Discovery of 6-(2,4-difluorophenoxy)-2-[3-hydroxy-1-(2-hydroxyethyl) propylamino]-8-methyl-8 H -pyrido[2,3- d ]pyrimidin-7-one (pamapimod) and 6-(2,4-difluorophenoxy)-8-methyl-2-(tetrahydro-2 H -pyran-4-ylamino)pyrido[2,3- d ]pyrimidin-7(8 H)-one (R1487) as orally bioavailable and highly selective inhibitors of p38α mitogen-activated protein kinase

Article abstract of DOI:10.1021/jm101423y

The development of a new series of p38α inhibitors resulted in the identification of two clinical candidates, one of which was advanced into a phase 2 clinical study for rheumatoid arthritis. The original lead, an lck inhibitor that also potently inhibited p38α, was a screening hit from our kinase inhibitor library. This manuscript describes the optimization of the lead to p38-selective examples with good pharmacokinetic properties.

Full text of DOI:10.1021/jm101423y

ARTICLE
pubs.acs.org/jmc
Discovery of 6-(2,4-Difluorophenoxy)-2-[3-hydroxy-1-(2-
hydroxyethyl)propylamino]-8-methyl-8H-pyrido[2,3-d]pyrimidin-7-
one (Pamapimod) and 6-(2,4-Difluorophenoxy)-8-methyl-
2-(tetrahydro-2H-pyran-4-ylamino)pyrido[2,3-d]pyrimidin-7(8H)-one
(R1487) as Orally Bioavailable and Highly Selective Inhibitors of p38r
Mitogen-Activated Protein Kinase†
David M. Goldstein,* Michael Soth,* Tobias Gabriel, Nolan Dewdney, Andreas Kuglstatter,
Humberto Arzeno, Jeffrey Chen, William Bingenheimer, Stacie A. Dalrymple, James Dunn, Robert Farrell,
Sandra Frauchiger, JoAnn La Fargue, Manjiri Ghate, Bradford Graves,‡ Ronald J. Hill, Fujun Li,
Renee Litman, Brad Loe, Joel McIntosh, Daniel McWeeney, Eva Papp, Jaehyeon Park, Harlan F. Reese,
Richard T. Roberts, David Rotstein, Bong San Pablo, Keshab Sarma, Martin Stahl,§ Man-Ling Sung,
Rebecca T. Suttman, Eric B. Sjogren, Yunchou Tan, Alejandra Trejo, Mary Welch, Paul Weller,
Brian R. Wong, and Hasim Zecic
Roche Palo Alto LLC, 3431 Hillview Avenue, Palo Alto, California 94304, United States
S Supporting Information
b
ABSTRACT: The development of a new series of p38R inhibitors resulted in the identification of
two clinical candidates, one of which was advanced into a phase 2 clinical study for rheumatoid
arthritis. The original lead, an lck inhibitor that also potently inhibited p38R, was a screening hit
from our kinase inhibitor library. This manuscript describes the optimization of the lead to p38-
selective examples with good pharmacokinetic properties.
’ INTRODUCTION
Lead compound 114(Chart 1) was initially designed to be an
inhibitor of lymphocyte specific protein tyrosine kinase (lck), a
member of the src kinase family. The lead was very potent against
lck but also showed significant potency against p38R (lck Kd =
0.2 nM, p38R Kd = 10 nM). Crystal structures of the lead bound
to each enzyme showed that, while the molecule binds similarly
to p38R and lck, there are important differences between the two
inhibitor complexes. This paper highlights our successful efforts
to engineer specificity toward p38R, which ultimately identified
2a (R1487, Chart 1)15 and 2b (pamapimod, Chart 1)16 as highly
selective and orally bioavailable p38R inhibitors that were
advanced into human clinical studies.17-19
p38 mitogen activated protein (MAP) kinases form a sub-
family of the well-characterized MAP kinase family of serine/
threonine kinases. p38 MAP kinases are widely expressed in
endothelial, immune, and inflammatory cells and play a central
role in regulating the production of cytokines including TNF-R,
IL-1β, and IL-6.1,2 Biologic agents targeting these cytokines or
their receptors have proven efficacious for inflammatory diseases
including rheumatoid arthritis (RA), psoriasis, and inflammatory
bowel disease (IBD).3,4
The p38 subfamily of MAP kinases includes four isoforms (R,
β, γ, and δ) that are encoded by separate genes. Recent analysis
of synovial tissue extracted from RA patients suggests that the R
and γ isoforms are overproduced within inflamed tissue and may
be the preferred targets for intervention in the disease.5-7 Several
excellent reviews describing the potential importance of p38R
MAP kinase (p38R) inhibitors as novel therapeutics for inflam-
matory diseases have been published.8-11
’ RESULTS AND DISCUSSION
Chemistry. Two general routes were adopted for the synth-
eses of pyrimidinopyridones containing varied 2-, 6- and 8-sub-
stitution; representative syntheses are shown in Schemes 1 and 2.
In the first route (Scheme 1), ethyl 4-chloro-2-methylthiopyr-
imidine-5-carboxylate was reacted with various amines to afford
intermediate esters 3, which were subsequently reduced to
We previously reported the design and synthesis of
4-azaindole12 and aminopyrazole13 based p38R inhibitors. In
this report, we describe our effort to optimize a lead compound
identified through screening of our kinase inhibitor library as
another piece of our overall strategy to advance multiple p38R
inhibitors with novel scaffolds.
Received: November 4, 2010
Published: March 04, 2011
r
2011 American Chemical Society
2255
dx.doi.org/10.1021/jm101423y J. Med. Chem. 2011, 54, 2255–2265
|

Journal of Medicinal Chemistry
ARTICLE
Chart 1
Scheme 1. Synthesis of Pyrimidinopyridonesa
a Reagents and conditions: (i) R2NH2, 55-100%; (ii) LiAlH4, 84-98%; (iii) MnO2, 68-99%; (iv) ArXCH2CO2Et, base, 22-99%; (v) mCPBA or
Oxone, 30-100%; (vi) R3NH2, base, 23-100%. 5a (R2 = Me) was a common intermediate for the preparations of 2a-j and 2v-ii. 7a (R2 = Me, XAr =
2,4-difluorophenoxy) was a common intermediate for the preparations of 2a,b and 2v-ii. In some cases, final products 2 were obtained as HCl salts. X =
O, S, CH2. For X = S, step v was omitted.
Scheme 2. Alternative Synthesis of Pyrimidinopyridonesa
a Reagents and conditions: (i) thiourea, NaOMe, MeOH/EtOH, Δ, 72%; (ii) MeI, K2CO3, acetone, 95%; (iii) EtOCOCH2OAr, K2CO3, NMP, 120 °C,
50%; (iv) R2X, base or DEAD, R2OH, PBu3, 40-100%; (v) mCPBA or Oxone, 63-100%; (vi) R3NH2, base, 49-67%. In some cases, final products 2
were obtained as HCl salts. For 2k (R2 = H), step iv was omitted. 10a (R1 = 2,4-difluoro) was a common intermediate for the preparations of 2k, 2o, 2p,
2r, 2s, and 2t.
alcohols 4 and then oxidized to 4-aminopyrimidine-5-carboxal-
dehydes 5. Aldehydes 5 were condensed with appropriately
substituted acetates to afford the core pyrimidinopyridones 6.
The methylthio groups of pyrimidinopyridones 6 were oxidized
to the corresponding sulfones 7, which were displaced with
amines to afford desired compounds 2. This route was the most
general and was required for 8-substituents more complex than
straight chains.
