Discovery of amido-benzisoxazoles as potent c-Kit inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.07.111

Deregulation of the receptor tyrosine kinase c-Kit is associated with an increasing number of human diseases, including certain cancers and mast cell diseases. Interference of c-Kit signaling with multi-kinase inhibitors has been shown clinically to successfully treat gastrointestinal stromal tumors and mastocytosis. Targeted therapy of c-Kit activity may provide therapeutic advantages against off-target effects for non-oncology applications. A new structural class of c-Kit inhibitors is described, including in vitro c-Kit potency, kinase selectivity, and the observed binding mode.

Full text of DOI:10.1016/j.bmcl.2008.07.111

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5115–5117
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of amido-benzisoxazoles as potent c-Kit inhibitors
a
a
a
a
a
Roxanne K. Kunz a, , Shannon Rumfelt , Ning Chen , Dawei Zhang , Andrew S. Tasker , Roland Bürli ,
Randall Hungate a, Violeta Yu b, Yen Nguyen b, Douglas A. Whittington c, Kristin L. Meagher c,
Matthew Plant d, Yanyan Tudor e, Michael Schrag f, Yang Xu f, Gordon Y. Ng d, Essa Hu a
*
a Department of Medicinal Chemistry, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
b Department of HTS Molecular Pharmacology, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
c Department of Molecular Structure, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
d Department of Inflammation, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
e Department of Molecular Sciences, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
f Department of Pharmacokinetics and Drug Metabolism, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320-1799, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Deregulation of the receptor tyrosine kinase c-Kit is associated with an increasing number of human dis-
eases, including certain cancers and mast cell diseases. Interference of c-Kit signaling with multi-kinase
inhibitors has been shown clinically to successfully treat gastrointestinal stromal tumors and mastocyto-
sis. Targeted therapy of c-Kit activity may provide therapeutic advantages against off-target effects for
non-oncology applications. A new structural class of c-Kit inhibitors is described, including in vitro
c-Kit potency, kinase selectivity, and the observed binding mode.
Received 17 June 2008
Revised 23 July 2008
Accepted 28 July 2008
Available online 31 July 2008
Keywords:
c-Kit
Ó 2008 Elsevier Ltd. All rights reserved.
Receptor tyrosine kinase
Mast cells
The receptor tyrosine kinase, c-Kit, is a transmembrane glyco-
protein, which is highly expressed on mast cells, as well as hema-
topoietic stem cells, germ cells, and gastrointestinal tract cells.1
Ligation of c-Kit to the cytokine stem cell factor (SCF) results in
dimerization, transphosphorylation, and initiation of multiple
downstream signaling pathways. Deregulation, overexpression, or
mutation of c-Kit signaling has been implicated in a variety of hu-
man tumors, including mast cell tumors, non-small cell lung tu-
mors, and gastrointestinal stomal tumors.2,3 Additionally, the
interaction between c-Kit and SCF regulates the survival, differen-
tiation, and proliferation of mast cells, which has implications in
inflammatory and autoimmune diseases.4,5
primary counter-screen to avoid acting through the MAP kinase
pathway.8 Preliminary in vivo metabolism studies of compound 1
revealed extensive hydrolysis of the lower trifluoromethyl amide
and formation of a potentially toxic aniline metabolite.9,10 While
bis-amide 1 represented a new structural class of c-Kit inhibitor,
a more selective compound, which avoided the potential formation
of reactive metabolites, was desired.11
Based on molecular modeling, we envisioned that the lower
amide in compound 1 could be replaced by a bicyclic benzisoxaz-
ole moiety. Substitution with this bioisostere would bypass the
amide hydrolysis metabolism pathway.12 We also chose to replace
Gleevec (also known as imatinib mesylate or STI-571) inhibits
the tyrosine kinases c-Abl, platelet-derived growth factor (PDGF),
and c-Kit, and is approved for the treatment of chronic myeloge-
nous leukemia and gastrointestinal stromal tumors.6 For non-
oncology applications, however, we desired a potent and more
selective c-Kit inhibitor for the treatment of mast cell-related
autoimmune or inflammatory diseases. As such, compound 1 was
identified through a screen of our kinase-preferred library. Gratify-
ingly, compound 1 exhibited greater c-Kit enzymatic potency com-
pared to Gleevec (Fig. 1)7 but was less selective against p38 kinase
(14-fold vs 236-fold, respectively), which was selected as our
N
N
O
Me
N
N
Me
N
N
N
H
HN
Gleevec
1
N
HN
CF3
HN
O
O
cKit IC50 = 0.058 μM
p38 IC50 = 13.69 μM
cKit IC50 = 0.014 μM
p38 IC50 = 0.199 μM
* Corresponding author. Tel.: +1 805 313 5370; fax: +1 805 480 3016.
