Oxindole derivatives as inhibitors of TAK1 kinase

Article abstract of DOI:10.1016/j.bmcl.2011.01.077

Several series of oxindole analogues were synthesized and screened for inhibitory activity against transforming growth factor-β-activating kinase 1 (TAK1). Modifications around several regions of the lead molecules were made, with a distal hydroxyl group

Full text of DOI:10.1016/j.bmcl.2011.01.077

Bioorganic & Medicinal Chemistry Letters 21 (2011) 1724–1727
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Oxindole derivatives as inhibitors of TAK1 kinase
Jeffrey W. Lockman a, , Matthew D. Reeder a, Rosann Robinson b, Patricia A. Ormonde b, Daniel M. Cimbora b,
⇑
Brandi L. Williams b, J. Adam Willardsen a
a Departments of Chemistry, Myrexis, Inc., 305 Chipeta Way, Salt Lake City, UT 84108, USA
b Discovery Biology, Myrexis, Inc., 305 Chipeta Way, Salt Lake City, UT 84108, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Several series of oxindole analogues were synthesized and screened for inhibitory activity against trans-
forming growth factor-b-activating kinase 1 (TAK1). Modifications around several regions of the lead
molecules were made, with a distal hydroxyl group in the D region being critical for activity. The most
potent compound 10 shows an IC50 of 8.9 nM against TAK1 in a biochemical enzyme assay, with com-
pounds 3 and 6 showing low micromolar cellular inhibition.
Received 9 December 2010
Revised 18 January 2011
Accepted 19 January 2011
Available online 26 January 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Kinase
Oxindole
TAK1
Inflammation is a crucial process for wound healing, but if not
done in an orderly fashion it can lead to unregulated cell prolifer-
ation and tumor growth.1,2 A key cellular factor linking inflamma-
(a)
HN
OH
N
O
O
tion and cancer is transcription factor NF-
jB. Knockout of
upstream NF- B regulator IKKb leads to a significant reduction in
j
O
O
tumor volume and number in a murine model of hepatocellular
carcinoma3 as well as a dramatic reduction in tumor incidence in
a colitis-associated cancer mouse model.4
S
NH2
OH
N
A critical component of the NF-jB-activating pathway is trans-
HO
O
NH2
O
forming growth factor-b-activating kinase 1 (TAK1, MAP3K7, EC
2.7.11.25). TAK1 associates with TAB1 and TAB2, and this complex
activates IKK after ubiquitination mediated by both TRAF6 and
Ubc13-Uev1A.5–7 Kinase-dead TAK1 has been shown to inhibit
2
1
O
(b)
B via both the TNFa8 and IL-19 pathways.
C
S
the activation of NF-
j
N
Small molecule inhibitors of TAK1 have been described (Scheme
1a). The natural product 5Z-7-oxozeaenol 1 is a potent, irreversible
ATP-competitive inhibitor of TAK1, with and IC50 = 8 nM. It acts to
D
N
N
OH
block NF-
jB activation in cells and prevents proinflammatory sig-
B
A
O
N
naling. Heterocyclic small molecule inhibitors of TAK1 have also
been described. The most potent of these thiophenecarboxamide
compounds, 2, has an IC50 = 50 nM and sub-micromolar cytotoxic-
ity across a panel of leukemia and lymphoma cell lines.10
High-throughput screening and elimination of promiscuous hits
led to the elucidation of one hit class: oxindoles as exemplified by
3. The molecule was subdivided into regions and SAR explored
(Scheme 1b).
H
3
Scheme 1. Reagents: (a) literature inhibitors 1, IC50 = 8nM; 2, IC50 = 50nM; (b) lead
molecule 3, IC50 = 46 nM.
Synthesis was undertaken using variations on literature meth-
ods (Scheme 2).11–13
Isatins were condensed with thiosemicarbazides using catalytic
sulfuric acid in ethanol at 120 °C for 15 min under 300 W of micro-
wave heating. The resulting semicarbazones were cyclized with
⇑
Corresponding author. Tel.: +1 801 214 7885.
E-mail address: jeffrey.lockman@myrexis.com (J.W. Lockman).
various
a-bromoesters in the presence of sodium acetate in
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.077

