Novel, potent and selective anilinoquinazoline and anilinopyrimidine inhibitors of p38 MAP kinase

Article abstract of DOI:10.1016/j.bmcl.2004.08.007

SAR studies led to the identification of 4-(3-benzoylamino-6-methyl- anilino)quinazolines as potent and selective inhibitors of p38 MAP kinase. Further optimisation led to the identification of a series of 4-(3-benzoylamino-6-methyl-anilino)pyrimidines as potent inhibitors of the p38 MAP kinase signalling pathway in vitro and in vivo. SAR studies led to the identification of 4-(3-benzoylamino-6-methyl-anilino)quinazolines as potent and selective inhibitors of p38 MAP kinase. Further optimisation led to the identification of a series of 4-(3-benzoylamino-6-methyl-anilino)pyrimidines as potent inhibitors of the p38 MAP kinase signalling pathway in vitro and in vivo.

Full text of DOI:10.1016/j.bmcl.2004.08.007

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5389–5394
Novel, potent and selective anilinoquinazoline and
anilinopyrimidine inhibitors of p38 MAP kinase
John G. Cumming,* Caroline L. McKenzie, Stuart G. Bowden, Douglas Campbell,
David J. Masters, Jason Breed and Philip J. Jewsbury
AstraZeneca, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK
Received 18 May 2004; revised 30 July 2004; accepted 5 August 2004
Available online 3 September 2004
Abstract—SAR studies led to the identification of 4-(3-benzoylamino-6-methyl-anilino)quinazolines as potent and selective inhibi-
tors of p38 MAP kinase. Further optimisation led to the identification of a series of 4-(3-benzoylamino-6-methyl-anilino)pyrimidines
as potent inhibitors of the p38 MAP kinase signalling pathway in vitro and in vivo.
Ó 2004 Elsevier Ltd. All rights reserved.
Rheumatoid arthritis (RA) is a chronic disease causing
joint pain, swelling and stiffness. Disease progression is
associated with destruction of cartilage and bone lead-
ing to serious disability. Recently, advances in the treat-
ment of RA have been made with the introduction of
biological disease modifying anti-rheumatic drugs
(DMARDs) that sequester the pro-inflammatory cyto-
kine TNF-a.1 Our aim was to identify an orally active
small molecule inhibitor of TNF-a and other pro-
inflammatory cytokines such as IL-1.
F
O
S
N
Cl
HN
N
MeO
MeO
N
H
N
N
1
2
We began our programme by screening a library of 6,7-
dimethoxy-4-anilinoquinazolines, typified by compound
2, which had been synthesised as part of a programme
aimed at finding inhibitors of the epidermal growth
factor receptor tyrosine kinase (EGFR).5 Compound
2 is a potent inhibitor of EGFR (IC50 5nM) and a
weak inhibitor of p38 (IC50 0.56lM). Over 300
analogues containing various combinations of substitu-
ents on the aniline phenyl ring were screened for their
ability to inhibit the kinase activity of human p38a.6
Only one compound, the 3-benzoylamino analogue 3a
(Table 1), showed significant potency against p38 and
selectivity with respect to EGFR. We proceeded with a
study of the SAR around the aniline portion of com-
pound 3a.
The mitogen-activated protein (MAP) kinase p38 was
identified as a member of a cell signalling pathway that
mediates the release of TNF-a and IL-1b from mono-
cytes stimulated with lipopolysaccharide (LPS).2 The
observation that the small molecule inhibitor of p38
MAP kinase SB 203580 (1) can reduce LPS-induced
TNF-a levels in vitro and in vivo has led to numerous
efforts to identify p38 inhibitors with potential for use
as DMARDs.3 In addition, research into the role of
p38 indicates that an inhibitor might have other clinical
uses, such as in inflammatory bowel disease (IBD), pso-
riasis, stroke, myocardial ischaemia, Alzheimerꢀs disease
and respiratory disease.3e,4
The anilinoquinazolines 37 were prepared by a three step
sequence (Scheme 1): reaction of the appropriate
3-nitroaniline 4 with 4-chloro-6,7-dimethoxyquinazoline
(5), reduction of the nitro then acylation of the ani-
line. Alternatively the steps could be reversed. Where
appropriate, a symmetrically substituted 1,3-phenyl-
ene diamine 6 was used, omitting the nitro reduction
step.
