Development of N-4,6-pyrimidine-N-alkyl-N′-phenyl ureas as orally active inhibitors of lymphocyte specific tyrosine kinase

Article abstract of DOI:10.1016/j.bmcl.2006.04.072

A new class of lymphocyte specific tyrosine kinase (lck) inhibitors based on an N-4,6-pyrimidine-N-alkyl-N′-phenyl urea scaffold is described. Many of these compounds showed low-nanomolar inhibition of lck kinase activity as well as IL-2 synthesis from Ju

Full text of DOI:10.1016/j.bmcl.2006.04.072

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3646–3650
Development of N-4,6-pyrimidine-N-alkyl-N0-phenyl ureas as orally
active inhibitors of lymphocyte specific tyrosine kinase
Jennifer A. Maier, Todd A. Brugel,* Mark Sabat, Adam Golebiowski,
Matthew J. Laufersweiler, John C. VanRens, Corey R. Hopkins, Biswanath De,
Lily C. Hsieh, Kimberly K. Brown, Vijayasurian Easwaran and Michael J. Janusz
Procter & Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason-Montgomery Rd. Mason, OH 45040, USA
Received 28 March 2006; revised 21 April 2006; accepted 24 April 2006
Available online 8 May 2006
Abstract—A new class of lymphocyte specific tyrosine kinase (lck) inhibitors based on an N-4,6-pyrimidine-N-alkyl-N0-phenyl urea
scaffold is described. Many of these compounds showed low-nanomolar inhibition of lck kinase activity as well as IL-2 synthesis from
Jurkat cells. One of these analogs, 7i, was shown to be orally efficacious by in vivo testing in a rat adjuvant-induced arthritis study.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Lymphocyte specific protein tyrosine kinase (Lck), one
of the eight members of the Src family of tyrosine ki-
nases, is expressed on T-lymphocytes and natural killer
cells. Lck plays a critical role in T-cell antigen receptor
(TCR) signal transduction. Upon presentation of anti-
gens associated with the MHC complex to the T-cell ex-
pressed CD4/CD8 co-receptors, auto-phosphorylation
of Lck occurs initiating additional kinase activity.1,2
The immunoreceptor tyrosine activation motif (ITAM)
of the f chain of the CD3 subunit is phosphorylated
by Lck, leading to binding of ZAP-70 through its SH2
domain. This in turn leads to phosphorylation of
ZAP-70 by Lck.3 Following activation of additional
signaling molecules within the T-cell, Ca2+ is released
inducing calcineurin activation and the ultimate produc-
tion of interleukin-2 (IL-2) and interferon-c (INF-c)
cytokines.4 Expression of these cytokines leads to fur-
ther activation and proliferation of T-lymphocytes char-
acterized as the immune response.
been shown to lack proper T-cell development and acti-
vation.5 Therefore, selective inhibition of Lck by small
molecule inhibitors would provide an orally bioavailable
therapy for suppressing cytokine production and ulti-
mately treating autoimmune disorders.
We have previously investigated a series of trisubstituted
ureas for the inhibition of TNF-a production. Using
X-ray co-crystallization studies, the potent analog 1
(Fig. 1) was shown to bind to p38a MAP kinase.6 The
compound adopted a pseudo-bicyclic conformation
through internal hydrogen bonding between the urea
NH and one of the pyrimidine nitrogens. Compounds
from this class containing the 2,4-pyrimidine
substitution were not active when screened against lck
HO
Cl
HN
The over-expression of pro-inflammatory cytokines has
been implicated in a number of autoimmune disorders
such as rheumatoid arthritis (RA), inflammatory bowel
disease (IBD), psoriasis, systemic lupus erythematosus
(SLE), and organ graft rejection. Lck ꢀ/ꢀ mice have
HN
N
N
N
N
HN
Cl
Cl
O
N
HN
F
N
O
F
1
2
Keywords: Lck kinase; IL-2 cytokine; Rheumatoid arthritis; Trisubsti-
tuted ureas.
TNF-a IC50 = 122 nM
Lck IC50 = >10 μM
Lck IC50 = 3.86 μM
*
Corresponding author. Tel.: +1 513 622 4498; fax: +1 513 622
3681; e-mail: brugel.ta@pg.com
Figure 1. Urea inhibitors of TNF-a and lck kinase.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.04.072

J. A. Maier et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3646–3650
3647
Table 1. In vitro activity for N-methyl ureas
kinase. However, the closely related analog 2 (Fig. 1),
containing the 4,6-pyrimidine substitution, was shown
to have moderate potency toward inhibition of lck
kinase activity with an IC50 of 3.86 lM.
