. In addition to basic
amines, a number of other substituents were tolerated
A diverse group of substituents were incorporated at
C-5 and C-6, and found to have improved enzymatic
potencies (Table 2, 13–20) compared to unsubstituted
6. Potency in cells, however, was static or worse, except
when electronegative substituents, such as fluoro (18)
and trifluoromethyl (20), were incorporated at C-6,
which presumably increased the lipophilicity of the mol-
ecules and thus improved permeability of the com-
pounds through the cell membrane. At C-7, fluorine
(21) was the only substituent that maintained activity.
Our program focused on VEGFR-2 affinity but consid-
ered concurrent affinity for VEGFR-1, PDGFRb, and
Table 1. Structure–activity relationship of substituents on ring D of the indazole benzimidazole
5'
R'
4'
N
NH
N
N
H
Compound
R0
VEGFR-214
VEGFR-1
IC50 (lM)
PDGFRb14
IC50 (lM)
FGFR-1
IC50 (lM)
HMVEC
EC50 (lM)
IC50 (lM)
2
3
H
3.3
0.72
>3.0
3.4
1.4
5.9
>3.0
>3.0
0.83
>10
N
N
N
O
O
N
4’-
4
5
0.16
0.13
0.078
0.15
0.083
0.17
0.26
0.050
0.044
0.028
0.008
0.096
0.048
0.17
0.26
0.17
0.69
0.50
1.4
0.10
0.16
0.097
0.097
0.22
7.0
5’-
5’-
5’-
0.082
0.048
0.018
0.21
CH3
N
6
N
H3C
7
N
N
CH3
5’-
5’-
8
N
O
N
N
9
0.44
1.9
0.098
0.27
0.27
0.30
5’-
5’-
N
10
N
CH3

C. M. McBride et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3595–3599
3597
Table 2. Structure–activity relationship of substituents on ring A of the indazole benzimidazole
N
N
N
NH
4
5
R
N
N
6
H
7
Compound
R
VEGFR-2
IC50 (lM)
VEGFR-1
IC50 (lM)
PDGFRb
IC50 (lM)
FGFR-1
IC50 (lM)
HMVEC
EC50 (lM)
6
11
12
13
14
H
4-OBn
4-NH(CO)NHt-Bu
0.078
0.66
0.028
0.063
0.004
0.003
0.028
0.69
1.8
0.048
0.094
0.003
0.008
0.015
0.097
0.46
0.24
0.070
0.030
0.044
0.44
0.083
0.008
5-OBn
5-OPh
HN
0.19
0.080
15
0.044
0.021
1.7
0.016
0.22
5-
16
17
18
19
20
21
22
23
5-NH(CO)NHt-Bu
5-CO2Me
6-F
0.005
0.031
0.028
0.021
0.052
0.33
0.001
0.015
0.010
0.011
0.002
0.15
0.081
0.063
0.31
0.001
0.073
0.028
0.023
0.002
0.20
0.35
0.13
0.033
0.22
0.004
1.4
6-OBn
6-CF3
0.18
0.062
2.3
7-F
7-OBn
7-NH(CO)NHt-Bu
>3.0
>3.0
>3.0
>3.0
>3.0
>3.0
>3.0
>3.0
1.8
—
FGFR-1 to be advantageous. For the indazole benz-
imidazole series, VEGFR-1 and FGFR-1 affinity
tracked consistently for most compounds and was typi-
cally better than VEGFR-2 affinity. While the com-
pounds did exhibit affinity for PDGFRb, they were
typically 2- to 39-fold less potent against PDGFRb than
VEGFR-2. Compound 14, however, showed 5-fold
selectivity for PDGFRb over VEGFR-2 suggesting the
5-phenoxy as a potential starting point for PDGFRb
selective compounds. Conversely, the aliphatic 5-cyclo-
hexylmethyl amine substituent (15) yielded significantly
lower affinity against PDGFRb while retaining potency
against the other RTKs. These observations of various
kinase selectivity profiles make this series an interesting
tool to begin examining the effect of such profiles on
cell-based activity and efficacy.
Figure 2. Docking model of compound 18 in VEGFR-2 homology
model. Surface of the binding site is colored by atom-type (Oxygen,
red; Nitrogen, blue; and Carbon, white). Hydrogen bonds to hinge
residues are labeled, as is the surface from the catalytic Lys866.
