Discovery and optimization of triazolopyridazines as potent and selective inhibitors of the c-Met kinase

Article abstract of DOI:10.1021/jm800043g

Tumorigenesis is a multistep process in which oncogenes play a key role in tumor formation, growth, and maintenance. MET was discovered as an oncogene that is activated by its ligand, hepatocyte growth factor. Deregulated signaling in the c-Met pathway ha

Full text of DOI:10.1021/jm800043g

J. Med. Chem. 2008, 51, 2879–2882
2879
Letters
Discovery and Optimization of
Triazolopyridazines as Potent and Selective
Inhibitors of the c-Met Kinase†
Brian K. Albrecht,*,‡ Jean-Christophe Harmange,‡
David Bauer,‡ Loren Berry,‡ Christiane Bode,‡
Alessandro A. Boezio,‡ April Chen,‡ Deborah Choquette,‡
Isabelle Dussault,§ Cary Fridrich,‡ Satoko Hirai,‡
Doug Hoffman,§ Jay F. Larrow,‡ Paula Kaplan-Lefko,§
Jasmine Lin,‡ Julia Lohman,‡ Alexander M. Long,‡
Jodi Moriguchi,§ Anne O’Connor,‡ Michele H. Potashman,‡
Monica Reese,§ Karen Rex,§ Aaron Siegmund,§
Kavita Shah,‡ Roman Shimanovich,‡ Stephanie K. Springer,‡
Yohannes Teffera,‡ Yajing Yang,§ Yihong Zhang,§ and
Steven F. Bellon‡
Figure 1. Reported c-Met inhibitors.
Amgen Inc., One Kendall Square, Building 1000, Cambridge,
Massachusetts 02139, and Amgen Inc., One Amgen Center DriVe,
Thousand Oaks, California 91320
ReceiVed January 17, 2008
Abstract: Tumorigenesis is a multistep process in which oncogenes
play a key role in tumor formation, growth, and maintenance. MET
was discovered as an oncogene that is activated by its ligand, hepatocyte
growth factor. Deregulated signaling in the c-Met pathway has been
observed in multiple tumor types. Herein we report the discovery of
potent and selective triazolopyridazine small molecules that inhibit
c-Met activity.
Figure 2. Analogues of triazolotriazine 2.
aim of the present work was to develop a potent, selective, ATP-
competitive orally bioavailable small molecule inhibitor of
c-Met.6
The receptor tyrosine kinase, c-Met, and its natural ligand,
hepatocyte growth factor (HGFa), are involved in cell prolifera-
tion, migration, and invasion and are essential for normal
embryonic development.1 However, when deregulated, the
c-Met/HGF pathway leads to tumorigenesis and metastasis.2 The
overexpression of c-Met and/or HGF, the amplification of
the MET gene, and mutations in the c-Met kinase domain have
been linked to human cancers.3 Recently it has been shown that
MET amplification occurs as a resistance mechanism in some
lung cancer patients that were initially responsive to gefitinib.4
Inhibition of c-Met activity in cell lines that reproduce this
resistance mechanism restored sensitivity to gefitinib. For these
reasons, c-Met small molecule kinase inhibitors have been
sought for therapeutic intervention.
Recently, we disclosed the structure of pyrimidinone 1 as a
potent (IC50 ) 10 nM) c-Met inhibitor (Figure 1).7 In an ongoing
effort to design novel inhibitors of the c-Met enzyme, we were
intrigued by a report from Sugen in which they showed that a
series of triazolotriazines of low molecular weight were potent
c-Met inhibitors.8 They reported that a representative example,
triazolotriazine 2, was shown to inhibit c-Met activity with an
IC50 of 6 nM.
Intrigued by the low molecular weight and unknown binding
mode of triazolotriazine 2 to c-Met, three structurally relevant
novel compounds were prepared and evaluated for their potency
against the c-Met enzyme (Figure 2). Since triazolopyridazine
3a had the greatest activity and was exquisitely selective against
other kinases,9 it was investigated further.
The cocrystal structure of 3a bound to the unphosphorylated
c-Met kinase domain revealed a bent “U-shaped” binding mode
with the inhibitor wrapped around Met1211 (Figure 3). A direct
hydrogen bond is formed between the backbone NH of Met1160
(linker) and the phenol-O with a distance of 3.0 Å. A second
hydrogen bond is mediated by a water molecule and bridges
the backbone carbonyl of Met1160 and the phenol-H. Other
notable interactions include a π-stacking interaction between
the triazolopyridazine core and Tyr1230 and a hydrogen bonding
interaction between N1 of the inhibitor and the backbone NH
of Asp1222.
Inhibition of the tyrosine kinase activity by an ATP-
competitive small molecule is a pharmacologically attractive
method that has been demonstrated for other tyrosine kinases.5
One limitation to small molecule kinase inhibitors is the
difficulty of obtaining specificity for the desired enzyme. The
†
Cocrystal structures of c-Met with 3a and 4 have been deposited in
the Protein Data Bank with access codes 3CCN and 3CD8, respectively.
* To whom correspondence should be addressed. Phone: 617-444-5166.
Fax: 617-577-9822. E-mail: brian.albrecht@amgen.com.
‡
Amgen Inc., MA.
Amgen Inc., CA.
§
a Abbreviations: ATP, adenosine triphosphate; HGF, hepatocyte growth
factor; Met1211/1260, methionine 1211/1260; Tyr1230, tyrosine 1230;
Asp1222, aspartic acid 1222; NADPH, nicotinamide adenine dinucleotide
phosphate; HATU, N,N,N′,N′-tetramethyl-O-(7-azabenzotriazol-1-yl)uro-
nium hexafluorophosphate.
Our previous crystallographic analysis of pyrimidinone 1
revealed a strikingly different mode of binding to the c-Met
active site (Figure 4). Instead of an overall bent shape,
10.1021/jm800043g CCC: $40.75
2008 American Chemical Society
Published on Web 04/22/2008