R
R1
N
N
HN
N
N
O
Recently, a patent application from Novartis disclosed
their efforts toward the development of a similar series
of trisubstituted ureas as protein kinase inhibitors.7 In
this manuscript, we would like to report the results of
our own independent efforts in the development of tri-
substituted ureas as lck inhibitors.
H
a
Compound
R
R1
Lck IC50 (lM)
2
F
2,6-di-Cl
2-Cl
2-Cl
3.86
1.82
3.13
7a
7b
F
Et2N
N
7c
7d
7e
7f
2,6-di-Cl
0.301
0.190
4.29
Synthesis of the 4,6-pyrimidine urea class of lck inhibi-
tors is exemplified in Scheme 1. Boc protection of 4-mor-
pholinoaniline (3) gave Boc-aniline 4. Addition of 4
to 4,6-dichloropyrimidine gave the anilino-pyrimidine
5. A second displacement with methyl amine using
microwave irradiation (250 W, CEM DiscoverTM micro-
wave synthesizer) led to the formation of diamino
pyrimidine 6. Urea formation was effected by reaction
of 5 with 2,6-dichlorophenylisocyanate. Finally, Boc
deprotection gave the desired urea 7c. The remaining
analogs in this report were synthesized using a modified
version of this route.
O
O
O
N
N
2-Cl-6-Me
2-Cl
N
N
N
2,6-di-Cl
0.006
N
N
N
7g
7h
7i
2-Cl-6-Me
2-Cl
0.009
0.708
0.087
0.028
0.096
0.418
>10
All compounds in this study were tested for inhibition of
lck kinase activity.8 When the N-2,6-dichlorophenyl
group was replaced with 2-chlorophenyl (7a), the lck
inhibition improved marginally (Table 1). Further
replacement of the 4-fluoroaniline with a 4-diet-
hylaminoaniline gave compound 7b which showed com-
parable lck inhibition to 2. Morpholinyl analogs 7c and
7d containing an N0-2,6-disubstituted phenyl substituent
showed greatly improved lck potency. However, the cor-
responding N0-2 substituted phenyl 7e showed much re-
duced activity. A similar trend was seen with the
corresponding N-methyl piperazine analogs 7f–7h,
although the compounds were relatively more potent
overall. We also observed sub-100 nM lck inhibition
for the 4-(2-diethylamino)-ethylether analogs 7i–7k.
An approximate 5-fold loss in potency was seen
when substituting an N0-2,6-difluorophenyl for the
N0-2,6-dichlorophenyl analog (7l vs 7i). As with the
N0-2-chlorophenyl analogs, a significant loss of activity
was seen with the N0-2-methoxyphenyl analog 7m.
O
N
2,6-di-Cl
2-Cl-6-Me
2-Cl,6-CF3
2,6-di-F
2-MeO
O
O
O
O
N
N
N
N
7j
7k
7l
7m
a Standard deviation for the assay was 15% or less.
in Table 2. The N-benzyl analog 8a containing an N0-di-
substituted phenyl group showed modest activity.
However, other N-benzyl analogs containing an
N0-chlorophenyl group (8b and 8c) lost all inhibitory
kinase activity. When the N-dimethylaminopropyl
Having explored some general SAR about the 6-pyrim-
idyl and urea N0-positions, we next turned our attention
to the urea N-position. We replaced the N-methyl group
with various alkyl and aryl substituents as summarized
Cl
N
N
N
N
NH2
NHBoc
N
N
HN
Boc
Boc
N
Cl
N
N
Cl
H
HN
N
O
a
b
c
d,e
N
O
N
O
N
O
N
O
N
O
3
4
5
6
7c
Scheme 1. Synthesis of 4,6-pyrimidine urea 7c. Reagents and conditions: (a) Boc2O, toluene, 110 ꢁC; (b) 4,6-dichloropyrimidine, NaH, THF, 65 ꢁC,
80%; (c) MeNH2, dioxane, microwave, 125 ꢁC, 15 min, 79%; (d) 2,6-dichlorophenylisocyanate, toluene, 110 ꢁC, 80%; (e) TFA, CH2Cl2, rt.

3648
J. A. Maier et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3646–3650
Table 3. 6-Pyrimidine substitution SAR
group was incorporated (8d–8g), kinase inhibition was
regained. The most potent analog, 8g, contained the
4-morpholinoaniline group. When the dimethylamino-
propyl group was replaced with a methylpiperazine
(8h), only minimal attenuation of activity was observed.
