Pyrazolo-pyrimidines: A novel heterocyclic scaffold for potent and selective p38α inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.03.019

The synthesis and structure-activity relationships (SAR) of p38α MAP kinase inhibitors based on a pyrazolo-pyrimidine scaffold are described. These studies led to the identification of compound 2x as a potent and selective inhibitor of p38α MAP kinase with excellent cellular potency toward the inhibition of TNFα production. Compound 2x was highly efficacious in vivo in inhibiting TNFα production in an acute murine model of TNFα production. X-ray co-crystallography of a pyrazolo-pyrimidine analog 2b bound to unphosphorylated p38α is also disclosed.

Full text of DOI:10.1016/j.bmcl.2008.03.019

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2652–2657
Pyrazolo-pyrimidines: A novel heterocyclic scaffold for potent
and selective p38a inhibitors
Jagabandhu Das,* Robert V. Moquin, Sidney Pitt, Rosemary Zhang, Ding Ren Shen,
Kim W. McIntyre, Kathleen Gillooly, Arthur M. Doweyko, John S. Sack,
Hongjian Zhang, Susan E. Kiefer, Kevin Kish, Murray McKinnon, Joel C. Barrish,
John H. Dodd, Gary L. Schieven and Katerina Leftheris
Bristol-Myers Squibb Research and Development, PO Box 4000, Princeton, NJ 08543-4000, USA
Received 20 February 2008; revised 5 March 2008; accepted 6 March 2008
Available online 10 March 2008
Abstract—The synthesis and structure–activity relationships (SAR) of p38a MAP kinase inhibitors based on a pyrazolo-pyrimidine
scaffold are described. These studies led to the identification of compound 2x as a potent and selective inhibitor of p38a MAP kinase
with excellent cellular potency toward the inhibition of TNFa production. Compound 2x was highly efficacious in vivo in inhibiting
TNFa production in an acute murine model of TNFa production. X-ray co-crystallography of a pyrazolo-pyrimidine analog 2b
bound to unphosphorylated p38a is also disclosed.
Ó 2008 Elsevier Ltd. All rights reserved.
The mitogen activated protein kinase p38 is a serine/
threonine kinase that exists in four isoform (a, b, c,
and d). Expression of p38 isoforms varies among differ-
ent cell types of the immune system and p38a is believed
to be the predominant isoform involved in inflamma-
tion.1 Activation of p38a leads to upregulation of proin-
flammatory cytokines such as TNFa and IL-1b.2 These
cytokines are associated with the onset of inflammatory
and autoimmune diseases such as rheumatoid arthritis
(RA), Crohn’s disease, psoriasis, and inflammatory bo-
wel disease (IBD).3
In this letter, we describe the synthesis and structure–
activity relationship (SAR) studies of the pyrazolo-
pyrimidines that led to the identification of compound
2x as a potent and selective p38a inhibitor. Analog 2x
was also highly efficacious in vivo in an acute pharmaco-
dynamic model of TNFa production in mice.
The synthesis of the pyrazolo-pyrimidine analogs fol-
lows some general routes as described in Schemes 1–3.
Preparation of analogs with a C-6 amino substituent
(2b–2j) utilized readily available 4,6-dihydroxy-1-phe-
nyl-pyrazolo-[3,4-d]pyrimidine 75 (Scheme 1) which
upon treatment with phosphorous pentachloride and
phosphorous oxychloride under reflux formed the corre-
sponding 4,6-dichloro-adduct 8. Reaction of 8 with
aniline 9 in absolute ethanol in the presence of N,N-
diisopropylethylamine formed exclusively the 4-anili-
no-adduct 10 which served as a common intermediate
for the preparation of 6-amino substituted analogs.
Accordingly, treatment of 10 with N-methyl-homopiper-
azine in isopropanol at elevated temperature formed
analog 2d in 60% yield.
Identification of potent and selective p38a inhibitors as
clinical candidates has therefore generated considerable
interest in the pharmaceutical industry.4 We recently
disclosed a series of 5-cyanopyrimidines 1 (Fig. 1) as po-
tent inhibitors of the p38a MAP kinase.4a In an attempt
to design novel chemotypes we envisioned that the
incorporation of a ring constraint through cyclization
of the 5-cyano and the 6-aminoalkyl functions of 5-
cyanopyrimidines should lead to a pyrazolo-pyrimidine
2 (Fig. 1) which may serve as a novel scaffold for design-
ing inhibitors.
