Hit-to-lead optimization and kinase selectivity of imidazo[1,2-a] quinoxalin-4-amine derived JNK1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.06.087

As the result of a rhJNK1 HTS, the imidazo[1,2-a]quinoxaline 1 was identified as a 1.6 μM rhJNK1 inhibitor. Optimization of this compound lead to AX13587 (rhJNK1 IC₅₀ = 160 nM) which was co-crystallized with JNK1 to identify key molecular interactions. Kinase profiling against 125+ kinases revealed AX13587 was an inhibitor of JNK, MAST3, and MAST4 whereas its methylene homolog AX14373 (native JNK1 IC₅₀ = 47 nM) was a highly specific JNK inhibitor.

Full text of DOI:10.1016/j.bmcl.2013.06.087

Bioorganic & Medicinal Chemistry Letters 23 (2013) 5217–5222
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Hit-to-lead optimization and kinase selectivity of
imidazo[1,2-a]quinoxalin-4-amine derived JNK1 inhibitors
Bei Li , Oana M. Cociorva , Tyzoon Nomanbhoy, Helge Weissig, Qiang Li, Kai Nakamura, Marek Liyanage,
Melissa C. Zhang, Ann Y. Shih, Arwin Aban, Yi Hu, Julia Cajica, Lan Pham, John W. Kozarich,
⇑
Kevin R. Shreder
ActivX Biosciences, Inc., 11025 N. Torrey Pines Road, La Jolla, CA 92037, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
As the result of a rhJNK1 HTS, the imidazo[1,2-a]quinoxaline 1 was identified as a 1.6 lM rhJNK1 inhib-
Received 14 May 2013
Revised 19 June 2013
Accepted 27 June 2013
Available online 8 July 2013
itor. Optimization of this compound lead to AX13587 (rhJNK1 IC50 = 160 nM) which was co-crystallized
with JNK1 to identify key molecular interactions. Kinase profiling against 125+ kinases revealed AX13587
was an inhibitor of JNK, MAST3, and MAST4 whereas its methylene homolog AX14373 (native JNK1
IC50 = 47 nM) was a highly specific JNK inhibitor.
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
JNK1 inhibitor
Imidazo[1,2-a]quinoxalin-4-amine
Kinase profiling
AX13587
AX14373
The c-Jun N-terminal kinases (JNKs) are serine/threonine ki-
nases and members of the extensive MAP Kinase family. In mam-
mals there are three JNK genes: JNK1 (four isoforms), JNK2 (four
isoforms) and JNK3 (two isoforms).1 The JNK1 and JNK2 genes
are widely expressed, whereas JNK3 expression is restricted to
the brain, heart and testis.2 JNK activation by extracellular stimuli,
such as stress or cytokines, leads to the phosphorylation of several
transcription factors and cellular substrates implicated in cell pro-
liferation, cell survival, cell death, DNA repair and metabolism.3 Be-
cause these pathways are related to the pathogenesis of several
diseases JNKs represent valuable targets in the development of
new therapies.4 Celgene has disclosed the SAR data, co-crystal
structures with JNK3, and in vivo data for the JNK inhibitors CC-
3595 and CC-930,6 the latter of which entered phase II clinical trials
for the treatment of idiopathic pulmonary fibrosis.
Percent inhibition was calculated as the normalized decrease in
the fragment intensities of probe-labeled peptides in samples incu-
bated with inhibitor compared to those without. In cases where
the tryptic active site peptide is redundant (e.g., the JNKs) % inhibi-
tion cannot be assigned to an individual kinase and multiple ki-
nases appear in a single row (Table 2). Considering the high
screening concentration of 30 lM, this analysis indicated 1–3 were
selective inhibitors of the JNK family with few off-targets.
The optimization of the most potent and JNK selective of these
hits (1) was conducted via the derivatization of 4-chloroimi-
dazo[1,2-a]quinoxaline10. This intermediate was treated with var-
ious cyclohexylamine derivatives (TEA (2 equiv), DMSO, 100 °C,
16 h) leading to compounds 4–7 (Table 1). This effort lead to the
trans-hydroxy derivative 7 and an accompanying 1.8- to 9.2-fold
increase in the rhJNK1 inhibitory activity relative to compounds
1–6. While progress was made with respect to potency, kinase pro-
filing indicated that little ground was made in tuning out off-target
kinases for compound 7 (Table 2). Nevertheless we moved forward
with 7 knowing that at some point we would have to improve JNK
selectivity.
