Discovery of N-phenyl nicotinamides as potent inhibitors of Kdr

Article abstract of DOI:10.1016/j.bmcl.2007.07.077

Inhibition of tumor-induced angiogenesis is a promising strategy in anticancer research. Neovascularization is a process required for both tumor growth and metastasis. Enhanced understanding of the underlying molecular mechanisms has led to the discovery

Full text of DOI:10.1016/j.bmcl.2007.07.077

Bioorganic & Medicinal Chemistry Letters 17 (2007) 6003–6008
Discovery of N-phenyl nicotinamides as potent inhibitors of Kdr
Celia Dominguez,a,* Leon Smith,f Qi Huang,a Chester Yuan,a Xiaohu Ouyang,h Lynn Cai,a
Paul Chen,g Joseph Kim,a Timothy Harvey,a Rashid Syed,a Tae-Seong Kim,a
Andrew Tasker,a,ꢀ Ling Wang,b Michael Zhang,b Angela Coxon,c James Bready,c
Charles Starnes,c Danlin Chen,c Yongmei Gan,c Sesha Neervannan,e Gondi Kumar,d
Anthony Polverinob and Richard Kendallb
aChemistry Research and Discovery, One Amgen Center Drive Thousand Oaks, CA 91320, USA
bCancer Biology, One Amgen Center Drive Thousand Oaks, CA 91320, USA
cCancer Pharmacology, One Amgen Center Drive Thousand Oaks, CA 91320, USA
dPharmacokinetics and Drug Metabolism, One Amgen Center Drive Thousand Oaks, CA 91320, USA
ePharmaceutics Amgen Inc., One Amgen Center Drive Thousand Oaks, CA 91320, USA
fBristol-Myers Squibb Pharmaceutical, Research Institute, PO Box 5400, Princeton, NJ 08543-5400, USA
g515 Oakbry Court Thousand Oaks, CA 91360, USA
h2802 Parsons Blvd #1H, Flushing, NY 11354, USA
Received 27 March 2007; revised 8 June 2007; accepted 17 July 2007
Available online 22 August 2007
Abstract—Inhibition of tumor-induced angiogenesis is a promising strategy in anticancer research. Neovascularization is a process
required for both tumor growth and metastasis. Enhanced understanding of the underlying molecular mechanisms has led to the
discovery of a variety of pharmaceutically attractive targets. Decades of investigation suggest that vascular endothelial growth fac-
tor (VEGF) and its receptors, in particular VEGFR2 or kinase insert-domain-containing receptor (Kdr), play a critical role in the
growth and survival of endothelial cells in newly forming vasculature. The clinical utility of inhibitors of this receptor tyrosine kinase
is currently under intense investigation. Herein we report our efforts in this arena.
Ó 2007 Elsevier Ltd. All rights reserved.
Angiogenesis, the development of new blood vessels, is a
crucial process required for development and mainte-
nance of various physiological and pathological condi-
tions. Under normal circumstances angiogenesis occurs
during embryonic development, wound healing, and
female menstruation. Pathological angiogenesis is asso-
ciated with a variety of disease states, including psoria-
sis, diabetic retinopathy, rheumatoid arthritis, chronic
inflammation, and cancer. Tumor cells promote angio-
genesis to allow proper nourishment and removal of
metabolic waste.1
mors in response to hypoxia. The function of VEGF is
mediated by two high affinity receptors expressed on vas-
cular endothelial cells, Fms-like tyrosine kinase (Flt1 or
VEGFR1) and Kdr (VEGFR2), and VEGFR3 (Flt4),
found on lymphatic vessels. Messenger RNA for both
Flt-1 and Kdr is up-regulated in tumor-associated endo-
thelial cells, but not in the surrounding vasculature.2,3
All are members of the receptor tyrosine kinase family,
and consist of an extracellular ligand binding domain, a
trans-membrane region, and an intracellular split kinase
domain. Ligand binding effects receptor dimerization,
an event that increases the intrinsic kinase activity of the
non-phosphorylated protein by trans autophosphoryla-
tion and leads to the initiation of a classical growth factor
receptor signal transduction cascade.4,5 Kdr is an attrac-
tive target because of its tissue expression specificity and
key role in angiogenesis. The importance of the VEGF
axis in cancer therapy is highlighted by the observed clin-
ical activity of Avastin, an antibody directed against the
VEGF ligand.6 The most advanced tyrosine kinase inhibi-
Vascular endothelial growth factor (VEGF) is a key pro-
angiogenic cytokine that is released by almost all solid tu-
Keywords: VEGF; VEGFR-2 inhibitor; Kdr inhibitor; Angiogenesis
inhibitor; Kinase inhibitor.
*
Corresponding author. Tel.: +1 310 342 5503; fax: +1 310 342 5519
(C.D.); e-mail: celia.dominguez@chdi-inc.org
Andrew Tasker has helped in the preparation of the manuscript.
ꢀ
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.07.077

