Identification and SAR of a new series of thieno[3,2-d]pyrimidines as Tpl2 kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.07.069

We report here the synthesis and SAR of a new series of thieno[3,2-d]pyrimidines as potent Tpl2 kinase inhibitors. The proposed binding mode suggests the potential flipped binding mode depending on the substitution. Biacore studies show evidence of binding of these molecules to the protein kinase. The kinome inhibition profile of these molecules suggests good selectivity.

Full text of DOI:10.1016/j.bmcl.2011.07.069

Bioorganic & Medicinal Chemistry Letters 21 (2011) 5952–5956
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification and SAR of a new series of thieno[3,2-d]pyrimidines as Tpl2
kinase inhibitors
Yike Ni a, , Ariamala Gopalsamy a, Derek Cole b, Yonghan Hu a, Rajiah Denny a, Manus Ipek a, Julie Liu c,
⇑
Julie Lee c, J. Perry Hall c, Michael Luong c, Jean-Baptiste Telliez c, Lih-Ling Lin c
a Medicinal Chemistry, Pfizer, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
b Medicinal Chemistry, Takeda San Diego, 10410 Science Center Drive, San Diego, CA 92121, USA
c Inflammation and Immunology, Pfizer, 200 Cambridge Park Drive, Cambridge, MA 02140, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We report here the synthesis and SAR of a new series of thieno[3,2-d]pyrimidines as potent Tpl2 kinase
inhibitors. The proposed binding mode suggests the potential flipped binding mode depending on the
substitution. Biacore studies show evidence of binding of these molecules to the protein kinase. The
kinome inhibition profile of these molecules suggests good selectivity.
Received 24 May 2011
Revised 18 July 2011
Accepted 20 July 2011
Available online 26 July 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Tpl2 kinase
Thieno[3,2-d]pyrimidine
Tpl2 (Tumor Progression Locus-2) kinase is a serine/threonine
kinase in the MAP3K family that activates the MEK-ERK pathway
Cl
F
leading to the production of TNF-
was also shown to play a role in TNF-
has been well documented that TNF- , a pro-inflammatory cyto-
a
and type 1 interferon.1 Tpl2
R1
signaling of TNF-a 2
. It
HN
HN
HN
N
a
H
N
H
N
CN
CN
R2
a
N
N
kine, is involved in the initiation and progression of inflammatory
diseases such as rheumatoid arthritis (RA).3 Tpl2 knock-out mice
show reduced lipopolysaccharide (LPS)-induced TNF production.4
Preliminary studies also demonstrated that Tpl2 knock-out mice
show reduced disease severity in the murine collagen-induced
arthritis (CIA) model.5 Therefore, inhibition of Tpl2 kinase has the
potential to be a novel and effective therapy for the treatment of
RA and other human inflammatory diseases.
Considering that several mitogen-activated protein kinase (MAP
kinase) inhibitors have advanced to clinical trials, the studies of
Tpl2 inhibition are still their infancy giving the fact that to date
only a handful of small molecule inhibitors of Tpl2 have been doc-
umented (Fig. 1). 1,7-Naphthyridine-3-carbonitriles 1 and quino-
N
1
2
O
H2N
O
O
N
N
N
NH
N
S
NH2
N
S
NH2
4
3
Figure 1. Literature reported Tpl2 inhibitors.
line-3-carbonitrile
2 are the earliest classes of compounds
reported as reversible and ATP-competitive Tpl2 kinase inhibitors.6
These Tpl2 inhibitors demonstrate a good correlation between cell-
free Tpl2 inhibition and p-MEK and TNF inhibition in human
monocytes, establishing a role of Tpl2 in TNF production in these
cells. Thieno[2,3-c]pyridines 3 and 4 are the other two classes of
Tpl2 inhibitors that were disclosed recently.7
Tpl2 remains a challenging target for structure-based or in silico
based drug design as no crystal structure of Tpl2 has been reported
to date. It has been hypothesized that the difficulty of obtaining a
crystal structure could be due to the instability of the active free
form which is rapidly degraded in vitro.7a Albeit the lacking of
crystal structure support, the low homology of Tpl2 to other ki-
nases brings a unique opportunity to achieve selectivity across
the kinome. We describe herein the identification and structure-
activity relationship (SAR) of a new thieno[3,2-d]pyrimidine series
⇑
Corresponding author. Tel.: +1 617 665 5673; fax: +1 617 665 5682.