Keywords: p38 MAP kinase inhibitors.
*
Corresponding author. Tel.: +44 1625 517755; fax: +44 1625
510823; e-mail: john.cumming@astrazeneca.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.007

5390
J. G. Cumming et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5389–5394
Table 1. p38a, EGFR and HWB inhibition of selected anilinoquinazolines
Compd
R2
R4
R6
X
p38a inhibition IC50, lMa
EGFR inhibition IC50, lM
HWB inhibition IC50, lMa
1
––
––
H
H
H
H
H
H
F
––
––
H
––
––
H
––
––
0.019 ( 0.001)
0.56 ( 0.15)
>33
0.005
5.6
2.0 ( 0.5)
>50
>50
2
3a
3b
3c
3d
3f
3g
3h
3i
H
0.141 ( 0.025)
0.054 ( 0.015)
0.088 ( 0.025)
0.078 ( 0.016)
0.518 ( 0.057)
2.09 ( 0.22)
F
F
H
0.49
11
>50
>50
H
Cl
Me
F
H
H
H
>33
nt
21 ( 2)
>50
>50
Cl
Me
H
H
H
H
3.2
2.8
H
H
0.69 ( 0.17)
>50
>50
H
H
H
H
H
H
Cl
H
H
H
0.615 ( 0.094)
0.212 ( 0.053)
0.047 ( 0.002)
0.031 ( 0.009)
1.69 ( 0.49)
2.2
3j
H
NMe2
NMe2
NMe2
NMe2
N-morpholino
>33
29
>50
>50
3k
3l
H
Cl
Me
H
H
>33
1.0
35 ( 9)
>50
16 ( 8)
3m
3n
Cl
H
Me
0.010 ( 0.001)
>33
a Values are geometric means of three experiments, standard deviation is given in parentheses (nt = not tested).
Again, the 6-chloro (3k) and 6-methyl (3l) analogues
were superior to the unsubstituted (3j) and 4-chloro
(3m) anilines in terms of p38 potency and selectivity.
Compounds 3a–m were then tested for their ability to
inhibit LPS-stimulated TNF-a production in human
whole blood (HWB).6 The analogues with a methyl in
the 6-position showed weak activity in this assay (e.g.,
3l, IC50 35lM), while the other compounds were not
active up to 50lM. This contrasts with compound 1,
which has IC50 values of 19nM against p38 and
2.0lM in the HWB assay. As investigations proceeded
in the bisamide series it was discovered that replacement
of the dimethylamino group with morpholino resulted in
compounds with improved HWB activity8 and so the
corresponding anilinoquinazoline hybrid 3n was pre-
pared. The p38 activity increased 3-fold and the activity
in the whole blood assay was slightly improved but still
poor (IC50 16lM). The physicochemical properties of 3n
are less than ideal as it is lipophilic (logD7.4 = 3.6) and
highly protein bound (98% in rat plasma) with low aque-
ous solubility (7lM). We reasoned that the addition of
side-chains containing basic amine groups5 would im-
prove the physical properties, and a reduction in plasma
protein binding would be reflected in an improvement in
whole blood activity.
Cl
N
MeO
MeO
R
N
5
NO2
H2N
a, b, c
or c, b, a
4
6
R4
NH
R6
HN
R
a, c
R2
or c, a
X
MeO
MeO
N
O
H2N
NH2
N
3
i
Scheme 1. Reagents and conditions: (a) 5, PrOH, HCl, 85–90°C; (b)
Fe, AcOH/H2O, 110°C; (c) ArCOCl, Et3N, CH2Cl2, rt.
The results (Table 1, compounds 3a–i) demonstrated
that substitution of the aniline ring with methyl or
chloro in the 4-position (3g, 3i, 3f) or with fluoro in
the 2-position (3h) led to a decrease in p38 activity rela-
tive to the parent 3a. The 4,6-di-fluoro analogue 3b had
improved p38 activity but its selectivity over EGFR
dropped to around 10-fold. The 6-chloro (3c) and 6-
methyl (3d) compounds showed good p38 activity
(<100nM) and reduced EGFR activity, the selectivity
of 3d being better than 400-fold.