6004
C. Dominguez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6003–6008
N
O
O
N
O
Cl
H
N
O
H
N
N
N
N
CF3
F
H
H
H
O
N
H
BAY 43-9006 (1)
SU11248 (2)
Figure 1. Approved Kdr inhibitors.
Table 1. Structure–activity relationship of the central template
Cl
Cl
HN
HN
N
N
O
R
NH
N
N
P-Kdr n = 5 Ki (lM)11
PTK 787 (3)
IC50= 37 nM
Compound
R
Figure 2. PTK 787.
4
0.008a
N
tors in clinical development are BAY 43-9006 (Bayer/OSI)
1 and SU11248, (Sugen/Pfizer) 2 both of which have re-
cently received FDA approval (Fig. 1).1
5
6
>10a
0.49
>10
N
Br
The phthalazine, PTK 787 3, (Fig. 2) appeared intrigu-
ing as it failed to dock into any models of Kdr derived
from crystallographic data in the Protein Data Bank.7,8
At first glance the aminophthalazine could be envi-
sioned to be capable of interacting with the hinge re-
gion of the protein in a classical donor–acceptor pair
fashion. This appeared, however, to be a very unlikely
binding mode due to steric clash of the chlorobenzene
with the C(8) proton of the phthalazine. The pyridylm-
ethyl pendant seemed likely to be functioning as a
hinge-region binding element, although it was unclear
how the remainder of the molecule would be accom-
modated. We envisaged breaking the phthalazine N–
N bond to yield an anthranilamide. This would retain
most of the features of the phthalazine, be straightfor-
ward to synthesize, and moreover would be amenable
to analogue work. Given the poor reactivity of an aro-
matic amine situated ortho to a carbonyl function, it
seemed a pyridine-based—either 2- or 4-halo substi-
tuted—central template would provide a more attrac-
tive alternative, wherein bond construction could be
effected via an aromatic nucleophilic substitution.
Although useful in a synthetic sense, the introduction
of this nitrogen atom represented a departure from
the carbocyclic nature of the phthalazine. We therefore
undertook a broad approach surveying a variety of 2-
amino-substituted carboxamides derived from readily
available starting materials.
N
N
7
N
8
0.14
0.12a
0.015
>10
N
S
9
10
11
a UP-Kdr Ki.
substitutions exemplified by the 6-methyl 5, or 5-bromo
6, proved less satisfactory. Transposition of the nitrogen
to yield the picolinamide 7 was very poorly tolerated.
Pyrimidine 8 suffered a 10-fold loss in activity when com-
pared to nicotinamide 4. Thiophene 9 displayed activity
comparable to the pyrimidine, and the fully carbocyclic
anthranilamide 10 suffered only a twofold loss with re-
spect to the original nicotinamide. Finally, saturation of
the ring system highlighted by the cyclohexane 11, re-
sulted in a decrease of inhibitory activity.
The structure–activity relationships around the central
template are highlighted in Table 1. Maintaining the 4-
chlorophenyl amide, and the 4-(aminomethyl)pyridyl
pendants as constants, the unsubstituted nicotinamide 4
proved to be potent inhibitor with a Ki of 8 nM. Further
The nature of the aniline substituent and the pendant
aminomethylpyridine was explored via focused libraries
holding the nicotinamide constant. As shown in Table 2