J. W. Lockman et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1724–1727
1725
Table 1
1
R
In vitro activities of compounds 3–19: zone A modifications
O
R
O
S
N
S
N
O
N
N
a, b
N
N
O
OH
O
N
R
O
N
H
H
N
H
3-42
R
Compound
E/Za
R
IC50
(
lM) or% inhibition at 10 lM
c
3
4
5
6
7
E
Z
E
Z
E
Z
Mix
E
Z
Amb
Amb
Mix
Mix
Mix
Amb
Mix
Mix
H
0.046
0.098
0.042
0.100
0.099
0.099
0.021
0.0089
0.032
3.0
9%
4%
35%
4%
10%
9%
0%
O
O
5-Me
5-Me
H
5-F
5-F
5-Cl
5-Br
5-Br
5-OCF3
5,7-Cl2
5-Cl-7-Me
4-Br-5-Me
4-Cl
N
H
N
H
43-46
8
9
1
Ar
R
O
10
11
12
13
14
15
16
17
18
19
Cl
O
d
Cl
O
O
N
N
H
H
47-51
4,7-Cl2
4-Cl-7-OMe
6-Cl-7-Me
H
N
N
Ar
a
Assignment of stereochemistry around C@N bond: E, Z, mix, or ambiguous.
N
e,f
O
O
Table 2
N
N
H
H
Activity of compounds 20–36: D-region modifications
O
52-55
S
N
Scheme 2. Reagents and conditions: (a) R-NHCONHNH2, H2SO4, EtOH; (b)
Br(R1)CHCOOEt, NaOAc, EtOH; (c) ArCHO, NaOMe, MeOH; (d) R1COAr, AcOH; (e)
(i) NaH, DMF; (ii) 2,4-dichloropyrimidine; (f) K2CO3, 2-ethoxyethanol.
N
N
R
O
N
Compound E/Za
R
IC50
(
lM) or% inhibition at 10 lM
ethanol to yield the oxindole products. Variations in the isatin lead
to changes in zones A and B, variations in the a-bromoester led to
change in zone C, and variations in the thiosemicarbazide led
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Amb 5-OPh-Ph
Amb 4-Me-Ph
Amb 4-OMe-Ph
0%
0%
>5
>5
>5
4%
17%
17%
6%
19%
>50
to changes in zone D. After purification, an E and Z isomer was ob-
served in many cases and assigned on the basis of 1-D and 2-D 1
H
E
Z
H
H
NMR. It could not be unambiguously identified which double bond
had changed in conformation, only that the D-ring was positioned
above the A-ring (defined as Z) or not (defined as E).
Amb 4-CF3-Ph
Amb 4-COOEt-Ph
Amb 3,4-OCH2O-Ph
E
Z
4-OBn-Ph
4-OBn-Ph
Different routes were employed to generate diversity around
the linker. Aldol condensations were utilized to generate alkenyl-
derivatives 43–46 and enones 47–51. Deprotonation of oxindoles
with NaH followed by addition of 2,4-dichloropyrimidine and sub-
sequent amination with various anilines led to compounds 52–55.
The primary in vitro assay used was a 33P-incorporation assay
utilizing a TAK1-TAB1 fusion protein.14 The potencies of active
compounds were confirmed with an IP kinase assay.15
Zone A tolerated changes at the 5-position of the oxindole
(Table 1). Similar biochemical activities were shown by the 5-
methyl, -bromo, -chloro, and -fluoro compounds, while the larger
5-CF3 derivative showed reduced inhibition and no further substi-
tution was tolerated. Both disubstituted derivatives and com-
pounds differently substituted around the A-ring were not active.
For most E/Z pairs there was little difference in the potencies of
the different conformations: only the 5-bromo compounds 10
Amb 4-N(Me)2Ph
Amb CH2(2-pyrrolidine) 30
Amb CH2(4-pyridine)
Amb 4-OH(3-pyridine)
Amb 2-OH-Ph
8.2
25
5.8
29
25
E
Z
3-OH-Ph
3-OH-Ph
a
Assignment of stereochemistry around C@N bond: E, Z, mix, or ambiguous.
4-phenol was critical to TAK1 inhibition. In order to determine
the function of this distal hydroxyl residue compounds were made
to specifically mimic possible modes of action (compounds 33–36).
These analogues suggest that the irreplaceable element of the hy-
droxyl group is its hydrogen-bond donating ability: compounds
containing substituents with similar electron donating abilities or
hydrogen-bond accepting capabilities are completely inactive
without the hydroxyl hydrogen. Additionally, the placement of
and 11 showed greater than 2-fold difference in IC50
.
The effects of varying substitution were similarly stark when
modifications were made to the D region (Table 2): a D-region