Substitution with a 4-pyridyl group (8i) conversely
caused a complete loss of activity.
Cl
N
N
HN
R
Cl
N
N
R1
O
H
Compound
R
R1
Lck IC50
(lM)
a
We also investigated various alkylamino substitutions
at the 6-pyrimidyl position to replace the anilines used
to this point (Table 3). The methylamine analog 9a, a
urea regioisomer of 7i, was inactive toward lck inhibi-
tion. The N-dimethylaminopropyl analog 9b showed
some modest activity, but still much reduced compared
to the corresponding 6-pyrimidyl aniline derivatives.
While maintaining the N-methyl-N0-2,6-dichlorophenyl
urea functionality constant, we looked at various 6-
alkylamino pyrimidine analogs (9c–9f). Several combi-
nations of linear and cyclic alkyl groups of various
tether lengths were tested, but none of these analogs
showed any appreciable activity. These results indicate
the importance of having an anilino substituent off the
6-position of the pyrimidine in order to produce lck
inhibition.
9a
9b
Me
Me
4-Et2N–(CH2)2OPh
>10
2.35
N
N
9c
9d
9e
9f
Me
Me
Me
Me
>10
5.99
>10
>10
N
O
O
N
a Standard deviation for the assay was 15% or less.
Next, we investigated an alternative substitution pattern
about the N0-phenyl group of the urea. An initial lead in
the N0-2,5-disubstituted phenyl analogs was the 5-meth-
oxy-2-methylphenyl derivative 10a (Table 4). We discov-
ered that demethylation of the methoxy group in 10a
Table 2. Urea N-alkyl and phenyl SAR
Cl
Table 4. N0-2,5-Disubstituted phenyl urea SAR
R
N
N
HN
R2
N
N
R1
O
R
R1
H
N
N
HN
N
H
N
O
a
Compound
R
F
R1
R2 Lck IC50 (lM)
8a
CH2Ph
Me 1.45
a
Compound
R
R1
Lck IC50 (lM)
O
O
N
N
10a
MeO
OH
2.58
8b
8c
8d
8e
8f
NEt2
NEt2
NEt2
CH2Ph
H
H
H
>10
>10
4.49
10b
10c
16a
16b
18a
18b
18c
0.003
0.004
0.004
0.032
0.149
0.052
0.003
N
N
N
OH
O
N
O
O
O
O
O
O
NH
O
N
N
N
N
N
N
N
N
NEt2
Me 0.971
Cl 0.229
Cl 0.155
Cl 0.230
Cl >10
O
N
O
N
8g
8h
8i
N
S
HN
O
O
O
O
N
N
N
HN
N
H
N
F3C
N
O
a Standard deviation for the assay was 15% or less.
a Standard deviation for the assay was 15% or less.

J. A. Maier et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3646–3650
3649
(BBr3, CH2Cl2, ꢀ78 ꢁC to rt) gave 5-hydroxy-2-methyl
derivative 10b which showed excellent activity in the
kinase assay. The substitution of the pyrimidyl 4-(2-
PK studies (Table 5). All compounds were tested for
inhibition of IL-2 production in a Jurkat cellular as-
say.10 Morpholinyl and piperazinyl analogs 7c and 7g
showed fair to good IL-2 inhibitory activity (2.90 and
0.749 lM). However, both of these compounds showed
poor aqueous solubility contributing to the reported
poor bioavailability for 7c. When the 4-(2-diethylamino-
ethoxy)-aniline was used (7i and 7j), the resulting com-
pounds again showed good to fair IL-2 activity, but
both compounds showed significantly improved aque-
ous solubility and good in vitro metabolism. Additional-
ly, 7i proved to be well absorbed orally in a rat PK study
(F% = 66%). The highly potent lck inhibitor 10b showed
equally strong potency in the IL-2 assay. The compound
also possessed good aqueous solubility. However,
metabolism studies showed a high intrinsic clearance
(CLint = 108 mL/min/kg)11 resulting in very poor oral
bioavailability. We then transitioned to the equally po-
tent amides 16a and 18c, which showed even greater rel-
ative inhibition of IL-2 production, but low solubility
prohibited further progression of these compounds into
PK studies.
diethylamino)-ethoxyaniline with
aniline (10c) did not result in any reduction in activity.
a
4-morpholinyl
Based on the success of the N0-5-hydroxy phenyl analogs
10b and 10c, we decided to extend the substitution at the
5-position of the N0-phenyl group to include amide func-
tionality. Scheme 2 shows the synthesis of amides 16 and
18. Starting with anilines 119 or 12, formation of the cor-
responding isocyanates was effected by heating the
hydrochloride salts of the amides in the presence of
phosgene in a sealed tube to give 13 and 14. The Boc-
protected ureas 15 and 17 were formed using standard
conditions (see Scheme 1). Deprotection of the Boc
group in 15 gave the desired amide urea 16. Nitro analog
17 was reduced (H2, Pd/C) and the resulting aniline acyl-
ated and Boc-deprotected to give the reverse amide 18.
