Development of pyrimidine-based inhibitors of Janus tyrosine kinase 3

Article abstract of DOI:10.1016/j.bmcl.2006.08.022

A new class of pyrimidine-based Janus tyrosine kinase 3 (JAK3) inhibitors are described. Many of these inhibitors showed low nanomolar activity against JAK3.

Full text of DOI:10.1016/j.bmcl.2006.08.022

Bioorganic & Medicinal Chemistry Letters 16 (2006) 5633–5638
Development of pyrimidine-based inhibitors
of Janus tyrosine kinase 3
Jack J. Chen,* Kumar D. Thakur,ꢀ Michael P. Clark,ꢁ Steven K. Laughlin,§
Kelly M. George, Roger G. Bookland, Jan R. Davis, Edward J. Cabrera,
Vijay Easwaran, Biswanath De and Y. George Zhang
Procter and Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason-Montgomery Rd., Mason, OH 45040, USA
Received 14 June 2006; accepted 2 August 2006
Available online 24 August 2006
Abstract—A new class of pyrimidine-based Janus tyrosine kinase 3 (JAK3) inhibitors are described. Many of these inhibitors
showed low nanomolar activity against JAK3.
Ó 2006 Elsevier Ltd. All rights reserved.
Janus tyrosine kinases (JAKs) are a small family of
structurally and functionally related non-receptor
tyrosine kinases, including JAK1, JAK2, JAK3, and
tyrosine kinase 2 (TYK2).1 JAK-mediated tyrosine
phosphorylations of cytokine receptors and signal trans-
ducers and activators of transcriptions (STATs) are the
important signal transduction pathways used by many
cytokines, growth factors, and interferons.2–4 Unlike
the rest of JAK family members that are widely
expressed in many mammalian tissues, JAK3 is predom-
inantly located in endoplasmic membranes of hemato-
poietic cells, and specifically associates with the
common cytokine receptor c chain (cc) which is a shared
component of the receptors for IL-2, IL-4, IL-7, IL-9,
and IL-15.5 The pivotal roles in signaling through cc-
containing cytokine receptors and unique tissue distri-
bution make JAK3 an ideal biological target to manage
the abnormal cytokine activities implicated in many can-
cer cells and inflammatory lymphocytes.5
of antibiotic undecylprodigiosin, effectively blocked
JAK3-dependent T-cell proliferation with an IC50 of
7.5 lM. This compound showed 2-fold greater specifici-
ty versus JAK2-dependent cell proliferation. CP-690,550
is a JAK3-inhibitory small molecule progressed to phase
II clinic trials for treatment of acute rejection following
kidney transplant surgeries.7 The oxindole 1 was devel-
oped by Aventis from a series designed originally for
CDK kinase.8 In spite of its inherited activities toward
CDK kinases, this compound is so far the only reported
molecule showing both low nanomolar potency and
near 20-fold JAK3/JAK2 selectivity in enzymatic assays
(see Fig. 1).
H
N
N
C11H23
N
O
N
NH
HCl
N
O
N
NH
NC
In comparison to patent claims, the chemistry publica-
tions pertaining to JAK3 inhibition are relatively scarce.
Stepkowski et al. reported that PNU156804,6 an analog
PNU156804
CP-690,550
H
N
Keywords: Tyrosine kinase inhibitor; JAK3; JAK2; Pyrimidine.
*
N
Corresponding author. E-mail: jack.01.chen@gmail.com
Current address: AstraZeneca Pharmaceuticals, 35 Gatehouse Drive,
O
H
N
H
N
ꢀ
N
N
N
Waltham, MA 02451, USA.
Current address: Metastatix, 1860 Montreal Road, Tucker, GA
N
N
ꢁ
30084, USA.
Current address: Girindus America, 8560 Reading Road, Cincinnati,
OH 45215, USA.
1
2
§
Figure 1. JAK3 inhibitors.
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.08.022