The second route was used for some straight-chain 8-substit-
uents and for the 8-unsubstituted analogue (Scheme 2). 2-Hy-
droxymethylene-3,3-diethoxypropionitrile potassium salt was
condensed with thiourea to afford pyrimidinethiol 8, which
2256
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Figure 1. X-ray crystal structure of the lck-1 complex (pink, 2.8 Å resolution, PDB accession code 3KMM) superimposed with structures of p38R
(yellow) in complex with compound 1 (A, 2.1 Å, 3FSF) and compound 11 (B, 2.0 Å, 3FSK).
was reacted with methyl iodide to afford intermediate 9. Com-
pound 9 was condensed with ethyl 2,4-difluorophenoxyacetate
to afford the core pyrimidinopyridone 10, which could be
alkylated at the 8-position with primary or isopropylalkyl halides
(attempted reactions of pyrimidinopyridones 10 with more
complex secondary halides or alcohols under Mitsunobu condi-
tions led to O-alkylation). The final steps of the sequence were
analogous to the first route.
Chart 2
Many varied syntheses were required for individual side chains
and are detailed in the Experimental Section and Supporting
Information.
Optimization of Kinase Selectivity. Our chemistry approach
began with compound 1, a lead identified from within the Roche
kinase library. Compound 1 was initially designed to inhibit lck (lck
Kd = 0.2 nM, p38R Kd = 10 nM). While lck shares relatively low
sequence homology with p38R in the active site (41%), both
enzymes contain threonine as the gatekeeper residue (Thr106 in
p38R, Thr316 in lck) in the ATP binding pocket, which allows access
to similarly sized hydrophobic binding domains. X-ray crystal
structures of compound 1 bound to each kinase confirmed similar
binding motifs in the active sites of both enzymes (Figure 1A).
The overlay of the two cocrystal structures revealed an
opportunity to alter the kinase selectivity profile in favor of
p38R over lck by taking advantage of the larger pocket in the
lower lip portion of p38R (amino acids 111-113). The lower lip
portion of p38R is larger than that of lck (amino acids 321-324)
as a consequence of p38R lacking a residue corresponding to
Gly322 in lck (Figure 1). This glycine residue, which is present in
345 human kinases, adopts a conformation favorable to an
intramolecular hydrogen bond with the hinge portion of the
protein (amino acids 317-320 in lck, amino acids 107-110 in
p38R), reducing the overall size of the binding pocket.
Exchange of the aniline moiety of compound 1 with a saturated
alkylamine increased the three-dimensional bulk of the molecule
and caused unfavorable steric interactions between the inhibitor
and the lower lip portion of lck kinase. This is exemplified by the
structure of p38R complexed with compound 1120 (Chart 2,
Figure 1B). Compound 11 showed dramatically improved
selectivity for p38R over lck (p38R Kd = 32 nM, lck Kd > 20 μM).
In addition to this modification, the central pyrimidinopyr-
imidone core was converted to a pyrimidinopyridone, as it was
found that pyrimidinopyridones were generally more potent for
p38R. The sum of these changes, exemplified by compound 2c,
effectively inverted a potent tyrosine kinase inhibitor class into
one with improved selectivity versus the lck family tyrosine
kinases and increased intrinsic potency for p38R (p38R Kd =
0.3 nM, lck Kd = 1800 nM). However, while these modifications
did improve selectivity, kinase selectivity profiling of compound
2c indicated that this compound still hits a large number of
kinases (Figure 2). Pan-kinase selectivity was a general problem
for this class of compounds. Compound 2c was further profiled
in a 14-day toxicity study and showed toxicity in spleen and
kidney, as well as infections. We assumed that much of the
toxicity was likely due to off-target kinase effects.
To further improve selectivity for p38R over other kinases, we
turned to optimal space-filling of the backpocket. The size of the
backpocket of kinases is defined by the gatekeeper residue, which
in p38R is Thr106 (Figure 3A). The gatekeeper is a larger residue
in many other kinases, thus offering the opportunity to design
selectivity for p38R by optimal filling of this backpocket.
Examination of the p38R-bound crystal structure of our pre-
viously disclosed azaindole inhibitor 12 (Chart 2),12 which does
effectively fill the backpocket (Figure 3B), inspired a simple
modification of our newer series.
A comparison of the crystal structures for 6-phenylpyridone 2c
and azaindole 12 clearly shows that the directly linked phenyl
side chain of 2c is not filling the same hydrophobic volume as the
phenyl side chain of 12 (Figure 3A,B). 6-Phenoxypyridones 2a
and 2d were subsequently designed to position the phenoxy ring
in the same space as the phenyl ring of 12. X-ray crystallography
confirmed that the binding motif enabled the phenoxy
2257
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Figure 2. Evolution of the series with respect to kinase selectivity. The kinase dendrograms represent % ligand displacement at 10 μM test compound
using the Ambit Biosciences kinase screen of >300 kinases; each branch corresponds to a different kinase. Large size dots represent 99-100%
displacement. Medium size dots represent 91-99% displacement, and small size dots represent 51-91% displacement.
the p-fluoro atom on the phenoxy moiety of 2a is within 3.2 and
3.1 Å of the peptide carbonyl carbons of L104 and V105,
respectively. This short distance may indicate a weak attractive
interaction, the like of which has been described and carefully
analyzed previously.21,22 Comparison of the in vitro enzyme IC50
of the nonfluorinated analogue 2d with the IC50 of 4-flouro
compound 2e (binding mode shown in Figure 3C) shows that
the 4-fluorine increases ligand potency about 8-fold (Table 2).
However, a similar comparison of the 2,4-difluoro example 2a to
the 2-fluoro analogue 2g does not show the same boost in
potency, suggesting that multiple factors influence the SAR for
this side chain.
Most importantly, the kinase selectivity of 6-phenoxypyridone
2a is greatly improved relative to the 6-phenylpyridones. After
initial screening of the compounds at 10 μM versus >300
kinases,23 the kinases inhibited were followed up with Kd
determinations (10-point curve) for confirmation of inhibition
and refinement of the selectivity profile. In the case of compound
2a, detailed Kd determinations on the kinases inhibited in the
single point screen confirmed submicromolar inhibition for only
10 out of >300 kinases tested (Figure 2, Table 1). High selectivity
for p38R over other kinases is general for this series (2-
alkylamino-6-phenoxypyrimidinopyridones), and this generality
allowed us much freedom to explore substituents for optimal
potency and druglike properties. Figure 2 also contains the kinase
selectivity profile for our ultimate phase 2 candidate, 2b, which shows
submicromolar Kd for only seven out of >300 kinases tested
(Figure 2, Table 1). Both compounds 2a and 2b bind to p38R
and p38β isoforms but do not generally bind to either the p38γ or
p38δ isoforms. Selectivity profiles for other key examples are similar.
SAR Studies. Tables 2-4 summarize the potency SAR for
substitution at the 6-, 8-, and 2-positions of our pyrimidinopyr-
idones. For most compounds, shifts in potency between our
enzyme assay and our HWB (human whole blood) assay were
relatively small (2- to 20-fold). Trends in SAR generally tracked
between the two assays.
As discussed earlier in the manuscript, pan-kinase selectivity
for our inhibitors is strongly dependent on the 6-substituent.
This position is also sensitive with respect to potency (Table 2).
6-Aryl substituents such as the o-chlorophenyl of compound 2c
generally confer good potency (although poor pan-kinase
selectivity). Many 6-phenoxy analogues (examples 2a, 2d-g)
Figure 3. X-ray crystal structures of p38R in complex with (A)
compound 11, 2.0 Å resolution, PDB accession number 3FSK, (B)
compound 12, 2.1 Å, 1OZ1, (C) compound 2d, 2.25 Å, 3FLZ, and (D)
2a, 1.9 Å, 3FLN.
substituent to optimally fill the hydrophobic pocket accessible in
p38R (Figure 3C,D). An alternative rotamer of the “gatekeeper”
Thr106 relative to that observed in enzyme bound to 2c deepens
this pocket to accommodate the phenoxy group. Interestingly,
2258
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Table 1. Kinase Selectivity of 2a and 2ba
Table 3. Effect of Substitution at the 8-Position
kinase assay Kd (nM)
kinase
P38R
2ab
2bc
1.3
0.2
29
P38β
120
>1000
260
ADCK3
CSNK1ε
DDR1
GAK
320
>1000
270
680
29
>1,000
>1,000
190
JNK1
JNK2
9
16
assay IC50 (nM)a
JNK3
3
19
compd
R2
p38R inhibition
HWBb
NLK
6
170
RSK2
RSK4
STK36
150
73
300
2a
2k
2l
Me
10 ( 4
18 ( 3
1 ( 0.1
9 ( 1
200 ( 87
84 ( 80
15 ( 7
150
H
330
>1,000
Et
a Kd values were determined for all kinases that demonstrated greater
than 85% inhibition at 10 μM. b Versus 317 kinases (290 wild type).
c Versus 353 kinases (321 wild type).
