Ke Zheng1, Sarah Iqbal, Pamela Hernandez, HaJeung Park, Philip V LoGrasso, Yangbo Feng. 1. Medicinal Chemistry, ‡Discovery Biology, §Crystallography/Modeling Facility, Translational Research Institute, and ∥Department of Molecular Therapeutics, Scripps Florida, The Scripps Research Institute , 130 Scripps Way, No. 2A1, Jupiter, Florida 33458, United States.
Abstract
The c-jun N-terminal kinase 3 (JNK3) is expressed primarily in the brain. Numerous reports have shown that inhibition of JNK3 is a promising strategy for treatment of neurodegeneration. The optimization of aminopyrazole-based JNK3 inhibitors with improved potency, isoform selectivity, and pharmacological properties by structure-activity relationship (SAR) studies utilizing biochemical and cell-based assays, and structure-based drug design is reported. These inhibitors had high selectivity over JNK1 and p38α, minimal cytotoxicity, potent inhibition of 6-OHDA-induced mitochondrial membrane potential dissipation and ROS generation, and good drug metabolism and pharmacokinetic (DMPK) properties for iv dosing. 26n was profiled against 464 kinases and was found to be highly selective hitting only seven kinases with >80% inhibition at 10 μM. Moreover, 26n showed good solubility, good brain penetration, and good DMPK properties. Finally, the crystal structure of 26k in complex with JNK3 was solved at 1.8 Å to explore the binding mode of aminopyrazole based JNK3 inhibitors.
The c-jun N-terminal kinase 3 (JNK3) is expressed primarily in the brain. Numerous reports have shown that inhibition of JNK3 is a promising strategy for treatment of neurodegeneration. The optimization of aminopyrazole-based JNK3 inhibitors with improved potency, isoform selectivity, and pharmacological properties by structure-activity relationship (SAR) studies utilizing biochemical and cell-based assays, and structure-based drug design is reported. These inhibitors had high selectivity over JNK1 and p38α, minimal cytotoxicity, potent inhibition of 6-OHDA-induced mitochondrial membrane potential dissipation and ROS generation, and good drug metabolism and pharmacokinetic (DMPK) properties for iv dosing. 26n was profiled against 464 kinases and was found to be highly selective hitting only seven kinases with >80% inhibition at 10 μM. Moreover, 26n showed good solubility, good brain penetration, and good DMPK properties. Finally, the crystal structure of 26k in complex with JNK3 was solved at 1.8 Å to explore the binding mode of aminopyrazole based JNK3 inhibitors.
As a member of the
mitogen-activated protein kinase (MAPK) family,
the c-Jun N-terminal kinases (JNKs) regulate the serine/threonine
phosphorylation of several transcription factors[1] when they are activated in response to various stimuli
such as oxidative stress, neurotoxins, cytokines, and fatty acids.[2−8] There are three humanJNK isoforms: JNK1, JNK2, and JNK3.[9−12] JNK1 and JNK2 are ubiquitously expressed in most tissues, while
JNK3 is primarily expressed in the brain and, to a lesser extent,
in the heart and testes.[1,10−17] Recent studies have shown that JNK3 plays a central role in the
brain to mediate neurodegeneration, such as β-amyloid processing
and neuronal apoptosis in Alzheimer’s disease,[18] as well as the mediation of neurotoxicity in rodent models
of Parkinson’s disease.[3,19−22] The selective expression of JNK3 in the brain, along with findings
that JNK3 knockout mice exhibit amelioration of neurodegeneration
in animal models of Parkinson and Alzheimer’s disease, makes
inhibiting this isoform a particularly promising therapeutic target
for neurodegenerative diseases.[18,23,24]Identifying potent and selective inhibitors of JNK3 may contribute
toward neuroprotective therapies with reduced untoward side effect
profiles if JNK1 inhibition has such negative effects. To date, however,
the development of specific small molecule inhibitors with high isoform
selectivity for JNK3, especially against JNK1, is still a relatively
untapped area as the three JNK isoforms share more than 90% sequence
identity in the ATP pocket. Many published JNK3 inhibitors are also
potent for JNK1, JNK2, and some inhibit p38α as well, because
of their high degree of amino acid sequence similarity, which might
lead to potential side effect profiles on immune and inflammatory
systems.[25,26] Thus, developing isoform-specificJNK3 inhibitors
as therapeutics has gained considerable interest over the past few
years despite most reports being centered on pan-JNK inhibitors.[25,27−45] Our previous effort toward the development of isoform selective
JNK3 inhibitors led to the identification of a class of aminopyrazolecompounds. These aminopyrazole-based JNK3 inhibitors not only had
high selectivity against p38α but also showed noticeable isoform
selectivity against JNK1 (compound SR-4326, 18.5-fold, Figure 1).[36] Herein, we describe
extensive SAR studies in a continuing effort to develop highly isoform
selective, efficacious, and pharmacologically viable JNK3 inhibitors
from this novel aminopyrazole scaffold. The focused medicinal chemistry
efforts led to the discovery of several highly potent and isoform
selective JNK3 inhibitors with an isoform selectivity of >50-fold
over JNK1. In addition, these JNK3 inhibitors generally had high selectivity
against the closely related protein kinase p38α, were potent
in protecting against ROS generation and mitochondrial dysfunction,
and were optimized to have good DMPK properties for topical use and/or
in iv dosing.
Figure 1
A previously disclosed
isoform selective JNK3 inhibitor SR-4326.
Chemistry
Several short routes were
used to synthesize inhibitors 6–8 which possessed different substituents on
the urea moiety (Scheme 1). Ullman coupling
of 4-nitro-1H-pyrazole 1 with 3-bromobenzoicacid 2 yielded acid 3. Amide formation to 3 gave intermediate 4, followed by a Pd/C mediated
hydrogenation to afford amine 5. This intermediate was
directly subjected to amide formation to furnish inhibitor 6. Inhibitor 7 was formed via a one-step coupling reaction
from amine 5 under microwave reaction conditions. Inhibitors 8a–l were synthesized by treating amine 5 with different isocyanates or with triphosgene and a secondary
amine, respectively.
Scheme 1
Synthesis of Inhibitors 6–8
Reagents and conditions: (a)
CuI, trans-N,N-dimethylcyclohexane-1,2-diamine,
Cs2CO3, DMF, 90 °C; (b) 6-methylpyridin-3-amine,
EDC, HOBt, DIEA, CH2Cl2, 25 °C; (c) Pd/C,
MeOH, H2; (d) K2CO3, THF or EDC,
HOBt, DIEA, CH2Cl2, 25 °C; (e) 2-chloro-1H-benzo[d]imidazole, EtOH, HCl, microwave,
140 °C; (f) isocyanate, CH2Cl2, 25 °C,
or secondary amine; triphosgene, 50 °C, or CDI, THF, microwave,
150 °C.