In order to rationalize the observed SAR from a 3D
structural perspective, a homology model for VEGFR-2
was built and used to dock representative ligands.15
Figure 2 shows the docking model for compound 18,
with three hydrogen bonds from the indazole benzimid-
azole core to the hinge domain and the 50-piperidinylpi-
peridine substituent pointing out into solvent. This
model suggests that one tautomer for the benzimidazole
substructure will be preferred in order to present a
hydrogen bond donor to the hinge domain. Substitution
on the C-4 position with a hydrogen bond donor as in
compound 12 will drive the tautomeric equilibrium to
the desired form, while substitution with a hydrogen
bond acceptor (13) will drive the equilibrium to the
undesired form. The SAR confirms this hypothesis.
The model further supports the observation that substi-
tutions in the 5- and 6-positions are well tolerated. The
6-position gives access to the well-known selectivity
pocket, which in Figure 2 is located behind Lys866.
The high affinity of 5-substituted compound 16 can be
rationalized by the observation that the tert-butylurea
will be oriented similar to the same substituent in Pfiz-
er’s FGFR1 inhibitor, PD173074,16,17 which was the li-
gand in the FGFR1 crystal structure used as the
template for the homology modeling. The additional
affinity is most likely due to the hydrophobic interac-

3598
C. M. McBride et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3595–3599
R'
O
H
H2N
N
1:1 EtOH:Toluene, 80˚C
+
R'
NH
R
N
H2N
N
H
R
N
N
H
Scheme 1. Synthetic scheme for indazole benzimidazoles.
tions of the tert-butyl group with the P-loop. The model
finally confirms that substituents on the 7-position of the
A-ring and the 40-position of the D-ring are not tolerat-
ed due to collisions with the protein.
model (Fig. 3). Administration of 30 or 100 mpk qd of
compound 18 for 14 days resulted in 50% and 88% tu-
mor growth inhibition in the KM12L4a colon tumor
model in nude mice and was well tolerated.
The observations above deal with the general features of
a proposed VEGFR-2 binding site. The kinase inhibi-
tion data for this series show varying selectivity profiles.
Since the RTKs investigated in this paper are very
homologous, the differences in SAR must be governed
by subtle differences in the respective active sites of these
kinases.
In conclusion, a versatile scaffold was developed which
had potent inhibitory effects against the receptor tyro-
sine kinases involved in angiogenesis and blood vessel
maintenance. In addition, compound 18 was shown to
have favorable pharmacokinetics and exhibit impressive
tumor growth inhibition.
The indazole benzimidazole core is easily assembled
from an indazole aldehyde and a phenylenediamine
(Scheme 1) in modest yields (30–50%). While some inda-
zole aldehydes are commercially available, most were
synthesized from the corresponding indoles18 via a
nitrosation-rearrangement in modest to good yields
(50–80%).19–22 Additional analogs (e.g., 13, 15–16, and
23) could be synthesized following construction of the
core indazole benzimidazole by carrying a nitro through
the reaction sequence from commercially available 4-, 5-
or 7-nitroindole. Subsequent catalytic reduction fol-
lowed by reaction with electrophiles gave the desired
products.
Acknowledgments
The authors acknowledge William McCrea for the syn-
thesis of the nitroaniline used to synthesize compound
10, Timothy Machajewski and Brandon Doughan for
providing elemental analysis data for compound 18,
and Helen Ye for performing the matrigel and tumor
xenograft experiments.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2006.03.069.
The pharmacokinetic properties of 18 were examined,
and it showed an oral bioavailability of 62%, a t1/2 of
253 min, a clearance of 71 mL/min/kg, a volume of dis-
tribution of 24.8 L/kg, and an AUC of 170 mg min/mL.
This compound was further evaluated in angiogenesis
and tumor xenograft models. All doses tested complete-
ly inhibited neovascularization in the murine matrigel
References and notes
1. Folkman, J. Nat. Med. 1995, 1, 27.
2. Kerbel, R. S. Carcinogenesis 2000, 21, 505.
3. Carmeliet, P. Nat. Med. 2003, 9, 653.
4. Ferrara, N. Oncologist 2004, 9, 2.
5. Blume-Jensen, P.; Hunter, T. Nature 2001, 411, 355.
6. ‘FDA Approves New Treatment for Advanced Kidney
Cancer,’ The Food and Drug Administration Press
Release, 2005.