2m
2n
2o
2p
2q
2r
cyclopropyl
(CH2)3OH
13 ( 13
14 ( 3
2 ( 0.5
20 ( 4
6 ( 0.5
1 ( 0.2
12 ( 2
7 ( 1
(CH2)2SO2CH3
308 ( 262
18 ( 7
(CH2)3SO2CH3
1,1-dioxotetrahydro-2H-thiopyran
CH2C(OH)(CH3)2
(S)-CH2CHOHCH2OH
CH2CN
11 ( 8
Table 2. Effect of Substitution at the 6-Position
84 ( 66
15 ( 9
2s
2t
2 ( 0.3
1 ( 0.1
28 ( 19
52 ( 23
2u
1-methanesulfonylpiperidinyl
a Mean ( SEM, n g 3. b Inhibition of TNF-R induced production
of IL-1β.
Many different side chains are tolerated at the 8-position of the
scaffold (Table 3), highlighting the fact that this vector points
toward solvent (Figures 3 and 4). The SAR around this position
was relatively flat. While other side chains did increase in vitro
potencies, none offered any clear advantages over the original
methyl group when considering metabolic stabilities and in vivo
exposures (data not shown).
As described earlier in the manuscript, 2-alkylamino side
chains show greater kinase selectivity than 2-anilino side chains;
beyond this general consideration, we had considerable room to
operate off the 2-position (Table 4). X-ray crystal structures of
p38R-ligand complexes revealed that the scaffold can shift to
allow optimal fitting for 2-substituents of varying size and
stereochemistry (Figure 4C-F). This accommodation of the
protein for changes in scaffold positioning lent much flexibility to
variation at the 2-position, which turned out to be the most
profitable one to explore in optimizing the physical properties of
the series. Many examples with excellent properties were dis-
covered; we highlight five of them in the next section.
assay IC50 (nM)a
compd
X
R1
p38R inhibition
HWBb
2a
2c
2d
2e
2f
O
20,40-diF
20-Cl
H
40-F
30-F
20-F
20-F
20-F
20,40-diF
10 ( 4
200 ( 87
bond
O
21 ( 10
106 ( 36
14 ( 6
85.7 ( 43
451 ( 215
547 ( 269
1219 ( 597
64.8 ( 28
731 ( 261
407 ( 166
713 ( 242
O
O
594 ( 211
15 ( 6
2g
2h
2i
O
S
111 ( 35
42 ( 15
40 ( 11
CH2
CH2
2j
a Mean ( SEM (standard error of the mean), n g 3. b Inhibition of TNF-
R induced production of IL-1β.
Profiling of 2a, 2b, 2cc, 2ee, and 2hh. These five analogues
from Table 4 were chosen for further studies, based on encouraging
activities and properties in an acute in vivo PD model (Table 5).
Compound 2cc, containing a chiral sulfone side chain, had
good HWB activity, reasonable solubility, and high permeability.
Further in vivo profiling in a rat model of lipopolysaccharide
(LPS) induced production of TNF-R showed that 2cc inhibited
cytokine production by 88% at a dose of 3 mg/kg. However, the
in vitro and in vivo stability of this compound was very high.
Compound 2cc in vivo showed low clearance in both rat and
monkey as well as long half-lives (∼24 h in monkey). Worries
have high potency (and good pan-kinase selectivity). 6-Phe-
nylthio and 6-benzyl analogues are typically less potent than
6-phenoxy analogues (compare 2h and 2i to 2g), and the
corresponding 6-sulfones are inactive (data not shown).
We chose to focus on phenoxy substituents, since they gave
rise to good potencies and promising kinase selectivity profiles.
Further SAR on the phenoxy group revealed a preference for
small ortho and para substituents over no substituents or meta-
substitution. o,p-Difluoro substitution also optimized metabolic
stability of this side chain (data not shown).
2259
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Figure 4. X-ray crystal structures of p38R in complex with (A) 2a, 1.9 Å resolution, PDB accession number 3FLN, (B) 2b, 2.1 Å, 3FLW, (C) 2cc, 2.3 Å,
3FLS, (D) 2dd, 1.9 Å, 3FLQ, (E) 2ii, 1.9 Å, 3FMH, and (F) 2hh, 1.7 Å, 3FMK.
2260
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
over drug accumulation prompted a deprioritization of this
compound.
excellent; however, the compound was discovered significantly
later than 2a and 2b. Therefore, compound 2hh was held as a
backup to 2a and 2b, which were already in the clinic.
Compound 2ee, containing an achiral acetyl piperidine side
chain, also performed well in all in vitro assays. This example had
good in vivo PK profiles and demonstrated good inhibition of
cytokine production in our rat LPS model (76% inhibition of
TNF-R production at 3 mg/kg). The decision to deprioritize this
compound was made based on a variety of toxicities observed
preclinically in vivo.
Compound 2hh, containing a novel chiral 2-tetrazole-contain-
ing side chain, was an exceptionally potent example in HWB and
demonstrated reasonable solubility and permeability. Com-
pound 2hh also demonstrated good inhibition of cytokine
production in our rat LPS model (90% inhibition of TNF-R
production at 3 mg/kg). The overall profile for 2hh was
From this series, 2a was the first molecule identified that
satisfied all of our selection criteria and was selected for advance-
ment into GLP studies (in fact, this decision predated the
discovery of many of the compounds described in this man-
uscript). Compound 2b was later identified as a second clinical
candidate.
Pharmacological Evaluation of 2a and 2b15. The in vitro
and in vivo pharmacology results for 2a and 2b are summarized
in Table 6. Both compounds potently inhibited cytokine produc-
tion in a variety of in vitro and in vivo models. Compounds 2a
and 2b inhibited production of TNFR by human monocytic cells
(THP-1) and inhibited production of IL-1β in human whole
blood (HWB) induced by LPS.
In rodent models of acute inflammation, both 2a and 2b were
dosed orally and evaluated for their ability to inhibit the produc-
tion of TNFR and IL-6 after ip injection of LPS into Hanover
Wistar rats. The compounds were administered to rats 0.5 h prior
to challenge with 50 μg/kg LPS. Serum was collected 1.5 h after
LPS injection for analysis of cytokines and concentrations of
drug. 2a and 2b demonstrated significant dose-dependent in-
hibition of serum TNFR and IL-6. The estimated effective dose
(ED) for TNFR inhibition is reported in Table 6.
Table 4. Effect of Substitution at the 2-Position
In an alternative rat model, the compounds were administered
to rats 0.5 h prior to challenge with TNFR. Serum was collected 2
h after ip injection of TNFR for analysis of cytokines and
concentration of drug. The compounds inhibited TNFR-in-
duced IL-6 at an ED50 ( ASE of 0.4-1.1 mg/kg. The plasma
concentrations of the p38 inhibitors associated with efficacy in
these acute rat in vivo cytokine assays were generally in close
agreement to their in vitro IC50 in human cells.