The synthesis of 6-unsubstituted analogs (2k–2r) fol-
lowed a similar route (Scheme 2). Accordingly, the reac-
tion of amino-pyrazole 6 with formamide6 at 190 °C
Keywords: p38a inhibitors; Pyrazolo-pyrimidines.
Corresponding author. Tel.: +1 609 252 5068; fax: +1 609 252
6804; e-mail: jagabandhu.das@bms.com
*
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.03.019

J. Das et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2652–2657
2653
Me
N
H
N
O
HN
N
5
O
N
N
cyclize
HN
6
NH2
Me
Me
1
p38α IC50 = 3 nM
hPBMC TNFα IC50 = 18 nM
Me
HN
N
H
N
O
Me
HN
O
H
4
N
N
R1
N
6
1
N
N
O
N
N
R2
N
N
2x
R3
p38α IC50 = 5 nM
hPBMC TNFα IC50 = 6 nM
2
Figure 1. Activities of cyanopyrimidine 1 and pyrazolo-pyrimidine 2x.
O
CN
NHNH2
NH2
NC
NC
OEt
N
a
b
NH2
N
N
+
NH2
N
5
4
6
3
Cl
N
OH
Me
N
c
N
d
N
N
H
+
N
N
N
Cl
OH
H2N
Me
N
O
8
9
7
Me
HN
Me
H
H
N
N
Me
HN
Me
O
O
e
f
N
N
6
N
N
N
N
N
Cl
N
N
N
Me
10
2d
Scheme 1. Reagents and conditions: (a) EtOH, 60 °C, 1 h, 60%; (b) sulfuric acid, 0 °C to rt, 7 h, 90%; (c) urea, 200 °C, 3 h, 70%; (d) POCl3, PCl5, D,
3 h, 93%; (e) i-Pr2NEt, EtOH, 60 °C, 4 h, 60%; (f) N-Me-homopiperazine, i-PrOH, 100 °C,3 h, 60%.

2654
J. Das et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2652–2657
O
OH
NH2
Cl
N
N
N
N
N
b
NH2
a
N
6
N
N
N
N
11
12
Me
Me
OEt
NHR1
HN
N
HN
N
O
c
N
O
d, e
N
N
N
6
N
N
2
13
Scheme 2. Reagents and conditions: (a) HCONH2, 190 °C, 3 h, 75%; (b) POCl3, 90 °C, 3 h, 85%; (c) ethyl-3-amino-4-methylbenzoate, EtOH, 140 °C,
1.5 h, 95%; (d) 1 N aq NaOH, THF–MeOH, rt, 16 h, 85%; (e) EDC, HOBt, THF–DMF, R2NH2, i-Pr2NEt, EtOH, 60 °C, 7 h, 55–100%.
formed 4-hydroxy-1-phenyl-pyrazolo-[3,4-d]pyrimidine
11. Reaction of 11 with phosphorous oxychloride at ele-
vated temperature afforded the 4-chloro derivative 12
which upon treatment with 2-methyl-5-carbethoxy-ani-
line formed 13. Saponification of ethyl ester 13 and sub-
sequent treatment of the corresponding acid with an
amine under standard coupling conditions afforded
analogs 2.
An alternate synthetic route was utilized for analogs
with a 6-amino substituent and is illustrated with the
synthesis of 2a (Scheme 3).7 Base-catalyzed condensa-
tion of phenyl hydrazine 3 and 2-amino-4,6-dichloro-
pyrimidine-5-carboxaldehyde 14 in refluxing THF
formed 4-chloropyrimidine 15 which upon reaction with
aniline 9 afforded analog 2a in 40% overall yield in two
steps.
The preliminary structure–activity optimization studies
of pyrazolo[3,4-d]pyrimidines (2) are outlined in Tables
1–3. Compounds were evaluated for their ability to inhi-
bit phosphorylation of substrate (myelin basic protein)
by recombinant human p38a.8 A peripheral blood
mononuclear cell-based assay (hPBMC) was used to
measure the ability of compounds to inhibit LPS-in-
duced TNFa production in human primary cells.8
Cl
Cl
N
N
OHC
Cl
a
N
PhNHNH2
3
N
N
NH2
+
N
NH2
15
14
Table 1 summarizes some of the salient SAR findings for
C-6 substitution in the pyrazolo-pyrimidine ring. The
amino analog 2a displayed excellent activity in the
p38a enzyme assay but was only moderately potent in
the cellular assay. The corresponding methylamino ana-
log 2b is roughly 5-fold less potent. Both enzyme and
cellular potencies of analog 2b could be significantly im-
proved by the attachment of a cyano ethyl side chain
(2c) to the methylamino group. A similar boost in cellu-
lar potency was observed upon introduction of a polar
pyrrolidinoethyl side chain (2h) in primary amine analog
2a. Homopiperazine analog 2d was identified as one of
the most potent analogs in the amine series. A wide vari-
ety of substituents are tolerated at the C-6 position.