The synthesis of the target compounds 9 (Scheme 2)
started from the 6-acetyloxyquinazoline intermediate 8,
which was synthesised according to the route outlined
in Scheme 1.
The SAR that was emerging showed close parallels with
that found in another chemical series, the bisamides
such as 7, which was under investigation in our labora-
tories.8 We decided to investigate this parallel SAR fur-
ther by preparing the ꢁhybridꢀ compounds 3j–m
containing the 3-dimethylaminobenzamide moiety
found in 7.
The incorporation of the basic side-chains maintained
the p38 activity and did lead to an increase in whole
blood activity (Table 2), the most active compound 9a
giving an HWB IC50 of 3.5lM. There was also a dra-
matic increase in aqueous solubility (240lM) and a
reduction in protein binding (83% in rat plasma),
although logD7.4 remained high (3.5).
N
OH
Our SAR studies indicated that a variety of substituents
could be tolerated at the 5- and the 6-positions of the
quinazoline, suggesting that this part of the molecule
was pointing out towards the solvent in the binding
pocket. This was later confirmed by our binding mode
H
N
H
N
O
O
7

J. G. Cumming et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5389–5394
5391
appropriate amine. Compounds with side-chains linked
via oxygen at the 2-position were prepared by reaction
of 11a with the appropriate alcohol in the presence of
potassium t-butoxide, but 10a failed to react under these
conditions.
O
NH
HN
N
O
N
O
N
O
N
MeO
8
The results for representative examples are shown in
Table 3.
a, b
Replacement of the quinazoline with pyrimidin-4-yl
(10b) or 6-chloropyrimidin-4-yl (10a) led to a reduction
in p38 activity. In contrast, the 2-chloropyrimidin-4-yl
analogue 11a maintained the p38 potency and also gave
a respectable whole blood IC50 of 11lM. Analogues
with 6-N-linked side-chains (such as 10c–e) were less
potent as p38 inhibitors. Pyrimidine analogues substi-
tuted in the 2-position with various alkoxy and amino
groups had good p38 activity (11b–j). The optimal sub-
stitution was found to be N-alkyl-piperidin-4-yl-oxy
(11i,j), which gave sub-micromolar activity in whole
blood. Analogues of several of the most interesting
compounds were prepared substituting a 3-morpholino-
phenyl (e.g., 12) or 3-morpholino-5-fluorophenyl (e.g.,
13) for the 2-morpholinopyridin-4-yl, with no significant
effect on the p38 or HWB potencies (compare 11j and
12, Table 3).
O
NH
HN
N
R
O
N
O
N
R
N
MeO
N
9
Scheme 2. Reagents and conditions: (a) NH3, MeOH, 50°C; (b)
R2NCH2CH2ClÆHCl, Cs2CO3, DMA, 100°C.
Table 2. p38a and HWB inhibition of selected anilinoquinazolines
Compd NR2
p38a inhibition
IC50, lMa
HWB inhibition
IC50, lMa
9a
9b
9c
9d
9e
(iPr)2N
0.052 ( 0.009)
0.024 ( 0.008)
0.039
3.5 ( 2.5)
7.8 ( 1.9)
6.3
Morpholin-4-yl
Pyrrolidin-1-yl
(Me)2N
0.035 ( 0.003)
0.056 ( 0.010)
6.9 ( 1.8)
4.6 ( 2.6)
Piperidin-1-yl
a Values are geometric means of three experiments, standard deviation
is given in parentheses.
O
HN
NH
N
N
O
X
N
studies (vide infra). This led us to ask if the quinazoline
could be replaced by a pyrimidine, with the advantage of
reducing the size of the molecule, which should lead to
an improvement in its physicochemical properties. The
6-chloropyrimidine analogue 10a was prepared by reac-
tion of the elaborated aniline with 4,6-dichloropyrim-
idine (Scheme 3).9 Palladium-catalysed reduction using
ammonium formate gave 10b. Reaction of the aniline
with 2,4-dichloropyrimidine proceeded regioselectively
to give 11a.10 Basic side-chains linked via nitrogen at
the 2- or 6-positions were incorporated by reaction of
the chloropyrimidine intermediate 10a or 11a with the
Y
12 X = (1-nPr-piperidin-4-yl)-O-; Y = H
13 X = SMe; Y = F
In order to investigate the selectivity of the two series for
p38 the representative anilinoquinazoline 9d and
anilinopyrimidine 12 were evaluated against a panel of
24 protein kinases.11 Both compounds showed excellent
selectivity for p38a and p38b, with no activity seen
against the other enzymes at 10lM.12
The more potent compounds in HWB from both series
were evaluated for oral pharmacokinetics in the rat
using a cassette dosing protocol (compounds were dosed
at ꢀ2mg/kg in a propyleneglycol formulation, n = 2 ani-
mals, data not shown).13 No plasma levels were seen
with compounds containing basic groups, however neu-
tral anilinopyrimidines 11a and 13 did show exposure.