E-mail address: kunz@amgen.com (R.K. Kunz).
Figure 1. In vitro potencies of Gleevec and bis-amide 1.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.07.111

5116
R. K. Kunz et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5115–5117
Cl
Cl
Cl
Cl
R1
MeO2C
MeO2C
MeO2C
F
a, b
c, d
H2N
F
HO
F
F
HO
e
R1
CHO
N
N
H
N
Cl
2
3
4
5
Y
N
Y
N
Cl
O
O
Cl
HO2C
X
X
N
H
NH2
f, g
O
N
R1
R1
h
N
N
H
N
H
6
7
Scheme 1. Reagents and conditions: (a) n-BuLi, THF, methylchloroformate, À78 to 23 °C (b) LDA, DMF, THF, À78 to 23 °C (50%, two steps); (c) hydroxylamine, EtOH; (d) N-
chlorosuccinimide, DMF, reflux (90%, two steps); (e) aniline, THF, 23 °C (54–78%); (f) NaH, DMF, 0 °C (40–100%); (g) LiOH, dioxane, reflux (13–100%); (h) aniline, HATU, Et3N,
DMF, reflux (ꢀ1–97%).
the C-6 methyl group of compound 1 with a chloro group based on
our in vivo metabolism studies, which showed benzylic hydroxyl-
ation and subsequent oxidation to an aldehyde metabolite. The
synthesis of amido-benzisoxazoles (i.e., 7, Scheme 1) began with
lithiation of 1-chloro-3-fluorobenzene (2) followed by treatment
with methylchloroformate to access the benzoate ester. ortho-Lith-
iation and formylation led to aldehyde 3, which was condensed
with hydroxylamine to give the intermediate oxime. Subsequent
chlorination with N-chlorosuccinimide gave compound 4 in good
yield for the four steps. Various anilines were introduced by dis-
placement of the chloride to provide hydroxyamidine 5. Closure
of the benzisoxazole ring13 was effected by treatment with sodium
hydride, which was followed by ester hydrolysis to give acid 6. Fi-
nally, a series of arylamides were prepared utilizing standard cou-
pling conditions.14
Initial structure–activity relationship (SAR) studies began with
substituted anilines (Table 1). A class of para-substituted com-
pounds was prepared (9–15), and it was found that increasing
the alkyl chain length from ethyl (9) to n-pentyl (12) resulted in
a 24-fold decrease in potency. Replacement of the ethyl group with
t-butyl, trifluoromethyl, or trifluoromethoxy (13–15) maintained
c-Kit potency, although p38 selectivity was diminished. Analogs
incorporating meta-substitution were examined next (16–19).
Although the intrinsic potency of m-nitrile compound 17 was sim-
ilar to that of Gleevec, the overall fold selectivity against p38 was
lower (90-fold vs 236-fold, respectively). Compound 19 was found
to have the best combination of c-Kit potency (3 nM) and fold
selectivity (113-fold). Compared to the original screening hit 1,
19 was nearly 5-fold more potent against c-Kit and 8-fold more
selective against p38. Disubstitution did not lead to an increase
in potency or selectivity (20).
Having identified the optimal arylamino-benzisoxazole frag-
ment, the m-trifluoromethyl group was kept constant while vary-
ing the aryl amide (Table 2). Substitution of the original
pyrimidine ring with an amino or acetamido-pyrimidine led to a
decrease in p38 selectivity (24 and 25). Similarly, amino- or acet-
amido-pyridines resulted in significant erosion in selectivity (22
and 23). Unsubstituted pyridine (26) or chloro-pyrimidine (27)
also did not yield an advantage over the parent pyrimidine com-
pound 19.