Additionally, 2a and 2b were analgesic in vivo. In a rat model of
inflammatory hyperalgesia, both compounds increased tolerance
to pressure in a dose-dependent manner, suggesting that p38
inhibition may also reduce the pain associated with inflamma-
tion. The anti-inflammatory activities of 2b in multiple cell, acute
and chronic in vivo models of inflammation have been recently
reported in greater detail.18
Pharmacokinetics and Metabolism of 2a. A summary of the
pharmacokinetic parameters for 2a is detailed in Table 7. In-
travenous systemic clearance (Cl) of 2a was variable across
species: low in monkey, medium in rat, and very high in dog
(exceeded hepatic blood flow). Volume of distribution at steady
state (Vdss) was moderate in all species studies. Oral bioavail-
ability of 10 mg/kg 2a in monkey, rat, and dog was 51.6%, 29.3%,
10.3%, respectively. Absorption was fairly slow but mostly
complete in rat and monkeys. Compound 2a was moderately
(90.4-95.6%) bound to rat, monkey, and dog plasma proteins
assay IC50 (nM)a
compd
R3
p38R inhibition
HWBb
2a
2b
2v
2w
2x
4-tetrahydropyranyl
CH(CH2CH2OH)2
Me
10 ( 4
14 ( 2
260 ( 75
4 ( 0.3
31 ( 8
67 ( 26
25 ( 8
158 ( 67
223 ( 74
5 ( 1
200 ( 87
100 ( 30
isopropyl
45 ( 30
(S)-3-tetrahydrofuranyl
CH2C(CH3)2OH
C(CH3)2CH2OH
CH(CH2OH)2
373 ( 396
2y
2z
2aa
189 ( 94
1487 ( 771
50 ( 12
>3000
2bb (CH2)2SO2CH3
2cc
(R)-CHCH3CH2SO2CH3
2dd (S)-CHCH3CH2SO2CH3
50 ( 20
1 ( 0.6
4 ( 0.6
15 ( 2
2 ( 0.7
11
2ee
2ff
N-acetyl-4-piperidinyl
30 ( 5
trans-4-hydroxycyclohexanyl
1,1-dioxotetrahydro-2H-thiopyran
68 ( 4
2gg
769 ( 309
9 ( 1
2hh (S)-CHCH3CH2(2-tetrazole)
2iic
(R)-CHCH3CH2(2-tetrazole)
318
a Mean ( SEM, n g 3. b Inhibition of TNF-R induced production of IL-
1β. c n = 1 data.
Table 5. Further Profiling of Most Promising Examples
parameter
2a
2b
170
2cc
18
2ee
14
2hh
23
aqueous solubility,a μg/mL
6
Caco2 permeability,b 106 cm/s
AB 33.0
BA 23.7
1.5
AB 2.74
BA 22.6
5.4
AB 13.5
BA 22.4
4.0
AB 17.9
BA 19.8
3.3
AB 30.1
BA 19.7
3.3
human protein binding, % free
rat LPS 3 mg/kg, % inhibition
68
77
88
76
90
a Phosphate buffer, pH 6.5. b Permeability in Caco-2 cells AB (apical to basolateral) and BA (basolateral to apical) movement of 10 μM test compound in
21 day cultured Caco-2 cells ((cm/s) ꢀ 10-6), pH 7.4, on both sides.
2261
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Table 6. Summary of Biological Efficacy of 2a and 2b
cell or animal model
in vitro IC50 (μM)
in vivo ED50 (mg/kg)/EC50 (μM) in ratsc
THP-1a
HWBa
LPSb
LPSb
TNFRb
Yeast
hyperalgesia
compd TNFR inhib IL-1β inhib
TNFR inhib
IL-6 inhib
IL-6 inhib
2a
2b
0.047 ( 0.01 0.17 ( 0.02 (0.8 ( 0.29)/(0.03 ( 0.008) (0.4 ( 0.07)/(0.01 ( 0.005) (0.4 ( 0.17)/(0.02 ( 0.01) 5.5 ( 1.63
0.025 ( 0.01 0.1 ( 0.03
(0.3 ( 0.07)/(0.12 ( 0.02)
(0.1 ( 0.03)/(0.05 ( 0.03)
(1.1 ( 0.33)/(0.27 ( 0.17) (20.4 ( 4.8)/(4.3 ( 0.67)
a The data are expressed as the combined mean IC50 ( SEM. b The data are expressed as the combined mean ED50/EC50 ( ASE (asymptotic standard
error). c Data represent results from two independent rat experiments.
Table 7. Single Dose Plasma Pharmacokinetic Parameters of
2a after Dosing in Animals
Table 8. Single Dose Plasma Pharmacokinetic Parameters of
2b after Dosing in Fed Animals
species, strain
species, strain
rat, Hanover-
Wistar
monkey,
dog,
rat, Hanover-
Wistar
monkey,
parameter
dose (mg/kg)
Cynomolgus
Beagle
Cynomolgus
dog, Beagle
10
10
10
dose (mg/kg)
gender
5
5
5
gender
female
HCl salt
60,a 3b
2.0
male
free base
4, 3b
4.67
1.56
22.5
29.4
51.6
0.183
2.73
95.4
female
HCl salt
4, 3b
0.5
female
56,a 56b
0.5
male
4, 3b
female
4, 3b
0.67
1.38
9.18
5.22
drug form
sample size (iv, po)
po Tmax (h)
sample size (iv, po)
po Tmax (h)
po Cmax (μg/mL)
po T1/2 (h)
1.50
po Cmax (μg/mL)
po T1/2 (h)
1.34
1.86
0.65
0.11
4.04
13.8, est
9.41, 72 h
ND
2.23
po AUC (μg h/mL)
3.13, 24 h
3
po AUC inf (μg h/mL)
2.35
0.31
(0 to last t)
3
po F (%)
29.3
10.3
po F (%)
80.1
1.26
3.56
82.9
>100
0.65
20.6
85.4
78.2
0.75
12.7
ND
iv Cl ((L/h)/kg)
iv Vdss (L/kg)
% protein bindingc
1.24
3.25
iv Cl ((L/h)/kg)
iv Vdss (L/kg)
% protein bindingc
1.94
4.65
95.6
90.4
a The iv (intravenous) values are based on mean plasma concentrations;
plasma from three rats per sample time. Formulation for iv dosing was
aqueous 40% w/v cyclodextrin. b The po (per os) values are based on
mean parameter; serial plasma samples obtained from each animal.
Formulation for po dosing was aqueous 1% hydroxypropyl methylcel-
lulose, 0.4% polysorbate 80, 0.9% benzyl alcohol, 0.04% citric acid
monohydrate, 0.24% sodium citrate dihydrate. c The in vitro binding of
[14C]2a to plasma protein at 37 °C was studied by ultrafiltration
methods.
a The iv values are based on mean plasma concentrations; plasma from
three rats per sample time. Formulation for iv dosing was aqueous 40%
w/v cyclodextrin. b The po values are based on mean parameter; serial
plasma samples obtained from each animal. Formulation for po dosing
was aqueous 1% hydroxypropylmethylcellulose, 0.4% polysorbate 80,
0.9% benzyl alcohol, 0.04% citric acid monohydrate, 0.24% sodium
citrate dihydrate. c The in vitro binding of [14C]2b to plasma protein at
37 °C was studied by ultrafiltration methods.
Compound 2b is neither a substrate nor a strong inhibitor of
CYP450 enzymes (IC50 > 29 μM).
and highly (98.5%, data not shown) bound to human plasma
proteins. Rat and monkey plasma contained mostly 2a but also
metabolites including diol 2b.
The human cytochrome P450 enzyme primarily responsible
for the transformation of 2a was 1A2. The IC50 values for
inhibition of human cytochrome P450s, 1A2, 2C9, 2C19, 2D6,
and 3A4 were all greater than 40 μM.
Pharmacokinetics and Metabolism of 2b. A summary of the
pharmacokinetic parameters for 2b is detailed in Table 8.