As the result of a high throughput screen of an internal com-
pound library against recombinant human JNK1 (rhJNK1), three
compounds with rhJNK1 IC50 values <10 lM (1–3, Table 1) were
found.7 These compounds were profiled against a panel of over
125 kinases using the desthiobiotin-tagged ATP probe AX99898
as previously described.9 After probe-labeling of PC3 lysate in the
To explore the role of the fused imidazole ring of 7, two com-
pounds (8 and 9 in Fig. 1) were synthesized from 4-chloropyrrol-
o[1,2-a]quinoxaline11 and 4-chloro-1,2-dihydroimidazo[1,2-a]
presence of inhibitors (30 lM) followed by trypsinization, kinase
active site peptides were identified and quantified using LC/MS.
quinoxaline12 in an analogous fashion to 7 (rhJNK1 IC50 = 0.91
lM).
By comparison to 7, compounds 8 and 9 had significantly attenu-
ated rhJNK1 inhibitory activities, with rhJNK1 IC50 values of
⇑
Corresponding author. Tel.: +1 858 526 2517; fax: +1 858 587 4878.
E-mail address: kevins@activx.com (K.R. Shreder).
These authors contributed equally to this work.
>100 lM and 73 lM respectively. This loss in potency indicated
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.06.087

5218
B. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5217–5222
Table 1
Table 2
rhJNK1 IC50 values of the HTS hits 1–3 and derivatives of 1
Kinase selectivity profiling (KiNativ™) of 1–3 and 7 in PC3 lysate showing % inhibition
of ATP probe labeling (AX9989) of the indicated kinase (leftmost column) at
a
screening concentration of 30
l
M
N
N
N
R
1
2
3
7
N
H
JNK1,JNK2,JNK3 80.4 41.6 46.2 97.5
Compound
R
rhJNK1 IC50
1.6
(lM)
MAST3
CDK9
<35
<35
<35
<35
50.7 <35
<35 90.7
1a
PI4KB
55.9 59.9 66.5 90.8
37.2 35.7 <35 79.8
CDK10
PITSLRE
PIP4K2C
PIP4K2C
PEK
2a
3a
6.7
8.4
<35
<35
<35 82.4
<35 59.4 40.3 51.9
<35 52.6 <35 50.9
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35 65.0
<35 41.0
<35 36.9
<35 38.9
<35 37.0
<35 49.6
<35 45.6
F
GCN2
4
5
8.0
2.4
F
GCN2
H
PCTAIRE1
TAO2
O
N
S
O
CK2α2
LKB1
H2N
HO
6
1.7
Data is shown in the form of a heat map and structures of compounds can be found
in Table 1. Over 100 additional kinases were screened but did not show significant
inhibition (i.e., <35%). More than one probe labeling site exists for PIP4K2C and
7a
0.91
a
Profiled against a panel of human kinases (see Table 2).
N
N
N
N
N
HO
HO
that the non-bridgehead nitrogen and aromaticity of the fused
imidazole ring present in 7 were critical JNK1 binding elements.
Accordingly we chose not to make changes to the core imi-
dazo[1,2-a]quinoxaline scaffold.
N
H
N
H
8
9
To further increase potency and selectivity, 7- and 8-subsituted
carboxy and carboxamide derivatives of 7 were explored. The syn-
theses of 8-substituted derivatives are shown in Scheme 1. When
methyl 3-fluoro-4-nitro-benzoate (10) and ethyl imidazole-2-car-
boxylate were heated in the presence of cesium carbonate com-
pound 11 was produced via a SNAr reaction. When the nitro
group of 11 was reduced to the corresponding amine with iron
powder in ethanol/water/acetic acid and then subjected to heating,
cyclization to form the tricyclic intermediate 12 occurred. Chlori-
nation of 12 to give the 4-Cl intermediate 13 was accomplished
using phosphoryl chloride and catalytic N,N-dimethylaniline. Com-
Figure 1. Structure of pyrrolo[1,2-a]quinoxaline (8) and 1,2-dihydroimidazo[1,2-
a]quinoxaline (9) analogs of 7.