Plasma clearances of representative basic compounds
11f and 12 were determined by i.v. dosing and found
to be greater than rat hepatic blood flow, suggesting that
oral bioavailability is being limited by first-pass elimina-
tion. Rat PK i.v. and oral PK profiles were determined
for 13 (Fig. 1). The clearance was 25mL/min/kg (i.e.,
about one third hepatic blood flow), volume of distribu-
tion at steady state was 0.8L/kg and oral bioavailability
was 26%. Compound 13 was selected for in vivo testing
in a rat LPS-stimulated TNF-a inhibition model and
showed significant activity 2h after dosing orally at
10mg/kg (77 4% and 51 8% inhibition in two sepa-
rate experiments).14
O
HN
N
O
H2N
NH
NH
N
N
a
O
N
O
N
N
R
b
10a (R = Cl)
10b (R = H)
10c-e
c
d
HN
NH
O
N
N
O
N
N
R
11a (R = Cl)
d or e
11b-j
Scheme 3. Reagents and conditions: (a) 4,6-dichloropyrimidine,
nBuOH, iPr2NEt, 120°C; (b) 2,4-Dichloropyrimidine, nBuOH, iPr2-
NEt, 120°C; (c) HCO2NH4, Pd/C, EtOH, rt; (d) R1R2NH, Pr2NEt,
i
nBuOH, 100°C; (e) R3OH, KOtBu, BuOH, 140°C.
t

5392
J. G. Cumming et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5389–5394
Table 3. p38a and HWB inhibition of selected compounds
Compd
R
p38a Inhibition IC50, 5lMa
HWB inhibition IC50, lMa
10a
10b
10c
10d
10e
11a
11b
11c
11d
11e
11f
11g
11h
11i
11j
12
Cl
H
0.244 ( 0.030)
0.217 ( 0.011)
0.287 ( 0.048)
0.456 ( 0.058)
1.220 ( 0.650)
0.019 ( 0.002)
0.098 ( 0.015)
0.058 ( 0.004)
0.083 ( 0.008)
0.111 ( 0.007)
0.079 ( 0.008)
0.037 ( 0.005)
0.033 ( 0.005)
0.045 ( 0.004)
0.055 ( 0.007)
0.049 ( 0.007)
0.014 ( 0.001)
>50
27 ( 13)
23 ( 9)
>50
Et2N(CH2)3NH–
(1-pyrrolidinyl)-(CH2)3NH–
4-Me-homopiperazin-1-yl
Cl
>50
11 ( 3)
>50
(1-morpholinyl)-(CH2)3NH–
(1-pyrrolidinyl)-(CH2)3NH–
(1-Me-piperidin-4-yl)-N(Me)–
Allyl-N(Me)–
16 ( 4)
5.3 ( 0.3)
>50
Me2NCH2C(Me)2CH2O–
Me2N(CH2)3O–
3.7 ( 1.9)
3.2 ( 1.7)
1.1 ( 0.4)
0.50 ( 0.27)
0.41 ( 0.21)
0.30 ( 0.06)
3.4 ( 1.5)
(1-Me-piperidin-3-yl)-CH2O–
(1-Me-piperidin-4-yl)-O–
(1-nPr-piperidin-4-yl)-O–
––
13
––
a Values are geometric means of three experiments, standard deviation is given in parentheses.
10.000
i.v.
p.o.
1.000
0.100
0.010
0.001
0
2
4
6
8
10
12
Time (h)
Figure 1. PK profile for compound 13 in rat; dosing at 1mg/kg i.v.
(n = 2) and at 1mg/kg p.o. (n = 2).