Synthesis of Inhibitors 6–8
Reagents and conditions: (a)
CuI, trans-N,N-dimethylcyclohexane-1,2-diamine,
Cs2CO3, DMF, 90 °C; (b) 6-methylpyridin-3-amine,
EDC, HOBt, DIEA, CH2Cl2, 25 °C; (c) Pd/C,
MeOH, H2; (d) K2CO3, THF or EDC,
HOBt, DIEA, CH2Cl2, 25 °C; (e) 2-chloro-1H-benzo[d]imidazole, EtOH, HCl, microwave,
140 °C; (f) isocyanate, CH2Cl2, 25 °C,
or secondary amine; triphosgene, 50 °C, or CDI, THF, microwave,
150 °C.The preparation of inhibitor 12 is described in Scheme 2. Treatment
of ethyl 1H-pyrazole-4-carboxylate 9 with 3-bromobenzoic acid 2 followed by amide
formation gave amide 10. Amide 10 was hydrolyzed
under basicconditions to yield acid 11, which was then
subjected to an amide formation reaction using EDC as the coupling
reagent to furnish the final product 12.
Reagents and conditions:
(a)
CuI, trans-N,N-dimethylcyclohexane-1,2-diamine,
Cs2CO3, DMF, 90 °C; (b) 6-methylpyridin-3-amine,
EDC, HOBt, DIEA, CH2Cl2, 25 °C; (c) LiOH,
THF/H2O (1:1), 100 °C; (d) 2-chloroaniline, EDC, HOBt,
DIEA, CH2Cl2, 25 °C.Inhibitors 16a–d, all of which
have a fluoro substitution on the middle phenyl ring, were synthesized
following procedures shown in Scheme 3. The
commercially available 4-nitro-1H-pyrazole 1 was N-arylated with ethyl 3-bromobenzoate 13 using standard Ullman procedures, followed by ester hydrolysis under
basicconditions to form acid 14. Amide formation followed
by Pd/C mediated reduction of 14 was applied to yield
amine 15. Finally, intermediate 15 was subjected
to a urea formation to furnish product 16.
Reagents and conditions:
(a)
CuI, trans-N,N-dimethylcyclohexane-1,2-diamine,
Cs2CO3, DMF, 90 °C; (b) LiOH, THF/H2O (1:1), 100 °C; (c) 6-methylpyridin-3-amine, EDC, HOBt,
DIEA, CH2Cl2, 25 °C; (d) Pd/C, MeOH, H2; (e) 1-chloro-2-isocyanatobenzene, CH2Cl2, 25 °C.The preparation of inhibitors 20–22 is outlined in Scheme 4. Ullman coupling
of 4-nitro-1H-pyrazole 1 with 1-bromo-3-iodobenzene
(for synthesis of compounds 20) or 3-bromobenzonitrile
(for synthesis of compounds 21) or 3-bromobenzoic acid
(for synthesis of compounds 22), followed by a Pd/C mediated
reduction, afforded the intermediate amine 18. These
amines were reacted with 1-chloro-4-isocyanobenzene in DCM to furnish
intermediates 19a−c. Target compounds 20 were prepared through a Suzuki coupling reaction between
intermediate 19a and a boronic acid pinacol estercatalyzed
by Pd(PPh3)4. Target compounds 21a and 21b were prepared through treatment of intermediate 19b with hydrazine monohydrate, followed by a cyclization
reaction with acetimidamide or acetic acid. Target compounds 22 were prepared through amide formation in the presence of
HATU in DMF. All final inhibitors were purified by reverse-phase preparative
HPLC.
Reagents and conditions: (a)
CuI, trans-N,N-dimethylcyclohexane-1,2-diamine,
Cs2CO3, DMF, 100 °C; (b) Pd/C, MeOH, H2; (c) 1-chloro-2-isocyanatobenzene, CH2Cl2, 25 °C; (d) R1B(OH)2, Ph(PPh3)4, K2CO3, dioxane/H2O (3:1), 95 °C; (e) N2H4·xH2O, EtOH, reflux ; (f) acetimidamide, EtOH,
NaOMe or acetic acid, propylphosphonic anhydride, TEA, DMF, 110 °C;
(g) R3NH2, EDC, HOBt, DIEA, CH2Cl2, 25 °C.Applications of N-substituted
pyrazole amines led to inhibitors 26. As shown in Scheme 5, the N1 alkylation
of 3-nitropyrazole with a primary iodo- or bromoalkane using potassium
carbonate as the base was first used to furnish N-substituted pyrazoles,
followed by reduction of the nitro group to give pyrazole amines 24. Amide formation between 24 and acid 3, followed by a Pd/C mediated reduction furnished amine 25. A one-pot urea formation of 25 with 1-chloro-4-isocyanobenzene
in DCM afforded the target inhibitor 26 which was purified
by reverse-phase preparative HPLC.
In the crystal structures published in our previous work, the aniline–urea
moiety (Figure 1) binds to the hydrophobic
selectivity pocket (hydrophobic pocket I) and is responsible for the
isoform selectivity of JNK3 over JNK1.[36] We therefore started our SAR investigations by replacing the aniline–urea
moiety in region A (Figure 1) with selected
bioisosters. As shown in Figure 2, very few
replacements to the urea moiety were tolerated at region A, as all
compounds hadJNK3 IC50 values of >10 μM. For
example,
replacement of the aniline NH group with a carbon (6a) or oxygen (6b) or a direct heterocyclic ring connection
(6c,d) all resulted in complete loss of
JNK3 inhibition. Similarly, structural changes to the other NH group
of urea were not tolerated either. For example, amidecompound 12a and pyrazolo[4,3-c]pyridinecompound 12b were completely inactive. Some other replacements of the
urea moiety were also investigated (such as compound 7; for more data see Supporting Information), but no improvement was observed. The observed results for modifications
to the urea moiety suggested that the ureacore was essential for
achieving a high JNK3 affinity. Therefore, this urea moiety will be
retained in all further optimizations.
Figure 2
Compounds with low JNK3 inhibition activity with IC50 >
10 μM.
A previously disclosed
isoform selective JNK3 inhibitor SR-4326.Compounds with low JNK3 inhibition activity with IC50 >
10 μM.We next investigated
a series of replacements and substitutions
on the phenyl group of the urea moiety. As shown in Table 1, replacing the aromatic group with an alkyl group
was not tolerated, as indicated by the significant loss of JNK inhibition
activity for 8a. Compound 8b, which had
no substitutions on the urea phenyl ring, showed a JNK3 inhibition
activity similar to the leadcompound SR-4326 but with almost no isoform
selectivity. In the case of chloro substitutions, both the meta- and
the para-Cl substitutions (8d and 8e) led
to a decrease in JNK affinity, which could be due to space limitation
inside the binding pocket. However, when a Cl substituent was introduced
at the ortho-position of the phenyl ring (8c), where
there should be enough space to accommodate a Cl group based on the
crystal structures,[36] higher JNK3 potency
was obtained probably because of the extra hydrophobic interactions
from the Cl group. More interestingly, some isoform selectivity (JNK3
over JNK1) was achieved as well (4.4-fold). On the other hand, fluoro
substitutions on this phenyl ring exhibited a significantly different
SAR profile. Compared to the leadcompound SR-4326 and compound 8c, a mono-F-substitution at all three positions resulted
in reduction of JNK3 inhibition activity, although these substitutions
could improve the isoform selectivity (compounds 8f, 8g, 8h). While the meta-F compound 8g and the para-F compound 8h showed a slight decrease
of JNK3 potency, a substitution at the ortho-position resulted in
around 3-fold loss of potency as compared to the unsubstituted analog 8b (Table 1).
