Aminopyridine carboxamides as c-Jun N-terminal kinase inhibitors: Targeting the gatekeeper residue and beyond

Article abstract of DOI:10.1016/j.bmcl.2006.08.097

The structure-activity relationships of 5,6-positions of aminopyridine carboxamide-based c-Jun N-terminal Kinase (JNK) inhibitors were explored to expand interaction with the kinase specificity and ribose-binding pockets. The syntheses of analogues and the impact of structural modification on in vitro potency and cellular activity are described.

Full text of DOI:10.1016/j.bmcl.2006.08.097

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5723–5730
Aminopyridine carboxamides as c-Jun N-terminal kinase inhibitors:
Targeting the gatekeeper residue and beyond
Gang Liu,* Hongyu Zhao, Bo Liu, Zhili Xin, Mei Liu, Christi Kosogof,
Bruce G. Szczepankiewicz, Sanyi Wang, Jill E. Clampit, Rebecca J. Gum,
Deanna L. Haasch, James M. Trevillyan and Hing L. Sham
Metabolic Disease Research, Global Pharmaceutical Research and Development, Abbott Laboratories,
Abbott Park, IL 60064-6101, USA
Received 12 July 2006; revised 21 August 2006; accepted 23 August 2006
Available online 12 September 2006
Abstract—The structure–activity relationships of 5,6-positions of aminopyridine carboxamide-based c-Jun N-terminal Kinase
(JNK) inhibitors were explored to expand interaction with the kinase specificity and ribose-binding pockets. The syntheses of
analogues and the impact of structural modification on in vitro potency and cellular activity are described.
Ó 2006 Elsevier Ltd. All rights reserved.
Dysregulation of the insulin receptor (IR) and reduced
tyrosine phosphorylation of the insulin receptor sub-
strate (IRS) proteins contribute significantly to periphe-
ral insulin resistance and b-cell failure.1 Increased serine
phosphorylation of IRS-1 is a common finding during
insulin resistance and type 2 diabetes.2 c-Jun N-terminal
kinase-1 (JNK-1), a member of the MAP kinase family,
associates with IRS-1 and phosphorylates Ser307.3 This
phosphorylation blocks the interaction between IRS-1
and IR, inhibits insulin action.4 Phosphorylated ser307
also targets IRS-1 for degradation by the proteasome
and affects the subcellular distribution of IRS-1. Target-
ed disruption of the JNK-1 gene in mice protects the
animals from diet-induced obesity, and results in de-
creased adiposity and enhanced secretion of adiponec-
tin.5 The JNK-1 KO mice maintain lower fasting
plasma glucose and insulin levels compared to wild-type
littermates and demonstrate greater insulin sensitivity in
both oral glucose and intraperitoneal insulin tolerance
tests.5 Small molecule JNK inhibitors can potentially of-
fer therapeutic utility for the treatment of diabetes and
insulin resistance.6 A number of JNK inhibitors have
recently appeared in the patent and primary literatures.7
Recently we disclosed a novel series of aminopyridine
carboxamides as potent, selective, ATP-competitive
pan-JNK inhibitors, as represented by compounds 1
and 2 (Fig. 1).8,9 The binding modes of this series of
JNK inhibitors have been determined by X-ray crystal-
lography. As shown in Figure 2, the 4-aminopyridine
amide of 2 displayed hydrogen bonds (as highlighted
by the dashed lines in magenta color) to the backbones
of amino acid Glu109 and Met111 as the ‘classical’
hinge interactions.
Closer examination of the structural complex also re-
vealed that the 5-cyano group of the pyridine pointed
to the gatekeeper residue Met108 (Fig. 2). It is well
known that this single residue in the ATP-binding pock-
et can act as a gatekeeper and control kinase selectivity
to a wide range of structurally unrelated compounds.10
Thus, modification of the 5-cyano group would provide
the possibility of expanding into the hydrophobic
O
O
NH2
NH2
O
N
N
S
O
O
H
N
N
N
O
N
O
H
2
O
1
IC50 = 0.019 μM vs. JNK1
IC50 = 0.12 μM vs. JNK2
IC50 = 0.045 μM vs. JNK1
IC50 = 0.16 μM vs. JNK2
Keywords: JNK inhibitors; Gatekeeper residue; Specificity pocket;
Ribose pocket.