C. Dominguez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6003–6008
6005
Table 2. Structure-activity relationship of nicotinamide aniline amide
Table 2 (continued)
R1
Compound
R1
P-Kdr n = 5 Ki (lM)11
HN
O
26
27
0.320
>6.66
N
NH
OH
N
Compound
R1
P-Kdr n = 5 Ki (lM)11
N
N
28
29
>6.66
12
0.002
Cl
>20
13
14
0.004
0.008
O
CF3
CF3
the aniline ring tolerated a modicum of substituents,—
preferably not a large substituent ortho to the nitrogen
linkage (compound 29)— including alkyl chains (17,
22, 26), phenyl 18, and carbocycles at the para position
(12). The meta-chloro 25 showed enhanced potency with
Ki of 2 nM, respectively. Disubstitution at both meta
and para positions with small groups was also well tol-
erated (13, 16). The addition of polar functionality typ-
ified by alcohols or basic amines resulted in a decrease in
activity (27, 28), suggesting that interactions between
this part of the molecule and the protein were hydropho-
bic in nature.
CF3
CF3
15
16
17
18
0.009
0.017
0.019
0.023
Cl
Cl
Variation of the aminomethylpyridine pendant provided
a narrow SAR and is outlined in Table 3. The piperonyl
Table 3. Structure–activity relationship of nicotinamides varying
pyridine pendant
O
O
19
20
0.010
0.023
HN
O
OCF3
NH
R2
N
Compound
R2
P-Kdr n = 5 Ki (lM)11
21
22
0.091
0.106
19
0.01
N
O
O
30
31
32
33
0.038
0.07
O
O
23
0.177/(0.029)
O
NH
N
0.135
24
25
0.259
0.002
O
F
>20
Cl
S

6006
C. Dominguez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6003–6008
methane 30 showed good activity as did indazole 31,
although an almost order of magnitude loss compared
to the starting pyridine. The para-fluoro benzyl ana-
logue 32 displays modest activity (Ki = 135 nM) while
benzothiophene 33 is a poor inhibitor.
The paraphenoxyphenyl amide 19 proved amenable to
cocrystallization with the enzyme (PDB Accession No.
2p2i). The protein adopts an Asp1046-Phe1047-
Gly1048 (DFG) out conformation (Fig. 3) wherein
the anilino ring reaches into an extended hydrophobic
pocket resulting from the movement of this section of
polypeptide chain. This conformation is not observed
in the presence of ATP or the classical quinazoline-
type inhibitors.8 Conversely, the pendant pyridyl
nitrogen forms a hydrogen bond with the hinge resi-
due Cys919, an interaction used by both ATP and
quinazolines. The conformational change in the pro-
tein exposes previously buried atoms in the ligand-
binding site. The NH of the aniline amide is
H-bonded with the conserved Glu885 from Helix C,
and the carbonyl oxygen forms an H-bond with
Asp1046 from the DFG loop. We can rationalize that
the loss of activity displayed by the pyrimidine 8 was
probably due to electrostatic repulsion from the oxy-
gen of Val914, and that both C(5) and C(6) substitu-
tion create a steric clash with the backbone of the
protein. The nitrogen atom of the transposed picoline
8 would result in an electrostatic repulsion with the
carboxylate of Glu885.
Figure 3. Compound 21 co-crystallized with Kdr (pdb 2p2i).
1400
1: Vehicle
1200
2: 37.5 mgs/kg
1000
3: 75 mgs/kg
800
p
< 0.0033
< 0.0001
< 0.0001
4: 150 mgs/kg
600
400
200
0
p
p
5
7
9
11
13
15
17
19
21
23
Time (days)
105
100
95
90
85
80
75
70
65
60
1 - Vehicle
2 - 37.5 mg/kg
3 - 75 mg/kg
4 - 150 mg/kg
0
2
4
6
8
10
12
14
16
18
20
Time (days)
Figure 4. Compound 4 in A431 xenograft dosed po/bid.