An extended kinase profile of benzisoxazole 19 was undertaken
to assess potential for off-target activity (Fig. 2). In addition to p38,
good selectivity over KDR was achieved (118-fold), as well as Tie-2
(>5
and cFMS was observed. Compound 19 was equipotent against the
most closely related receptor tyrosine kinase PDGF- (4 nM). This
lM). Moderate (20-fold or greater) selectivity against Src, Lck,
a
result is not entirely surprising given the sequence homology be-
tween the active site residues of c-Kit and PDGF.15 Indeed, Gleevec
is nearly equipotent against PDGF as compound 19 and it has been
suggested that Gleevec may have added potential for treatment of
diseases regulated through PDGF signaling.16
Table 1
SAR of substituted arylamino-benzisoxazoles represented by general structure 8
N
O
Cl
N
N
H
R3
R2
Table 2
O
SAR of m-CF3 amido-benzisoxazoles represented by general structure 21
8
N
N
H
Y
N
R1
O
Cl
Compound R1 R2
R3
c-Kit IC50
M)
p38 IC50
M)
Fold
selectivity
X
N
H
(
l
(
l
O
N
9
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
t-Bu
CN
OCF3
CF3
CF3
Et
Pr
Bu
Pent
t-Bu
CF3
0.007
0.007
0.028
0.170
0.008
0.011
0.399
0.189
0.296
1.83
0.190
0.217
0.187
0.050
5.34
57
27
11
11
24
20
17
10
91
39
113
24
21
10
11
12
13
14
15
16
17
18
19
20
N
H
CF3
p38 IC50
Compound
X
Y
c-Kit IC50
(
l
M)
(
l
M)
Fold selectivity
19
22
23
24
25
26
27
N
H
0.003
0.002
0.001
0.001
0.004
0.003
0.020
0.339
0.010
0.013
0.015
0.182
0.065
0.285
113
5
OCF3 0.011
CH
CH
N
N
CH
N
NH2
NHAc
NH2
NHAc
H
H
H
H
H
F
0.005
0.059
0.004
0.003
0.011
13
15
46
22
14
0.156
0.339
0.265
CI

R. K. Kunz et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5115–5117
5117
formed between compound 19 and the KDR mutant is a 2.9-Å
hydrogen bond between the benzisoxazole nitrogen and the back-
bone amide of Asp810. Although the potential exists for a hydro-
gen bond between Glu-640 and the arylamino-benzisoxazole
hydrogen, it is not observed in the KDR mutant co-crystal struc-
ture. The equivalent interaction is present in the c-Kit/Gleevec
structure (with the amide hydrogen of Gleevec) and this discrep-
ancy could be a result of subtle structural differences between
the KDR mutant and wild-type c-Kit.
IC50 (μM)
Kinase
N
cKit
p38
KDR
Tie-2
Src
Lck
cFMS
PDGF-α
0.003
0.339
0.353
> 5.00
0.117
0.075
0.071
0.004
O
O
Cl
N
N
H
19
N
N
H
CF3
In summary, a new class of c-Kit inhibitors has been described,
which exhibits superior in vitro potencies compared to Gleevec.
Compound 19, exhibiting the greatest selectivity against p38,
was nearly 20-fold more potent against c-Kit as compared to Glee-
vec. Improvement in off-target selectivity against p38 was
achieved by modification of the arylamino group.
Figure 2. Kinase selectivity of benzisoxazole 19.
References and notes
1. Nocka, K.; Majumder, S.; Chabot, B.; Ray, P.; Cervone, M.; Bernstein, A.; Besmer,
P. Gene Dev.. 1989, 3, 816.
2. Heinrich, M. C.; Blanke, C. D.; Druker, B. J.; Corless, C. L. J. Clin. Oncol. 2002, 20,
1692.
3. Lennartsson, J.; Jelacic, T.; Linnekin, D.; Shivakrupa, R. Stem Cells 2005, 23, 16.
4. Eklund, K. K. Immunol. Rev. 2007, 217, 38.
5. Marshall, J. Nat. Rev. 2004, 4, 787.
6. Capdeville, R.; Buchdunger, E.; Zimmerman, J.; Matter, A. Nat. Rev. Drug Discov.
2002, 1, 493.