E-mail address: yike.ni@pfizer.com (Y. Ni).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.07.069

Y. Ni et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5952–5956
5953
Table 3
O
OH
S
Substitutent pattern and inhibition of Tpl2 activity
S
R2
N
Br
X
N
N
S
N
R1
5
Figure 2. High throughput screening (HTS) hit.
Compound
R1
X
R2
Tpl2 IC50 (lM)
19
20
I
I
S
S
CH2CO2Et
CH2CO2H
5.1
18
Table 1
Profile of 5.
N
21
22
S
S
CH2CO2H
CH2CO2H
22
OMe
TPL2 IC50
MW
TPSA
(
lM)
4.8
381
62
Cl
S
>100
c log P
4.4
Ligand efficiency
Aqueous solubility (lg/ml at pH 7.4)
Rat liver microsome stability (minute)
0.35
>100
>30
23
24
S
CH2CO2H
CH2CO2H
>100
34
N
O
Cl
O
3. Carboxylic acid isosteres
25
O
O
O
CH2CO2H
10
NH
O
OH
O
26
27
28
CH2CO2H
CH2CO2H
CH2CO2H
6.2
2.3
1.0
NH
NH
2. Linker replacement
S
S
N
Br
N
1. Phenyl substitution
Figure 3. Medicinal chemistry strategy.
N
O
NH
29
30
NH
CH2CO2H
CH2CO2H
>100
>100
Cl
Cl
Table 2
SAR of phenyl substitution on inhibition of Tpl2 activity
NMe
O
O
S
R2
R1
replacement of the sulfur linker to eliminate potential side effect
such as oxidation of the sulfur associated with the aryl thioether,
and (3) identification of an optimal isostere of the carboxylic acid
which has potential permeability issue.
S
N
N
Compound
R1
R2
Tpl2 IC50 (lM)
Initial SAR of phenyl substitution indicated that relatively small
groups on the phenyl ring can be tolerated (Table 2, compounds 6–
11). Large substitution in the R1 position like –SMe, phenyl led to a
dramatic loss of potency (compounds 12, 13). Interestingly, more
polar hydrogen-bond donor groups, such as amide, directly at-
tached to the phenyl ring also resulted in good potency (compound
14). Our docking experiments suggest that these hydrogen-bond
donor groups make a hydrogen-bond interaction with the Asp car-
boxylic acid residue in the DFG motif. However, hydrogen-bond
acceptor or other polar groups may lead to steric or electrostatic
clashes with the side chain (data not shown). Moving the substitu-
tion from 4- to 3-position resulted in a similar level or slight loss of
potency (9 vs 8, 15 vs 14). It is unclear at this stage why the methyl
substituted amide resulted in compound 16 being inactive. The re-
versed amide 17 provided the highest potency in this series with a
6
7
8
9
10
11
12
13
14
15
16
17
18
4-F
4-Cl
H
H
H
H
H
H
H
H
H
H
H
H
Et
8.1
1.2
1.7
3.5
1.1
1.7
34
>100
1.5
4-Me
3-Me
4-CN
4-OMe
4-SMe
4-Ph
4-CONH2
3-CONH2
4-CONHMe
4-NHCOMe
4-Cl
7.1
>100
0.18
>100
as Tpl2 inhibitors following traditional SAR approach and molecu-
lar modeling studies.