Table 1
Biochemical
IC50 Values
for JNK3 and JNK1 for SAR Studies of the Urea Moietya
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
No inhibition up to 10 μM.
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.No inhibition up to 10 μM.The unusual SAR observed for F-substituted compounds
in Table 1 might be due to the F-bonding/interactions.[46,47] Different types of interactions between fluorine atoms with protein
functional groups have been reported (not found in other halogens),
and in some protein environments, compounds with a fluorine substitution
behave very differently from other halogen-substituted compounds.[47] Such fluorine interactions include F–H
hydrogen bonds, π-interactions with aromatic or guanidine groups,
multipolar interactions, and/or F-nitrogen interactions (halogen bonding).[47] For the aminopyrazole based JNK inhibitors,
these special F-interactions could result in a change of inhibitor
conformations in the pocket or a twist of the orientation of the phenyl
ring inside hydrophobic pocket I. These changes in conformation or
orientation could produce unfavorable binding of the inhibitor in
the ATP pocket, and therefore cause a decrease in inhibition of JNK
activity for compounds 8f–h.Significant loss of potency also occurred with replacement of the o-Cl-phenyl moiety by an o-Cl-benzyl group
(8i, Table 1). This result indicated
that the hydrophobic selectivity pocket might not have enough room
to tolerate a larger group or that the phenyl ring of the benzyl group
could not assume a conformation/orientation that could give maximal
hydrophobic interactions. The effect of substitutions on the phenyl
urea NH was also studied. Unfortunately, any substitution to this
NH group greatly reduced the JNK3 potency (Table 1). For example, a simple methyl substitution resulted in compound 8j which exhibited a JNK3 IC50 of 3063 nM. Larger
substituents yielded even lower JNK3 inhibitory activity (compounds 8k and 8l, IC50 > 10 μM).
As
is demonstrated in the crystal structure (see Figure 4), this urea NH is involved in H-bonding interactions with
protein residues, and its alkylation is predicted to reduce the JNK3
inhibitory activity.
Figure 4
X-ray crystal structure of 26k in JNK3.
The feasibility of applying some minor
structural modifications
to the central phenyl group attached to the pyrazole N1 (region B,
Figure 1) was explored next. Our preliminary
SAR showed that a methyl substitution was tolerated here.[36] We envisioned that introduction of a fluoro
substitution on the benzene ring might introduce additional interactions
that could perturb the JNK3 activity and selectivity, as well as improve
DMPK properties. Therefore, compounds with a simple fluoro substitution
at different positions of this benzene ring were investigated (Table 2). Although improvement in isoform selectivity was
achieved (compared to SR-4326 and 8c) for the 2-fluoro
substitution (16a) and the 6-fluoro substitution (16d), loss of JNK3 inhibition potency was observed for all
fluoro-substituted compounds 16a–d, especially for the 3-fluoro-derivative 16b, which
showed an IC50 value against JNK3 in the micromolar range
(4588 nM). The reduction in JNK inhibitory potency could be due to
the F-bondings/interactions[46,47] previously mentioned
and as observed in compounds 8f–h (Table 1). In addition, the 2-F analog 16a exhibited potent inhibition of cytochrome P450 isoforms
CYP2C9 and CYP2D6 (see Table 5). Because of
these detrimental features, F-substitutions on this benzene ring will
not be applied in future optimizations.
Table 2
Biochemical
IC50 Values
for JNK3 and JNK1 for SAR Studies for the Middle Phenyl Moietya
cmpd
R
JNK3 IC50 (nM)
JNK1 IC50 (nM)
JNK1/JNK3
16a
2-F
80
2369
29.6
16b
3-F
4588
N/Ib
−
16c
4-F
71
180
2.5
16d
6-F
230
3691
16.1
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
No inhibition up to 10 μM.
Table 5
Inhibitor
Selectivity and in Vitro
DMPK Dataa
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
Not determined.
No inhibition up to 10 μM.
With fluoro-substitution on the
middle phenyl ring at Region B.
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.No inhibition up to 10 μM.The crystal structure published in our previous work
indicated
that these compounds were type I kinase inhibitors and showed that
the amidecarbonyl and the aromatic ring (in the amide moiety) did
not have any significant interactions with the protein (only an H-bond
interaction was observed between the amide NH and the protein backbone).[36] Thus, structural modifications in this part
will likely not affect the binding mode significantly. To test this
observation, a series of replacements for the amide moiety in region
C (Figure 1) with some amide bioisosters were
investigated. As shown in Figure 3, the pyridine
and pyrimidine based structures (20a–b), which have heteroatoms on the aromatic ring to form H-bonds, showed
no JNK inhibitory activity. The triazole (21a) and the
oxodiazole (21b) based amide isosters, which mimic the
binding of the anilinoamide in region C, also gave no inhibition for
JNK3 and JNK1. Moreover, alkylation of the amide NH (22a) led to complete loss of JNK3 inhibitory activity. These results
indicated that the amide structure was critical for efficient binding
and the amide moiety −CONH– was indispensable for this
scaffold to achieve a high JNK3 affinity.
Figure 3
Compounds with low JNK3
inhibition activity with IC50 > 10 μM.
Compounds with low JNK3
inhibition activity with IC50 > 10 μM.SAR studies of the amide moiety are a very important
part of our
efforts to improve both JNK3 inhibitory potency and isoform selectivity
(against JNK1) for this aminopyrazole-based scaffold. The effect of
substitution patterns on the pyridine ring for leadcompound 8c was first explored. As shown in Table 3, moving the methyl group from 4′-position to 2′-position
reduced JNK3 inhibitory potency by 3.7-fold but it reduced the JNK1
potency as much as 15.8-fold providing a benefit for the overall JNK3/JNK1
isoform selectivity (22b vs 8c). The dimethylpyridinecompound 22c showed a JNK3 potency similar to that of
the monosubstituted analogue 22b but a reduced isoform
selectivity. Nevertheless, leadcompound 8c still had
the best JNK3 potency (38 nM). While the 2′-pyridine analogs 22d and 22e exhibited a significant loss of JNK3
inhibition potency, the 4′-pyridinecompound, 22f, presented just a slightly lower JNK3 potency (98 nM vs 38 nM) but
a better isoform selectivity (28-fold vs 4.4-fold, 22f vs 8c). The dimethyl-4′-pyridinecompound 22g exhibited an even higher isoform selectivity (35-fold),
although with a sacrifice in JNK3 potency. Interestingly, the pyrimidineamide22h showed a similar JNK3 potency but had isoform
selectivity slightly better than compound 8c. It is also
important to note that the benzyl type amide 22i gave
a weaker JNK3 inhibition but improved isoform selectivity as compared
to the corresponding aniline type amide 8c and SR-4326.