C. Dominguez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6003–6008
6007
Compounds 4 and 25 were potent in VEGF-driven HU-
VEC proliferation assays (8 nM and 40 nM, respec-
tively), displayed good selectivity in bFGF-driven
HUVEC proliferation with IC50s greater than 1 lM,
and demonstrated good pharmacokinetics properties in
the rat (compound (4) 100 mg/kg 45% F, Cl = 4.4 L/h/
kg,AUC 20,000 ngh/mL). Accordingly, nude mice bear-
ing an A431 human cervical epidermoid carcinoma
xenograft (Fig. 4) were treated orally twice daily with
compound 4. A statistically significant delay in disease
progression is realized in a dose responsive manner. This
is accompanied by statistically insignificant and non-
dose responsive loss of body weights.
nitrocarboxylates were first condensed with the appro-
priate aniline, then reductively aminated with an
aldehyde. Thiophene 9 was prepared from methyl 3-
aminothiophene-2-carboxylate via a reductive amina-
tion, ester hydrolysis, and condensation sequence.
Synthesis of 2-amino-nicotinamide analogues in Table
2 was carried out in a reverse fashion adopting a library
format. Beginning with 2-chloronicotinic acid, displace-
ment of the halogen with pyridylmethylamine followed
by EDCI-mediated coupling with the appropriate ani-
line provided products in an average yield of 77%
(Scheme 2).9 All analogues were characterized by
1
LCMS, H NMR, and mass spectrometry.
Scheme 1 illustrates the general synthesis of 2-aminonic-
otinamides, 3-aminopicolinamide, and the synthesis of
thiophene amide 9.9 Depending on the availability of
the starting material, pyridinecarboxylates bearing an
activated halogen were first condensed with the appro-
priate aniline, then subjected to nucleophilic aromatic
substitution with the pendant amine. Alternatively, 2-
In conclusion, we have identified several potent and effi-
cacious 2-aminonicotinamide Kdr inhibitors possessing
good PKDM properties. These compounds bind to the
enzyme in a DFG-out conformation. Subsequent to this
work, a research group at Novartis presented similar
ideas.10 Further work in this area will be reported in
due course.
OH
R1
X
H2N
O
A.
i)
Y
NH2
X = CH or N; Y = CH
X = CH, Y = N
R1
R1
HN
OH
HN
R1
X
X
X
Y
i)
X
Y
H2N
O
O
ii)
O
B.
R'
N
H
NO2
NH2
X = CH or N; Y = CH
X = CH, Y = N
X = CH or N; Y = CH
X = CH, Y = N
O
O
O
O
S
S
HN
O
Cl
S
iii)
iv)
C.
NH
NH2
NH
N
N
Scheme 1. Synthesis of 2-nicotinamides, 3-aminopicolinamide, aa nd thiophene amide. Reagents and conditions: (i) A—1: aniline, EDC, HOBT,
TEA, CH2Cl2, rt 38%, B—1: EDC, HOBT, TEA, CH2Cl2, 2: Fe, NH4Cl, EtOH/H2O, rt; (ii) R0-CHO, NaBH3CN, 15%; (iii) C—1: 4-
Pyridylcarboxaldehyde, NaBH(OAc)3, DCE, 75 °C, 6 h 70%, (iv) 1—LiOH, H2O, DioxaneÆMeOHÆH2O (3/1/1) 150 °C 5 min, 2—EDC, HOBT, p-
chlroaniline, DMF, DIPEA, 130 °C, 6 min, 15% over 2 steps.
R1
OH
OH
HN
R1
H2N
O
i.
ii.
O
O
Z
Z
N
NH
R2
N
Cl
Z
N
NH
R2
Z = H, 5-Br, 6-Me, 6-Cl
O
R2 =
N
NH
N
O
S
F
Scheme 2. Synthesis of 5 and 6-substituted 2-amino nicotinamides. Reagents and condition: (i) R2NH2, sealed tube, 130 °C, 38%; (ii) EDC, DIEA,
77%.