*
Figure 1. Selected 4-aminopyridine carboxamide JNK inhibitors from
our laboratories.
Corresponding author. Tel.: +1 847 935 1224; fax: +1 847 938
1674; e-mail: gang.liu@abbott.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.097

5724
G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
acetylenes to provide 5-alkynyl analogues 7a–f,14 while
Ile106
Suzuki coupling of 6 using vinyl boronic acids generated
5-alkenyl derivatives 8a–d. Interestingly, Sonogashira
Met108
coupling of
6 with trimethylsilylacetylene yielded
TMS-enyne 9, which was desilylated to give enyne 8e.
A second carbonylation of 6 yielded the ester 10.
Subsequent hydrolysis and coupling afforded the amides
11a–b.
Met77
The 6-chloropyridine intermediate was prepared accord-
ing to the procedure shown in Scheme 2. The readily
available 2-hydroxy-6-chloropyridine 1215 was activated
as a triflate. Selective palladium-catalyzed carboxylation
under 60 psi of carbon monoxide at room temperature
yielded a methyl ester, which was saponified to give acid
13. Condensation of 13 with p-methylsulfonylbenzyl-
amine provided the versatile intermediate 14. Suzuki-
type coupling of 6-chloropyridine 14 with aryl/alkenyl
boronic acids yielded most of the 6-carbon analogues
15a–p.
Figure 2. X-ray crystal structure of 2 (in brown color) in the JNK1
active site overlaid with ATP (in gray color).
specificity pocket beyond the gatekeeper residue. The
amino acid sequences in the specificity pocket are some-
what different among the three JNK isoforms, for exam-
ple, Ile106 and Met77 in JNK1 are mutated to Leu and
Leu in JNK2, respectively. 4-(4-Pyrimidinyl)-5-pheny-
limidazole derivative 3a was reported as a potent pan-
JNK inhibitor with interaction to the specificity pocket
(Fig. 3).11 The recently disclosed 6-anilinoimdazole-
based JNK3 inhibitor 3b12 and aniniliobipyrine-based
JNK3 inhibitor 3c13 were shown to occupy the
specificity pocket of JNK3 via the hydrophobic aniline
groups, while achieving some degree of selectivity for
JNK3 over JNK1 (Fig. 3).
The aromatic nucleophilic substitution for forming the
6-ether analogues did not work with alkoxides even
under forced conditions in a microwave reactor. We
found that a copper-catalyzed ether formation proto-
col16 worked relatively well to provide several analogues
(16a–e) for SAR evaluation (Scheme 2). 6-Amino-
substituted pyridine modifications (17a–r) could be
readily generated with various amines under micro-
wave-assisted SNAr conditions using DMA as a solvent.
The 6-chloropyridine could also be further carboxylated
at elevated temperature to provide 6-carboxylate, which
was converted to amide 18.
Additionally, further extension of the 6-ethoxy group of
2 could increase the interaction with the ribose triphos-
phate-binding sites of the ATP pocket. The extensive
structural modification of the 5,6-positions of the pyri-
dine core of 2 to locate and optimize these potential
interactions is described in this communication. These
interactions, if realized, were expected to further im-
prove the potency and general kinase selectivity of the
resulting inhibitors.
A 6-bromopyridine was needed for the desired Heck-type
coupling. The previously disclosed 6-ethoxyester 199 was
converted to 6-bromopyridine 20 in low yield as shown in
Scheme 3. Hydrolysis and coupling with amine provided
6-bromopyridine amide 21. Heck coupling of 21 with
acrylate was successful to give a tert-butyl ester, which
was deprotected to yield acid 15q. Hydrogenation of the
olefin led to the propoinic acid 15r.