Replacement of the amino group with hydrogen (2e),
Me
HN
Me
H
N
NHMe
Me
H2N
O
9
O
N
6
N
1
N
b
N
NH2
2a
Scheme 3. Reagents and conditions: (a) THF, Et3N, D, 20 min, 78%;
(b) EtOH–DMF, 146 °C, 0.5–1 h, 50%.

J. Das et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2652–2657
Table 3. SAR for side chain carboxamide modification
2655
Table 1. SAR for C-6 substitution
Me
HN
Me
HN
H
N
H
N
Me
R1
O
O
N
N
N
N
R2
N
N
N
N
Compound R2
p38a IC50 hPBMC TNFa
(nM)
a
Compound
R1
H
p38a
IC50 (nM)
hPBMC TNFa
IC50 (nM)
b
IC50 (nM)
a
b
2a
2b
2c
NH2
NHMe
N(Me)CH2CH2CN
3
14
5
193
513
43
2s
2t
7
4
635
19
OMe
Me
Bn
2e
2u
2v
2w
10
67
—
184
240
21
N
3-Pyridyl
n-Bu
1480
>2000
2d
8
16
N
Me
N
2e
2f
H
10
11
18
67
>2000
>2000
2x
2y
2z
5
3
8
6
O
OMe
Me
2g
H
N
N
N
64
<8
2h
4
36
NH
N
O
N
N
2i
2j
6
27
N
Et
a n = 4, variation in individual values, <20%.
b n = 3, variation in individual values, <25%.
12
290
a n = 4, variation in individual values, <20%.
b n = 3, variation in individual values, <25%.
nally, the replacement of the methoxy group with a pyrr-
olidino-ethoxy substituent (2i) resulted in significant
improvement in both biochemical and cellular potencies
and is consistent with our earlier observation with C-6
amino substituted analogs (2b and 2h). Because of its
excellent potency, the C-6 unsubstituted analog 2e was
selected for further SAR optimization.
alkoxy (2f), or alkyl (2g) retained most of the p38a en-
zyme activity. However, analogs 2f and 2g were found
to be significantly less potent in the cell-based assay. Fi-
Table 2. SAR for N-1 substitution
In order to further improve the cellular potency in the
methyl carboxamide series, we turned our attention to
the effect of substitution at the N-1 position of the pyr-
azolo-pyrimidine ring. Some salient SAR findings on
analog 2e are reported in Table 2. Replacement of the
phenyl group with alkyl groups showed mixed results.
The N-1 Me analog 2k was about 25-fold less potent
than the phenyl derivative 2e, while the tert-butyl analog
2l retained the potency in vitro. However, analog 2l is
significantly less potent than compound 2e in the cellular
assay. Similarly the phenyl ring could be replaced with a
benzyl (2m) with some attenuation in cellular potency.
However, the replacement of the phenyl ring with a pyr-
idyl ring (2n) resulted in several fold drop in both en-
zyme and cellular activities. In contrast, substitution(s)
on the phenyl ring (2o–2r) was tolerated. The 2-F-phenyl
analog 2o, and the 2,4,6-trimethylphenyl analog 2q were
equipotent to 2e in both the biochemical and cellular
assays.
Me
H
N
HN
N
Me
O
N
N
N
R3
Compound R3
p38a
IC50 (nM) IC50 (nM)
hPBMC TNFa
b
a
2k
2l
Me
tert-Butyl
Ph
253
13
10
13
67
9
—
>250
67
2e
2m
2n
2o
2p
2q
2r
CH2Ph
2-Pyridyl
2-F-phenyl
292
801
50
2,6-Di-Cl-phenyl
2,4,6-Tri-Me-phenyl
4-F-phenyl
6
>250
50
220
24
9
a n = 4, variation in individual values, <20%.
b n = 3, variation in individual values, <25%.