7. ‘FDA Approves Pfizer’s Sutent for Kidney and Bowel
Cancers,’ Pfizer, Inc., 2006.
8. Hecht, J. R.; Trarbach, T.; Jaeger, E.; Hainsworth, J.;
Wolff, R.; Lloyd, K.; Bodoky, G.; Borner, M.; Laurent,
D.; Jacques, C. In 41st ASCO Annual Meeting; Orlando,
FL, 2005; Vol. 3.
9. Antonios-McCrea, W. R.; Frazier, K. A.; Jazan, E. M.;
Machajewski, T. D.; McBride, C. M.; Pecchi, S.; Ren-
howe, P. A.; Shafer, C. M.; Taylor, C. Tetrahedron Lett.
2006, 47, 657.
10. Lopes de Menezes, D. E.; Peng, J.; Garrett, E. N.; Louie,
S. G.; Lee, S. H.; Wiesmann, M.; Tang, Y.; Shephard, L.;
Goldbeck, C.; Oei, Y.; Ye, H.; Aukerman, S. L.; Heise, C.
Clin. Cancer Res. 2005, 11, 5281.
Figure 3. Mice were implanted with FGF supplemented matrigel and
administered compound 18 daily for 7 days. Hemoglobin concentra-
tions were determined in the excised matrigel plugs.

C. M. McBride et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3595–3599
3599
11. Shafer, C. M. Abstracts of Papers, 228th ACS National
Meeting, Philadelphia, PA, United States, August 22–26,
2004; 2004; ORGN-248.
12. The lead compound from this series, CHIR-258, is
currently in Phase I clinical trials.
13. During the course of this project, two patent applica-
tions on indazole benzimidazole compounds as tyrosine
kinase and serine–threonine kinase inhibitors were pub-
lished (WO 2001002369 and WO 2001053268). The
compounds described in those publications, however,
were significantly different so as to not affect our
program.
14. For both of the biochemical VEGFR2 and PDGFRb
assays, the highly conserved mouse homologues have been
used.
15. The VEGFR-2 homology model was built using Chemical
Computing Group’s MOE software. Default settings were
used in the alignment and homology modeling module and
the FGFR1 crystal structure (2FGI) from the Brookhaven
Protein Data Bank was used as a template. Compound 18
was built and optimized in the active site of the homology
model using the Flo+apr2003 software from Thistlesoft.
16. Supplemental materials contains an overlay of compound
16 and PD173074.
D.; Schlessinger, J.; Hubbard, S. R. EMBO J. 1998, 17,
5896.
18. Anderson, W. K.; Gopalsamy, A.; Reddy, P. S. J. Med.
Chem. 1994, 37, 1955.
19. Buchi, G.; Lee, G. C. M.; Yang, D.; Tannenbaum, S. R.
J. Am. Chem. Soc. 1986, 108, 4115.
20. Allen, D. G.; Eldred, C. D.; Judkins, B. D.; Mitchell, W.
L. In PCT Int. Appl.: WO 9749699, 1997; pp. 40.
21. 6-Fluoroindazole aldehyde (3.25 g, 19.8 mmol, 1.0 eq)
and 4-(4-piperidylpiperidyl)benzene-1,2-diamine (5.41 g,
19.7 mmol, 1.0 eq) were dissolved in toluene and EtOH
(5:1), and heated at 70 °C under N2. After 2 h, the reaction
was opened to the atmosphere, and air was bubbled into
the reaction mixture using a fritted gas inlet tube for
20 min. After heating overnight at 70 °C, the reaction
mixture was concentrated to give a brown solid which was
purified by reverse phase HPLC. The TFA salt was
freebased by dissolving in EtOAc and washing the organic
layer with NaHCO3 (aq, satd), water, and brine. The
organic layer was then dried over Na2SO4, filtered, and
concentrated. The HCl salt was added by taking up the
free base in 1:1 CH3CN:H2O and adding HCl (2 M) until
the solid dissolved. The solution was lyophilized to give 18
as a yellow solid (2.1 g, 63%).
17. Mohammadi, M.; Froum, S.; Hamby, J. M.; Schroeder,
M. C.; Panek, R. L.; Lu, G. H.; Eliseenkova, A. V.; Green,
22. Further synthetic details and characterization for 18 are
included in the Supplemental Material.