pound 13 was converted to 14 by treatment with trans-4-aminocy-
clohexanol in DMSO at 120 °C. Hydrolysis of 14 was carried out
using aqueous potassium hydroxide in THF to give 15. Reaction
of 15 with primary amines using the carbodiimide EDC at room
temperature yielded 8-position amides. 7-subsituted carboxy and
carboxamide derivatives were synthesized in a similar manner
starting from methyl 4-fluoro-3-nitro-benzoate. The rhJNK1 inhib-
O
O
O
O
N
O
N
N
N
N
N
O
F
N
b
c
O
d
O
a
O
EtO2C
N
H
Cl
O2N
O2N
10
11
12
13
O
O
O
N
N
HO
N
N
HO
N
N
N
HO
N
OH
NHR
f
e
O
N
H
N
H
N
N
H
14
15
Scheme 1. Generalized synthesis of JNK1 inhibitors: (a) ethyl 1H-imidazole-2-carboxylate, Cs2CO3, DMF, 55 °C, 70%; (b) iron powder, EtOH:AcOH:H2O (2:2:1), sonication, 4 h,
62%; (c) POCl3, N,N-dimethylaniline, reflux, 80%; (d) trans-4-aminocyclohexanol, TEA, DMSO, 120C, 5 h, 95%.; (e) KOH/H2O, THF, 80 °C, 4 h, then HCl, 85%; (f) NH2R, HOBt, EDC,
DMF, rt, 4 h. 7-Substituted derivatives were synthesized in a similar manner from methyl 4-fluoro-3-nitro-benzoate.

B. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5217–5222
5219
Table 3
rhJNK1 IC50 values of 7- and 8-carboxy and carboxamide derivatives of compound 7
Table 4
rhJNK1 IC50 values of 8-carboxamide derivatives of compound 7
O
A
B
R8
R7
N
HO
N
N
N
HO
N
N
N
H
N
H
C
N
H
Compound
R7
R8
rhJNK1 IC50 (lM)
A
B
C
rhJNK1 IC50 (lM)
15a
16
H
-COOH
H
-CONH2
H
0.31
1.5
0.57
4.17
23a
19a
24
Me
H
Et
H
Me
H
Et
H
CH2OH
CH2CH2CH3
iPr
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
1.98
0.58
1.7
0.14
4.9
-COOH
H
-CONH2
17a
18
25a
26
H
O
CH2OH
H
H
H
H
H
H
Me
Et
27a
28a
29
30
31
0.16
0.2
19a
H
0.58
1.5
N
H
0.35
0.36
0.62
0.69
0.27
0.16
0.092
0.077
0.18
1.22
0.43
0.25
0.19
0.16
O
Ph
CH2CH2OH
CH2OCH3
Me
20
H
N
H
32
33a
34
Et
O
35a
36a
37
CH2CH2
CH2CH2CH2
21
H
0.68
N
H
CH2CH2CH2CH2
CH2CH2CH2CH2CH2
CH2CH2CH2CH2CH2CH2
CH2CH2
38
39
40
41
O
Me
Cl
F
22
H
0.49
N
H
CH2CH2
CH2CH2
AX13587
a
a
Profiled against a panel of human kinases (see Table 5).
Profiled against a panel of human kinases (see Table 5).
itory activity of various paired 7- and 8- substituted derivatives of
compound 7 were then compared (Table 3). In 3 out of 4 pairs, 8-
substitution yielded better rhJNK1 inhibitory activity than the 7-
substituted analog (compare 15 vs. 16, 17 vs. 18, 19 vs. 20). For
the remaining pair, both compounds 21 and 22 had similar rhJNK1
IC50 values. Finally, all 8-substituted derivatives in Table 3 were
more potent than compound 7.
tives thereof (25 and 27) were more potent than their enantiomers
(23, 24, and 26, respectively), by as much as 31-fold (compare 26
versus 27). Other derivatives of compound 19 (28–32) were inves-
tigated, but none found as potent as 25 (B = Et, rhJNK1 IC50 = 140 -
nM). Two achiral a,a,-gem-disubstituted benzylamides (33 and 34)
were examined and only one such compound (34, A = B = Et,
rhJNK1 IC50 = 160 nM) was of similar potency as 25. Using the
rhJNK1 assay, two spirocyclic compounds 35 (spirocyclopropyl,
rhJNK1 IC50 = 92 nM) and 36 (spirocyclobutyl, rhJNK1 IC50 = 77 nM)
were found to be the two most potent rhJNK1 inhibitors in this
study. Neither increasing the spiroalkyl ring size (37–39) nor deriv-
The above data prompted us to further investigate 8-benzyla-
mide derivatives of compound 7 (Table 4). When chiral
tuted benzylamides were examined (19, 23–27), it was found
that amides based on (R)- -methylbenzylamine (19) and deriva-
a-substi-
a
Figure 2. Crystal structure of AX13587 in the JNK1 active site (PDB ID: 4L7F).