Tpl2 IC50 of 0.18 lM. Replacement of the carboxylic acid with ester
proved to be unfruitful as evidenced by the loss of activity of com-
Thienopyrimidine 5 (Fig. 2, Tpl2 IC50 = 4.8 lM) was identified as
our lead compound following the high through-put screening of a
compound library. Initial profile revealed that this compound has
low micromolar Tpl2 potency, decent calculated properties and
very good aqueous solubility (Table 1). Initial ADME profile of 5
also showed it is stable in rat liver microsomes.
pound 18 when compared with compound 7.
The two iodo intermediates 19 and 20 used for preparing the fi-
nal targets were also tested in the enzymatic assay. It is interesting
to see that the ester 19 is three fold more potent than the acid 20
(Table 3), unlike the activity relationship between 7 and 18. Given
the difficulty of obtaining a crystal structure of Tpl2, we used a
homology model to gain some insight into the SAR and guide our
As demonstrated in Figure 3, our medicinal chemistry strategy
included (1) exploration of substitution on the phenyl ring, (2)

5954
Y. Ni et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5952–5956
medicinal chemistry effort. Two binding modes were proposed in
the literature for the thienopyridine series.7 Based on the two pro-
posed binding modes, it was hypothesized that the majority of tar-
gets 6–18 adopted the binding mode showed in Figure 4 in which
the 3-N of the thieno[3,2-d]pyrimidine binds to the hinge region.
The carboxylic acid chain extends to the solvent pocket and the
substituted phenyl ring points to the Lys region. In contrast, when
the phenyl ring was replaced with iodo, the binding mode could be
flipped as indicated in Figure 5 where the 1-N of the thieno[3,2-
d]pyrimidine interacts with the hinge NH (Ala), iodo on the thio-
phene ring points to the solvent pocket and the polar chain extends
to the Lys region. In this case, the relative hydrophobic pocket near
the Lys residue would favor the ester instead of the free carboxylic
acid which is supported by the enzymatic potency of compounds
19 and 20.
Figure 4. Model of 10 bound to homology model of Tpl2.
Figure 5. Flipped binding mode depending on the substitution.
n-BuLi, I2
EtO
HS
TEA,EtOH
COOEt
Cl
N
Cl
N
S
N
-78 oC to rt
2 h
O
S
S
S
12 h
N
N
N
I
I
31
19
2N LiOH,
HO
HO
Ar(BOH)2
S
N
S
EtOH:THF(1:1),
Pd catalyst
dioxane
O
70 0C, 2 h
O
S
S
N
N
I
Ar
N
20
6 - 22
Scheme 1. Synthesis aryl or heteroaryl substituted thienopyrimidines.
EtO
O
S
nBuLi, Br2
-78 oC to rt
2 h
HS
TEA,EtOH,
rt, 12 h
COOEt
Cl
Cl
N
HN
S
S
S
N
N
N
Br
Br
DMF
N
N
32
33
HO
O
2N LiOH, EtOH,
60-80 °C, 4 h
EtO
S
S
O
S
S
N
N
N
N
N
N
34
23
Scheme 2. Synthesis cyclic amino analog 23.

Y. Ni et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5952–5956
5955
OH
R
O
OEt
X
B
O
OH
HX
OEt
Pd(PPh3)4, CsF,
NaHCO3, dioxane,
H2O, reflux, 3 h
Cl
N
Cl
N
NaH,DMF,30 min
100°Cmicrowave
R
S
R
S
S
N
N
N
I
N
35
31
X = O, N
HO
X
2N LiOH,THF
4h, reflux
O
S
R
N
N
X = O, N
24 ~ 30
Scheme 3. Synthesis of oxygen- and nitrogen-linker analogs.
