Pyrazolone based TGFβR1 kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.10.108

Interruption of TGFβ signaling through inhibition of the TGFβR1 kinase domain may prove to have beneficial effect in both fibrotic and oncological diseases. Herein we describe the SAR of a novel series of TGFβR1 kinase inhibitors containing a pyrazolone core. Most TGFβR1 kinase inhibitors described to date contain a core five-membered ring bearing N as H-bond acceptor. Described herein is a novel strategy to replace the core structure with pyrazolone ring, in which the carbonyl group is designed as an H-bond acceptor to interact with catalytic Lys 232.

Full text of DOI:10.1016/j.bmcl.2009.10.108

Bioorganic & Medicinal Chemistry Letters 20 (2010) 326–329
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Pyrazolone based TGFbR1 kinase inhibitors
*
Kevin Guckian , Mary Beth Carter, Edward Yin-Shiang Lin, Michael Choi, Lihong Sun,
P. Ann Boriack-Sjodin, Claudio Chuaqui, Benjamin Lane, Kam Cheung, Leona Ling, Wen-Cherng Lee
Biogen Idec Inc., 14 Cambridge Center, Cambridge, MA 02142, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Interruption of TGFb signaling through inhibition of the TGFbR1 kinase domain may prove to have ben-
eficial effect in both fibrotic and oncological diseases. Herein we describe the SAR of a novel series of
TGFbR1 kinase inhibitors containing a pyrazolone core. Most TGFbR1 kinase inhibitors described to date
contain a core five-membered ring bearing N as H-bond acceptor. Described herein is a novel strategy to
replace the core structure with pyrazolone ring, in which the carbonyl group is designed as an H-bond
acceptor to interact with catalytic Lys 232.
Received 26 August 2009
Revised 21 October 2009
Accepted 26 October 2009
Available online 29 October 2009
Keywords:
TGFb
Ó 2009 Elsevier Ltd. All rights reserved.
Pyrazolone
Kinase inhibitors
TGFb signaling is involved in a wide variety of physiological
processes. Inhibition of TGFb signaling has been shown to provide
therapeutic benefit in preclinical models in fibrosis and oncol-
ogy.1,2 Currently, several antibodies and antisense oligonucleotides
that interrupt TGFb signaling are progressing through clinical trials
for treatment in both fibrotic disease and oncology. Inhibition of
the kinase domain of TGFbR1 which is responsible for propagation
of the downstream signal after TGFb binding to the receptor repre-
sents another strategy for TGFb pathway inhibition. We therefore
sought to design selective ATP competitive inhibitors of TGFb sig-
naling by inhibition of the TGFbR1 kinase domain.
We and others have previously described vicinal bisaryl hetero-
cycles as potent and selective inhibitors of TGFbR1 (Fig. 1).1,3–6 In
an effort to expand diversity and improve the in vivo properties
of this class of inhibitors the nature of the heterocycle core was
investigated. Through modeling studies the pyrazolone heterocycle
was identified as a promising five-membered heterocyclic core.
The carbonyl group of pyrazolone was designed to capture the
HB interaction with Lys 232. Herein we describe the synthesis,
SAR, structural characterization, and in vivo properties of this class
of TGFbR1 kinase inhibitors.
addition, the core heterocycle itself makes a H-bonding interaction
with Lys 232. Finally, in all members of this class of inhibitors de-
scribed to date the moiety occupying the hydrophobic pocket
needs to contain a basic nitrogen in the 2-position (2-pyridyl in
most cases) to maintain good potency. Structural evidence sug-
gests that the pyridyl nitrogen is part of a complex set of interac-
tions between a bound water molecule, the protein, and the
inhibitor. Through modeling it was predicted that the carbonyl of
the pyrazolone core may obviate the need for the basic nitrogen
in the 2-position, by interacting with both Lys 232 and the bound
water molecule. Therefore, a series of 4-substituted pyrazolones
were synthesized to test this hypothesis.
The consensus binding mode for the vicinal bisaryl heterocycle
class of inhibitors is well understood from co-crystal structures of
both pyrazole6 and imidazole (manuscript in preparation) based
scaffolds. These studies show that one of the aromatic arms
protrudes into the hydrophobic pocket formed due to the small
serine gatekeeper found in TGFbR1 kinase. The other arm extends
toward the hinge where at least one H-bonding contact is made. In
* Corresponding author.