Intravenous systemic clearance of 2b was low to moderate across
species Compound 2b was well absorbed and had high oral
bioavailability in the rat, monkey, and dog (62%-80%). Term-
inal elimination half-life (T1/2) was moderate (∼4 h) in rodents
and high (>9 h) in monkeys and dogs. The elimination of 2b was
by metabolic clearance in the rat and by metabolic and renal
clearance in the monkey. Compound 2b was weakly (82.9% and
85.4%, respectively) bound to rat and monkey plasma proteins
and moderately bound (94.6%, data not shown) to human
plasma proteins.
’ CONCLUSION
Our chemistry approach began with a molecule identified
from within the Roche kinase library. A compound that was
initially designed to inhibit lck was later demonstrated to inhibit
p38R. An X-ray crystal structure of this lead compound con-
firmed similar binding motifs in the active sites of both enzymes
and revealed opportunities to alter the kinase selectivity profile in
favor of p38R. Incorporation of saturated alkylamines at the
2-position increased the three-dimensional bulk of the molecule
and caused unfavorable steric interactions between the inhibitor
and kinases containing a glycine insertion in the lower lip region.
To further enhance the kinase selectivity profile, an additional
template modification was made to expand the volume of the
back hydrophobic pocket of p38R filled by the inhibitor series by
insertion of an oxygen atom between the core scaffold and the
aromatic ring. Indeed, an alternative rotamer of the “gatekeeper”
2262
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
Thr106 deepens this pocket to accommodate the phenoxy
group. To our satisfaction, in vitro studies demonstrated that
optimized inhibitors 2a and 2b were highly potent and highly
selective inhibitors of p38R. Further profiling of these inhibitors
in a range of human cell assays, in vivo PK/PD and in vivo efficacy
studies, pharmacokinetic parameters across multiple species, and
safety studies supported the selection criteria for advancement of
both 2a and 2b into human clinical trials. Compound 2b was
ultimately selected for advancement into phase 2 studies over 2a
based primarily on its superior pharmacokinetic profile in
humans.
ppm 2.51 (s, 3 H), 2.99 (d, J = 4.91 Hz, 3 H), 8.50 (s, 1 H), 8.63 (d,
J = 3.78 Hz, 1 H), 9.74 (s, 1 H). MS (EI/CI) m/z: 184 (M þ H).
6-(2,4-Difluorophenoxy)-2-methanesulfonyl-8-methyl-8H-
pyrido[2,3-d]pyrimidin-7-one (7a). A mixture of 4-methylamino-2-
methylthiopyrimidine-5-carboxaldehyde (5a) (4.8 g, 26 mmol) and
methyl 2,4-difluorophenoxyacetate (5.4 g, 29 mmol) in 50 mL of
1-methyl-2-pyrrolidinone was treated with potassium carbonate (6.0 g,
44 mmol). The reaction mixture was stirred at 120 °C for 12 h. Additional
methyl 2,4-difluorophenoxyacetate (2.5 g, 13 mmol) and potassium
carbonate (2.5 g, 18 mmol) were added. The mixture was stirred at
120 °C for an additional 6 h, then cooled to room temperature. Water
(100 mL) was added, and the solution was stirred for 45 min and then
filtered. The resulting solid was washed three times with water, added to
75 mL of ethyl acetate, and the resulting mixture was stirred for 1 h. The
solid was then isolated by filtration and dried to yield 6.1 g (69%) of
6-(2,4-difluorophenoxy)-8-methyl-2-methylsulfanyl-8H-pyrido[2,3-
d]pyrimidin-7-one (6a). 1H NMR (300 MHz, DMSO-d6) δ ppm 2.59 (s,
3 H), 3.66 (s, 3 H), 7.05-7.20 (m, 1 H), 7.35 (td, J = 9.35, 5.72 Hz, 1 H),
7.41 (s, 1 H), 7.51 (ddd, J = 11.44, 8.77, 3.05 Hz, 1 H), 8.82 (s, 1 H).
A solution of 6-(2,4-difluorophenoxy)-8-methyl-2-methylsulfanyl-
8H-pyrido[2,3-d]pyrimidin-7-one (50 g, 0.15 mol) in 420 mL of
dichloromethane was treated with formic acid (17 g, 0.37 mol), followed
by portionwise addition of a 30% aqueous hydrogen peroxide solution
(71 g, 0.63 mol), then stirred at 38 °C until reaction completion as
confirmed by area normalized HPLC. The resulting light yellow solution
was cooled to <25 °C. Then 32 mL of water was added, and layers were
separated. The organic layer was washed with an aqueous NaOH
solution (5 g of NaOH in 53 mL of H2O). Then dichloromethane
was exchanged to 160 mL of tert-butyl methyl ether. The solution was
cooled to <5 °C, and precipitate was isolated by filtration, washed with
36 mL of tert-butyl methyl ether, and dried on the filter to afford 38 g
(71%) of 6-(2,4-difluorophenoxy)-2-methanesulfonyl-8-methyl-8H-
pyrido[2,3-d]pyrimidin-7-one. Mp 182.2-183.0 °C. 1H NMR
(DMSO-d6) δ 3.47 (s, 3 H), 3.74 (s, 3 H), 7.17-7.28 (m, 1 H),
7.40-7.52 (m, 2 H), 7.61 (ddd, J = 11.21, 8.76, 3.01 Hz, 1 H), 9.18 (s, 1
H). MS (EI/CI) m/z: 368 (M þ H).
’ EXPERIMENTAL SECTION
Chemistry. Reagents and solvents were obtained from commercial
suppliers and were used without further purification. Flash chromatog-
raphy was performed with Merck silica gel 60 (230-400 mesh), and
reaction progress was determined by thin-layer chromatography (TLC)
on silica gel plates. Visualization was done with UV light (254 nm) or
iodine. Yields were based on purified compounds and were not
1
optimized. H NMR measurements were recorded at 300 MHz using
a Bruker AMX 300 instrument in DMSO-d6 or CDCl3 with tetra-
methylsilane as the internal standard. Melting points recorded were
uncorrected. Purity and characterization of compounds were established
by a combination of elemental analyses and NMR analytical techniques,
and the purity was >95% by NMR for final products 2a and 2b.
Elemental analyses on final products 2a and 2b were within (0.4%
for C, H, N.
4-Methylamino-2-methylthiopyrimidine-5-carboxaldehyde
(5a). Methylamine in ethanol (33%, 35 mL, 280 mmol) was slowly added
to a solution of ethyl 4-chloro-2-methylthiopyrimidine-5-carboxylate (20 g,
86 mmol) in 250 mL of dichloromethane at 0 °C. The solution was stirred
for 30 min. Then 150 mL of waterwas added, and the layers were separated.
The organic layer was dried over MgSO4, filtered, and concentrated to 19 g
(97%) of ethyl 4-methylamino-2-methylthiopyrimidine-5-carboxylate (3a)
as a white solid. 1H NMR (300 MHz, DMSO-d6) δ ppm 1.30 (t, J = 7.25
Hz, 3 H), 2.49 (s, 3 H), 2.98 (d, J =4.58 Hz, 3H), 4.28 (q, J = 7.25 Hz, 2 H),
8.19-8.28 (m, 1 H), 8.52 (s, 1 H).
A mixture of lithium aluminum hydride (8.2 g, 210 mmol) in 300 mL
of dry tetrahydrofuran at 5 °C was treated dropwise with a solution of
ethyl 4-methylamino-2-methylthiopyrimidine-5-carboxylate (46 g, 200
mmol) in 450 mL of dry tetrahydrofuran. The reaction mixture was
stirred for 15 min, and then 18 mL of water was added dropwise with
caution. The mixture was stirred for 30 min. Then 8.5 mL of a 15%
aqueous NaOH solution was added dropwise, followed by 25.5 mL of
water. The resulting suspension was stirred for 17 h at room temperature
and then filtered. The filter residue was sequentially washed with two
100 mL portions of tetrahydrofuran, and the combined filtrate and
washings were concentrated to a residue. The residue was suspended in
200 mL of a 2:1 ethyl acetate/hexane solution, and a yellow solid was
isolated by filtration and dried to provide 32.7 g (88%) of 4-methyl-
amino-2-methylthiopyrimidine-5-methanol (4a). 1H NMR (300 MHz,
DMSO-d6) δ ppm 2.43 (s, 3 H) 2.86 (d, J = 4.58 Hz, 3 H), 4.31 (d, J =
5.34 Hz, 2 H), 5.10 (t, J = 5.34 Hz, 1 H), 6.82 (d, J = 4.58 Hz, 1 H), 7.82
(s, 1 H).