A more pronounced improvement in potency was ob-
served during the SAR investigation of the carboxamide

2656
J. Das et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2652–2657
modification in the pendant aniline ring (Table 3). It
should be noted that the relative positions of the methyl,
carboxamide, and the amino group in the pendant ani-
line ring were critical for the p38a enzyme activity. Re-
moval or transposition of the methyl and the
carboxamide groups resulted in several orders of magni-
tude loss in biochemical potency (data not shown). The
primary carboxamide analog 2s was significantly less
potent in the cellular assay despite retaining the bio-
chemical activity relative to the methyl amide 2e. In con-
trast a substantial increase in both biochemical and
cellular potencies was observed upon the substitution
of the methyl amide with a methoxy amide (2t). The
replacement of the methyl amide with either a benzyl
(2u), pyridyl (2v) or long chain alkyl amide (2w) was det-
rimental for enzyme activity. A significant improvement
in biochemical potency and a more dramatic increase in
cellular potency were observed upon replacement of the
methyl amide with five-membered heterocyclic amides,
most notably iso-oxazole (2x), and N–Et pyrazole (2z).
Analogs 2x and 2z were thus identified as two of the
most potent p38a inhibitors in this series.
4.1 h). The in vivo efficacy of 2x was demonstrated in
an acute pharmacodynamic model of LPS-induced
TNFa production in mice. Mice were dosed orally with
2x at 5 mg/kg, 5 h prior to LPS administration and the
plasma TNFa level was measured 90 min after the
LPS challenge. In this model, compound 2x inhibited
circulating TNFa level by >70%.8
To understand the binding mode of this novel pyrazolo-
pyrimidine class of p38a inhibitors, compound 2b was
co-crystallized with purified, unphosphorylated p38a
MAP kinase and the X-ray structure of the complex
was determined.8,9 The key binding interactions between
compound 2b and the p38a enzyme are illustrated in
Figure 2.
The X-ray structure of 2b complexed with p38a reveals a
combination of H-bonding and hydrophobic interac-
tions. Consistent with our previous findings with related
p38a inhibitors, the pendant carboxamido-aniline ring
occupies an angular hydrophobic pocket and makes
key hydrogen-bond interactions. More specifically, the
aniline NH is in H-bonding distance with Thr106
˚
Based on its p38a inhibitory potency, acceptable liability
profile, and in vitro metabolic stability, analog 2x was
selected for further evaluation. Compound 2x was char-
acterized as a highly potent inhibitor of human p38a en-
zyme with a Ki value of 0.2 nM, excellent metabolic
stability with microsomal metabolic rates of 0.000 and
0.005 nmol/min/mg protein, respectively, in humans
and rats. Analog 2x also displayed a clean profile
against CYP 450 inhibition with IC50 values of
>40 lM for 1A2, 2C9, 2C19, 2D6, and 3A4 isozymes.
The kinase selectivity profile of 2x was determined
against several receptor and non-receptor tyrosine ki-
nases, as well as serine/threonine kinases (Table 4).
Compound 2x was shown to be at least 1000-fold selec-
tive over 20 different kinases. Furthermore, a close ana-
log in this series, compound 2b was found to be a
selective p38a inhibitor (IC50 = 14 nM) over p38c and
d (IC50s > 30 lM). No selectivity was observed in the
inhibition of p38b isoform (IC50 = 12 nM).
(2.01 A) and the carboxamido NH is engaged in an H-
˚
bond interaction with Glu71 (2.07 A). The hinge region
residue, Met109 NH forms an H-bond with 20 position
nitrogen of the pyrazole. This 20 pyrazole nitrogen cor-
responds to the nitrile N of the cyanopyrimidine 1.
The phenyl ring attached to the pyrazolo-pyrimidine
ring is skewed out of plane. This orientation may ex-
plain some of the SAR observed regarding substitution
at the ortho-position of the pendant phenyl ring.
In conclusion, we have identified a series of exquisitely
potent and selective p38a inhibitors using a novel pyraz-
olo-pyrimidine scaffold. SAR optimization on different
fragments of this ring system led to the identification
of analog 2x as one of the most potent p38a inhibitor
in this series with single digit nanomolar potency for
the inhibition of TNFa release in human peripheral
blood mononuclear cells (PBMCs). Oral efficacy was
The pharmacokinetic profile of compound 2x was deter-
mined in mice. Upon oral dosing at 10 mg/kg, analog 2x
was determined to be highly bioavailable (F% 60) with a
low clearance rate (0.3 L/kg/h), high volume of distribu-
tion (1.3 L/kg), and acceptable terminal half-life (t1/2
=
Table 4. Kinase selectivity profile of 2x
Kinase
IC50 (lM)
Kinase
IC50 (lM)
p38a
KDR
Akt
0.005
>10
>50
>50
>10
>50
>30
>50
>5
Jak3
Lck
>50
>50
>50
>50
>30
>50
46
FGFR1
SYK
CaMKII
Cdk2
Itk
MK2
ERK
PKA
PKCa
PKCd
PKCs
PKCf
Raf
FGF
188
40
GSK3
HER1
IGF-1R
Figure 2. Binding interactions between 2b and unphosphorylated p38a
based on X-ray crystallographic analysis. Hydrogen-bond distances
are given in angstroms with key protein residues labeled.