E-mail address: kevin.guckian@biogenidec.com (K. Guckian).
Figure 1. Structures of the literature described TGFb inhibitors.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.108

K. Guckian et al. / Bioorg. Med. Chem. Lett. 20 (2010) 326–329
327
Table 2
Effect of substitution position on TGFpRl Kinase inhibition
Compds
R2
R3
R4
Ki (nM)
PAI-Luc IC50 (lM)
10
11
12
13
14
15
16
17
–OH
–H
–H
–NH2
–H
–H
–H
–OH
–H
–H
–NH2
–H
–H
–H
–OH
–H
–H
–NH2
–H
–H
77
53
>10
>10
nd
nd
>10
nd
Scheme 1. Reagents and conditions: (a) EDC, DIPEA, N,O-dimethylhydroxylamine,
DCM/DMF; (b) LDA, ethyl acetate, THF, À78 °C; (c) dimethylhydrazine, pyridine, 90°
o/n; (d) Pd(OAc)2, CsOAc, DMF, tri-(furyl)phosphine, aryl(heteroaryl)Br or aryl(het-
950
952
130
640
450
45
eroaryl)I 125°, o/n; (e) NBS, DCM,
D; (f) Pd(Dppf)2, dioxane, 2 M NaCO3, aryl(het-
eroaryl boronic acid.
–CN
–H
–H
–CN
>10
>10
Assay description in Ref. 7.
Synthesis of this class of inhibitors proceeded as shown in
Scheme 1. Starting from the quinazoline acid 1 the Weinreb amide
was formed. LDA deprotonation of ethyl acetate and displacement
of the Weinreb amide gave ketoester 2. Heating of ketoester 2 and
dimethylhydrazine in pyridine gave pyrazolone 3 in good yield.
Analog synthesis was accomplished by two separate routes. The
first involves bromination of pyrazolone 3 with NBS and subse-
quent Suzuki coupling to give the final compound. The second
more direct route was C–H activation of the pyrazolone with Pd
in the presence of an Aryl bromide or iodide to give the final pyr-
azolone compounds.
Table 3
Effect of meta substitution on TGFbRI kinase activity
Compds
R3
Ki (
l
M)
PAI-Luc IC50 (lM)
A summary of TGFbR1 kinase inhibition activity by these ana-
logs is reported in Tables 1–4. For reference SB-431542 (Fig. 1)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
–CH2CH3
–CHCH2
–CH(CH3)2
–SCH3
–OCH3
–CF3
–CH2OCH3
–CH2OH
–CH2NH2
–CH2CN
–C(O)CH3
–C(O)NH2
–C(O)OCH3
–F
0.006
0.012
>2800
0.060
0.020
0.111
0.035
0.370
>2800
0.024
0.410
>2800
0.440
0.078
0.019
0.012
0.019
0.920
>2800
>2800
0.39
0.56
nd
>10
1.4
1.1
3.69
nd
nd
4.4
nd
nd
nd
Table 1
Effect of pyridyl nitrogen on TGFpRl kinase inhibition
0.99
1.9
0.22
>10
nd
Compds
Ar
TGFb Ki (
lM)
TGFb PAI-Luc IC50 (lM)
–CI
–Br
–NHSO2Me
–NHC(O)CH3
–NHSO2CH2CH3
–NHC(O)NHCH3
4
0.011
1.70
nd
nd
Assay description in Ref. 7.
5
6
0.005
>2.80
1.03
nd
has a 17 nM Ki in the biochemical assay and a 250 nM IC50 in the
PAI-Luc cellular assay. Table 1 shows the influence of basic nitro-
gens on TGFbRI kinase inhibition when placed in the hydrophobic
pocket relative to their aromatic counterparts. The 2-pyridyl deriv-
ative compound 4 was found to be sevenfold more potent than
phenyl and the 2-pyridyl-3-methyl derivative was found to be only
fourfold more potent than m-tolyl. A similar trend was seen in a
previously reported series with a pyrazole core.8 Introduction of
a basic nitrogen in the 3 or 4-position results in loss of inhibition.
These data prompted further exploration of simple aromatics.
We first sought to define the optimal position for substitution on
the aromatic ring. Table 2 shows that meta substitution of –OH, –
NH2, and –CN is superior to substitution in either the ortho or para
position therefore the meta position was chosen for further
exploration.