In summary, for pyridine and/or pyrimidine based amides, 3′-N
derivatives gave better JNK3 potency but not satisfactory isoform
selectivity (8c and 22h). On the other hand,
4′-N analogs exhibited good isoform selectivity but lower JNK3
inhibition activity (22f and 22g). Nevertheless,
no compounds from this class could present both good JNK3 inhibition
potency (IC50 < 50 nM) and excellent isoform selectivity
(>50-fold).
Table 3
Biochemical IC50 Values
for JNK3 and JNK1 for SAR Studies for the Amide Moietya
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
No inhibition up to 10 μM.
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.No inhibition up to 10 μM.Although the benzamide moiety (region C, Figure 1) extends away from hydrophobic pocket II and is
exposed to
solvents, results in Table 3 indicated that
structures in region Ccan still significantly affect both JNK3 inhibitory
activity and isoform selectivity. Considering that pyridine/pyrimidineamidecompounds might also have cytotoxicity and poor microsomal stabilities
(see Table 5 and Table 6), we expanded SAR studies for this part to other amide structures.
Thus, amides based on five-membered heterocyclic ring systems were
investigated. As shown in Table 4, the N-methylpyrazolecompounds 26a–d showed a similar potency and isoform selectivity compared
to leadcompound SR-4326. Application of an N-methyl-triazole
(26e) resulted in a 5-fold reduction in JNK3 inhibitory
activity. Compounds having larger substituents on the NH of pyrazole
groups gave better properties. For example, an N-isopropyl
substitution on the pyrazole ring yielded a JNK3 inhibitor (26f) having better isoform selectivity compared to the N-methyl analog (26a).
Table 6
Cytotoxicity and Cell Based Potency
Data for Selected Compounds
cytotoxicity in SHSY5Y cells after 48 h
cmpd
% cell viability at 10 μM
% cell viability at 30 μM
in-cell Western,a SHSY5Y IC50 (nM)
inhibition
of 6-OHDA induced cell death, SHSY5Y IC50 (nM)
inhibition
of 6-OHDA induced mitochondrial membrane
depolarization,SHSY5Y IC50 (nM)
22f
7
5
866
2976
40
22g
24
5
2331
ndb
ndb
22i
80
56
905
13
17
26f
84
16
3250
ndb
130
26j
91
98
1436
568
25
26g
105
97
N/Ic
N/Ic
N/Ic
26k
112
97
N/Ic
N/Ic
31
26n
92
98
1895
281
4
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
Not determined.
No inhibition up to 10 μM.
Table 4
Biochemical IC50 Values
for JNK3 and JNK1 for SAR Studies for the Amide Moietya
IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.
No inhibition at 10 μM.
Remarkably, those
analogs with a substitution containing a dimethylamino,
diethylamino, piperidine, or pyrrolidine group (26g–k) exhibited high JNK3 potency (IC50 of 21–34
nM) and good to excellent isoform selectivity (22- to 60-fold over
JNK1). One compound, 26k, which had a substituent containing
a pyrrolidine moiety, stood out and showed the highest JNK3 potency
(IC50 < 1 nM) and isoform selectivity (>500-fold
over
JNK1 and >210-fold over JNK2; see Table 5). Careful inspection
of the crystal structure of 26k indicated the potential
for a hydrogen bond between the
carbonyl of N89 in JNK3 with the pyrrolidinyl NH of 26k with an H-bond distance of 2.9 Å (Figure 4). This potential for
tight interaction with optimal hydrogen bond distance and optimal
H-bond angle was likely the reason for the potent inhibition of 26k (the energy minimized state of a computational model showed
the N-pyrrolidinyl of 26k could rotate
∼90° to make a H-bond with N89). In contrast, 26j, which contained the six-member pipyridinyl ring and was not optimized
for H-bond distance and angle, had weaker inhibition to JNK3 than 26k. This observation is likely true for all of the other
analogs presented in Table 4. Excellent JNK3
isoform selective inhibitors were also obtained from amidescontaining
no aromatic groups (alkylamides, 26l–q). As shown in Table 4, all the inhibitors
prepared gave excellent isoform selectivity and fair to excellent
JNK3 inhibitory activity. The much better JNK3 potency and the higher
isoform selectivity (JNK3 over JNK1) for these compounds containing
a side chain amino group could be due to extra H-bond interactions
from the amino group, and a combination of these interactions and
the hydrophobic interaction inside hydrophobic pocket I might be more
favorable for JNK3 than for JNK1. These interactions are indicated
in the crystal structure of 26k with JNK3 (Figure 4).IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.No inhibition at 10 μM.To better understand the observed SAR for these JNK3
isoform selective
inhibitors, the crystal structure for 26k with JNK3 was
solved at 1.8 Å (Figure 4). This crystal
structure showed that 26k (the R-enantiomer)
interacted with JNK3 through H-bonding with the hinge residue M149
with a bond distance of 2.7–3.2 Å. The hydrogen bond between
the benzamide NH of 26k and the main-chain atoms of M149
in JNK3 would explain why substitution on the amide NH (22a) resulted in no JNK3 affinity (shown as yellow dotted lines). An
extra H-bond also occurred between the urea NH (on the aniline side)
and the side chain of residue K93 bridged by a water molecule. The
existence of this interaction explains why N-alkylation of the urea
NH led to loss of JNK3 inhibitory activity (compounds 8j, 8k, and 8l, Table 1). As mentioned, the N-pyrrolidinylpyrazole group
made optimal H-bond interactions based on length and angle with residue
N89, and the rest of the ring structure of 26k made tight
hydrophobic interactions with residues in ATP binding pocket. The
center phenyl ring made hydrophobic interactions with side chain of
I70 and backbone of A151. The central pyrazole ring also made hydrophobic
interactions with the side chains of A91 and V196.The most
significant hydrophobic interactions occurred at the terminal o-Cl-phenyl ring (attached to the urea moiety) and were
tightly sandwiched between the side chains of K93 and M146. Nearby
hydrophobic residues V78, M115, L126, L144, and L206 construct a hydrophobic
pocket for the o-Cl-phenyl ring. These structural
observations are consistent with the observed SAR results indicating
that there was almost no JNK3 affinity if the urea moiety was replaced
or the urea NH was alkylated. This structure also explains why there
was a reduction of JNK3 inhibitory activity when the phenyl ring was
substituted with a Cl on its meta- or para-position (8d and 8e), since there was not enough space inside this
hydrophobic pocket. The isoform selectivity of JNK3 over JNK1 for
aminopyrazole-based JNK3 inhibitors was also mainly due to interactions
in this pocket, which has been documented in our previous publications.[36]X-ray crystal structure of 26k in JNK3.To evaluate the selectivity of
these aminopyrazole based JNK3 inhibitors,
several selected compounds were subjected to additional counter screening
against JNK2 and p38. As shown in Table 5,
almost all aminopyrazolecompounds had no inhibition on the most closely
related kinase p38 at a concentration of 10 μM with the exception
of the N-alkyl substituted pyrazole amide 26f (3623 nM for p38). Moreover, these compounds had inhibitory activity
for JNK2 with potency similar to that for JNK3 with only a few exceptions
(22f, 22g, 22i, 26k, and 26l). To further evaluate the general kinase selectivity
of these aminopyrazole based JNK3 inhibitors, compound 26n was subjected to a profiling study at a concentration of 10 μM
in the full panel KINOMEscan (Ambit, San Diego, CA),[48−50] a high throughput method for screening kinase inhibitors against
a panel of 464 kinases. The TREEspot maps in Figure 5 revealed that 26n possessed an extremely high
selectivity with significant inhibitions (>80% at 10 μM)
observed
only for JNK1/2/3, Clk2, haspin, Mek3, and Ysk4 (see Supporting Information for detailed profiling data).