2880 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 10
Letters
Scheme 1a
Figure 3. Cocrystal structures of triazolopyridazine 3a and c-Met.
a Reagents and conditions: (a) glycolic acid, KOH, DMSO, 160°C, then
HCl to pH ∼3; (b) HATU, iPr2NEt, DMF, 50°C; (c) p-TsOH, MeOH, 55°C.
t
(d) 4, 10a-l, 10n: RB(OH)2, Pd (0), Bu3P or dppf, DMF or dioxane, ∆.
10m: RSnMe3, Pd(0), X-Phos, dioxane, ∆.
Figure 4. Cocrystal structures of pyimidone 1 and c-Met.
Figure 6. Cocrystal structures of triazolopyridazine 4 and c-Met.
were directly attached to the triazolopyridazine core of the
molecule. In addition to the cocrystal structure, metabolite
identification revealed that the C6-phenyl ring was prone to
metabolism. Incubation of triazolopyridazine 4 with rat and
human liver microsomes in the presence of NADPH qualita-
tively yielded C6-phenylarene oxidation products as the major
metabolites. With the proof of concept in hand, an effort to
explore the SAR around 4 was initiated.
On the basis of the above rationale, we necessitated a general
method that allowed for the rapid preparation of C-6 aryl
analogues. The synthesis began by treating 4-chloro-7-meth-
oxyquinoline 5 with glycolic acid in the presence of KOH to
afford substituted acetic acid derivative 6 (Scheme 1). HATU
coupling of acid 6 and 1-(6-chloropyridazin-3-yl)hydrazine 7
yielded hydrazide 8, which was dehydratively cyclized under
mildly acidic conditions. Triazolopyridazine 9 contained an
activated triazolopyridazine chloride, which was an effective
coupling partner in a variety of palladium catalyzed reactions
affording analogues 10.
The SAR commenced by the introduction of fluorine on the
phenyl ring in an effort to block the observed metabolic
activation and modulate π-stacking interactions with the hope
of increased potency (Table 1). Both meta and para substitutions
(10a/b) were potent in the enzyme assay with a substantial shift
in the cellular assay. The ortho-substituted fluorine analogue
10c was not as well tolerated by the c-Met enzyme. In addition,
both the 3,4-difluoro 10d and the 3,5-difluoro 10e analogues
were potent against c-Met enzyme activity but again showed a
Figure 5. Overlay of structures 1 and 3a and proposed hybrid 4.
pyrimidinone 1 adopts an extended conformation. In addition,
1 utilizes a quinoline to bind to Met1160 instead of a phenol.
The cocrystal structures of c-Met with 1 and 3a were aligned
to see how the inhibitors are situated relative to one another in
the c-Met binding pocket (Figure 5). The alignment of the two
structures shows that the quinoline of 1 and the phenol of 3a
occupy the same area of the protein and they both make a
donor-acceptor interaction with Met1160. We sought to
capitalize on this overlap and design a novel c-Met inhibitor
through the formation of a hybrid structure that contains the
triazolopyridazine core of 3a and the quinoline portion of 1.
Because the C-6 methoxy group on the quinoline was only 2.0
Å away from the C-6 phenyl group on 3a, we omitted it from
the hybrid product. The outcome of this exercise was the
formation of triazolopyridazine quinoline 4, which was an
efficient inhibitor of the c-Met enzyme with good cellular
activity (Table 1, entry 1).
The cocrystal structure of 4 and c-Met confirmed that 4 binds
the way it was envisioned (Figure 6). On the basis of the
cocrystal structure, we rationalized that modifications of the C-6
phenyl group on the triazolopyridazine core could modulate the
π-stacking interactions with Tyr1230 allowing for increased
potency. For this reason, aromatic and heteroaromatic groups