7
>2.80
nd
8
9
0.078
0.020
>10.0
3.05
Assay description in Ref. 7.

328
K. Guckian et al. / Bioorg. Med. Chem. Lett. 20 (2010) 326–329
Table 4
Effect of heteroaromatics on TGFpRI kinase activity
Compds
Ar
TGFb Ki (
l
M)
TGFb PAI-Luc IC50 (lM)
38
0.190
>10.0
39
40
0.073
0.365
>10.0
nd
41
42
43
0.054
0.64
1.87
>10.0
nd
nd
44
1.12
nd
Assay description in Ref. 7.
A large series of meta substituted inhibitors were synthesized
(Table 3) in which the size, shape and polarity of the substituents
were varied. Ethyl and ethylene substitution are the most potent
compounds in the series and are approximately equipotent with
the 2-pyridyl-3-methyl derivative described in Table 1. In general
this position was very sensitive to the nature of substitution with
preference for small hydrophobic substituents without branching
on the alpha carbon.
Figure 2. (A) Co-crystal structure of compound 24 and the kinase domain of
TGFbR1 (PDB ID 3KCF). (B) Overlay of compound 24 with a published co-crystal
structure of a vicinal biaryl pyrazole.6
Several heterocycles including bicycles were incorporated to
further explore the hydrophobic pocket (Table 4). Small heterocy-
cles were well tolerated with 3-substituted furan and thiophene
approximately equipotent with phenyl. Surprisingly, 2-substitu-
tion of furan and thiophene results in a significant loss in potency.
Finally, incorporation of bicycles leads to a large loss in potency
regardless of substitution.
The cellular potency of these analogs is also shown in Tables 1–
4. The data show that at least a one order of magnitude loss in po-
tency is seen when measuring TGFbR1 kinase inhibition inside the
cell. The SAR seen in the cellular assay roughly tracks to that seen
in the biochemical assay. We chose to take several of the com-
pounds which showed cellular potency less than 500 nM and study
them in vivo.
To confirm the binding hypothesis a co-crystal structure was
obtained with compound 24 (PDB ID 3KCF). As predicted the car-
bonyl oxygen of the pyrazolone ring interacts with Lys 232 and
the nitrogen of the quinoxaline ring binds to the backbone of His
283 (Fig. 2A). Because the water molecule in the hydrophobic
pocket is no longer hydrogen bonded to the inhibitor, the water
is pushed down further than that seen in co-crystal structures with
a 2-pyridyl group in the hydrophobic pocket (Fig. 2B), although the
hydrogen bonds between the water molecule and the protein res-
idues (Glu 245, Tyr 249, and Asp 351) are maintained. The struc-
tural rearrangement that ensues does lead to a slightly larger
hydrophobic pocket, which explains the tolerability of slightly lar-
ger groups in the meta position (Table 3) than has been seen in
other vicinal biaryl TGFb inhibitors.6
Oral bioavailability in general was poor in this series and suf-
fered from high clearance. One exception, compound 33 showed
34% oral bioavailability at 5 mg/kg in rat with a 3 hour half life.10
Compound 33 was further studied in a surrogate efficacy model
developed to monitor TGFbR1 kinase inhibition using imaging. At
5 mg/kg in mice a 43% inhibition of TGFb signaling was seen rela-
tive to vehicle control.11
The pyrazolone series described above has a novel pharmaco-
phore pattern relative to currently published TGFbR1 kinase inhib-
itors that allows replacement of the ubiquitous nitrogen H-bond
acceptor within the core with a carbonyl without significant loss
in biochemical potency. Although the cellular potency and oral bio-
availability in general were poor relative to other series, compound
33 showed significant TGFbR1 kinase inhibition in a surrogate effi-
cacy model in mice.
The selectivity of this class of inhibitors was tested by measur-
ing the percent inhibition at 10 lM of compound 18 in a panel of
218 kinases at Upstate. Only two kinases, ALK4 at 76% and PKC
alpha at 58% showed inhibition over 50%. These data suggest that
members of this class of inhibitors can be highly selective TGFbR1
kinase inhibitors.
Acknowledgment
We thank Dr. Cathy Sukow for assistance in generating Figure 2.