It is of interest to investigate the SAR of different heterocycles in
addition to the phenyl analogs. Unfortunately, as demonstrated in
Table 3, replacement of the phenyl portion with various heterocy-
cles resulted in significant loss of potency (compounds 21 and
22).8 Cyclic amines were also used as surrogates to the phenyl ring
and all of these amino analogs led to a loss of activity (compound
23). Given the potential side effect of the sulfur linker, we next
turned out attention to replacing it with an oxygen linker. Although
the para-chlorophenyl analogs with oxygen linker was much less
potent than with sulfur linker (compound 24 vs 7), the reversed
amide analog 25 delivered improved potency with the new linker.
The N-acetylaniline fragment in 25 would trigger the aniline struc-
tural alert,9 which led us to modify this fragment. We were de-
lighted to observe that modification to rigidify the N-acetylaniline
forming a bicycle further improved the potency toward Tpl2 (com-
pounds 26–28). The indazole analog 28 proved to be the most po-
tent compound bearing the oxygen linker with a Tpl2 IC50 of
1.0 lM. Efforts trying to identify an optimal bioisosteres for the car-
boxylic acid were unfruitful. Analogs bearing ester, amide or het-
erocycles instead of the free carboxylic acid all turned out to be
inactive (data not shown).10 Replacement of the sulfur linker with
nitrogen linker was also tried. All analogs bearing the nitrogen link-
ers proved to be inactive indicating electronic conflict in the linker
region with more basic nitrogen (compounds 29–30).
The general synthesis of the analogs 6–22 is highlighted in
Scheme 1. Treatment of the commercially available 4-chlorothie-
nopyrimidine with n-butyllithium followed by iodination provided
4-chloro-6-iodothieno[3,2-d]pyrimidine 31, which reacted with
ethyl 2-mercaptoacetate to give analog 19. Hydrolysis of 19 affor-
ded acid 20, which coupled with various boronic acids or boronic
esters provided the desired analogs.
A typical synthesis of cyclic amine substituted thienopyrimi-
dines was demonstrated in Scheme 2. Failure of coupling between
the iodo analog 19 and secondary amines led us to switch to the
Figure 6. In solution competitive binding assay. Tpl2 (50 nM) was either injected alone (solid black lines) or in the presence of excess compound (50 lM, dashed lines). (A)
Tpl2 competition in the presence of the reference compound (dashed line, positive control). Tpl2 binding in the presence of Staurosporine (dotted line, negative control). (B)
Compound 5 and (C) compounds 7, 8 and 10 prevented Tpl2 binding to the reference compound surface. Data shown for each compound are representative of at least three
independent surfaces where the same Tpl2 binding or competition was observed.

5956
Y. Ni et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5952–5956
Table 4
Kinase selectivity profile (%inhibition at 1 lM)
Kinase
CDK2 cyclinA
Aurora A
CHK1
CHK2
SRC
EGFR
FGFR1
GSK3B
JAK3
LCK
17
28
Kinase
17
28
46.8
27.3
MK2
3.3
50.5
88.5
MARK1
23.0
1.3
À0.8
À7.8
MET
9.8
10.8
17.0
29.0
NEK2
3.3
À3.8
À1.3
PAK4
11.0
11.0
3.3
PKA
À3.3
6.8
50.5
56.5
ROCK1
17.0
26.0
43.3
29.5
CK1a1
À1.3
4.5
13.8
6.0
P38
5.3
À2.3
MST2
27.3
À3.3
25.3
21.8
À3.5
À9.0
À0.8
use of bromo intermediate 33 which was synthesized from 32 fol-
lowing the same sequence as in Scheme 1. Coupling of 33 to 4-
methylpiperidine led to the desired product 34 which was hydro-
lyzed to provide analog 23.
Acknowledgment
The authors thank the Analytical Division and Sample Logistics
at Pfizer Cambridge for their assistance on the project.
To synthesize the oxygen- or nitrogen-linker analogs, interme-
diate 31 was coupled with corresponding boronic acids or boronic
esters to give 35, which reacted with 2-hydroxyl or 2-amino
acetate, followed by hydrolysis to provide the desired products
24–30 (Scheme 3).