7. c-Kit intrinsic activity was measured by c-Kit-homogeneous time resolved
fluorescent assay.
8. Reported activity of Gleevec against c-Kit and p38 was measured internally.
9. Compound 1 was dosed intravenously at 5 mg/kg in rats. After a period of 1, 4,
and 8 h, the plasma, bile, and urine were sampled and analyzed for presence of
metabolites by LC–MS/MS.
10. Zhu, Z.; Espenson, J. H. J. Org. Chem. 1995, 60, 1326.
11. For in vivo validation of mast-cell activation using structurally distinct
chemotypes to that described here, see: (a) Hu, E.; Tasker, A.; White, R. D.;
Kunz, R. K.; Human, J.; Chen, N.; Bürli, R.; Hungate, R.; Novak, P.; Itano, A.;
Zhang, X.; Yu, V.; Nguyen, Y.; Tudor, Y.; Plant, M.; Flynn, S.; Xu, Y.; Meagher, K.
L.; Whittington, D. A.; Ng, G. Y. J. Med. Chem. 2008, 51, 3065; (b) Chen, N.; Bürli,
R. W.; Neira, S.; Hungate, R.; Zhang, D.; Yu, V.; Nguyen, Y.; Tudor, Y.; Plant, M.;
Flynn, S.; Meagher, K. L.; Lee, M. R.; Zhang, X.; Itano, A.; Schrag, M.; Xu, Y.; Ng,
G. Y.; Hu, E. Bioorg. Med. Chem. Lett. 2008, 18, 4137.
Figure 3. Co-crystal structure of compound 19 and V916T KDR mutant enzyme.
Oxygen atoms, red; nitrogen atoms, blue; chlorine atoms, orange; fluorine atoms,
magenta; sulfur, yellow. The KDR mutant residues are labeled in black, and the
corresponding wild-type c-Kit residues are shown in blue.
To gain a better understanding of the binding mode of com-
pound 19, a co-crystal structure was obtained with a KDR mutant
protein where the gatekeeper residue was mutated from valine to
threonine to mimic the active site of c-Kit. The 2.9-Å resolution
structure is shown below in Fig. 3.17 Consistent with the Gleevec/
c-Kit co-crystal structure,18,19 benzisoxazole 19 binds to the DFG-
out form of the enzyme. The gatekeeper threonine is engaged in
a 3.2 Å hydrogen bonding interaction with the amide hydrogen,
similar to the Gleevec aminopyrimidine–Thr-670 hydrogen bond
in c-Kit. The pyrimidine ring forms a hydrogen bond with the back-
bone amide adjacent to the Cys-673 residue in the linker region,
which mirrors the hydrogen bond formed between this residue
in c-Kit and the pyridine nitrogen of Gleevec. The third contact
12. An in vitro metabolite identification study of compound 23 was conducted in
rat hepatocytes at 20
lM at 37 °C, which produced no evidence of
benzisoxazole ring-opening.
13. Angell, R. M.; Baldwin, I. R.; Bamborough, P.; Deboeck, N. M.; Longstaff, T.;
Swanson, S. WO2004010995.
14. Unoptimized yields.
15. Manning, G.; Whyte, D. B.; Martinez, R.; Hunter, T.; Sudarsanam, S. Science
2002, 298, 1912.
16. Bunchdunger, E.; Cioffi, C. L.; Law, N.; Stover, D.; Ohno-Jones, S.; Druker, B. J.;
Lydon, N. B. J. Pharm. Exp. Ther. 2000, 295, 139.
17. PDB deposition code number for the co-crystal structure is 3DTW.
18. Mol, C. D.; Lim, K. B.; Vandana Sridhar, H.-Z.; Chien, E. Y. T.; Sang, B.-C.;
Nowakowski, J.; Kassel, D. B.; Cronin, C. N.; McRee, D. E. J. Biol. Chem. 2003, 278,
31461.
19. Mol, C. D.; Dougan, D. R.; Schneider, T. R.; Skene, R. J.; Kraus, M. L.; Schiebe, D.
N.; Snell, G. P.; Zou, H.; Sang, B.-C.; Wilson, K. P. J. Biol. Chem. 2004, 279, 31655.