Extrapolating from the PK study gives a plasma concen-
tration of 13 at 2h after a 10mg/kg dose of approxi-
mately 1lM, assuming dose linearity. Assuming
equivalent potency in rat and human whole blood this
suggests that a plasma concentration of approximately
one third of the in vitro IC50 is sufficient to give = 50%
inhibition in the rat in vivo LPS model. Enhanced sensi-
tivity in LPS challenge models has been suggested for
other p38 inhibitors.15
Figure 2. Crystal structure of p38 with compound 9c bound. The
electron density of 9c is shown in purple. Protein residues 71, 106 to
110 and 168 to 170 (the DFG motif) are shown and hydrogen bonds to
9c are indicated with dashed lines.
Met109. Other orientations of the ligands in the ATP
binding site also failed to provide convincing structures.
There have been several reports on the use of X-ray crys-
tallographic studies to determine the binding mode of
p38 inhibitors.16 Compounds structurally similar to 1
bind to the ATP binding site of unactivated p38a
MAP kinase with the pyridine nitrogen making a hydro-
gen bond to the backbone amide NH of Met109.16b This
hydrogen bond is also seen between the corresponding
residue in other protein kinases and N1 of the adenine
ring of ATP, and also the N1 of the quinazoline ring
of compounds such as 2.17 During the early stage of
our programme, we performed modelling studies using
the published X-ray crystal structure of apo-p38,18 but
were unable to accommodate structures 3 and 9 into
the ATP binding site while maintaining a hydrogen
bond between the quinazoline N1 and the amide of
Subsequently we co-crystallised unactivated p38a with
inhibitor 9c (Table 2, Scheme 2) and derived an X-ray
˚
crystal structure at 2.2A resolution (Figs. 2 and 3).
The structure reveals that the quinazoline N1 also forms
a hydrogen bond with the backbone amide NH of
Met109 (Fig. 2) and the aniline ring overlaps the aryl
ring of 1 (4-F-Ph in SB 203580, 3-I-Ph in this struc-
ture),16b buried into the hydrophobic region formed by
the carbon atoms of residues including Lys53 and
Thr106 (Fig. 3). However, in our structure, there are fur-
ther hydrogen bonds between the amide NH of 9c and
the carboxylate side chain of Glu71, and between the
amide carbonyl of 9c and the backbone amide NH of
Asp168. The 2-pyrrolidinylethoxy group of 9c points

J. G. Cumming et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5389–5394
5393
physicochemical properties and potency in human
whole blood. Further optimisation led to 2-alkoxy-4-
anilinopyrimidines, such as 12, with sub-micromolar po-
tency in whole blood, however these analogues showed
high clearance and poor bioavailability in rats. Oral
exposure in rats was seen with neutral analogues, and
the most potent of these in whole blood (13, IC50
3.4lM) showed reasonable rat pharmacokinetics and
activity in an in vivo model at 10mg/kg p.o. X-ray crys-
tallographic studies have revealed the binding mode,
which shows a large movement of the DFG motif to
accommodate the aniline 5-substituent. These series of
inhibitors show potential for the development of an oral
treatment of rheumatoid arthritis and other inflamma-
tory diseases.
Acknowledgements
The authors thank our colleagues Lucy Ashman, Carol
Biddulph, Karen Burns, Chris Heapy, Brian Ellershaw
and Janet Hayward for developing and carrying out
the biological assays, Giles Hassall for producing the
protein used in the crystallographic studies, Jon Read
for providing Figures 2 and 3 and Ken Page for per-
forming the PK studies.
Figure 3. Overlay of complexes of compound 9c (blue ball and stick,
yellow protein backbone with the side-chains of residues Thr106,
Lys53, Glu71 and Phe169 shown in yellow ball and stick) and SB
203580 (orange with orange protein backbone)16b showing induced
movement of DFG motif to accommodate 9c.
out towards the solvent and has weak electron density in
the structure suggesting it has unrestricted movement.
The SAR studies described above show that this group
does not contribute to kinase activity. The most signifi-
cant feature of the structure is the large induced move-
ment of the conserved kinase DFG motif (residues
168–170 in p38a) to accommodate the aniline 5-substit-
uent. A previously reported X-ray structure of an ani-
linoquinazoline structurally similar to 2 bound to p38
does not include this conformational change.17 At the
time this induced fit was unprecedented and accounts
for the failure of the docking studies. Similar induced
movements have since been reported for other chemical
series bound to p38,3b,16c and for other kinase com-
plexes,19 and been associated with slow binding kinet-
ics,16c suggesting the structural rearrangement required
for the binding event is rare. No detailed kinetic analyses
have been done for this series of compounds though
slow binding kinetics might be expected by analogy.