Figure 5
TREEspot interaction maps for 26n (also coded
as SR-11935)
in the panel profiling of 464 kinases.
Inhibition of four selected cytochrome P450 isoforms (1A2, 2C9,
2D6, and 3A4) was also tested at 10 μM, and results demonstrated
that most of these aminopyrazole-based JNK3 inhibitors had generally
low inhibition against all four enzymes. Generally, the pyrazole-amidecompounds 26f, 26j, 26g, 26k presented lower inhibition for P450 isoforms compared
to those pyridineamide based compounds (22e, 22f, 22i, 16a). The alkylamide based compounds
gave even better P450 inhibition profiles. Compounds 26l and 26n exhibited almost no inhibition for all four
selected isoforms even when tested at 10 μM. It is noteworthy
that the F-substituted compound 16a exhibited potent
inhibitions for isoforms CYP2C9 and CYP2D6.IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.Not determined.No inhibition up to 10 μM.With fluoro-substitution on the
middle phenyl ring at Region B.The microsomal stability of these compounds in both human and mouse
liver microsomes were evaluated. Briefly, the microsomal stabilities
of most inhibitors based on the pyridine amide (16a, 22f, 22g, 22i) or the N-alkyl substituted pyrazoleamide (26f) were low with
the exception of the F-substituted compound 16a (half-life
was 50 min in human and 86 min in mouse). The presence of a secondary
aminepyrrolidine ring or piperidine ring at the amide part was clearly
a favorable factor for metabolic stability. For example, inhibitors 26g and 26l, which had no secondary amine substitutions,
exhibited a stability of t1/2 = 35–40
min in human microsomes and only 4–6 min in mouse microsomes,
while the corresponding inhibitors containing a secondary amine moiety
(26j, 26k, 26n) had a much
greater microsomal stability (t1/2 = 113–408
min in human and t1/2 = 33–72 min
in mouse). The aqueous solubility was also tested for these compounds
at both pH 3.5 and pH 7.4. As shown in Table 5, while the solubility for compounds containing the pyridineamide
(22f, 22g, 22i) and N-alkyl substituted pyrazoleamide 26f were
low, inhibitors with substituents containing a secondary or tertiary
amine exhibited excellent solubility at both pH 3.5 and pH 7.4. Indeed, 26n had excellent solubility at both pH 3.5 and pH 7.4 (Table 5).TREEspot interaction maps for 26n (also coded
as SR-11935)
in the panel profiling of 464 kinases.The cytotoxicity in SHSY5Ycells was evaluated for some selected
lead inhibitors. As shown in Table 6, pyridineamide based inhibitors 22f and 22g showed significant cytotoxicity at both 10
and 30 μM, while the benzylamide based inhibitor 22i and the simple alkyl substituted pyrazoleamide 26f exhibited
significant toxicity only at 30 μM. Remarkably, other pyrazoleamide
and alkylamide based compounds were all nontoxic even at a high concentration
of 30 μM in SHSY5Ycells, indicating that they were relatively
safe JNK3 inhibitors. This was especially true for 26n which had >90% cell viability at both 10 and 30 μM.Some compounds were also evaluated in vitro for their ability to
inhibit the phosphorylation of c-Jun in neuronal
SHSY5Ycells. In-cell Western blot analysis is a direct measure of
JNK activity and was employed to investigate the ability of these
JNK3 isoform selective inhibitors to inhibit c-Jun phosphorylation.[20,36,41,51,52] As shown in Table 6, most of the compounds had an IC50 of around 1 μM
or higher in this assay, indicating that they were reasonable JNK
inhibitors. It is noteworthy that there was a large right shift between
JNK3 biochemical potency and cell-based potency for these isoform
selective JNK3 inhibitors (Table 6). The lack
of JNK1 inhibition and the cell types might be part of the reasons.[27,53] Indeed, because c-Jun phosphorylation is dependent on all three
isoforms and SHSY5Ycells contain all three isoforms, the uninhibited
JNK1 activity in these cells was the likely reason for the high cell-based
IC50 values. In contrast, these same compounds showed good
cell potency in other types of cell assays (Table 6) which were not as dependent on JNK1 nuclear activity for
substrate phosphorylation but rather more centered on the role of
JNK3 in mitochondrial dysfunction.[3,4,51,52]IC50 values are the mean
of two or more experiments (with triplicate replicates for each experiment)
with errors within 80% of the mean.Not determined.No inhibition up to 10 μM.These compounds showed protection from 6-OHDA-induced
mitochondrial
membrane depolarization and cell death in SHSY5Ycells, both of which
are regulated and promoted by JNK activation.[54] Improved cell potency was obtained in these two assays as compared
to that of the in-cell Western assay. Interestingly, very good cell
potency was seen in the mitochondrial membrane potential assays (IC50 < 50 nM) for most compounds (Table 6). The lower IC50 value in mitochondrial membrane potential
assays compared to that in in-cell Western blot assays for these compounds
(Table 6) could be due to the fact that JNK2/3
has a greater contribution to mitochondrial dysfunction than JNK1.