4-Methylamino-2-methylthiopyrimidine-5-methanol (20 g, 110
mmol) and 1 L of dichloromethane were combined with stirring and
treated with manganese dioxide (87 g, 1.0 mol). The resulting suspen-
sion was stirred for 24 h and then filtered through Celite. The filter
residue was washed with 100 mL of dichloromethane, and the combined
filtrate and washings were concentrated to give 15.8 g (80%) of
4-methylamino-2-methylthiopyrimidine-5-carboxaldehyde as a white
solid. Mp 101.7-102.3 °C . 1H NMR (300 MHz, DMSO-d6) δ
6-(2,4-Difluorophenoxy)-8-methyl-2-(tetrahydro-2H-pyr-
an-4-ylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (2a). A sus-
pension of 6-(2,4-difluorophenoxy)-2-methanesulfonyl-8-methyl-8H-
pyrido[2,3-d]pyrimidin-7-one (7a) (55 g, 150 mmol), sodium carbo-
nate (16.5 g, 156 mmol), and 4-aminotetrahydropyran (22.00 g, 217.5
mmol) in 150 mL of toluene was stirred at 80 °C overnight. The reaction
mixture was diluted with 500 mL of toluene, and the hot suspension was
filtered. The filtrate was allowed to cool, and precipitated solid was
isolated by filtration, washing with 100 mL of water and then 100 mL
of a 1:1 methanol/water solution, to afford 53 g (90%) of 6-(2,4-
difluorophenoxy)-8-methyl-2-(tetrahydro-2H-pyran-4-ylamino)pyrido-
[2,3-d]pyrimidin-7(8H)-one. Mp 185.7-187.5 °C. 1H NMR (DMSO-
d6) δ 1.56 (qd, J = 11.83, 4.20 Hz, 2 H), 1.87 (br s, 2 H), 3.41 (t, J = 10.68
Hz, 2 H), 3.49-3.64 (m, 3 H), 3.89 (d, J = 11.06 Hz, 2 H), 4.02 (br s, 1
H), 6.97-7.10 (m, 1 H), 7.18 (td, J = 9.16, 5.72 Hz, 1 H), 7.35-7.53 (m,
2 H), 7.57-7.85 (m, 1 H), 8.57 (s, 1 H). MS (EI/CI) m/z:
389 (M þ H).
6-(2,4-Difluorophenoxy)-2-[3-hydroxy-1-(2-hydroxy-ethyl)-
propylamino]-8-methyl-8H-pyrido[2,3-d]pyrimidin-7-one (2b).
A 1.0 M solution of lithium aluminum hydride in tetrahydrofuran solution
(571 mL, 571 mmol), under mechanical stirring at less than -10 °C, was
slowly treated with 3-aminopentanedioic acid dimethyl ester (50.0 g,
285 mmol) in 50 mL of tetrahydrofuran, keeping the temperature under
20 °C. The reaction mixture was stirred at room temperature for 18 h, then
cooled to less than -10 °C. Water (41 mL, 2.3 mol) was slowly added as the
mixture was vigorously stirred, keeping the temperature under 20 °C.
Diethylamine (33 mL, 320 mmol) was added to the mixture, which was
then stirred at room temperature for 1 h. The solids were removed by
2263
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
filtration and the filtrate was concentrated under vacuum at 45 °C to afford
13.7 g (40%) of 3-aminopentane-1,5-diol as an oil. 1H NMR (CD3OD) δ
3.67 (m, 4 H), 3.02 (tt, J = 8.11, 4.8 Hz, 1 H), 1.60 (m, 4 H).
well). Test compounds were dissolved in DMSO and then diluted with
the culture medium such that the final DMSO concentration was 5%.
Then 25 μL aliquots of test solution or medium with DMSO (control)
were added to each well. The cells were incubated for 30 min at 37 °C.
LPS (Sigma, St. Louis, MO) was added to the wells at a final
concentration of 0.5 μg/mL, and cells were incubated for an additional
2 h. At the end of the incubation period, culture supernatants were
collected and the amount of TNF-R present was determined by a
specific trapping ELISA assay using two anti-TNF-R antibodies (2TNF-
H22 and 2TNF-H34) as described.16
Hexamethyldisilazane (20 mL, 96 mmol) and chlorotrimethylsilane
(0.5 mL, 4 mmol) were added to 3-aminopentane-1,5-diol (9.6 g, 81
mmol) in 40 mL of tetrahydrofuran. The mixture was stirred at reflux for
18 h. Solvents were removed under vacuum at 45 °C, and the remaining
oil was distilled under vacuum at 85-107 °C to afford 17.7 g (83%) of
3-trimethylsilanyloxy-1-(2-trimethylsilanyloxyethyl)propylamine as an
oil. 1H NMR (CDCl3) δ 3.65 (m, 4 H), 2.97 (tt, J = 8.39, 4.3 Hz, 1 H),
1.51 (m, 6 H), 0.05 (s, 18 H).
Inhibition of TNFr and IL-1β Biosynthesis in LPS-Stimu-
lated Peripheral Blood Mononuclear Cells. Blood was collected
from healthy volunteers (drug-free for 2 weeks) into vacutainers
(Becton Dickinson, Mountainview, CA) containing 19 units/mL so-
dium heparin. Peripheral blood mononuclear cells (PBMC) were
separated using a Histopaque-1077 gradient (Sigma, H-8889). Briefly,
whole blood was diluted 1:1 in Dulbecco’s phosphate-buffered saline
(Gibco, no. 14190-136) and layered onto the histopaque solution
followed by centrifugation at 400g for 20 min at room temperature.
Cells were collected and washed once in cold PBS followed by red cell
lysis with lysing media (Sigma no. R-7757, lot no. 46H2373). A final cell
suspension was prepared at 2 ꢀ 106 per mL in RPMI 1640 (Gibco BRL,
A solution of 6-(2,4-difluorophenoxy)-2-methanesulfonyl-8-methyl-
8H-pyrido[2,3-d]pyrimidin-7-one (7a) (10.0 g, 27.2 mmol), 3-tri-
methylsilanyloxy-1-(2-trimethylsilanyloxyethyl)propylamine (9.4 g, 36
mmol), and N,O-bis(trimethylsilyl)acetamide (3.55 mL, 14.4 mmol) in
20 mL of tetrahydrofuran was stirred at reflux overnight. The reaction
mixture was concentrated under vacuum and the residue stirred in a
mixture of 25 mL of tert-butyl methyl ether and an aqueous K2HPO4
solution (2 g in 50 mL of water). Precipitate was isolated by filtration and
sequentially washed with water, 20 mL of a 30:70 methanol/water
mixture, and 20 mL of toluene to afford 10.3 g (93%) of 6-(2,4-
difluorophenoxy)-2-[3-hydroxy-1-(2-hydroxyethyl)propylamino]-8-
methyl-8H-pyrido[2,3-d]pyrimidin-7-one. Mp 157.4-158.2 °C. 1H
NMR (DMSO-d6) δ 1.62-1.86 (m, 4 H), 3.41-3.53 (m, 4 H), 3.53-
3.65 (m, 3 H), 4.17-4.35 (m, 1 H), 4.40 (t, J = 5.05 Hz, 2 H), 7.03 (d, J =
7.07 Hz, 1 H), 7.10-7.25 (m, 1 H), 7.37-7.53 (m, 2 H), 7.59 (d, J = 8.59
Hz, 1 H), 8.50-8.61 (m, 1 H). MS (EI/CI) m/z: 407 (M þ H).