>50
>25
>40
>40

J. Das et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2652–2657
2657
4. (a) Liu, C.; Wrobleski, S. T.; Lin, J.; Gillooly, K. M.;
Mcintyre, K.; Pitt, S.; Shen, D. R.; Shuster, D. J.; Zhang,
H.; Doweyko, A. M.; Sack, J. S.; Barrish, J. C.; Dodd, J.
H.; Schieven, G. L.; Leftheris, K. J. Med. Chem. 2005, 48,
6261; (b) Hynes, J.; Leftheris, K. Curr. Top. Med. Chem.
2005, 5, 967; (c) Miwatashi, S.; Arikawa, Y.; Kotani, E.;
Miyamoto, M.; Naruo, K.; Kimura, H.; Tanaka, T.; Asahi,
S.; Ohkawa, S. J. Med. Chem. 2005, 48, 5966; (d) Goldstein,
D. M.; Gabriel, T. Curr. Top. Med. Chem. 2005, 5, 1017; (e)
Diller, D. J.; Lin, T. H.; Metzger, A. Curr. Top. Med.
Chem. 2005, 5, 953; (f) Dominguez, C.; Powers, D. A.;
Tamayo, N. Curr. Opin. Drug Discov. Devel. 2005, 8, 421.
5. Cheng, C. C.; Robins, R. K. J. Org. Chem. 1958, 23, 852.
6. Cheng, C. C.; Robins, R. K. J. Org. Chem. 1956, 21, 1240.
7. Seela, F.; Steker, H. Helv. Chim. Acta 1986, 69, 1602.
8. For a description of the biological assays, and X-ray crystal
structure determination protocol, see Hynes, J., Jr.; Dyck-
man, A. D.; Lin, S.; Wrobleski, S. T.; Wu, H.; Gillooly, K.
M.; Lonial, H.; Loo, D.; McIntyre, K. W.; Pitt, S.; Shen, D.
R.; Shuster, D. J.; Zhang, X.; Behnia, K.; Marathe, P. H.;
Doweyko, A.; Barrish, J.; Dodd, J.; Schieven, G.; Leftheris,
K. J. Med. Chem. 2008, 51, 4.
also demonstrated with this analog in an acute murine
model of TNFa inhibition. In addition, the molecular
basis for p38a inhibition of this class of inhibitor was
established via X-ray crystal structure determination of
an enzyme–inhibitor complex.
References and notes
1. (a) Hale, K. K.; Trollinger, D.; Rihanek, M.; Manthey, C.
L. J. Immunol. 1999, 162, 4246; (b) Allen, M.; Svensson, L.;
Roach, M.; Hambor, J.; McNeish, J.; Gabel, C. A. J. Exp.
Med. 2000, 191, 859; (c) Fearns, C.; Kline, L.; Gram, H.; Di
Padova, F.; Zurini, M.; Han, J.; Ulevitch, R. J.
J. Leukocyte Biol. 2000, 67, 705.
2. (a) Schieven, G. L. Curr. Top. Med. Chem. 2005, 5, 921; (b)
Saklatvala, J. Curr. Opin. Pharm. 2004, 4, 372; (c) Kumar, S.;
Boehm, J.; Lee, J. C. Nat. Rev. Drug Discov. 2003, 2, 717; (d)
Adams, J. L.; Badge, A. M.; Kumar, S.; Lee, J. C. Prog. Med.
Chem. 2001, 38, 1; (e) Lee, J. C.; Laydon, J. T.; McDonnell, P.
C.; Gallagher, T. F.; Kumar, S.; Green, D.; McNulty, D.;
Blumenthal, M. J.; Heys, J. R. Nature 1994, 372, 739.
3. (a) Wagner, G.; Laufer, S. Med. Res. Rev. 2006, 26, 1; (b)
Westra, J.; Limburg, P. C. Mini-Rev. Med. Chem. 2006, 6, 867.
9. The X-ray coordinates have been deposited with RCSB
Protein Data Bank Database (RCSB ID Code rcsb046732
and PDB ID Code 3CG2).