5634
J. J. Chen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5633–5638
Our interest in JAK3 aroused from pursuing small
molecular therapeutics for autoimmune diseases.
Screening the internal kinase compound collections
and hit follow-up resulted in 1,4-disubstituted pyrimi-
dines, represented by structure 2. Further chemistry ef-
forts pursuing this lead produced a novel class of
potent and selective JAK3 inhibitors.
(e.g., 13 and 14). In spite of low nanomolar IC50
against JAK3, this series in general is more potent
toward JAK2. In our hand, the benchmark CP-
690,550 showed no selectivity in favor of JAK2
although it has been reported to be 20-fold more
selective toward JAK3.10 And the oxindole 1 is
around 16-fold more selective toward JAK3.
Compound 2 and its analogs were easily synthesized
with the sequences outlined in Scheme 1, starting from
commercially available 4-chloro-2-thiomethyl-pyrimi-
dine 3. The ‘flipped’ version (8) of 2 was prepared in a
similar manner (Scheme 2).
Compound 17 was prepared in search for alterna-
tive benzimidazole moieties using the sequences
outlined in Scheme 3. Equipped with a 6-cyano
benzimidazole and a small alkyl amine, this com-
pound maintains the potency of the azabenzimidaz-
ole analogs and is 3-fold more selective in favor of
JAK3.
The inhibitory activity of this series against JAK3
and JAK2 was assessed by the kinase-Glo lumines-
cent assay (Table 1) with both CP-690,550 and the
To further augment the JAK3/JAK2 selectivity,
structural extension was added to the 6-positions
of the pyrimidine core via the sequences outlined
in Scheme 4, furnishing the tri-substituted pyrimdine
series.
oxindole
1
compound as benchmark references.9
The results suggested that both enzymes prefer the
benzimidazole moiety at the 2-position of the pyrim-
idine core. Switching sequence alignment of amine
and benzimidazole usually results in at least sevenfold
potency loss (e.g., 8). The presence of 2-nitrogen
atom in the benzimidazole moiety is crucial since its
removal resulted in complete loss of activity (e.g.,
15). And 6-azabenzoimidazole regioisomers (e.g., 2)
are usually 20- to 30-fold more potent than the 5-
aza regioisomers (e.g., 5). Methylation on the 2-posi-
tion of the benzimidazole (e.g., 16) or the nitrogen of
the amine moieties (e.g., 12) resulted in great reduc-
tion of activity. The a-methyl group on the amine
side seems also critical since both its removal and
restraint caused almost complete loss of potency
The results (Table 2) unfortunately suggested that nei-
ther JAK3 nor JAK2 seems sensitive toward the changes
at the 6-position, the modifications nevertheless provide
the opportunities for improving the pharmacokinetic
profile of this class in future.
To further investigate the in vivo behaviors of the
pyrimidine class, selected compounds were tested in
TF-1 cell proliferation assays that assess their inhibitory
activities toward JAK3 or JAK2-dependent signaling
transductions (Table 3).11 In this measurement,
H
N
S
N
3
Cl
S
N
a
N
N
4
N
N
H
N
b, c
H
N
N
N
N
N
+
N
N
N
N
2
5
i
Scheme 1. Reagents and conditions: (a) (S)-1-phenylethanamine, Pr2NEt, Bu4NI, DMA, microwave (MW), 100 °C, 30 min, 88%; (b) mCPBA,
DCM, 0 °C, 10 min; (c) 5-azabenzimidazole, K2CO3, DMF, 140 °C, 76%.
N
N
S
N
Cl
S
N
N
S
N
N
a
+
N
N
N
N
N
3
6
7
N
N
H
N
b, c
H
N
N
N
N
N
+
N
N
N
N
8
9
Scheme 2. Reagents and conditions: (a) 5-azabenzimidazole, NaH, Bu4NI, DMA, rt, 49%. (b) mCPBA, DCM-MeOH, 0 °C, 10 min; (c) (S)-a-ethyl
i
benzylamine, Pr2NEt, DMA, MW, 100 °C, 30 min, 70%.