Biological Methods. p38 MAP Kinase Inhibition. HTS Assay.
HTS was performed using a Zymark robotic system and a 96-well,
radioactive filtration binding format. The enzyme was preincubated with
test compounds (40 μM) for 10 min at 30 °C, following which the
reaction was initiated by adding [33P]ATP (250 μM) and myelin basic
protein (150 μM). Following an additional incubation for 20 min at
30 °C the reaction was terminated, and the radiolabeled product was
captured on nitrocellulose filtration plate, washed, and counted for
radioactivity. Compounds that displayed >40% inhibition of the enzyme
activity were retested at multiple concentrations.
In Vitro Assay. Inhibition of human recombinant active p38R [5 nM]
was tested by measuring the incorporation of 33P from γ-[33P]ATP [50
μM, 1 μCi] into myelin basic protein (MBP) [35 μM]. The assay was
performed in a 96-well microtiter plate. In a 40 μL reaction volume, an
amount of 26 μL of diluted p38R [5 nM] in assay dilution buffer (ADB)
(20 mM MOPS, pH 7.4, 40 mM MgCl2, 1 mM DTT, 25 mM β-glycerol
phosphate, 5 mM EGTA, 1 mM sodium orthovanadate) was preincu-
bated at room temperature for 10 min with 4 μL of test compounds [0-
100 μM] dissolved in dimethylsulfoxide (DMSO) [10%]. The kinase
reaction was initiated by the addition of 10 μL of assay mix, containing
ADB and γ-[33P]ATP [50 μM, 1 μCi] and MBP [35 μM]. After 30 min
at 30 °C the reactions were stopped by the removal of 25 μL of reaction
sample onto 200 μL of 0.75% phosphoric acid (H3PO4) containing acid
prewetted Multiscreen 96-well phosphocellulose filter plate (Millipore).
The plate membranes were washed three times with 200 μL of 0.75%
H3PO4 for the removal of the free radionucleotide. Then 50 μL of
Microscint-20 scintillant (Packard) was added to each well. The plate
was sealed with a plastic film and counted in a Packard Topcount
microplate scintillation counter. The 50% inhibitory value was calculated
by fitting the data to the equation
no. 11875-085), 10% FBS, 2 mM L-glutamine, and 5.5 ꢀ 10-5
M
β-mercaptoethanol. Dilutions of test compounds were predispensed in
25 μL aliquots (before addition of cells) into round-bottom 96-well
plates (U bottom TC plate; Costar no. 3799). The starting concentra-
tion was 100 μM in 5% DMSO, and six half log serial dilutions were
made. After the addition of 200 μL of cell suspension and 25 μL of 5 μg/
mL lipopolysaccharide (LPS, E. coli; 0127; B8; catalog no. L3129, Sigma
Chem Co., St. Louis, MO) in medium, the final DMSO concentration
was 0.5%. Aliquots of test compounds were diluted an additional 10-fold,
and the final LPS concentration was 500 ng/mL. PBMC suspension
(200 μL per sample) was preincubated with the diluted compound for
30 min at 37 °C followed by addition of 25 μL of LPS dissolved in RPMI
(Gibco BRL, no. 11875-085) to produce a final LPS concentration of
0.5 μg/mL. Negative control wells received RPMI alone. Plates were
incubated for 2 or 5 h at 37 °C in a 5% CO2 atmosphere for
determination of TNFR or IL-1β, respectively, followed by centrifuga-
tion at 150g for 10 min. Supernatants were collected from each sample
and stored in polypropylene plates at 4 °C. TNF-R and IL-1β were
determined by Elisa (Antibody Solutions, Palo Alto, CA) following the
manufacturer’s instructions. Optical density of each sample was read at
450 nm with a reference of 650 nm. Cytokine concentrations were
determined from a standard curve using Molecular Devices SoftMax Pro.
The percent inhibition was calculated for each concentration tested, and
an IC50 curve was constructed using Xlfit software.
Phamacokinetics. Female Wistar/Han (CRL:WI) Rats (Charles
River, Hollister, CA) weighing between 180 and 220 g were used. Beagle
dogs (∼5-10 kg) and cynomologous monkeys (5-8 kg) were residents
of the facility animal colony. Animals were allowed free access to a
standard laboratory chow and tap water and were housed in a constant
temperature-humidity environment. All studies were conducted under
IACUC approved protocols. Animals were administered either single iv
bolus doses (50% cyclodextran/water) or single oral suspension doses
prepared in aqueous vehicle containing 0.9% NaCl, 0.5% sodium
carboxymethyl cellulose, 0.4% polysorbate 80, and 0.9% benzyl alcohol.
Blood was collected from each animal with anesthesia at typically 0.08,
0.25, 0.5,1, 2, 3,4, 6, 8, and 24 h after dosing into tubes containing EDTA.
Plasma concentrations of 6f and/or 6x were determined by a liquid
chromatography/tandem mass spectrometry method (LC/MS/MS). 6f
or 6x and an internal standard were extracted from samples by protein
precipitation using acetonitrile. The supernatant was diluted using 0.1%
formic acid and analyzed using liquid chromatography (LC) with
tandem mass spectrometric detection (MS/MS). The standard curve
1=½Iꢁ
fractional activity ¼
IC50 þ 1
Inhibition of TNF Biosynthesis. THP1 cells were suspended in
culture medium [RPMI (Gibco-BRL, Gailthersburg, MD) containing
15% fetal bovine serum and 0.02 mM 2-mercaptoethanol, at 2.5 ꢀ 106
cells/mL, and then plated in a 96-well plate (0.2 mL aliquots in each
2264
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265

Journal of Medicinal Chemistry
ARTICLE
range for this assay is from 5.00 to 5000 ng/mL for 6f and 6x using a
plasma sample volume of 0.100 mL.
(8) Goldstein, D. M.; Gabriel, T. Pathway to the clinic: inhibition of
P38 MAP kinase. A review of ten chemotypes selected for development.
Curr. Top. Med. Chem. 2005, 5, 1017–1029.
(9) Pettus, L. H.; Wurz, R. P. Small molecule p38 MAP kinase
inhibitors for the treatment of inflammatory diseases: novel structures
and developments during 2006-2008. Curr. Top. Med. Chem. 2008,
8, 1452–1467.
(10) Goldstein, D. G.; Kuglstatter, A.; Lou, Y.; Soth, M. J. Selective
p38a inhibitors clinically evaluated for the treatment of chronic inflam-
matory disorders. J. Med. Chem. 2010, 53, 2345–2353.
Plasma Protein Binding. The in vitro binding of [14C]6x to
plasma protein at 37 °C was studied by ultrafiltration methods
Protein Crystallography. X-ray crystal structures of p38R-
ligand complexes where determined following published procedures.12
Atomic coordinates and structure factor amplitudes have been deposited
to the RCSB PDB along with detailed statistics on data collection and
structure refinement.
(11) Hammaker, D.; Firestein, G. S. “Go upstream, young man”:
lessons learned from the p38 saga. Ann. Rheum. Dis. 2010, 69, i77–i82.
(12) Trejo, A.; Arzeno, H.; Browner, M. F.; Chanda, S.; Cheng, S.;
Comer, D. D.; Dalrymple, S. A.; Dunten, P.; La Fargue, J.; Lovejoy, B.;
Freire-Moar, J.; Lim, J.; McIntosh, J.; Miller, J.; Papp, E.; Reuter, D.;
Roberts, R.; Sanpablo, F.; Saunders, J.; Song, K.; Villasenor, A.; Warren,
S. D.; Welch, M.; Weller, P.; Whiteley, P, E.; Zeng, L.; Goldstein, D. M.
Design and synthesis of 4-azaindoles as inhibitors of p38 MAP kinase.