5220
B. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5217–5222
highlighted in the JNK1 ATP binding site to illustrate several
favorable interactions with the imidazo[1,2-a]quinoxaline scaf-
fold, the trans-4-hydroxy-hexylamine group, and the 4-fluoro-
Glu73
Glu109
phenyl-cyclopropyl-amide
moiety:
dual
hydrogen
bond
F
Lys55
O
H
interactions with the hinge region (Met111), hydrogen bonding
with the solvent exposed Gln117, a bridged hydrogen bond be-
tween a water molecule and Asn156, and a hydrogen bond with
the catalytic Lys55. Lipophilic interactions are present between
the phenyl ring of the scaffold and the ribose-binding pocket as
well as in an induced-fit binding pocket formed by Arg69 and
Glu73 and occupied by the fluorophenyl group. This induced-fit
pocket is not found in any other JNK co-crystal structures and
is absent in the apo structures of JNK isoforms (e.g., PDB ID
1UKH, not shown). The positioning of the inhibitor within this
pocket also enables additional hydrophobic interactions between
the b-carbon of Asp169 and the cyclopropyl ring of the 8-amide
substituent. In total, these interactions explain aspects of the
structure–activity relationship of our compound series: the
importance of the trans-hydroxy group, the improvement of po-
tency when a C8 benzylamide was introduced, and the limited
steric tolerance for spirocyclic groups significantly larger than
the cyclopropyl group.
H
N
Arg69
O
Met111
N
N
N
H
N
Asp169
Asn156
H2O
O
H
Gln117
Scheme 2. Schematic drawing showing the key molecular interactions of the
AX13587-JNK1 co-crystal structure.
atization of 35 in the para-position (C = Me, Cl, F) resulted in an in-
crease in the rhJNK1 inhibitory activity.
To understand better the details of how this class of inhibitor
binds, a co-crystal structure of AX13587 in the JNK1 ATP-binding
site was obtained (Fig. 2 and Scheme 2). Selected residues are
Table 5
Kinase selectivity profiling (KiNativ™) of selected JNK1 inhibitors in PC3 lysate showing % inhibition of ATP probe labeling (AX9989) of the indicated kinase (leftmost column) at a
screening concentration of 10
l
M
15
17
23
19
25
27
28
33
35
36
JNK1,JNK2,JNK3
99.1
<35
97.5
<35
90.4
98.6
91.3
96.0
94.2
79.1
78.1
95.6
<35
<35
<35
69.0
<35
<35
44.4
<35
<35
<35
51.7
<35
<35
<35
<35
<35
<35
<35
<35
39.1
37.0
<35
<35
86.5
76.3
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
98.2
97.0
56.2
42.0
42.6
<35
40.4
55.3
51.0
<35
<35
<35
<35
<35
61.7
54.0
<35
<35
38.4
<35
<35
<35
45.3
43.2
<35
<35
36.9
<35
<35
<35
<35
38.7
<35
98.6
96.2
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
45.2
38.5
<35
<35
<35
<35
<35
<35
<35
<35
46.9
48.7
<35
<35
<35
<35
<35
<35
<35
98.5
97.9
46.6
<35
<35
<35
51.0
44.5
48.9
<35
<35
<35
<35
<35
59.4
56.4
<35
<35
<35
<35
<35
<35
40.1
44.7
<35
<35
45.6
<35
<35
<35
<35
<35
<35
95.3
90.2
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
98.3
99.3
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
36.6
<35
<35
<35
<35
<35
<35
<35
<35
<35
99.2
98.9
<35
<35
<35
<35
<35
38.9
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
35.9
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
99.2
99.2
59.1
49.8
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
98.7
97.6
<35
<35
44.3
<35
<35
<35
39.5
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
35.8
<35
<35
<35
<35
<35
<35
<35
<35
<35
99.5
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
<35
MAST3
CDK10
CDK9
CK1δ,CK1ε
GSK3β
PI4KB
PIP4K2C
PIP4K2C
PITSLRE
PFTAIRE1
PRPK
CDK2
CK1α
GCN2
GCN2
p38δ,p38γ
91.3
97.9
81.2
98.6
99.2
90.1
94.6
99.3
80.1
86.6
59.1
57.1
54.8
41.7
57.7
71.1
37.5
73.1
<35
67.0
<35
<35
<35
<35
<35
40.6
<35
48.9
<35
<35
PCTAIRE1
PEK
PI4KA,PI4KAP2
PKD2
TAO2
AMPKα1,AMPKα2
AMPKα1,AMPKα2
CaMK2α,CaMK2β,CaMK2δ,CaMK2γ
CaMK2γ
CHK2
CK2α2
LKB1
MPSK1
ZC1/HGK,ZC2/TNIK,ZC3/MINK
IKKε,TBK1
PIK3CB
37.1
Data is shown in the form of a heat map and compound structures can be found in Tables 3 and 4. Other kinases that gave <35% inhibition but are not shown in this table can
be found in Ref. 13. More than one probe labeling site exists for PIP4K2C, GCN2, and AMPK
1/AMPK 2, so these kinases appear twice on this table.