Modification of the 5-cyano group of 2 was achieved
starting from commercially available dichloropyridine
4 (Scheme 1). One of the chlorine was displaced with
an ethoxy group, while the other one was subsequently
subject to a palladium-mediated carbonylation to give
methyl ester 5. Selective iodination of the 5-position of
the pyridine, saponification of the ester, followed by
condensation between the resulting acid and amine cre-
ated 5-iodopyridine 6. This key intermediate 6 could
then undergo Sonogashira-type coupling with terminal
The inhibitory activity of JNK inhibitors discussed in
this communication was evaluated in JNK1 and JNK2
kinase assays measuring inhibition of phosphorylation
of a synthetic peptide substrate at Km ATP concentra-
tion for each enzyme as described previously.8 Most
of the active analogues were also assayed for their
capacity to inhibit c-Jun phosphorylation in HepG2
cells.8
H
NH
N
N
N
NH
N
H
N
N
N
N
O
O
H
NH
Cl
N
N
N
H
O
3a (Merck)
IC50 ~ 0.002 μM
vs. JNK1 & 3
3b (Astra-Zeneca)
IC50 = 0.071 μM vs. JNK1
IC50 = 0.0014 μM vs. JNK3
3c (Astra-Zeneca)
Cl
N
IC50 = 0.384 μM vs. JNK1
IC50 = 0.007 μM vs. JNK3
H
Cl
Figure 3. Selected JNK inhibitors from the literature.

G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
5725
NH2
N
O
S
NH2
NH2
I
a,b
c,d,e
O
H
N
O
O
N
O
Cl
N
4
Cl
S
O
O
5
NH2
6
R
R
O
f
f
O
H
N
N
O
NH2
O
S
7a-f
O
NH2
O
S
O
g
h
O
H
N
TMS
H
N
N
O
N
O
O
9
8a-e O
NH2
O
O
NH2
O
O
S
i
j,k
S
NHR
O
H
O
O
H
N
N
N
O
N
O
11a-b
O
10
O
Scheme 1. Reagents and conditions: (a) EtONa, EtOH, 150 °C, 2 h; (b) CO, PdCl2(dppf), MeOH, 100 °C, 4 h, 63% over two steps; (c) chloramine-T,
NaI, AcOH, rt, 30 min; (d) LiOH, MeOH/THF, rt; (e) p-methylsulfonylbenzyl amine, TBTU, Et3N, DMF, rt, 81% over three steps; (f) alkynes,
PdCl2(PPh3)2, CuI, Et3N, DMF, 120 °C, 20 min, ꢀ30%; (g) alkenyl boronic acids, Pd(PPh3)4, Na2CO3, DMF/THF/H2O, 120 °C, 20 min, ꢀ50%; (h)
TBAF, THF, rt, 80%; (i) CO, PdCl2(dppf), MeOH, 60 °C, 16 h, 80%; (j) LiOH, MeOH/THF, rt; (k) H2NR, TBTU, Et3N, DMF, rt, ꢀ65% over two
steps.
Scheme 2. Reagents and conditions: (a) Tf2O, pyridine, CH2Cl2, rt, 6 h, 69%; (b) 60 psi CO, Et3N, MeOH, PdCl2(dppf), rt, 16 h; (c) LiOH, MeOH/
THF, rt, 82% over two steps; (d) p-methanesulfonylbenzylamine, TBTU, Et3N, DMF, rt, 88%; (e) RB(OH)2, PdCl2(PPh3)2, Na2CO3, DME/EtOH/
H2O (7:2:3), 160 °C, 10 min, 70–80%; (f) ROH, CuI, 1,10-phenanthroline, Cs2CO3, 110 °C, 24 h, 60–70%; (g) RNH2, DMA, 160 °C, 20 min, 35–80%;
(h) 60 psi CO, Et3N, MeOH, PdCl2(dppf), 100 °C, 16 h; (i) 3N aq NaOH, MeOH, rt, 45% over two steps; (j) i-PrNH2, TBTU, Et3N, DMF, rt, 85%.
Using the readily available 5-iodopyridine 6, we tried to
replace the 5-cyano group of 2 with alkynes, hoping that
further extension of the alkyne would reach into the ki-
nase specificity pocket. The results, as shown in Table 1,
were generally disappointing. Only the terminal alkyne
7a possessed a meaningful potency, any other substitu-
ents (non-polar or polar, aliphatic or aromatic) off the
alkyne led to essentially inactive compounds (7b–f).