Letters
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 10 2881
Table 1. SAR of Triazolopyridazine Scaffold against c-Met.
position of the C-6 phenyl substituent projected toward solvent;
therefore, polar functionality was incorporated at this position.
Initially it was found that the p-methylbenzamide 10g was well
tolerated and potent in the cellular assay. The potency of 10g
could be improved to low single-digit nanomolar in the cellular
assay by incorporation of a fluorine or chlorine atom ortho to
the amide (10h or 10i, respectively).
In addition to substituted phenyl rings, five-membered
heterocycles were also explored. It was found that the 2-thiophe-
nyl analogue 10j was much more potent than the corresponding
2-furanyl analogue 10n. Although the 2- and 3-thiophenyl
analogues 10j/k were virtually equipotent in the enzyme assay,
10j was significantly more potent in the cellular assay at 6 nM.
Incorporation of a single methyl group at the thiophene 4
position (10l) showed an increase in potency in the enzyme and
cellular assays. Methylisothiazole 10m was prepared to optimize
the pharmacokinetic properties of thiophene 10l while still
maintaining cellular activity at 2 nM.
The pharmacokinetic profile of selected analogues was
evaluated (Table 2). Compound 4, trifluoro analogue 10f, and
methylisothiazole analogue 10m had desirable pharmacokinetics.
Chlorobenzamide analogue 10i was intrinsically stable in liver
microsomes and yet was rapidly cleared from the plasma
compartment in vivo.10 Methylthiophene 10l was metabolically
unstable and rapidly cleared in vivo. Interestingly, substitution
of the original triazolopyridazine 4 with three fluorine atoms
did not improve the microsomal clearance; yet the in vivo
clearance was markedly improved. Although 4 and 10m had a
higher bioavailability, the overall exposure of 10f was higher.
Analogue 10f possessed the best overall profile (PK/potency)
and was a candidate for our mouse pharmacodynamic assay.
Compound 10f was screened against a panel of tyrosine and
serine/threonine kinases. Impressively, 10f was found to be
highly selective for c-Met over a variety of kinases (>10 µM
against KDR, Lck, Src, IGF1R, Btk, Tie2, p38, Jnk2, CDK5,
Erk1, PKBR, PKAR, Msk1, Jak2, Abl, cKit, Aur2).
The inhibition of HGF-mediated c-Met phosphorylation in
mouse liver was evaluated. 10f was administered to mice by
oral gavage (3, 10, 30 mg/kg). Six hours postdose, human HGF
was injected iv to phosphorylate c-Met in the liver. The livers
were harvested, and c-Met phosphorylation was quantified. Oral
treatment of 10f led to a dose-dependent inhibition of HGF-
mediated c-Met phosphorylation with an approximate ED90 of
30 mg/kg and a corresponding plasma concentration of 6.7 µM
(Figure 7).
In summary, through the use of structural biology we were
able to devise a novel inhibitor of c-Met (4, IC50 ) 9 nM).
Although numerous potent analogues were prepared, analogue
10f possessed the most desirable profile. Furthermore, it was
determined that 10f was a potent inhibitor of HGF-mediated
c-Met phosphorylation in a mouse pharmacodynamic assay. The
inhibition of c-Met phosphorylation in this pharmacodynamic
model and the exquisite c-Met selectivity warrant future studies
a n g 2. b Inhibition of kinase activity. c Inhibition of HGF-mediated
c-Met phosphorylation in PC3 cells. See Supporting Information.
shift in the cellular assay. Interestingly, the trifluoro analogue
10f was potent in the enzyme and cellular assay with IC50
10 nM. The cocrystal structure of 4 also indicated that the para
<
Table 2. Pharmacokinetic Profile of Selected Compounds
Cl,a µL/min/mg
d
compd
RLM
MLM
Cl, L/h/kg
Vss, L/kg
T1/2, h
AUC0f∞
,
ng·h/mL
F,d
%
4
131
190
77
122
156
61
653
173
0.37c
0.058b
6.0c
0.38c
0.152b
3.8c
1.0c
3.5b
0.7c
0.3c
2.58b
2517
7840
ND
ND
5100
43
22
ND
ND
59
10f
10i
10l
10m
>1800
3.7c
1.0c
420
0.24b
0.35b
a In vitro (RLM ) rat liver microsomes; MLM ) mouse liver microsomes). In vivo experiments were carried out with male Sprague-Dawley rats (n )
3). b iv, 0.25 mg/kg (DMSO). c iv, 0.5 mg/kg (DMSO). d po, 2 mg/kg.