a
a

B. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5217–5222
5221
Kinase profiling of selected JNK1 inhibitors of different struc-
tural subtypes (from Tables 3 and 4, indicated by an asterisk)
was conducted at a screening concentration of 10
M (Table 5).13
Compounds with rhJNK1 IC50 values <600 nM highly inhibited
ATP probe labeling of the JNKs (i.e., >97%). In addition, when com-
pared 19 and 23 exhibited the same trend of enantioselective bind-
ing observed in the rhJNK1 assay. Interestingly, we found
derivatives lacking an 8-benzylamide group (8-COOH: 15 and 8-
CONH2: 17) to be highly promiscuous, inhibiting >7 kinases by
F
O
N
HO
N
N
l
N
H
n
N
H
AX13587
AX14373
(n = 0)
(n = 1)
Figure 3. Structures of AX13587 and AX14373.
>90% as well as many other kinases in the
10 M screening concentration, >90% inhibition is equivalent to
an IC50 value <1 M). Unexpectedly, chiral and achiral 8-benzla-
lM range (note: at a
l
Table 7
Native JNK1 and JNK2 IC50 values for selected compounds in this study
l
mides (23, 19, 25, 27, 28, 33, 35, and 36) carried MAST314 as a com-
mon off target. The spirobutyl derivative 36, identified as the most
potent rhJNK1 inhibitor, was also found to be a very potent MAST3
inhibitor. Using a more expansive interrogation of the kinome in
HuH-7 lysate, selected 8-benzylamide derivatives were examined
and found to be inhibitors of MAST 1/2 and MAST4 (see Table 6).
Considering the lack of off-targets other than MAST kinases, these
inhibitors are best described as dual JNK/MAST kinase inhibitors.
The ability to find inhibitors fortuitously for unexplored kinases
(e.g., MAST3) has been cited as one of the many benefits of kinase
profiling.15
Because MAST3 was a persistent off-target for the 8-benzyla-
mide series, we synthesized and explored 8-phenethylamides,
including AX14373, a methylene homolog of AX13587 (Fig. 3). Ki-
nase profiling found this simple modification dialed out all MAST
kinase inhibitory activity (Table 5 and data not shown). Further-
more AX14373 was found to be a highly JNK-selective inhibitor;
no significant off-targets were seen when profiled against over
200 kinases.
Compound
Native JNK1 IC50 (nM)
Native JNK2 IC50 (nM)
16
25
27
34
41
AX13587
AX14373
82
33
67
59
140
41
47
n.d.
100
135
>190
>190
n.d.
120
Jurkat lysate was used as a source of native JNK1 and JNK2. n.d. = not determined.
highly selective JNK inhibitor. Kinase profiling was used in all
phases of this effort from the analysis of hits, understanding the
SAR of unintended and unwanted off targets, and finally the confir-
mation of selectivity for the intended target. Future work describ-
ing the optimization and characterization of the 8-phenethylamide
series will be published in due course.