6008
C. Dominguez et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6003–6008
relationship of a new class of potent VEGF receptor
References and notes
tyrosine kinase inhibitors. J. Med. Chem. 1999, 42(26),
5369–5389.
1. Keren, P.; Zhu, Z. Development of angiogenesis inhibitors
to vascular endothelial growth factor receptor 2: current
status and future perspective. Frontiers in Bioscience 2005,
10, 1415–1439.
2. Kerbel, R. S. Tumor angiogenesis: past, present and the
near future. Carcinogenesis 2000, 21, 505–515.
3. Makinen, T.; Jussila, L.; Karpanen, T.; Keitunen, M. I.;
Pulkkanen, K. J.; Kauppinen, R.; Jackson, D. G.; Kubo,
H.; Nishikawa, S. I.; Yla-Herttuala, S.; Alitalo, K.
Inhibition of lymphangiogenesis with resulting lymphe-
dema in transgenic mice expressing soluble VEGF recep-
tor-3. Nat. Med. 2001, 7, 199–205.
4. Veikkola, T.; Karkkaninen, M.; Claesson-Welsh, L.;
Alitalo, K. Regulation of angiogenesis via vascular endo-
thelial growth factor receptors. Cancer Res. 2000, 60, 203–
212.
9. Chen, G., Booker, S., Cai, G. Croghan, M., Dipeitro, L.,
Dominguez, C., Elbaum, D., Germain, J., Huang, Q.,
Kim, J., Kim, T-S., Patel, V., Smith, L., Tasker, A., Xi,
N., Xu, S., Yuan, C. Preparation of Substituted Arylam-
ines and Methods of Use as Anti-tumor Agents. WO 02/
055501.
10. Manley, P.W., Bold, G. 2-Amino-Nicotinamide Deriva-
tives and Their Use as VEGF-receptor Tyrosine Kinase
Inhibitors. WO 01/55114.
11. Kinase reaction: [polyphor 96-well black round-bottomed
plate; Costar 3365] Using BIO-MEK FX 60 lL kinase
reaction buffer was added to 10 lL of P-Kdr (diluted in
KRB at 42.5 nM) in all wells except those in column 12
with a final concentration of 5 nM for Kdr enzyme (Stock
concentration of Kdr enzyme = 27 lM). This was fol-
lowed by 1.6 lL of inhibitor, which was incubated for
30 min at room temperature on a shaking platform. The
reaction was started with addition of 20 lL of 8.925 lM
ATP to the assay plates per well (final concentration of
ATP 1.05 lM). The reaction mixture was incubated for
60 min at room temperature on a shaking platform.
HTRF reaction: [Polyphor 96-well black round-bottomed
plate; Costar 3365]. To Costar 3356 plate were added
80 lL HTRF buffer, SA-APC (0.2 lL/80 lL), and Eu-
PT66 antibody (0.2 lL/80 lL), which was diluted with
EU-PT66 70X in HTRF buffer. 5 lL of the kinase reaction
was sampled using Biomek-FX and incubated for 30 min
at room temperature on a shaking platform, then read on
Ruby-Star. KRB: 20 mM Tris–HCl, pH 7.5, 10 mM
MgCl2, 5 mM MnCl2, 100 mM NaCl, 1.5 mM EGTA,
1 mM DTT, 0.2 mM SOV, and 20 lg /mL BSA (40 lL to
100 mL buffer/60 lL to 150 mL buffer). HTRF buffer:
50 mM Tris–HCL, pH 7.5, 100 mM NaCl, 0.1% BSA, and
0.05% Tween 20.
5. Neufeld, G.; Cohen, T.; Gengrinovich, S.; Poltorak, Z.
Vascular endothelial growth factor (VEGF) and its
receptors. FASEB J. 1999, 13, 9–22.
6. Hurwitz, H.; Fehrenbacher, L.; Novotny, W.; Cartwright,
T.; Hainsworth, J.; Heim, W.; Berlin, J.; Baron, A.;
Griffing, S.; Holmgren, E.; Ferrara, N.; Fyfe, G.; Rogers,
B.; Ross, R.; Kabbinavar, F. Bevacizumab plus irinotec-
an, fluorouracil, and leucovorin for metastatic colorectal
cancer. N. Engl. J. Med. 2004, 350(23), 2335–2342.
7. McTigue, M. A.; Wickersham, J. A.; Pinko, C.; Showalter,
R. E.; Parast, C. V.; Tempczyk-Russell, A.; Gehring, M.
R.; Mroczkowski, B.; Kan, C.-C.; Villafranca, J. E.;
Appelt, K. Crystal structure of the kinase domain of
vascular endothelial growth factor receptor 2: a key
enzyme in angiogenesis. Structure 1999, 7, 319–3308.
8. Hennequin, L. F.; Thomas, A. P.; Johnstone, C.; Stokes,
E. S. E.; Ple, P. A.; Lohmann, J.-J. M.; Ogilvie, D. J.;
Dukes, M.; Wedge, S. R.; Curwen, J. O.; Kendrew, J.;
Lambert-van der Brempt, C. Design and structure–activity