2882 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 10
Letters
(4) Engelman, J. A.; Zejnullahu, K.; Mitsudomi, T.; Song, Y.; Hyland,
C.; Park, J. O.; Lindeman, N.; Gale, C.-M.; Zhao, X.; Christensen, J.;
Kosaka, T.; Holmes, A. J.; Rogers, A. M.; Cappuzzo, F.; Mok, T.;
Lee, C.; Johnson, B. E.; Cantley, L. C.; Jaenne, P. A. MET
amplification leads to gefitinib resistance in lung cancer by activating
ERBB3 signaling. Science 2007, 316, 1039–1043.
(5) Baselga, J. Targeting tyrosine kinases in cancer: the second wave.
Science 2006, 312, 1175–1178.
(6) (a) For some recent examples see the following: Christensen, J. G.;
Schreck, R.; Burrows, J.; Kuruganti, P.; Chan, E.; Le, P.; Chen, J.;
Wang, X.; Ruslim, L.; Blake, R.; Lipson, K. E.; Ramphal, J.; Do, S.;
Cui, J. J.; Cherrington, J. M.; Mendel, D. B. A selective small molecule
inhibitor of c-Met kinase inhibits c-Met-dependent phenotypes in vitro
and exhibits cytoreductive antitumor activity in vivo. Cancer Res.
2003, 63, 7345-7355. (b) Smolen, G. A.; Sordella, R.; Muir, B.;
Mohapatra, G.; Barmettler, A.; Archibald, H.; Kim, W. J.; Okimoto,
R. A.; Bell, D. W.; Sgroi, D. C.; Christensen, J. G.; Settleman, J.;
Haber, D. A. Amplification of MET may identify a subset of cancers
with extreme sensitivity to the selective tyrosine kinase inhibitor PHA-
665752. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 2316–2321. (c) Zou,
H. Y.; Li, Q.; Lee, J. H.; Arango, M. E.; McDonnell, S. R.; Yamazaki,
S.; Koudriakova, T. B.; Alton, G.; Cui, J. J.; Kung, P.-P.; Nambu,
M. D.; Los, G.; Bender, S. L.; Mroczkowski, B.; Christensen, J. G.
An orally available small-molecule inhibitor of c-Met, PF-2341066,
exhibits cytoreductive antitumor efficacy through antiproliferative and
antiangiogenic mechanisms. Cancer Res. 2007, 67, 4408–4417.
(7) Bellon, S. F.; Kaplan-Lefko, P.; Yang, Y.; Zhang, Y.; Moriguchi, J.;
Rex, K.; Johnson, C. W.; Rose, P. E.; Long, A. M.; O’Connor, A. B.;
Gu, Y.; Coxon, A.; Kim, T.-S.; Tasker, A.; Burgess, T. L.; Dussault,
I. c-Met inhibitors with novel binding mode show activity against
several hereditary papillary renal cell carcinoma related mutations.
J. Biol. Chem. 2008, 283, 2675–2683.
(8) Zhang, F.-J.; Vojkovsky, T.; Huang, P.; Liang, C.; Do, S. H.; Koenig,
M.; Cui, J. Preparation of Triazolotriazines as c-Met Modulators for
Treating Cancer. WO2005010005, 2005.
Figure 7. Effect of 10f on HGF-mediated c-Met phosphorylation at
6 h (black circles correspond to plasma concentrations of 10f).
for this series of triazolopyridazines in cancer disease models.
These studies will be reported in due course.
Supporting Information Available: Analytical data and ex-
perimental protocols. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Giordano, S.; Ponzetto, C.; Di Renzo, M. F.; Cooper, C. S.; Comoglio,
P. M. Tyrosine kinase receptor indistinguishable from the c-met
protein. Nature 1989, 339, 155–156.
(2) Birchmeier, C.; Birchmeier, W.; Gherardi, E.; Vande Woude, G. F.
Met, metastasis, motility and more. Nat. ReV. Mol. Cell Biol. 2003, 4,
915–925.
(9) >25 µM against all kinases tested, including KDR, IGF1R, Tie2, Lck,
Jak3, BTK, p38R, PKBR, PKAR, Aur1/2, Abl.
(10) A qualitative bile-duct cannulated study in male Sprauge-Dawley rats
with this compound revealed that parent and metabolites were being
excreted in the bile after 8 h. Further details will be discussed
elsewhere.
(3) Dharmawardana, P. G.; Giubellino, A.; Bottaro, D. P. Hereditary
papillary renal carcinoma type I. Curr. Mol. Med. 2004, 4, 855–868.
JM800043G