References and notes
Because our mass spectrometry-based kinase profiling platform
cannot deconvolute % inhibition values for individual JNK isoforms,
an alternate assay was developed. To determine native JNK1 and
JNK2 IC50 values the immunoprecipitation of JNK1 and JNK2 which
had been desthiobiotin-tagged with the ATP probe AX950916 was
used. Labeling was conducted in the presence and absence of
inhibitors followed by immunoprecipitation with avidin. Probe-la-
beled proteins were eluted, resolved by 1D-SDS PAGE, and trans-
ferred to nitrocellulose. Western blot quantification with JNK1 or
JNK2 specific antibodies could then be used to calculate % inhibi-
tion of desthiobiotin labeling and subsequently JNK1 or JNK2 IC50
values. This analysis was conducted for selected compounds in this
study (see Table 7). Interestingly, in all cases, the native JNK1 IC50
value was lower than that measured with the recombinant form. A
result that suggests differences in potency may arise when assays
do not take place in the full complement of lysate.
1. Gupta, S.; Barrett, T.; Whitmarsh, A. J.; Julie Cavanagh, J.; Sluss, H. K.; Dérijard,
B.; Davis, R. J EMBO J. 1996, 15, 2760.
2. Martin, J. H.; Mohit, A. A.; Miller, C. A. Brain Res. Mol. Brain Res. 1996, 35, 47.
3. (a) Leppa, S.; Bohmann, D. Oncogene 1999, 18, 6158; (b) Weston, C. R.; Davis, R.
J. Curr. Opin. Genet. Dev. 2002, 12, 14.
4. (a) Manning, A. M.; Davis, R. J. Nat. Rev. Drug Disc. 2003, 2, 554; (b) Siddiqui, M.
A.; Reddy, P. A. J. Med. Chem. 2010, 53, 3005.
5. Plantevin Krenitsky, V.; Delgado, M.; Nadolny, L.; Sahasrabudhe, K.; Ayala, L.;
Clareen, S. S.; Hilgraf, R.; Albers, R.; Kois, A.; Hughes, K.; Wright, J.;
Nowakowski, J.; Sudbeck, E.; Ghosh, S.; Bahmanyar, S.; Chamberlain, P.; Muir,
J.; Cathers, B. E.; Giegel, D.; Xu, L.; Celeridad, M.; Moghaddam, M.; Khatsenko,
O.; Omholt, P.; Katz, J.; Pai, S.; Fan, R.; Tang, Y.; Shirley, M. A.; Benish, B.; Blease,
K.; Raymon, H.; Bhagwat, S.; Henderson, I.; Cole, A. G.; Bennett, B.; Satoh, Y.
Bioorg. Med. Chem. Lett. 2012, 22, 1427.
6. Plantevin Krenitsky, V.; Nadolny, L.; Delgado, M.; Ayala, L.; Clareen, S. S.;
Hilgraf, R.; Albers, R.; Hegde, S.; D’Sidocky, N.; Sapienza, J.; Wright, J.;
McCarrick, M.; Bahmanyar, S.; Chamberlain, P.; Delker, S. L.; Muir, J.; Giegel,
D.; Xu, L.; Celeridad, M.; Lachowitzer, J.; Bennett, B.; Moghaddam, M.;
Khatsenko, O.; Katz, J.; Fan, R.; Bai, A.; Tang, Y.; Shirley, M. A.; Benish, B.;
Bodine, T.; Blease, K.; Raymon, H.; Cathers, B. E.; Satoh, Y. Bioorg. Med. Chem.
Lett. 2012, 22, 1433.
In summary, lead optimization of the hit 1 resulted in both the
development of novel JNK/MAST kinase inhibitors and AX14373, a
7. rhJNK1 IC50 assay: Recombinant histidine-tagged, full-length human, JNK1
(Invitrogen) expressed in insect cells and activated by in vitro phosphorylation
by GST-tagged MAP2K7 was used in the assay. Briefly, in a volume of 30 lL, 70–
90 ng of JNK1 enzyme was incubated in the presence or absence of compound
at varying concentrations for 1 h at 30 °C in 50 mM Tris, pH 7.5, 10 mM
Table 6
MAST family off target activity for selected JNK inhibitors at a screening concentration
of 10 lM in HuH-7 lysate
magnesium acetate, 1 mM EGTA, 0.1
96, Millipore), 5 M ATP, and 1 mM dithiothreitol. Following the completion of
the kinase reaction an equal volume of KinaseGlo or KinaseGlo Plus luciferase
reagent (Promega) was added and the luminescence was read using
luminescence plate reader within 5–10 min. Compound activity was
expressed as inhibition relative to maximal inhibition observed at the
lG/lL human recombinant ATF2 (aa19-
l
a
MAST3
IC50 (μM)
MAST 1/2
MAST3
MAST4
compound
% inhibition % inhibition % inhibition
%
19
25
n.d.
n.d.