This interpretation is consistent with the in vivo observation of
Zhao and Herdegen where minimal to no JNK1 activity is seen on the
mitochondria in neurodegenerative models.[54] Thus, a small dose of these JNK2/3 isoform selective inhibitors
is sufficient to inhibit JNK2/3 activity and to rescue mitochondrial
membrane potential.The high cell potency for isoform selective
JNK2/3 inhibitors was
further confirmed in our mitochondrial ROS inhibition assays in SHSY5Ycells. As shown in Figure 6, compound 26n significantly reduced the ROS generated by 6-OHDA (35
μM) at a concentration as low as 40 nM (IC50 <
40 nM). Similar cell potency in this specific assay was also observed
for compound 26j (IC50 < 40 nM), showing
this phenomenon is shared among the compounds in this class. Indeed,
in contrast to biochemical assays, IC50 determinations
of functional end points in cell-based assays often have steep slopes
indicative of the functional readout that is being measured as opposed
to the direct substrate competitive inhibition seen in a biochemical
assay, and this could be why the cell-based IC50 is at
the biochemical IC50 for 26k and lower than
that for 26n. We have seen similar behavior in our aminopyrimidineclass[37] where the inhibition of ROS generation
(∼1 nM) was lower than that of the biochemical IC50 value (50 nM). Thus, it may be that only a small amount of JNK3
and JNK2 inhibition goes a long way toward shutting down ROS generation
and JNK mitochondrial function. Like 26n, the highlighted
compound in the Kamenecka et al. study (9l, SR-3562)
had exquisite selectivity over a panel of kinases. Thus, while we
cannot rule out off-target kinase activity as a possible explanation
for the lower IC50 in the cell-based assays, it is unlikely
given the high selectivity of these compounds. Finally, we cannot
rule out alternative interpretations such as other signaling pathways
that do not have a kinase component contributing to ROS or MMP changes
in the mitochondria, so this remains a possibility for the greater
potency in ROS generation and MMP dissipation for these compounds.
Figure 6
Inhibition
of mitochondrial ROS generation by inhibitor 26n in SHSY5Y
cells. SHSY5Y cells were treated with 35 μM 6-OHDA
for 5 h, and mitochondrial ROS was measured by normalized MitoSOX
fluorescence in the presence or absence of 40 nM, 370 nM, and 3 μM
inhibitor. Statistical significance (p < 0.05)
between control, untreated group, and 6-OHDA treated group is shown
by ∗. Significance (p < 0.05) between 6-OHDA-treated
groups and different concentrations of the inhibitor and 6-OHDA-treated
groups is shown by ∗∗.
Inhibition
of mitochondrial ROS generation by inhibitor 26n in SHSY5Ycells. SHSY5Ycells were treated with 35 μM 6-OHDA
for 5 h, and mitochondrial ROS was measured by normalized MitoSOX
fluorescence in the presence or absence of 40 nM, 370 nM, and 3 μM
inhibitor. Statistical significance (p < 0.05)
between control, untreated group, and 6-OHDA treated group is shown
by ∗. Significance (p < 0.05) between 6-OHDA-treated
groups and different concentrations of the inhibitor and 6-OHDA-treated
groups is shown by ∗∗.In order to evaluate the feasibility of using these JNK3
isoform
selective inhibitors for in vivo applications, in vivo pharmacokinetic
(PK) properties for selected compounds were studied in mice. As demonstrated
in Table 7, these select leadcompounds showed
good PK properties in iv dosing. They had good Cmax and AUC (area under curve) values and fair half-lives (∼1–1.5
h). The clearance (Cl) and the volume of distribution (Vd) properties were also good for 22i, 26j, and 26k but were only fair for compound 26n. On the other hand, these compounds showed poor PK properties
in oral dosing. For example, all of them exhibited almost no oral
bioavailability (F values are all ≤1%). Therefore,
the focus of future optimizations will be to improve oral PK properties
if oral administration is preferred. However, these compounds, as
JNK2/3 inhibitors for topical applications and/or in iv dosing, may
be very useful to protect against retinal degeneration in glaucoma[55] and could conceivably be used in this indication.
Finally, since JNK3 selective inhibitors have the potential to be
utilized in CNS disorders, we measured the brain penetration properties
of 26n. After ip dosing (10 mg/kg in mice) brain levels
at 2 h after dosing was approximately 55% of that found in the plasma.
In addition, we also measured the brain concentration of 26n at 2 and 6 h after po and iv dosing. Because of the poor oral bioavailability
of 26n (see Table 7) the brain
levels at 2 and 6 h were below the lower limit of quantitation (∼5
nM) after oral dosing but were found to be 60 and 40 nM at 2 and 6
h, respectively, after iv dosing of 2 mg/kg (Table 8).
Table 7
In Vivo PK Data on Mice for Selected
Lead JNK3 Inhibitorsa
cmpd
Cmaxa (μM), iv
AUC a (μM·h), iv
T1/2a (h), iv
Cl a ((mL/min)/kg), iv
Vda (L/kg), iv
Fa (%) po
22i
2.4
1.8
1.0
10
0.6
1.0
26j
3.2
1.5
1.6
12
0.7
0.4
26k
1.9
1.1
1.1
17
0.9
0.6
26n
1.7
1.0
1.5
21
1.2
0.5
Dosed at
0.5 mg/kg for iv and at
3 mg/kg for po. All data were based on three determinations.
Table 8
Brain Concentration
of 26n in Micea
dosing
2 h
6 h
iv (2 mg/kg)
60 nM
40 nM
po (5 mg/kg)
<5 nM
<5 nM
Data were generated
from three determinations.
Dosed at
0.5 mg/kg for iv and at
3 mg/kg for po. All data were based on three determinations.Data were generated
from three determinations.
Conclusion
We have developed a series of aminopyrazoleamide based JNK3 isoform
selective inhibitors starting from the initial leadcompound SR-4326.
SAR investigation and optimization successfully yielded potent JNK3
inhibitors with high selectivity over p38 kinase, as well as high
isoform selectivity against JNK1 and in some cases over JNK2. For
example, compound 26n had an isoform selectivity of >50-fold
for JNK3 against JNK1 and were potent in all three types of cell-based
assays in SHSY5Ycells (in-cell Western target modulation, inhibition
of 6-OHDA induced cell death, and inhibition of 6-OHDA induced mitochondrial
membrane depolarization). Another lead inhibitor 26k was
highly potent and selective with an IC50 value against
JNK3 in the subnanomolar range and possessed an isoform selectivity
of >500-fold over JNK1 and >20000-fold against p38α. In
addition, 26n and 26k were noncytotoxic,
potently inhibited
mitochondrial dysfunction, and possessed good PK properties in iv
dosing. In addition, 26n showed a plasma/brain ratio
of ∼2:1. The X-ray crystal structure of 26k in
JNK3 revealed a binding mode that can very well support the observed
SAR, and this structure was also helpful for designing potent and
selective JNK3 inhibitors. Future optimizations for these aminopyrazole-based
compounds will be mainly focused on improving the pharmacokinetic
properties in order to obtain better JNK3 isoform selective inhibitors
for oral dosing. These studies will be reported in due course.