1726
J. W. Lockman et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1724–1727
Table 3
4-position of the thiazolone, however larger substituents were
not potent.
Efforts were also targeted towards oxindole compounds with
greater diversity in the linker, C, and D regions (Table 4). Due its
critical nature within the SAR of 21–36, all analogues in these ser-
ies included a hydroxyl group or alternative hydrogen-bond donor
in the area replacing the D region. Styrene-type compound 44
demonstrated weak activity, although the activity was eliminated
when the trihyroxybenzene group was replaced by a disubstituted
ring (43) or an imidazole (45–46).
Activity of compounds 37–42a: C-region modifications
1
R
z
Y
X
N
N
N
R
OH
O
N
H
More success was found when the styrene-like core was homol-
ogated to form an enone structure. The pattern of substitution
about the distal phenyl ring was found to be critical, with 3-OH
Compound
R
X
Y
Z
R1
IC50
(lM) or% inhibition at 10 lM
37
38
39
40
41
42
H
H
Cl
Cl
Cl
Cl
S
N
S
S
S
S
CH
N
CH
CH
CH
CH
O
H
O
O
O
O
H
H
Me
Et
i-Pr
Ph
0%
0%
0.63
>5
>5
4%
(48, 4.1
more activity (47, 0.56
tion) and unsubstituted compound (50, >5
l
M) showing slight activity, 3,4-OH2 showing significantly
M), and both the 2,4-OH2 (49, 20% inhibi-
M) showing little
l
l
activity. Unexpectedly, conversion of the trisubstituted enone to
a tetrasubstituted compound dramatically decreased potency (51,
39% inhibition). Additional screenings on these compounds, how-
ever, showed them to not be competitive with ATP. Compounds
were also made utilizing a 2-aminophenyl-4-oxindoylpyrimidine
core. These compounds (52–55) showed weak activity.
After the identification of potent TAK1 inhibitors in a biochem-
ical assay, efforts were shifted to determining the cellular activities
of these compounds. Due to the dramatic decreases in aqueous sol-
a
All compounds other than 38 showed ambiguous E/Z stereochemistry.
the hydroxyl substituent is key to the compounds’ inhibition:
movement of the group away from the 4-position leads to inactive
compounds. Lastly, the lack of activity for 33 suggests that it exists
predominantly as the pyridone rather than the 4-hydroxypyridine.
Limited SAR was conducted around the C region (Table 3). The
carbonyl of the thiazolone was critical, as the analogues of 3
replacing the 5-thiazolone with a thiazoline, 37, or a 1,2,4-triazole,
38, were both inactive. Methyl substitution was tolerated at the
ubility upon 5-substitution (PBS solubility at pH 7.4: 3 = 65
lM,
10 = 3.1 M), 3 and 6 were chosen for cellular work despite their
l
lessened potencies compared to 9–11. Cellular TAK1 inhibition
was measured utilizing an ELISA assay to detect TAK1 autophosp-
Table 4
Activity of compounds 43–55: Thiazolone replacements
R
O
N
Compound
R
IC50 (lM) or% inhibition at 10 lM
43
44
45
46
3,4-(OH)2Ph (E)
2,3,4-(OH)3Ph (E)
4-imidazoyl (E)
4-imidazoyl (Z)
>5
1.7
32%
20%
1
R
R
O
Cl
O
N
H
Compound
R
H
H
H
H
Me
R1
IC50
0.56
4.1
21%
>5
(lM) or% inhibition at 10 lM
47
48
49
50
51
3,4-(OH)2
3-OH
2,4-(OH)2
H
3,4-(OH)2
39%
N
N
N
R
O
N
Compound
R
% inhibition at 10 lM
52
53
54
55
4-OH
3-OH
2-OH
2,4-(OH)2
39%
36%
52%
43%