Table 4 shows the results of kinase inhibition experi-
ments for five of these analogs. Activities in the
3–149 nM range were observed for the various deriva-
tives, indicating that even larger, more extended ureas
were well tolerated in the assay. Perhaps most interest-
ing was the 10-fold difference in potency between the
two amide isomers 16b and 18c.
Urea 7i was progressed further into an in vivo rat adju-
vant-induced arthritis study to test for oral efficacy. We
observed that 7i significantly (p < 0.01) inhibited hind
paw swelling when administered orally twice daily at
25 mg/kg for 7 days by 63% compared to vehicle-treated
control animals.14
Compounds from above that showed promising lck
inhibition were progressed into further in vitro and
In summary, we have reported a class of trisubstituted
ureas that have shown nanomolar activity for inhibiting
lck kinase activity. Six of the most potent analogs were
further tested for inhibition of IL-2 cytokine production,
showing a wide range of potencies. The urea analog that
showed the best profile from an in vitro activity and
pharmacokinetic stand point was tested for in vivo effi-
cacy. Urea 7i showed significant reduction in hind paw
swelling when tested in a rat adjuvant-induced arthritis
study. Therefore, this class of lck inhibitors shows prom-
ise as an orally bioavailable treatment for inflammatory
disorders.
a
b
H2N
R
OCN
R
13 R = C(O)NHR2
14 R = NO2
11 R = C(O)NHR2
12 R = NO2
O
N
N
HN
N
R3
N
N
HN
R
d, e, c
H
HN
Ar
N
O
R1N
Ar
N
O
15 R = C(O)NHR2, R1 = Boc
16 R = C(O)NHR2, R1 = H
17 R = NO2
c
18
Acknowledgments
Scheme 2. Reagents and conditions: (a) (i)—1 M HCl, Et2O; (ii)—
phosgene, toluene, 100 ꢁC; (b) 4,6-diaminopyrimidine, CH2Cl2; (c)
TFA, CH2Cl2; (d) 17, H2, Pd/C, MeOH; (e) NaOH (aq), CH2Cl2,
R3C(O)Cl.
We thank M. M. Belkin for solubility and formulation
studies, J. A. Troutman, C. A. Cruze, and W. P.
Schwecke for metabolism and PK studies, C. M. Dun-
away, M. Martin, and P. H. Zoutendam for analytical
chemistry support, and J. T. Roesgen for kinase assay
support.
Table 5. IL-2 inhibition and pharmacokinetic data for select ureas
a
Compound IL-2 IC50 Solubility CLint
b
Cmax
F
(lM)
(lg/mL) (mL/min/kg) (ng/mL) (%)
7c
7g
2.90
0.749
0.862
3.10
3
2
48
—
9
—
394
—
3
0
—
66
—
1
References and notes
7i
38
34
41
1
40
25
1. Shaw, A. S.; Amrein, K. E.; Hammond, C.; Stern, D. F.;
Sefton, B. M.; Rose, J. K. Cell 1989, 59, 627.
2. Rudd, C. E.; Treyillyan, J. M.; Dasgupta, J. D.; Wong, L.
L.; Schlossman, S. F. Proc. Natl. Acad. Sci. U.S.A. 1988,
85, 5190.
3. Duplay, P.; Thome, M.; Herve, F.; Acuto, O. J. Exp. Med.
1994, 179, 1163.
4. Weiss, A.; Littman, D. Cell 1994, 76, 263.
7j
10b
16a
18c
0.064
<0.010
<0.010
108
—
—
—
—
—
3
—
a IL-2 synthesis inhibition measured from Jurkat cells.
b Intrinsic clearance after 1 h for compounds (1 lM concentration)
tested in cryopreserved rat hepatocyte suspensions.