75.6
36.9
>86
76
90.8
87.5
97.2
80.6
97.7
93.1
76.1
79.8
>88
75.6
>88
78.7
1.01
1.4
maximal dose. The luminescence readouts from 3 replicates per concentration
were averaged to produce a single curve. IC50 values were calculated from this
curve using curve fitting software (GraphPad Prism).
8. Structure of the desthiobiotin-ATP probe AX9989
H
27
0.29
2.4
N
28
O
O
P
- O
O
35
0.24
0.74
N
H
O
O
N
O
NH2
AX13587
N
3
N
% Inhibition values are shown in the form of a heat map and compound structures
can be found in Table 4. Also shown also are MAST3 IC50 values calculated from the
corresponding MAST3 % inhibition value. n.d. = not determined.
N
HO
OH
.

5222
B. Li et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5217–5222
9. (a) Patricelli, M. P.; Szardenings, A. K.; Liyanage, M.; Nomanbhoy, T. K.; Wu, M.;
Weissig, H.; Aban, A.; Chun, D.; Tanner, S.; Kozarich, J. W. Biochemistry 2007, 46,
350; (b) Patricelli, M. P.; Nomanbhoy, T. K.; Wu, J.; Brown, H.; Zhou, D.; Zhang,
J.; Jagannathan, S.; Aban, A.; Okerberg, E.; Herring, C.; Nordin, B.; Weissig, H.;
Yang, Q.; Lee, J-D.; Gray, N. S.; Kozarich, J. W. Chem. Biol. 2011, 18, 699.
10.. Deleuze-Masquefa, C.; Moarbess, G.; Khier, S.; David, N.; Gayraud-Paniagua, S.;
Bressolle, F.; Pinguet, F.; Bonnet, P-A. Eur. J. Med. Chem. 2009, 44, 3406.
11.. Guillon, J.; Moreau, S.; Mouray, E.; Sinou, V.; Forfar, I.; Fabre, S. B.; Desplat, V.;
Millet, P.; Parzy, D.; Jarry, C.; Grellier, P. Bioorg. Med. Chem. 2008, 16, 9133.
12.. Heine, H. W.; Brooker, A. C. J. Org. Chem. 1962, 27, 2943.
13. Other kinases that gave <35% inhibition but are not shown in Table 5: AGK,
ANPb, ATR, CASK, CDK11, CDK8, CDK5, CHK2, CLK2, CSK, DNAPK, eEF2K, EGFR,
EphA1, EphA2, EphA7, EphB2, EphB4, Erk1, Erk2, FAK, FER, FYN, SRC, YES, ILK,
IRAK1, IRAK4, KHS1, KHS2, LOK, MAP2K1, MAP2K2, MAP2K3, MAP2K4,
MAP2K6, MAP3K2, MAP3K3, MAP3K4, MARK2, MARK3, MARK4, MET, MLK3,
RON, RSK2 domain2, SLK, SNRK, SRC, SRPK1, SRPK2, TAO1, TAO3, TLK1, TLK2,
Wnk1, Wnk2, Wnk4, ZAK.
14. (a) Garland, P.; Quraishe, S.; French, P.; O’Connor, V. Brain Res. 2008, 1195, 12;
(b) Sun, L.; Gu, S.; Li, X.; Sun, Y.; Zheng, D.; Yu, K.; Ji, C.; Tang, R.; Xie, Y.; Mao, Y.
Mol. Biol. (Mosk) 2006, 40, 808.
15. Miduturu, C. V.; Deng, X.; Kwiatkowski, N.; Yang, W.; Brault, L.;
Filippakopoulos, P.; Chung, E.; Yang, Q.; Schwaller, J.; Knapp, S.; King, R. W.;
Lee, J. D.; Herrgard, S.; Zarrinkar, P.; Gray, N. S. Chem. Biol. 2011, 18, 868.
16. Structure of the desthiobiotin-ATP probe AX9509
H
N
O
O
P
O
O
O
N
N
H
N
H
O
NH2
- O
N
O
3
N
N
HO
OH
MRCKb, MST1, MST2, MST4, YSK1, NDR2, NEK6, NEK7, NEK8, NEK9, p38
a,
.
p38b, p70S6K, p70S6Kb, PDK1, PHK 2, PKC , PKCb, PKR, PRP4, ROCK1, ROCK2,
c
a