References and notes
1. (a) Etanercept: Jarvis, B.; Faulds, D. Drugs 1999, 57, 945;
(b) Infliximab: Maini, R.; St. Clair, E. W.; Breedveld, F.;
Furst, D.; Kalden, J.; Weisman, M.; Smolen, J.; Emery,
P.; Harriman, G.; Feldmann, M.; Lipsky, P. Lancet 1999,
354, 1932.
2. Lee, J. C.; Laydon, J. T.; McDonnell, P. C.; Gallagher, T.
F.; Kumar, S.; Green, D.; McNulty, D.; Blumenthal, M.
J.; Heys, J. R.; Landvatter, S. W.; Strickler, J. E.;
McLaughlin, M. M.; Siemens, I. R.; Fisher, S. M.; Livi,
G. P.; White, J. R.; Adams, J. L.; Young, P. R. Nature
1994, 372, 739.
3. (a) Boehm, J. C.; Adams, J. L. Exp. Opin. Ther. Patents
2000, 10, 25; (b) Regan, J.; Capolino, A.; Cirillo, P. F.;
Gilmore, T.; Graham, A. G.; Hickey, E. R.; Kroe, R. R.;
Madwed, J.; Moriak, M.; Nelson, R.; Pargellis, C. A.;
Swinamer, A.; Torcellini, C.; Tsang, M.; Moss, N. J. Med.
Chem. 2002, 45, 2994; (c) Jackson, P. F.; Bullington, J. L.
Curr. Top. Med. Chem. 2002, 2, 1011; (d) Cirillo, P. F.;
Pargellis, C.; Regan, J. Curr. Top. Med. Chem. 2002, 2,
1021; (e) Kumar, S.; Boehm, J.; Lee, J. C. Nature Rev.
Drug Disc. 2003, 2, 717.
4. (a) Lee, J. C.; Kassis, S.; Kumar, S.; Badger, A.; Adams, J.
L. Pharmacol. Ther. 1999, 82, 389; (b) Armstrong, S. C.
Cardiovascular Res. 2004, 61, 427; (c) de Boer, W. Expert
Opin. Invest. Drugs 2003, 12, 1067; (d) Hele, D. J.; Belvis,
M. G. Expert Opin. Invest. Drugs 2003, 12, 5; (e) Irusen,
E.; Matthews, J. G.; Takahashi, A.; Barnes, P. J.; Chung,
K. F.; Adcock, I. M. J. Allergy Clin. Immunol. 2002, 109,
649.
Although attempts to co-crystallise p38 with an anilino-
pyrimidine inhibitor have to date been unsuccessful, it
appears likely that the binding mode will be similar to
that of 9c, at least in the benzoylaminoaniline portion.
However, the observed SAR (Table 3) suggests that
the pyrimidine cannot simply overlay onto the pyrim-
idine portion of 9c since there is a preference for 2- over
6-substitution, and 2-substitution on the quinazolinone
cannot be accommodated in the 9c binding mode. In-
stead it is likely that the pyrimidine ring in compounds
11b–j is rotated 180° around the N1 nitrogen to expose
the R group towards solvent.
5. Barker, A. J.; Gibson, K. H.; Grundy, W.; Godfrey, A. A.;
Barlow, J. J.; Healy, M. P.; Woodburn, J. R.; Ashton, S.
E.; Curry, B. J.; Scarlett, L.; Henthorn, L.; Richards, L.
Bioorg. Med. Chem. Lett. 2001, 11, 1911.
6. For in vitro assay experimental procedures see Ref. 7.
7. Cumming, J. G. WO 0020402, 2000; Chem. Abstr. 2000,
132, 265207.