We also examined some 5-alkene-substituted pyridines
as shown in Table 1. 5-Vinyl pyridine 8a is about 3-fold
less active than the 5-ethynyl analogue 7a. Further dec-
oration of the olefin with either aromatic (8b) or aliphat-
ic group (8c) seemed to have little impact on the potency
of resulting compounds. Interestingly, we found that
acrylonitrile (8d) and but-1-en-3-ynyl (8e) substituted
pyridines possessing progressively improved JNK inhib-
itory potency with encouraging cellular potency for 8d
as well. Replacing the 5-cyano group with a primary
amide (11a) yielded a somewhat active analogue, howev-
er, with much reduced cellular activity, while a propyl
amide replacement led to a virtually inactive analogue
11b.
For the modification of the 6-position of the pyridine,
we started with 3-substituted phenyl groups as a way
to extend into the sugar pocket. As shown in Table 2,
a hydrophobic 3-chloro phenyl (15a) led to a 100-fold
decrease of potency from 2, while 3-hydroxyphenyl
(15b) and 3-methoxyphenyl (15c) derivatives exhibited
sub-micromolar potency. A 3-methanesulfonylphenyl
analogue 15d is only 2-fold less active than 3. However,
its cellular potency was much worse.

5726
G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
NH2
NH2
NH2
N
N
O
S
N
N
a
b,c
O
H
N
O
O
N
O
N
Br
Br
N
O
O
19
20
21
O
NH2
NH2
O
S
O
S
N
d,e
f
O
O
H
H
N
N
N
R
N
R
15q
15r
O
O
Scheme 3. Reagents and conditions: (a) POBr3, pyridinium hydrobromide, H3PO4, 125 °C, 25 min, 27%; (b) LiOH, MeOH/THF, rt; (c)
p-methanesulfonylbenzylamine, TBTU, Et3N, DMF, rt, 90% over two steps; (d) tert-butyl acrylate, Pd(OAc)2, P(o-tolyl)3, Et3N, 100 °C, over night,
22%; (e) TFA/CH2Cl2, rt, 57%; (f) 10% Pd/C, H2, MeOH, rt, 100%.
Table 1. SAR of 5-pyridine carboxamide modifications
NH2
N
NH2
R
NH2
N
O
O
O
S
O
S
O
S
R
NHR
O
H
N
O
O
H
H
N
N
O
N
O
8a-e
O
7a-f
11a-b
O
O
a
a
Compound
R
JNK1 IC50 (lM)
JNK2 IC50 (lM)
Pc-Jun IC50 (lM)
7a
H
0.35 ( 0.17)
0.30 ( 0.03)
1.89
7b
7c
>10
NT
NT
3.43 ( 0.73)
1.09 ( 0.15)
2.24 ( 1.04)
1.55 ( 0.35)
1.18 ( 0.08)
2.43 ( 0.53)
2.93 ( 0.53)
0.41 ( 0.12)
0.079 ( 0.001)
0.318 ( 0.088)
5.25
>10
NT
S
7d
7e
NT
5.7
OH
OH
4.61 ( 0.51)
4.32 ( 0.22)
1.46 ( 0.58)
4.13 ( 0.43)
4.16 ( 2.56)
1.79 ( 0.39)
0.30
7.36
5.24
4.63
6.92
>10
0.807
1.4
7f
OH
8a
8b
8c
H
8d
8e
N
11a
11b
H
0.969
7.54
NT
NT
a Values are geometric means of at least two experiments, standard error is given in parentheses. NT, not tested.
Heteroaryls were also evaluated for accessing the ribose
pocket off the 6-position. 3-Thiophene yielded fairly po-
tent analogue 15e, and 3-furanyl analogue 15f is only 4-
fold less potent than 2. 2-Hydroxymethyl-substituted
(15h) and 2-benzylaminomethyl-substituted thiophene
(15i) are progressively less potent than the unsubstituted
15e. It is interesting to note that 3-carboxy-2-thiophene
derived 15g is fairly potent in vitro, however, its cellular
potency suffered due to the charge present in the molecule.
The six-membered heterocycles such as pyridines (15j–k)
and pyrimidine (15l) were roughly equipotent as 2-furanyl
analogue 15f with much reduced cellular potency.