K. Guckian et al. / Bioorg. Med. Chem. Lett. 20 (2010) 326–329
329
8. Sawyer, S. J.; Anderson, B. D.; Beight, D. W.; Campbell, R. M.; Jones, M. L.;
Herron, D. K.; Lampe, J. W.; McCowan, J. R.; McMillan, W. T.; Mort, N.; Parsons,
S.; Smith, E. C. R.; Vieth, M.; Wier, L. C.; Yan, L.; Zhang, F.; Yingling, J. M. J. Med.
Chem. 2003, 36, 3953.
9. Fu, K.; Corbley, M. J.; Sun, L.; Friedman, J. E.; Shan, F.; Papadatos, J. L.; Costa, D.;
Lutterodt, F.; Sweigard, H.; Bowes, S.; Boriack-Sjodin, P.-A.; Arduini, R. M.; Sun,
D.; Newman, M. N.; Zhang, X.; Mead, J. N.; Chuaqui, C. E.; Cheung, H.-K.; Zhang,
X.; Cornebise, M.; Carter, M.; Josiah, S.; Singh, J.; Lee, W.-C.; Gill, A.; Ling, L. E.
Arterio. Thromb. Vasc. Biol. 2008, 28, 665.
10. PK parameters were determined by oral and iv administration of compound 33
at 5 mg/kg using 20% captisol as the vehicle. Eight time points were taken per
arm and fitting of the curves and calculation of the PK parameters was
performed with WinNonlin (Pharsight).
11. To evaluate the impact of ALK5 inhibitors on the TGFb signaling pathway, Balb/
C mice were injected with an adenoviral TGFb inducible luciferase reporter
intravenously. One day following adenoviral reporter administration, mice
were first treated with a single oral dose of inhibitors or vehicle, and then
References and notes
1. Yingling, J. M.; Blanchard, K. L.; Sawyer, J. S. Nat. Rev. Drug Disc. 2004, 3, 1011.
2. Bierie, B.; Moses, H. L. Nat. Rev. Cancer 2006, 6, 506.
3. Kim, D. -K.; Jang, Y.; Lee, H. S.; Park, H. J.; Yoo, J. J. Med. Chem. 2007, 50, 3143.
4. Li, H. Y.; Wang, Y.; Heap, C. R.; King, C. H. R.; Mundla, S. R.; Voss, M.; Clawson, D.
K.; Yan, L.; Campbell, R. M.; Anderson, B. D.; Wagner, J. R.; Britt, K.; Lu, K. X.;
McMillen, W. T.; Yingling, J. M. J. Med. Chem. 2006, 49, 2138.
5. Sawyer, J. S.; Beight, D. W.; Britt, K. S.; Anderson, B. D.; Campbell, R. M.;
Goodson, T.; Herron, D. K.; Li, H. Y.; McMillen, W. T.; Mort, N.; Parsons, S.;
Smith, E. C. R.; Wagner, J. R.; Yan, L.; Zhang, F. M.; Yingling, J. M. Bioorg. Med.
Chem. Lett. 2004, 14, 3581.
6. Singh, J.; Chuaqui, C. E.; Boriack-Sjodin, P. A.; Lee, W. C.; Pontz, T.; Corbley, M. J.;
Cheung, H. K.; Arduini, R. M.; Mead, J. N.; Newman, M. N.; Papadatos, J. L.;
Bowes, S.; Josiah, S.; Ling, L. E. Bioorg. Med. Chem. Lett. 2003, 13, 4355.
7. Ki values were determined using the Cheng–Prusoff equation from IC50s for
displacement of a labeled TGFbRI kinase inhibitor, HTS446284, 4-(3-pyridin-2-
yl-1H-pyrazol-4-yl)-quinoline6, from purified recombinant TGFbRI kinase
domain. Inhibition of cellular TGFbRI activity was determined by incubation
of the test compound with HepG2 cells harboring a TGFb signaling reporter,
PAI-luciferase. These assays are described in detail in Ref. 9.
challenged with 3
lg of recombinant human TGFb protein injected into the
peritoneal cavity 1 h later. The abilities of ALK5 inhibitors to block TGFb
inducible luciferase reporter expression in the animals were assessed at real-
time non-invasively using the bioluminescence animal imaging IVISTM system
according to manufacturer instructions (Caliper Life Sciences).