Experimental Section
Commercially
available reagents and anhydrous solvents were used
without further purification unless otherwise specified. Thin layer
chromatography (TLC) analyses were performed with precoated silica
gel 60 F254. The mass spectra were recorded by LC/MS with Finnigan
LCQ Advantage MAX spectrometer of Thermo Electron. Flash chromatography
was performed on prepacked columns of silica gel (230–400 mesh,
40–63 μm) by CombiFlash with EtOAc/hexane or MeOH/DCM
as eluents. The preparative HPLC was performed on SunFire C18 OBD 10 μm (30 mm × 250 mm) with CH3CN + 50%
MeOH/H2O + 0.1% TFA as eluents to purify the targeted compounds.
Analytic HPLC was performed on Agilent Technologies 1200 series with
CH3CN (solvent B)/H2O + 0.9% CH3CN
+ 0.1% TFA (solvent A) as eluents, and the targeted products were
detected by UV in the detection range of 215–310 nm. All compounds
were determined to be >95% pure by this method. NMR spectra were
recorded
with a Bruker 400 MHz spectrometer at ambient temperature with the
residual solvent peaks as internal standards. The line positions of
multiplets were given in ppm (δ), and the coupling constants
(J) were given in hertz. The high-resolution mass
spectra (HRMS, electrospray ionization) experiments were performed
with Thermo Finnigan Orbitrap mass analyzer. Data were acquired in
the positive ion mode at resolving power of 100 000 at m/z 400. Calibration was performed with
an external calibration mixture immediately prior to analysis.
General Synthetic
Procedures
A mixture of 4-nitro-1H-pyrazole 1 (10 mmol), 3-bromobenzoic acid 2 (20 mmol),
CuI (2.0 mmol), trans-N,N-dimethylcyclohexane-1,2-diamine (4.0
mmol), and Cs2CO3 (30 mmol) in DMF (20 mL) was
purged with argon and stirred for 12 h at 100 °C in a sealed
tube. The reaction mixture was cooled, and filtered through a pad
of silica gel, and rinsed with EtOAc. The resulting solution was concentrated
in vacuo to yield a crude residue which was purified by chromatography
on silica gel (EtOAc/hexane) to provide 3-(4-nitro-1H-pyrazol-1-yl)benzoic acid 3.To a solution of
3-(4-nitro-1H-pyrazol-1-yl)benzoic acid 3 (5.0 mmol) in CH2Cl2 (20 mL) were added EDC
(10 mmol), HOBt (10 mmol), and diisopropylethylamine (15 mmol), and
the mixture was stirred for 30 min. Then the 6-methylpyridin-3-amine
(5.5 mmol) was added, and the resulting mixture was stirred overnight.
Water (50 mL) was added to the reaction mixture and extracted with
EtOAc (2 × 100 mL). The resulting solution was concentrated in
vacuo to yield a crude product (4).This intermediate
was hydrogenated in anhydrous methanol (100 mL)
in the presence of 10% Pd/C (1.0 g) under a balloon of hydrogen for
3 h. The mixture was filtered through a Celite pad and evaporated.
The residue was purified by chromatography on silica gel (dichloromethane/methanol)
to give the 3-(4-amino-1H-pyrazol-1-yl)-N-(6-methylpyridin-3-yl)benzamide 5.1-Chloro-2-isocyanatobenzene
(0.12 mmol) was added to 3-(4-amino-1H-pyrazol-1-yl)-N-(6-methylpyridin-3-yl)benzamide
(0.1 mmol) in CH2Cl2 (1.0 mL) at room temperature
and stirred for 1 h. The solvent was evaporated and the residue was
purified by reverse-phase preparative HPLC to give 3-(4-(3-(2-chlorophenyl)ureido)-1H-pyrazol-1-yl)-N-(6-methylpyridin-3-yl)benzamide 8c as a white powder.
Homogeneous time-resolved
fluorescence assay–enzyme inhibition studies were performed
in 384-well polystyrene homogeneous time-resolved fluorescence plates
(Grainier) for 1 h at ambient temperature (∼22 °C) with
0.5 μM biotinylated FL-ATF-2, 1.25 μM ATP, 0.75 nM activated
JNK3α1 or JNK2 or JNK1 (with a control in the absence of kinase
for determining the basal signal) in 10 μL volumes containing
the final concentrations of the following: 50 mM Hepes, pH 7.0, 2.5
mM MgCl2, 0.1 mg/mL bovine serum albumin, 1 mM dl-dithiothreitol, 0.01% Triton X-100 (all from Sigma-Aldrich), and
5% DMSO (with or without compound). A 10-point titration of all compounds
was carried out in 3-fold dilutions from 10 μM to 0.5 nM. After
22 min, the kinase reaction was terminated by addition of 10 μL
of quenching solution (1× Lance buffer, detection reagents, streptavidin-xlAPC
(200 nM), and europium cryptate-labeled rabbit polyclonal anti-phospho-ATF-2
(1 nM) were from Cis-Bio). The homogeneous time-resolved fluorescence
signal was detected using an EnVision plate reader 1 h after quenching.
The data from four different experiments were averaged and presented
as the mean ± SD. IC50 values were determined by fitting
the data to the equation for a four-parameter logistic.[3,4,37,51]
Cell Culture
SHSY5Ycells (ATCC) were grown at 37 °C
and 5% CO2 in DMEM/F:12 (Invitrogen) supplemented with
10% fetal bovine serum (FBS) and penicillin/streptomycin. To ensure
that the cells were actively growing, only cells at ∼80% confluency
and between passages five and 15 were used in the experiments. The
cells were serum starved for 24 h in DMEM/F:12 medium containing 2%
FBS before any treatment.
Mitochondrial Membrane Depolarization
Mitochondrial
membrane depolarization was monitored by MitoTracker Orange CMTRos
(Invitrogen) fluorescence. SHSY5Ycells (60 000 cells/well)
were seeded in black walled, clear bottomed 96-well plates. Cells
were incubated with compounds for 30 min before the addition of 35
μM 6-hydroxydopamine (6-OHDA) for 4 h. After the incubation,
cells were stained with 500 nM MitoTracker Orange CMTRos for 30 min
under growth conditions. Cells were washed twice in Hank’s
buffer salt solution (HBSS), and placed in prewarmed HBSS for fluorescent
recordings. Fluorescence was detected at 576 nm (exciting at 554 nm)
on a SpectraMax e5 plate reader (Molecular Devices). Mitochondrial
membrane depolarization was normalized to cell abundance by staining
the cells with Hoechst 33342 (excitation, 350 nm; emission, 450 nm).
Cell Viability
Cell viability of SHSY5Ycells was monitored
by MTT assay (Cayman Chemical). Cells (60 000 cells/well) were
seeded in a 96-well plate (clear bottom) and treated as described
in the text. At the culmination of each treatment the cells were treated
with the MTT reagent. Absorbance was monitored in a SpectraMax e5
plate reader (Molecular Devices).