J. J. Chen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5633–5638
Table 1. JAK3 and JAK2 kinase data for di-substituted pyrimidine analogs
5635
R1
N
R2
N
a
JAK3 IC50 (nM)
a
JAK2 IC50 (nM)
Compound
R1
R2
CP-690550
1
—
—
—
—
13b
38
11b
600
H
N
2
10
11
12
13
14
21
92
4
8
H
N
F
H
N
Cl
N
N
1928
860
51
N
N
153
>104
7218
H
N
>104
9097
H
N
N
N
H
N
N
5
8
700
145
51
49
N
N
H
N
N
H
N
15
>104
>104
N
N
N
N
N
H
N
16
17
>104
>104
124
N
N
H
N
45
NC
O
a SD for enzyme assays were typically 25% of the mean or less.
b Mean, n > 10.
N
H
H
N
S
N
N
N
N
a
b, c, d
O
O
3
N
N
NC
17
Scheme 3. Reagents and conditions: (a) (S)-1-methoxypropan-2-amine, iPr2NEt, Bu4NI, DMA, MW, 100 °C, 30 min, 88%; (b) mCPBA, DCM, 0 °C,
10 min; (c) 3-amino-4-nitrobenzonitrile, KtOBu, DMF, MW, 100 °C, 15 min, 76%; (d) SnCl2Æ2H2O, DMF, (MeO)3CH, MW, 120 °C, 30 min, 20%.
CP-690,550 is about 10-fold more selective toward
JAK3, whereas oxindole 1 showed no selectivity. The
pyrimidine compound 17 showed 5-fold JAK3/JAK2
selectivity.
In summary, we have reported a novel series of pyrimi-
dine-based Janus tyrosine kinase 3 inhibitors. Many of
the compounds in this series showed strong potency in
the JAK3 enzymatic assay below the 100 nM level.

5636
J. J. Chen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5633–5638
S
N
Cl
S
N
X
S
N
X
a
R2
b
R2
N
N
or c
N
Cl
Cl
R3
N
e
X
N
N
R2
5-Aza regioisomers
+
N
N
d
R3
N
N
X=N, O
f, g
N
X
R2
N
NC
R3
Scheme 4. Reagents and conditions: (a) amine or phenol, iPr2NEt, Bu4NI, DMA, MW; (b) R2B(OH)2, Pd(Ph3P)4, dioxane, K2CO3, MW; (c) amine,
iPr2Net, DMF, MW; (d) mCPBA, DCM, 0 °C, 10 min; (e) 5-azabenzimidazole, K2CO3, DMF, 140 °C; (f) 3-amino-4-nitrobenzonitrile, KtOBu,
DMF, MW, 100 °C; (g) SnCl2Æ2H2O, DMF, TMOF, MW, 120 °C, 30 min.
Table 2. JAK3 and JAK2 kinase data for tri-substituted pyrimidine analogs
R1
N
X
R2
N
X=O, N
R3
a
JAK3 IC50 (nM)
a
JAK2 IC50 (nM)
Compound
R1
X-R2
R3
N
N
H
N
Ph
Ph
18
19b
61
7b
N
N
N
N
H
N
19
20
21
22
23
19
Cl
N
N
O
27
120
131
331
17
68
F
N
N
O
F
H
N
N
N
Cl
113
279
N
N
H
N
Ph
N
H
N
Ph
Ph
24
25
180
60
59
OH
H
N
16
O
(continued on next page)