In summary, SAR studies around initial lead 3a led to
the discovery of two series of potent p38 inhibitors with
excellent selectivity over other kinases. Incorporation of
basic side-chains gave compounds with improved

5394
J. G. Cumming et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5389–5394
8. Brown, D. S.; Belfield, A. J.; Brown, G. R.; Campbell, D.;
obtained 60min after LPS challenge and plasma isolated
and stored at À80°C. TNFa concentration was measured
by ELISA (R&D Systems rat TNF-a Quantikine kit,
catalogue no. SRTA00). For each compound/dose the
percentage inhibition of TNF-a was calculated as a
percentage of controls.
Foubister, A.; Masters, D. J.; Pike, K. G.; Snelson, W. L.;
Wells, S. L. Bioorg. Med. Chem. Lett. 2004, 14, previous
communication in this issue. doi:10.1016/j.bmcl.2004.
08.006.
9. Cumming, J. G. WO 0127089, 2000; Chem. Abstr. 2000,
134, 295834.
10. Breault, G. A.; Ellston, R. P. A.; Green, S.; James, S. R.;
Jewsbury, P. J.; Midgley, C. J.; Pauptit, R. A.; Minshull,
C. A.; Tucker, J. A.; Pease, J. E. Bioorg. Med. Chem. Lett.
2003, 13, 2961.
11. For further information on kinases and assay protocols
see: Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P.
Biochem. J. 2000, 351, 95.
12. Work carried out on behalf of AstraZeneca by Sir Philip
Cohen, Division of Signal Transduction Therapy, Univer-
sity of Dundee, Dundee, UK. The kinase panel used was
the Core panel shown in Table 1 of Ref. 11.
13. For an example of this type of method see: Korfmacher,
W. A.; Cox, K. A.; Ng, K. J.; Veals, J.; Hsieh, Y.;
Wainhaus, S.; Broske, L.; Prelusky, D.; Nomeir, A.; White,
R. E. Rapid Commun. Mass Spectrom. 2001, 15, 335.
14. Procedure for rat in vivo LPS model: Female Alderley
Park rats (five per group) were dosed orally with the test
compound (10mg/kg; formulated in 1:1 v/v PEG/water;
0.2mL per 100g body weight) 60min prior to challenge
with LPS administered intravenously [LPS E.Coli
0111:B4; Sigma L-4130; 30lg in 0.2mL sterile physiolog-
ical saline (Phoenix Pharma Ltd)]. Control animals
received vehicle alone prior to LPS challenge. Blood was
15. Zimmitti, C. S.; Shwartz, R.; Torcellini, C. A.; Pargellis,
C. A.; Madwed, J. B.; Weldon, S. M. Poster #367,
American College of Rheumatology Annual Scientific
Meeting, San Francisco, CA, 11–15 November 2001.
16. (a) Wilson, K. P.; McCaffrey, P. G.; Hsiao, K.; Pazhani-
samy, S.; Galullo, V.; Bemis, G. W.; Fitzgibbon, M. J.;
Caron, P. R.; Murcko, M. A.; Su, M. S.-S. Chem. Biol.
1997, 4, 423; (b) Tong, L.; Pav, S.; White, D. M.; Rogers,
S.; Crane, K. M.; Cywin, C. L.; Brown, M. L.; Pargellis,
C. A. Nat. Struct. Biol. 1997, 4, 311; (c) Pargellis, C.;
Tong, L.; Churchill, L.; Cirillo, P. F.; Gilmore, T.;
Graham, A. G.; Grob, P. M.; Hickey, E. R.; Moss, N.;
Pav, S.; Regan, J. Nat. Struct. Biol. 2002, 9, 268; (d)
Fitzgerald, C. E.; Patel, S. B.; Becker, J. W.; Cameron, P.
M.; Zaller, D.; Pikounis, V. B.; OꢀKeefe, S. J.; Scapin, G.
Nat. Struct. Biol. 2003, 10, 764.
17. Shewchuk, L.; Hassell, A.; Wisely, B.; Rocque, W.;
Holmes, W.; Veal, J.; Kuyper, L. F. J. Med. Chem.
2000, 43, 133.
18. Wilson, K. P.; G. W.; Fitzgibbon, M. J.; Caron, P. R.;
Griffith, J. P.; Chen, W.; McCaffrey, P. G.; Chambers, S.
P.; Su, M. S.-S. J. Biol. Chem. 1996, 271, 27696.
19. Schindler, T.; Bornmann, W.; Pellicena, P.; Miller, W. T.;
Clarkson, B.; Kuriyan, J. Science 2000, 289, 1938.