Measurement of Compound
Cytotoxicity
Cytotoxicity of
the compounds was monitored by MTT assay (Cayman Chemical). SHSY5Ycells (60 000 cells/well) were seeded in a 96-well plate (clear
bottom) and incubated with different concentrations (0–30 μM)
of compounds for 48 h. At the culmination of the treatment the cells
were treated with the MTT reagent. Absorbance was monitored in a SpectraMax
e5 plate reader (Molecular Devices).
In Cell Western Assay
SHSY5Ycells (60 000 cells/well)
were plated in a clear-bottomed Packard View black 96-well plate in
100 μL of 10% FBS DMEM:F12 medium and were allowed to attach
overnight. Next day, the cells were treated with the compounds for
1 h prior to induction of the JNK pathway activation. The cells were
treated with 35 μM 6-OHDA for 4–5 h. Cells were then
fixed in 4% paraformaldehyde in PBS for 20 min at room temperature
with no shaking. They were then washed once with 0.1 M glycine to
neutralize paraformaldehyde for 5 min. Cells were permeabilized with
0.2% Triton X-100 in PBS for 20 min at room temperature on orbital
shaker after which they were washed once with PBS for 5 min. They
were then incubated with Licor blocking buffer in PBS (1:1 dilution
in PBS) for 1–1.5 h rocking at room temperature. Cells were
incubated with primary antibody p-c-Jun S63 Ab (Cell Signaling no.
9261) 1:100 dilution in Licor blocking buffer overnight at 4 °C.
Next day, they were washed twice with PBS–0.1% Tween 20 (PBST)
washing solution for 5 min each at room temperature on the orbital
shaker, followed by one wash with Licor blocking buffer containing
0.05% Tween-20 for 5 min on the shaker at room temperature. The cells
were then incubated with secondary antibody goat anti-rabbit IR800
(1:500 dilution) for 1 h at room temperature in the dark (covered
the plate with foil) in Licor blocking buffer-containing Tween-20.
Following this, cells were washed twice with PBST for 5 min each at
room temperature and then once with Licor blocking buffer-containing
0.05% Tween-20. The wells were then incubated with ToPro 3 stain (nucleic
acid staining), diluted 1:4000 in Licor blocking buffer or Licor blocking
buffer with 0.05% Tween-20 for 30 min at room temperature in the dark.
Finally the plates were washed twice with PBS and analyzed using the
Odyssey LICOR infrared scanner.[3,4,20,36,37,51,52,56]
Crystallization
Purification of
JNK3 39-402 and its
crystallization with AMPPNP was done following previously published
procedure.[36,37,57] Compound 26k was soaked into the crystal by adding
2 mM compound into the crystallization drop and incubating for 24
h. The crystal was then transferred to a mounting loop, and excess
soaking solution was removed and flash frozen by plunging into liquid
nitrogen.
Data Collection and Refinement
A
diffraction data set
was collected at LS-CAT beamline 21-ID-G (APS) using Marmosiac 300
CCD detector (Mar Research). The data set was processed with autoProc
with XDS as the data reduction engine. The PDB code 1JNK was used as the
molecular replacement model. Quick molecular replacement using Phaser
in Phenix suite properly positoned JNK3, 26k was identified
as positive densities in ATP binding pocket. Restraints and coordintates
for the compound was generated using eLBOW in Phenix suite and incorporated
into the JNK3coordinate using the graphics program Coot. The model
was then refined using autoBuster with TLS (translation, libration,
and screw-motion), water update, and unknown ligand search options
turned on. The model was manually inspected and adjusted after each
refinement cycle using Coot. The refinement was completed after the
free and crystallographic R-factors stabilized. Data
processing and refinement statistics are given in Supporting Information. Structural analysis and figure preparations
were done using PyMol. The coordinate and the structure factor are
deposited to wwPDB with the PDB code 4WH2.
Authors: R Jeffrey Neitz; Andrei W Konradi; Hing L Sham; Wes Zmolek; Karina Wong; Ann Qin; Colin Lorentzen; David Nakamura; Kevin P Quinn; John-Michael Sauer; Kyle Powell; Lany Ruslim; David Chereau; Zhao Ren; John Anderson; Frédérique Bard; Ted A Yednock; Irene Griswold-Prenner Journal: Bioorg Med Chem Lett Date: 2011-04-27 Impact factor: 2.823
Authors: Richard M Angell; Francis L Atkinson; Murray J Brown; Tsu Tshen Chuang; John A Christopher; Maria Cichy-Knight; Allison K Dunn; Kendra E Hightower; Susanna Malkakorpi; James R Musgrave; Margarete Neu; Paul Rowland; Robyn L Shea; Jeffery L Smith; Donald O Somers; Sonia A Thomas; Gladstone Thompson; Ruolan Wang Journal: Bioorg Med Chem Lett Date: 2006-12-15 Impact factor: 2.823
Authors: Britt-Marie Swahn; Fernando Huerta; Elisabet Kallin; Jonas Malmström; Tatjana Weigelt; Jenny Viklund; Patrick Womack; Yafeng Xue; Liselotte Ohberg Journal: Bioorg Med Chem Lett Date: 2005-11-15 Impact factor: 2.823
Authors: Candice E Crocker; Susan Khan; Michael D Cameron; Harold A Robertson; George S Robertson; Philip Lograsso Journal: ACS Chem Neurosci Date: 2011-04-20 Impact factor: 4.418
Authors: Sung Ok Yoon; Dong Ju Park; Jae Cheon Ryu; Hatice Gulcin Ozer; Chhavy Tep; Yong Jae Shin; Tae Hee Lim; Lucia Pastorino; Ajaya J Kunwar; James C Walton; Alan H Nagahara; Kun Ping Lu; Randy J Nelson; Mark H Tuszynski; Kun Huang Journal: Neuron Date: 2012-09-06 Impact factor: 17.173
Authors: Ke Zheng; Chul Min Park; Sarah Iqbal; Pamela Hernandez; HaJeung Park; Philip V LoGrasso; Yangbo Feng Journal: ACS Med Chem Lett Date: 2015-03-02 Impact factor: 4.345
Authors: Brian C Bales; Victoria Cotero; Dan E Meyer; Jeannette C Roberts; Monica Rodriguez-Silva; Tiberiu M Siclovan; Jeremy W Chambers; Michael J Rishel Journal: ACS Med Chem Lett Date: 2022-09-28 Impact factor: 4.632
Authors: Sandeep Rana; Yogesh A Sonawane; Margaret A Taylor; Smitha Kizhake; Muhammad Zahid; Amarnath Natarajan Journal: Bioorg Med Chem Lett Date: 2018-10-15 Impact factor: 2.823
Figure 1
Scheme 1
Scheme 2
Scheme 3
Scheme 4
Scheme 5
Figure 2
Figure 4
Figure 3
Figure 5
Figure 6