J. J. Chen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5633–5638
5637
Table 2 (continued)
a
JAK3 IC50 (nM)
a
JAK2 IC50 (nM)
Compound
R1
X-R2
R3
H
N
Ph
26
62
20
N
H
N
Ph
Ph
27
28
470
49
167
12
N
H
N
OMe
OMe
H
N
Ph
Ph
Ph
29
30
31
32
245
42
77
18
N
N
H
N
CN
H
N
167
165
56
CN
H
N
446
O
O
N
N
H
N
Ph
33
34
31
83
29
28
NC
O
O
H
N
F
H
N
35
51
141
O
O
a SD for enzyme assays were typically 25% of the mean or less.
b Mean, n > 10.
Table 3. Inhibition of proliferation of TF-1 cells, induced by either
IL-3 (for JAK2 activation) or IL-4 (for JAK3 activation), by
pyrimidine analogs
References and notes
1. Silvennoinen, O.; Ihle, J. N.; Schlessinger, J.; Levy, D. E.
Nature 1993, 366, 583.
2. Thomis, D.; Berg, L. Curr. Opin. Immunol. 1997, 9, 541.
3. O’Shea, J. J.; Pesu, M.; Borie, D. C.; Changelian, P. S.
Nat. Rev. Drug Discov. 2004, 3, 555.
Compound
Inhibition of IL-4
induced TF-1 cell
proliferation
Inhibition of IL-3 induced
TF-1 cell proliferation
a
IC50 (nM)
a
IC50 (nM)
4. Notarangelo, L.; Giliani, S.; Mazza, C. Immunol. Rev.
2000, 178, 39.
5. O’Shea, J. J. Ann. Rheumatic Dis. 2004, 63, ii67.
6. Stepkowski, S. M.; Erwin-Cohen, R. A.; Behbod, F.;
Wang, M.-E.; Qu, X.; Tejpal, N.; Nagy, Z. S.; Kahan, B.
D.; Kirken, R. A. Blood 2002, 99, 680.
7. Kudlacz, E.; Perry, B.; Sawyer, P.; Conklyn, M.;
McCurdy, S.; Brissette, W.; Flanagan, M.; Changelian,
P. Am. J. Transplant. 2004, 4, 51.
8. Adams, C.; Aldous, D. J.; Amendola, S.; Bamborough,
P.; Bright, C.; Crowe, S.; Eastwood, P.; Fenton, G.;
Foster, M.; Harrison, T. K. P.; King, S.; Lai, J.;
Lawrence, C.; Letallec, J.-P.; McCarthy, C.; Moorcroft,
N.; Page, K.; Rao, S.; Redford, J.; Sadiq, S.; Smith, K.;
Souness, J. E.; Thurairatnam, S.; Vine, M.; Wyman, B.
Bioorg. Med. Chem. Lett. 2003, 13, 3105.
CP-690,550
80b
600
335
2010
90
800b
500
635
1400
440
580
610
600
30
1
10
11
17
19
25
33
35
370
400
170
20
a SD for TF-1 cellular assay were typically 25% of the mean or less.
b Mean, n > 10.
And few analogs showed promising selectivity for JAK3
relative to JAK2 in cellular assays.
9. Promega’s Kinase-GloTM Luminescent Kinase Assay is a
homogeneous non-radioactive method of determining the
activity of purified kinases by quantifying the amount of
ATP remaining in solution following a kinase reaction.
The assay procedure involves addition of a single reagent
Acknowledgment
The author thank Dr. Adam Golebiowski for proof-
reading the manuscript.

5638
J. J. Chen et al. / Bioorg. Med. Chem. Lett. 16 (2006) 5633–5638
(Kinase-GloÒ Reagent) directly to a completed kinase
nant. Sigma #I 4269) at 1 nM was used to stimulate the
cells. Proliferation of the TF-1 cells was used to measure
the responses through JAK2 (IL-3) or JAK3 (IL-4) and
the inhibition of such response by specific compounds.
Five replicates per compound were seeded at a concen-
tration of 5000 cells/well in a 100 ll volume. Five different
concentrations of experimental compounds (5 replicates
each) were used to determine their EC50. After 4 days of
incubation, cell proliferation was measured during the last
18 h of incubation, using an ELISA (BrdU, chemilumi-
nescence) kit (Roche Diagnostics, Indianapolis, IN). The
relative luminescence units (RLU) were measured using a
Packard Fusion instrument (Perkin Elmer). The RLU
were used to calculate EC50s (Prism4 software).
reaction. This addition results in the generation of a
luminescent signal correlated with the amount of ATP
present and inversely proportional to the amount of
kinase activity.
10. CP-690,550, reported to possess 20:1 selectivity favoring
JAK3 over JAK2 (Ref. 7), showed virtually no selectivity
in a separate Kinase ProfilerTM Assays (radiometric filter
binding assays) run by Upstate Cell Signaling Solutions
(JAK3 IC50 = 5 nM, JAK2 IC50 = 6 nM).
11. All compounds were tested using an assay developed with
TF-1 (1, 2) cells (CRL-2003 from ATCC) which prolifer-
ate in response to GM-CSF, IL-3 and IL-4. IL-3 (Human
recombinant Sigma #I 1646) or IL-4 (Human recombi-