G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
Table 2. SAR of 6-pyridine modification with carbon-linked side chains
5727
NH2
O
N
S
O
H
N
N
R
O
15a-r, 18
a
a
Compound
R
JNK1 IC50 (lM)
JNK2 IC50 (lM)
Pc-Jun IC50 (lM)
Cl
OH
O
15a
1.18 ( 0.35)
0.84
1.2
15b
15c
15d
15e
15f
0.20 ( 0.08)
0.43 ( 0.20)
0.043 ( 0.004)
0.081 ( 0.005)
0.27 ( 0.07)
0.24 ( 0.15)
0.60 ( 0.38)
0.097 ( 0.039)
0.39 ( 0.12)
0.53 ( 0.11)
9.4
1.8
O O
S
10.5
0.61
4.3
S
O
S
OH
O
15g
0.25 ( 0.05)
0.16 ( 0.10)
>10
15h
15i
15j
0.57 ( 0.20)
1.99 ( 0.49)
0.32 ( 0.06)
0.68 ( 0.27)
NT
10.8
8.01
7.3
S
HO
S
NH
N
0.34 ( 0.01)
15k
15l
0.53 ( 0.12)
0.62 ( 0.06)
0.66 ( 0.06)
0.86 ( 0.02)
NT
N
N
>10.0
N
15m
15n
0.048 ( 0.010)
0.15 ( 0.04)
0.23 ( 0.07)
0.26 ( 0.05)
0.89
5.5
15o
15p
1.45 ( 0.35)
0.17 ( 0.08)
1.29 ( 0.19)
0.24 ( 0.08)
0.35
5.3
OH
OH
OH
15q
15r
18
0.30 ( 0.03)
1.64 ( 0.86)
0.43 ( 0.19)
0.43 ( 0.03)
6.92 ( 2.70)
0.12 ( 0.02)
>2.5
NT
2.9
O
O
H
N
O
a Values are geometric means of at least two experiments, standard error is given in parentheses. NT, not tested.

5728
G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
Table 3. SAR of 6-pyridine modification with amine/ether linked side chains
NH2
NH2
N
O
S
O
S
N
N
O
H
N
O
H
N
R
R
N
O
N
H
17a-r
16a-e
O
O
a
a
Compound
R
JNK1 IC50 (lM)
JNK2 IC50 (lM)
Pc-Jun IC50 (lM)
16a
0.015 ( 0.003)
0.038 ( 0.06)
0.25
16b
16c
16d
0.023 ( 0.006)
0.088 ( 0.044)
0.39 ( 0.03)
0.047 ( 0.013)
0.37 ( 0.05)
0.61 ( 0.04)
0.56
0.70
0.49
O
16e
0.42 ( 0.06)
0.45 ( 0.08)
NT
17a
17b
17c
0.067 ( 0.024)
0.032 ( 0.009)
0.085 ( 0.029)
0.084 ( 0.006)
0.069 ( 0.009)
0.16 ( 0.01)
0.49
1.2
0.62
17d
17e
17f
17g
17h
17i
0.096 ( 0.006)
0.82
0.29 ( 0.12)
4.7
0.48
NT
2.7
0.26
1.0
H
N
NH
2
0.093 ( 0.036)
0.066 ( 0.001)
0.059 ( 0.001)
NT
>10
>10
>10
O
O
0.087 ( 0.026)
0.081 ( 0.015)
NH
2
H
N
O
17j
0.054 ( 0.002)
0.093 ( 0.021)
0.12 ( 0.04)
1.6 ( 0.5)
7.3
0.18 ( 0.03)
0.15 ( 0.02)
0.21 ( 0.06)
NT
8.3
OH
OH
O
17k
17l
>10
5.9
NH2
17n
17o
NT
NT
9.2
N
17p
17q
17r
0.022 ( 0.005)
0.18 ( 0.05)
1.5 ( 0.3)
0.030 ( 0.015)
0.32 ( 0.14)
1.4
2.6
HO
>10
NT
>10
HO
HO
OH
OH
17m
0.076 ( 0.002)
0.10 ( 0.02)
a Values are geometric means of at least two experiments, standard error is given in parentheses. NT, not tested.

G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
5729
Olefinic substitutions off the 6-position pyridine offered
some interesting analogues. trans-Propene derivative
15m is only 2-fold less potent than lead compound 1
in both enzymatic and cellular assays. In comparison,
the cis-isomer 15n is 3- to 8-fold less potent in vitro
and in cells, respectively. Styrene substitution 15o
showed a 20-fold decrease in potency. Interestingly,
15o was shown to be more potent in HepG2 cells than
in the enzymatic assay. The possibility of cellular toxic-
ity could not be ruled out although the cells appeared
healthy during the cellular assay. Polar groups append-
ed to the trans-double bond, such as hydroxyethyl 15p
or carboxylic acid 15q, provided fairly potent analogues.