References and notes
1. Xiao, N.; Eidenschenk, C.; Krebs, P.; Brandl, K.; Blasius, A. L.; Xia, Y.;
Khovananth, K.; Smart, N. G.; Beutler, B. J. Immunol. 2009, 183, 7975.
2. (a) George, D.; Salmeron, A. Curr. Topics Med. Chem. 2009, 9, 611. and references
therein.; (b) Hall, J. P.; Kurdi, Y.; Hsu, S.; Cuozzo, J.; Liu, J.; Telliez, J. –B.; Seidl, K.
J.; Winkler, A.; Hu, Y.; Green, N.; Askew, G. R.; Tam, S.; Clark, J. D.; Lin, L. –L. J.
Biol. Chem. 2007, 282, 33295.
3. (a) Choy, E. H.; Panayi, G. S. N. Eng. J. Med. 2001, 344, 907; (b) Feldmann, M.;
Maini, R. N. Annu. Rev. Immunol. 2001, 19, 163; (c) Smolen, J. S.; Steiner, G. Nat.
Rev. Drug Disc. 2003, 2, 473; (d) Furst, D. E. Clin. Therapeutics 2004, 26, 1960; (e)
Simmons, D. L. Drug Discovery Today 2006, 11, 210.
4. (a) Dumitru, C. D.; Ceci, J. D.; Tsatasnis, C.; Kontoyiannis, D.; Stamatakis, K.; Lin,
J. H.; Patriotis, C.; Jenkins, N. A.; Copeland, N. G.; Kollias, G.; Tsichlis, P. N. Cell
2000, 103, 1071; (b) Sugimoto, K.; Ohata, M.; Miyoshi, J.; Ishizaki, H.; Tsuboi,
N.; Masuda, A.; Yoshikai, Y.; Takamoto, M.; Sugane, K.; Matsuo, S.; Shimada, Y.;
Matsuguchi, T. J. Clin. Invest. 2004, 114, 857.
Several compounds (compounds 5, 7, 8, 10) were selected to
investigate their binding to Tpl2. A Biacore-based competitive
binding assay consisted of a reference compound immobilized on
a sensor chip surface (250–470 RU) by amine coupling to measure
Tpl2 binding in the presence or absence of compound. Tpl2 binds
the reference compound as shown in Figure 6, solid lines by the in-
crease in RU during the injection time of 2 min (Fig. 6A and C) or
1 min. (Fig. 6B). In the presence of excess soluble reference com-
pound, no Tpl2 binding was observed (Fig. 6A, dashed line). As ex-
pected, Staurosporine (Fig. 6A, dotted line) did not compete for the
binding of Tpl2. In the presence of compound 5 (Fig. 6B) or com-
pounds 7, 8 or 10 (Fig. 6C) no Tpl2 binding was observed. Thus,
the compounds 5, 7, 8 and 10 bind Tpl2 and prevent Tpl2 binding
to the reference compound surface. However, initial test of these
compounds in cell assay (human monocyte) measuring the inhibi-
tion of phosphorylation of ERK did not show measurable inhibition,
which might be due to the poor permeability associated with the
carboxylic acids.
Compounds 17 and 28 were selected for kinase selectivity pro-
filing with an invitrogen panel of kinases. As shown in Table 4,
these compounds showed an overall good selectivity profile. Of
the 39 kinases tested, only one kinase (Aurora A) showed >60%
inhibition.