J. Med. Chem. 2003, 46, 4702–4713.
(13) Goldstein, D. M.; Alfredson, T.; Browner, M. F.; Dalrymple,
S. A.; Dunn, J.; Freire-Moar, J.; Harris, S.; Labadie, S. S.; La Fargue, J.;
Lapierre, J. M.; Li, F.; McWeeney, D.; Papp, E.; Ramesha, C.; Roberts,
R.; Rotstein, D.; Sanpablo, B. F.; So, O.-Y.; Talamas, F. X.; Tao, W.;
Trejo, A.; Villasenor, A.; Welch, T.; Weller, P.; Whiteley, P.; Young, K.;
Zipfel, S. Discovery of S-[5-amino-1-(4-fluorophenyl)-1H-pyrazol-4-
yl]-[3-(2,3-dihydroxypropoxy)phenyl]methanone (RO3201195), an
orally bioavailable and highly selective inhibitor of p38 map kinase.
J. Med. Chem. 2006, 49, 1562–1575.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures for
b
intermediate 10a and final compounds 2c-ii. This material is
available free of charge via the Internet at http://pubs.acs.org.
Accession Codes
†Structures are deposited in the RCSB Protein Data Bank under
3KMM, 3FLN, 3FLQ, 3FLS, 3FLW, 3FLZ, 3FMH, 3FMK, 3FSF
and 3FSK.
’ AUTHOR INFORMATION
Corresponding Author
*For D.M.G.: phone, 973-235-4008; fax, 973-235-6263; e-mail,
David-M.Goldstein@roche.com. For M.S.: phone, 973-235-
2717; fax, 973-235-6263; e-mail, Michael.soth@roche.com.
(14) Harris, W.; Hill, C. H.; Smith, I. E. D. Preparation of Pyrimi-
dopyrimidinones as T-Cell Tyrosine Kinase Inhibitors. PCT Int. Appl.
WO 2000024744 A1, 2000.
(15) Chen, J. J.; Dunn, J. P.; Goldstein, D. M.; Stahl, C. M.
Preparation of 2,6-disubstituted 7-oxopyrido[2,3-d]pyrimidines for
treating p38 mediated disorders. WO2002064594, 2002.
Present Addresses
‡Hoffmann-La Roche Inc., Nutley, NJ.
§F. Hoffmann-La Roche Ltd., Basel, Switzerland.
’ ABBREVIATIONS USED
(16) Hill, R. J.; Dabbagh, K.; Phippard, D.; Li, C.; Suttmann, R. T.;
Welch, M.; Papp, E.; Song, K. W.; Chang, K.-c.; Leaffer, D.; Kim, Y.-N.;
Roberts, R. T.; Zabka, T. S.; Aud, D.; Dal Porto, J.; Manning, A. M.;
Peng, S. L.; Goldstein, D. M.; Wong, B. R. Pamapimod, a novel p38
mitogen-activated protein kinase inhibitor: preclinical analysis of efficacy
and selectivity. J. Pharmacol. Exp. Ther. 2008, 327 (3), 610–619.
(17) Alten, R. E.; Zerbini, C.; Jeka, S.; Irazoque, F.; Khatib, F.;
Emery, P.; Bertasso, A.; Rabbia, M.; Caulfield, J. P. Efficacy and safety of
pamapimod in patients with active rheumatoid arthritis receiving stable
methotrexate therapy. Ann. Rheum. Dis. 2010, 69, 364–367.
(18) Cohen, S. B.; Cheng, T. T.; Chindalore, V.; Damjanov, N.;
Burgos-Vargas, R.; Delora, P.; Zimany, K.; Travers, H.; Caulfield, J. P.
Evaluation of the efficacy and safety of pamapimod, a p38 MAP kinase
inhibitor, in a double-blind, methotrexate-controlled study of patients
with active rheumatoid arthritis. Arthritis Rheum. 2009, 60, 335–344.
(19) Genovese, M. C. Inhibition of p38: Has the fat lady sung?
Arthritis Rheum. 2009, 60, 317–320.
IBD, inflammatory bowel disease; lck, lymphocyte specific pro-
tein tyrosine kinase; LPS, lipopolysaccharide; MAP, mitogen
activated protein; RA, rheumatoid arthritis; SAR, structure-
activity relationship
’ REFERENCES
(1) Schett, G.; Zwerina, J.; Firestein, G. The p38 mitogen-activated
protein kinase (MAPK) pathway in rheumatoid arthritis. Ann. Rheum.
Dis. 2008, 67, 909–916.
(2) Westra, J.; Limburg, P. C. p38 mitogen-activated protein kinase
(MAPK) in rheumatoid arthritis. Mini-Rev. Med. Chem. 2006,
6, 867–874.
(3) Schett, G.; Stach, C.; Zwerina, J.; Voll, R.; Manger, B. How
antirheumatic drugs protect joints from damage in rheumatoid arthritis.
Arthritis Rheum. 2008, 58, 2936–2948.
(20) Dunn, J. P.; Fisher, L. E.; Goldstein, D. M.; Harris, W.; Hill, C. H.;
Smith, I. E. D.; Welch, T. R. Preparation of Alkylamino Substituted Bicyclic
Nitrogen Heterocycles for Pharmaceutical Use as Inhibitors of p38 Protein
Kinase. PCT Int. Appl. WO 2001029042 A1, 2001.
(21) Paulini, R.; M€uller, K.; Diederich, F. Orthogonal multipolar
interactions in structural chemistry and biology. Angew. Chem., Int. Ed.
2005, 44, 1788–1805.
(4) Strand, V.; Singh, J. A. Improved health-related quality of life
with effective disease-modifying antirheumatic drugs: evidence from
randomized controlled trials. Am. J. Manag. Care 2008, 14, 234–254.
(5) Hale, K. K.; Trollinger, D.; Rihanek, M.; Manthey, C. L. Differ-
ential expression and activation of p38 mitogen-activated protein kinase
alpha, beta, gamma, and delta in inflammatory cell lineages. J. Immunol.
1999, 162, 4246–4252.
(22) M€uller, K.; Faeh, C.; Diederich, F. Fluorine in pharmaceuticals:
looking beyond intuition. Science 2007, 317, 1881–1886.
(6) Korb, A.; Tohidast-Akrad, M.; Cetin, E.; Axmann, R.; Smolen, J.;
Schett, G. Differential tissue expression and activation of p38 MAPK
alpha, beta, gamma, and delta isoforms in rheumatoid arthritis. Arthritis
Rheum. 2006, 54, 2745–2756.
(7) Schett, G.; Tohidast-Akrad, M.; Smolen, J. S.; Schmid, B. J.;
Steiner, C. W.; Bitzan, P.; Zenz, P.; Redlich, K.; Xu, Q.; Steiner, G.
Activation, differential localization, and regulation of the stress-activated
protein kinases, extracellular signal-regulated kinase, c-JUN N-terminal
kinase, and p38 mitogen-activated protein kinase, in synovial tissue and
cells in rheumatoid arthritis. Arthritis Rheum. 2000, 43, 2501–2512.
(23) Fabian, M. A. B.; William, H.; Treiber, D. K.; Atteridge, C. E.;
Azimioara, M. D.; Benedetti, M. G.; Carter, T. A.; Ciceri, P.; Edeen, P. T.;
Floyd, M.; Ford, J. M.; Galvin, M.; Gerlach, J. L.; Grotzfeld, R. M.; Herrgard,
S.; Insko, D. E.; Insko, M. A.; Lai, A. G.; Lelias, J.; Mehta, S. A.; Milanov, Z. V.;
Velasco, A. M.; Wodicka, L. M.; Patel, H. K.; Zarrinkar, P. P.; Lockhart, D. J. A
small molecule-kinase interaction map for clinical kinase inhibitors. Nat.
Biotechnol. 2005, 23, 329–336.
2265
dx.doi.org/10.1021/jm101423y |J. Med. Chem. 2011, 54, 2255–2265