Saturation of the propenic acid (15r) reduced the inhib-
itory potency by about 5-fold. Among the several 6-car-
boxamide derivatives, only the isopropyl amide 18
possessed sub-micromolar IC50 versus JNK1&2.
cavity within the ATP-binding pocket.17 Kinases with
a threonine at this position are readily targeted by a
diverse classes of small molecule inhibitors that can
access this natural pocket, including the kinase inhibi-
tors currently in clinical use (Gleevec, Iressa, Tarceva).
Having a larger methionine as the gatekeeper residue
in JNK made it intrinsically more difficult to reach the
‘specificity pocket’.
The second reason may have something to do with the 4-
aminopyridine carboxamide template we were using to
access the ‘specificity pocket’. The particular vector pro-
vided by extension off the 5-position of pyridine 1 may
not be as favorable as X-ray structure and molecular
modeling had suggested. The 6-anilinoimdazole-based
JNK3 inhibitor 3b (Fig. 3) binds to JNK3 in an in-
duced-fit manner, with the gatekeeper residue Met146
(JNK3 numbering) moved out of way to accommodate
the aniline portion of the molecule in the ‘specificity
pocket’. Such precedent suggests that with appropriate
molecular template, it is possible to target the ‘specifici-
ty’ pocket with methionine as the gatekeeper residue.
More productive SAR came from the 6-amino or 6-
ether analogues (Table 3). Isopropoxy compound 16a
is equipotent to the lead compound 2 in vitro, and 2-fold
more active in c-Jun phosphorylation assay than 2.
Some other smaller alkoxides seemed to work well too,
such as cyclopropylmethoxy 16b and isopropylmethoxy
16c. In comparison, larger alkoxide 16d and phenoxide
16e led to slightly weaker analogues.
In summary, we have investigated structural modifica-
tions to reach both the kinase specificity and ribose pock-
ets of JNK1 using our proprietary 4-aminopyridine
carboxamide template. Although several modifications
were identified as equivalent in terms of in vitro and cel-
lular potency to lead compound 2, we have yet to find
the extensions that optimize interaction with either the
kinase specificity pocket or the ribose pocket. The in-
sights gained from this work will facilitate future optimi-
zation efforts in other domains of the lead compounds.
Due to the relative ease of synthesis of 6-amino ana-
logues, more functional groups were explored with this
set of modification. Similar to the ether analogues, small
hydrophobic amines were favored to give inhibitors
(17a–d) with double-digit nanomolar IC50 values. The
benzylamine and phenethyl amine derivatives (17e–f)
showed reduced potency against JNK1&2. Neutral po-
lar extensions (17g–l) off the ethylamine were generally
well tolerated, such as urea, amides, alcohol, and ether.
Basic groups, such as primary amine (17n) or tertiary
amine (17o) off the ethylamine, led to analogues with
substantially reduced inhibitory potency. Alpha-substit-
uents off the 6-amino group were introduced to provide
a means for accessing the sugar pocket in the presence of
hydrophilic terminal groups. (S)-alaninol derivative 17p
is among the most potent analogues within the series,
while the (R)-enantiomer (17q) is 6-fold less potent. Fur-
ther extension with (S)-leucinol (17r) led to 50-fold drop
of potency, presumably due to unfavorable steric inter-
action. A symmetrical diol branch led to a fairly potent
analogue 17m. For 6-aminopyridine analogues, there
seemed to be a larger discrepancy between in vitro
IC50s and their cellular activity except the more lipophil-
ic 17a, 17c–d. Increased hydrophilicity could potentially
hinder the effective diffusion of the inhibitors into
HepG2 cells.
Acknowledgment
The authors thank Dr. Charles Hutchins for generating
the graphic shown in Figure 2.
References and notes
1. Dresner, A.; Laurent, D.; Marcucci, M.; Griffin, M. E.;
Dufour, S.; Cline, G. W.; Slezak, L. A.; Andersen, D. K.;
Hundal, R. S.; Rothman, D. L.; Petersen, K. F.; Shulman,
G. I. J. Clin. Invest. 1999, 103, 253.
2. Virkamaki, A.; Ueki, K.; Kahn, C. R. J. Clin. Invest. 1999,
103, 931.