In summary, a new series of thieno[3,2-d]pyrimidines has been
identified as potent and selective Tpl2 kinase inhibitors following
traditional SAR approach and molecular modeling studies. The
5. In house data.
6. (a) Wu, J.; Green, N.; Hotchandani, R.; Hu, Y.; Condon, J.; Huang, A.; Kaila, N.; Li,
H.-Q.; Guler, S.; Li, W.; Tam, S. Y.; Wang, Q.; Pelker, J.; Marusic, S.; Hsu, S.; Hall,
J. P.; Telliez, J.-B.; Cui, J.; Lin, L.-L. Bioorg. Med. Chem. Lett. 2009, 19, 3485; (b)
Green, N.; Hu, Y.; Janz, K.; Li, H.-Q.; Kaila, N.; Guler, S.; Thomason, J.; Joseph-
McCarthy, D.; Tam, S. Y.; Hotchandani, R.; Wu, J.; Huang, A.; Wang, Q.; Leung,
L.; Pelker, J.; Marusic, S.; Hsu, S.; Telliez, J.-B.; Hall, J. P.; Cuozzo, J. W.; Lin, L.-L. J.
Med. Chem. 2007, 50, 4728; (c) Kaila, N.; Green, N.; Li, H.-Q.; Hu, Y.; Janz, K.;
Gavrin, L. K.; Thomason, J.; Tam, S.; Powell, D.; Cuozzo, J.; Hall, J. P.; Telliez, J.-
B.; Hsu, S.; Nickerson-Nutter, C.; Wang, Q.; Lin, L.-L. Bioorg. Med. Chem. 2007,
15, 6425; (d) Hu, Y.; Green, N.; Gavrin, L. K.; Janz, K.; Kaila, N.; Li, H.-Q.;
Thomason, J. R.; Cuozzo, J. W.; Hall, J. P.; Hsu, S.; Nickerson-Nutter, C.; Telliez,
J.-B.; Lin, L.-L.; Tam, S. Bioorg. Med. Chem. Lett. 2006, 16, 6067; (e) Gavrin, L. K.;
Green, N.; Hu, Y.; Janz, K.; Kaila, N.; Li, H.-Q.; Tam, S. Y.; Thomason, J. R.;
Gopalsamy, A.; Ciszewski, G.; Cuozzo, J. W.; Hall, J. P.; Hsu, S.; Telliez, J.-B.; Lin,
L.-L. Bioorg. Med. Chem. Lett. 2005, 15, 5288.
7. (a) Cusack, K.; Allen, H.; Bischoff, A.; Clabbers, A.; Dixon, R.; Fix-Stenzel, S.;
Friedman, M.; Gaumont, Y.; George, D.; Gordon, T.; Grongsaard, P.; Janssen, B.;
Jia, Y.; Moskey, M.; Quinn, C.; Salmeron, A.; Thomas, C.; Wallace, G.; Wishart,
N.; Yu, Z. Bioorg. Med. Chem. Lett. 2009, 19, 1722; (b) George, D.; Friedman, M.;
Allen, H.; Argiriadi, M.; Barberis, C.; Bischoff, A.; Clabbers, A.; Cusack, K.; Dixon,
R.; Fix-Stenzel, S.; Gordon, T.; Janssen, B.; Jia, Y.; Moskey, M.; Quinn, C.;
Salmeron, J.-A.; Wishart, N.; Woller, K.; Yu, Z. Bioorg. Med. Chem. Lett. 2008, 18,
4952.
most potent compound was found to have
a Tpl2 IC50 of
0.18 M. Proposed binding mode suggested that the binding mode
l
8. Several other heterocycles such as 3-pyridyl, 4-chloro-3-pyridyl, 5-
methylthiophenyl, 3,5-dimethylisoxazol analogs were also prepared, but all
resulted in the loss of activity (Tpl2 IC50 >100 lM).
9. Blagg, J. Burger’s Medicinal Chemistry Drug Discovery and Development, seventh
ed; Wiley, 2010. Vol. 2, pp. 1–34.
could be flipped depending on the substitution. Primarily the car-
boxylic acid chain extends to the solvent pocket and the substi-
tuted phenyl ring points to the Lys region to engage interactions
in the hydrophobic pocket. Biacore studies showed evidence of
these molecules binding to the protein.
10. The amides tried include N-methylacetamide, N,N-dimethylacetamide;
isosteric heterocycles include pyridine, pyrimidine and imidazole.