J. W. Lockman et al. / Bioorg. Med. Chem. Lett. 21 (2011) 1724–1727
1727
orylation on Thr187.16 Both 3 and 6 inhibited TAK1 in cells with an
IC50 of 11–12 M. This cellular shift typical is for ATP-competitive
inhibition of protein kinases, as the biochemical assay was con-
ducted at an ATP concentration of 25 M (twice substrate KM)
and intracellular ATP is approximately 1–10 mM. Unexpectedly,
treatment with neither compound 3 nor 6 leads to significant cyto-
References and notes
l
1. Coussens, L. M.; Werb, Z. Nature (London) 2002, 420, 860.
2. Balkwill, F.; Coussens, L. M. Nature (London) 2004, 431, 405.
3. Nature (London) 2004, 431, 461.
4. Greten, F. R.; Eckmann, L.; Greten, T. F.; Park, J. M.; Li, Z. W.; Egan, L. J.; Kagnoff,
M. F.; Karin, M. Cell 2004, 118, 285.
l
5. Wang, C.; Dend, L.; Hong, M.; Akkaraju, G. R.; Inoue, J.; Chen, Z. J. Nature
(London) 2001, 412, 346.
toxicity in HeLa cells (TC50 >100 lM). Due to the unexpected lack
of cytotoxicity exhibited by compounds 3 and 6, no additional bio-
chemical work was done to determine the selectivity of any tested
compounds for TAK1 over other kinases.
6. Hofer-Wabinek, R.; Schmid, J. A.; Stehlik, C.; Binder, B. R.; Lipp, J.; de Martin, R.
J. Biol. Chem. 2000, 275, 22064.
7. Kanayama, A.; Seth, R. B.; Sun, L.; Ea, C.; Hong, M.; Shaito, A.; Chiu, Y.; Deng, L.;
Chen, Z. J. Mol. Cell. 2004, 15, 535.
In conclusion, several series of oxindole-based inhibitors of
TAK1 were synthesized and screened. Only 5-subsitution on the
oxindole was tolerated and the compounds required the hydro-
gen-bond donating ability of a free hydroxyl group at the terminus
of the D region. Compounds 3 and 6 showed low micromolar cel-
lular inhibition of TAK1, although no significant cytotoxicity. These
data raise doubts to the role TAK1 plays in cellular viability and
consequently its attractiveness as an oncology therapeutic target,
although the possibility that compounds with higher solubility
and potency will affect cellular viability cannot be completely
discounted.
8. Sakurai, H.; Miyoshi, H.; Toriumi, W.; Sugita, T. J. Biol. Chem. 1999, 274,
10641.
9. Ninomiya-Tsuji, J.; Kajino, T.; Ono, K.; Ohtomo, T.; Matsumoto, M.; Shiina, M.;
Mihara, M.; Tsuchiya, M.; Matsumoto, K. J. Biol. Chem. 2003, 278, 18484.
10. Byth, K, Palakurthi, S., Yu, L., Zhang, Q. Int. Patent Appl. WO/2008/007072.
11. Karali, N.; Gursoy, A.; Terzioglu, N.; Ozkirimli, S.; Ozer, H.; Ekinci, A. C. Arch.
Pharm. Pharm. Med. Chem. 1998, 331, 254.
12. Karali, N.; Terzioglu, N.; Gursoy, A. Arch. Pharm. Pharm. Med. Chem. 2002, 335,
374.
13. All synthesized compounds were characterized by reverse phase-HPLC/MS and
1H NMR.
14. Inhibition of recombinant TAK1-TAB1 enzymatic activity was determined with
a
33P filter plate assay using 25
lM ATP and histone H1 as substrate.
15. HeLa cells overexpressing TAK1 and TAB1 were stimulated with TNF
a
in the
absence or presence of TAK1 inhibitors and TAK1 kinase activity was measured
by immunoprecipitating TAK1–TAB1 from the cell lysates and performing an
in vitro kinase reaction.
Acknowledgments
16. Cells overexpressing TAK1 and TAB1 were stimulated with TNFa in the absence
We thank the Myrexis, Inc. analytical chemistry group for their
work on the purification and characterization of described com-
pounds, and Drs. Kraig Yager, Robert Carlson, and Mark Anderson
for their support of this work.
or presence of TAK1 inhibitors and TAK1 autophosphorylation was measured
with an in-cell ELISA assay. Cells were fixed and stained with anti-phospho-
TAK1 (T187) antibody followed by HRP-conjugated secondary antibody and
HRP activity was measured.