3650
J. A. Maier et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3646–3650
5. (a) van Oers, N. S. C.; Lowinkropf, B.; Finlay, D.;
in RPMI-1640 containing 10% fetal bovine serum and
1% antibiotics in the log growth phase (2 · 105 to
1 · 106 cells/ml) were harvested and incubated in tripli-
cate in 96-well plates for 30 min at 37 ꢁC in the presence
or absence of various concentrations of Lck inhibitors.
The cell–inhibitor mixture was then transferred into the
wells of the anti CD3e-coated 96-well plates, and PMA
was added to the wells at a final concentration of 10 ng/
ml (1 ng/well). The plates were incubated overnight at
37 ꢁC. The amount of IL-2 released into the culture
media was measured by ELISA (R&D Systems) and the
viability of the cells was determined using the MTS
assay (Promega, Madison, WI).
Connolly, K.; Weiss, A. Immunity 1996, 5, 429; (b)
Molina, T. J.; Kishigara, K.; Siderovski, D. P.; van Ewijk,
W.; Narendran, A.; Timms, E.; Wakeham, A.; Paige, C. J.;
Hartman, K.-U.; Veilette, A.; Davison, D.; Mak, T. W.
Nature 1992, 357, 161.
6. Brugel, T. A.; Maier, J. A.; Clark, M. P.; Sabat, M.;
Golebiowski, A.; Bookland, R. G.; Laufersweiler, M. J.;
Laughlin, S. K.; VanRens, J. C.; De, B.; Hsieh, L. C.;
Mekel, M. J.; Janusz, M. J. Bioorg. Med. Chem. Lett.
2006, article accepted for publication.
7. Ding, Q.; Gary, N. S.; Li, B.; Liu, Y.; Sim, T.; Uno, T.;
Zhang, G.; Pissot Soldermann, C.; Breitenstein, W.; Bold,
G.; Caravatti, G.; Furet, P.; Guagnano, V.; Lang, M.;
Manley, P. W.; Schoepfer, J.; Spanka, C. WO 2006000420.
Chem. Abstr. 2006, 144, 108347.
8. The ability of compounds to inhibit human Lck enzyme
(IC50) was determined using the commercially available
ProFlour Src-family Kinase Assay (Promega Corporation,
Madison, WI; cat. #1271). The assay was performed
according to manufacturer’s instructions, with 2 nM
recombinant human active Lck (Upstate Cell Signaling
Solutions, Charlottesville, VA; cat. #14-442) at an ATP
concentration of 10 lM. Experiments were conducted in
triplicate with an n = 2.
11. Metabolic stability was assessed based on the extent of test
article depletion over time. In vitro intrinsic clearance
(CLint) was calculated from the first-order depletion rate
constant and appropriate biological scaling factors.12,13
12. Davies, B.; Morris, T. Pharm. Res. 1993, 10, 1093.
13. Venkatakrishnan, K.; von Moltke, L. L.; Obach, R. S.;
Greenblatt, D. J. Curr. Drug Metab. 2003, 4, 423.
14. Adjuvant-induced arthritis was induced by a single sub-
cutaneous injection of a 5 mg/kg dose of a suspension of
Mycobacterium butyricum (Difco, Detroit, MI) in light
mineral oil into the base of the tail of male Lewis rats
(weights 190–220 g). Arthritis was assessed by the swelling
in the hind paws by water displacement using a plethys-
mometer (Stoelting Co, Wood Dale, IL). Dosing of
compound or vehicle (0.5% carboxymethyl cellulose with
1% Tween 80) was started once joint swelling was readily
apparent (therapeutic protocol) which was about 10 days
after injection of adjuvant. Compound 7i was adminis-
tered orally twice daily at 10 or 25 mg/kg for 1 week (day
10–17). The percent inhibition of paw swelling for
compound 7i-treated animals was calculated relative to
the vehicle controls and statistical significance was deter-
mined using an analysis of covariance.
9. Aniline 11 was synthesized by amidation and subsequent
hydrogenation of the commercially available 4-methyl-3-
nitro-benzoyl chloride.
10. IL-2 release was measured by stimulating the Jurkat E6-
1 T cell line (human acute T cell leukemia, ATCC,
Manassas, VA) with monoclonal anti-human CD3e
antibody and phorbol myristate acetate (PMA) (Sigma,
St. Louis, MO). Flat-bottommed 96-well plates were
pre-coated with 400 ng/well of anti-human CD3e mouse
antibody UCHT1 (R&D systems, Minneapolis, MN)
and incubated for 2 h at 37 ꢁC. Jurkat cells maintained