3. Aguirre, V.; Uchida, T.; Yenush, L.; Davis, R. J.; White,
M. F. J. Biol. Chem. 2000, 275, 9047.
4. Aguirre, V.; Werner, E. D.; Giraud, J.; Lee, Y.; Shoelson,
S. E.; White, M. F. J. Biol. Chem. 2002, 277, 1531.
5. Hirosumi, J.; Tuncman, G.; Chang, L.; Gorgun, C. Z.;
Uysal, K. T.; Maeda, K.; Karin, M.; Hotamisligil, G. S.
Nature 2002, 420, 333.
6. Manning, A. M.; Davis, R. J. Nat. Rev. Drug Disc. 2003,
2, 554.
7. Liu, G.; Rondinone, C. M. Curr. Opin. Investig. Drugs
2005, 6, 979.
8. Szczepankiewicz, B. G.; Kosogof, C.; Nelson, L. T. J.; Liu,
G.; Liu, B.; Zhao, H.; Serby, M. D.; Xin, Z.; Liu, M.;
Gum, R. J.; Haasch, D. L.; Wang, S.; Clampit, J. E.;
Johnson, E. F.; Lubben, T. H.; Stashko, M. A.; Olejnic-
zak, E. T.; Sun, C.; Dorwin, S. A.; Haskins, K.;
There are two possible reasons behind the challenges
associated with targeting the gatekeeper residue of
JNK kinases. As mentioned earlier, the gatekeeper resi-
due in the ATP-binding pocket has been shown to con-
trol kinase selectivity to a wide range of structurally
unrelated compounds. This residue is conserved as a
threonine or larger amino acids in the human kinome,
and structural analysis has shown that the size of this
gatekeeper residue restricts access to a pre-existing

5730
G. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5723–5730
Abad-Zapatero, C.; Fry, E. H.; Hutchins, C. W.; Sham,
H. L.; Rondinone, C. M.; Trevillyan, J. M. J. Med. Chem.
2006, 49, 3563.
13. Swahn, B.-M.; Xue, Y.; Arzel, E.; Kallin, E.; Magnus, A.;
Plobeck, N.; Viklund, J. Bioorg. Med. Chem. Lett. 2006,
16, 1397.
9. Zhao, H.; Serby, M. D.; Xin, Z.; Szczepankiewicz, B. G.;
Liu, M.; Abad-Zapatero, C.; Fry, E. H.; Johnson, E. F.;
Wang, S.; Pederson, T.; Gum, R. J.; Haasch, D. E.;
Clampit, J. E.; Rondinone, C.; Trevillyan, J. M.; Sham, H.
L.; Liu, G. J. Med. Chem. 2006, 49, 4455.
10. Blencke, S.; Zech, B.; Engkvist, O.; Greff, Z.; Orfi, L.;
Horvath, Z.; Keri, G.; Ullrich, A.; Daub, H. Chem. Biol.
2004, 11, 691.
14. Final compounds were isolated using either flash
chromatography on silica gel or reversed-phase HPLC
on a Waters C18 column with a gradient of 5% to
100% acetonitrile in 0.1% aqueous trifluoroacetic acid.
All analogues were is full agreement with the
proposed structures by 1H NMR, MS, and HPLC
(>95% purity).
15. Mittelbach, M.; Junek, H. Liebigs Ann. Chem. 1986, 3,
533.
16. Wolter, M.; Nordmann, G.; Job, G. E.; Buchwald, S. L.
Org. Lett. 2002, 4, 973.
11. Scapin, G.; Patel, S. B.; Lisnock, J.; Becker, J. W.;
LoGrasso, P. V. Chem. Biol. 2003, 10, 705.
12. Swahn, B.-M.; Huerta, F.; Kallin, E.; Malmstro¨m, J.;
¨
Weigelt, T.; Viklund, J.; Womack, P.; Xue, Y.; Ohberg, L.
Bioorg. Med. Chem. Lett. 2005, 15, 5095.
17. Liu, Y.; Shah, K.; Yang, F.; Witucki, L.; Shoket, K. M.
Bioorg. Med. Chem. 1998, 6, 1219.