Synthesis and structure-activity relationship of aminopyrimidine IKK2 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.04.062

The synthesis and structure-activity relationship of a novel series of aminopyrimidines are exemplified. Results of key compounds from within this series in the E-selectin reporter cell assay are also reported.

Full text of DOI:10.1016/j.bmcl.2008.04.062

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3622–3627
Synthesis and structure–activity relationship of aminopyrimidine
IKK2 inhibitors
Alistair H. Bingham, Richard J. Davenport,*, Richard Fosbeary, Lewis Gowers,
Roland L. Knight, Christopher Lowe, David A. Owen, David M. Parry and Will R. Pitt
UCB, Granta Park, Great Abington, Cambridge CB21 6GS, UK
Received 1 April 2008; revised 23 April 2008; accepted 25 April 2008
Available online 29 April 2008
Abstract—The synthesis and structure–activity relationship of a novel series of aminopyrimidines are exemplified. Results of key
compounds from within this series in the E-selectin reporter cell assay are also reported.
Ó 2008 Elsevier Ltd. All rights reserved.
H
The NF-jB pathway is important in regulating the
N
N
S
expression of cellular genes that are involved in the con-
trol of the immune and inflammatory response.1 The
activation of NF-jB induces the expression of more
than 150 genes2 such as cytokines (TNF, IL-1, IL-6),
chemokines (IL-8, MCP-1), cell adhesion molecules
(ICAM-1, VCAM-1) and proteases. Remarkably, NF-
jB induces the production of proteins,3 for example,
TNF, that are themselves able to stimulate NF-jB,
hence leading to an amplification of any physiological
effect of the NF-jB pathway.
N
S
N
O
N
N
N
O
O
HN
O
O
S
N
SPC839
1
IKK2 62nM
IKK1 13000nM
IKK2 64nM
IKK1 850nM
N
H
Figure 1. The IKK2 inhibitor SPC839 and our initial lead compound
1.
Activation of NF-jB is mediated by the increase in the
activation of two kinases, IKK1 and IKK2.4–7 IKK2
(À/À) knockout mice data8–11 have shown that IKK2
is more critical than IKK1 in activating the NF-jB
pathway for the inflammatory response, whilst data
from IKK1 (À/À) knockout mice have indicated a role
for IKK1 in skin and skeletal development. Hence small
molecule inhibition of IKK2 with selectivity over IKK1
could lead to novel treatments for some cancers, and
autoimmune inflammatory diseases, such as rheumatoid
arthritis.12 In 2004, it was reported that SPC839 (see
Fig. 1), a selective IKK2 inhibitor, was undergoing
pre-clinical development after showing good efficacy in
cancer and rheumatoid arthritis animal models.13,14 Re-
cent reports suggest that this compound has now pro-
gressed into Phase I clinical trials targeting an ultimate
endpoint of haematological malignancies.15 This long
time frame required to progress SPC839 exemplifies
the challenges of finding novel IKK2 inhibitor. Within
our laboratories we have investigated a novel series of
aminopyrimidines that show inhibitory action against
IKK2. In a previous communication,14 we disclosed a
structure–activity study of variations in the aminoben-
zothiazole group of structure 1. In this letter we disclose
the results of a study into the structure–activity relation-
ships of the 4-phenyl substituents of the aminopyrimi-
dine structure 1 (Fig. 1).
The lead compound 1 was found to have an IKK2 IC50
of 64 nM, with reasonable selectivity for IKK2 over
IKK1 (IKK1 IC50 850 nM). Unfortunately, these prom-
ising IKK2 primary data did not translate well into the
E-selectin reporter cell assay16 giving a very modest IC50
Keywords: IKK2; Aminopyrimidine; Inflammation; Kinase.
*
Corresponding author. E-mail: rdavenport@takedacam.com
Present address: Medicinal Chemistry, Takeda Cambridge,
Cambridge Science Park, Cambridge CB4 0PA, UK.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.04.062

A. H. Bingham et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3622–3627
3623
plex with two other similar kinases. Since these ana-
logues showed a consistent binding mode in all these
crystal structures, we made the assumption that the
binding of 1 in IKK2 would be equivalent. In this con-
Cys99
N
H
O
H
N
formation, the aminopyrimidine-nitrogen forms
a
N
S
N
hydrogen bond with the backbone NH of Cys99 and
the aminopyrimidine NH forms a hydrogen bond with
the backbone carbonyl of the same residue. It is also be-
lieved that the sulfonylpiperazine portion of our lead
structure overlays with the ribose-phosphate portion of
the natural ligand ATP.
N
N
O
O
S
HN
The purpose of this investigation was to probe the sul-
fonamide portion of our lead compound to optimize
binding potency, which should also in turn increase cel-
lular potency. We were also interested to see if other
functional groups, such as amides, reverse amides/sul-
fonamides, or just a bond could act as sulfonamide bio-
isoteres, potentially offering different IKK2 activity,
selectivity and cellular profiles.
Figure 2. Predicted binding mode of compound 1 in IKK2.
H
Me2N
N
N
S
O
O
N
N
a
c
5, 72-100%
The analogues were synthesized by several routes. The
directly-linked compounds 5 were synthesized in a con-
vergent manner. Firstly, the corresponding arylketone
2 (Scheme 1) was elaborated to the enone 3 using
DMF–DMA. Secondly, commercially available amino
benzothiazole was converted to benzothiazole guanidine
4 using cyanamide and nitric acid (85–98%). The enone 3
was combined with benzothiazole guanidine 4 under ba-
sic conditions to yield the desired aminopyrimidines 5.
2
R
3, 87-100%
R
R
NH2
H
N
NH2
S
b
NH2+
-
S
NO3
N
4, 85%
N
Scheme 1. Synthesis of arylaminopyrimidine analogues. Reagents: (a)
DMF–DMA; (b) NC–NH2, HNO3; (c) DMF, NaOH.
Amides and reverse sulfonamides were synthesized from
commercially available 4-nitroacetophenone (Scheme 2).
Elaboration to the enaminone 6 was performed using
DMF–DMA, followed by cyclisation to the aminopy-
rimidine 7 under basic conditions. The nitro group
value of 5700 nM. No IKK2 crystal structure is avail-
able, so a previously published homology model17 was
used to assist analogue design. This predicted binding
mode is shown inFigure 2 and shows analogue 1 in com-
H
Me2N
N
N
S
O
O
O
O
N
O
N
a
b, 4
N+
N+
7
N+
H
6
N
N
O
S
O
O
N
N
H
9
N
N
S
d
e
HN
R1
N
N
c
O
H
N
N
8
S
N
NH2
N
O
R1
10
HN
S
O
Scheme 2. Synthesis of arylaminopyrimidine analogues. Reagents: (a) DMF–DMA; (b) DMF, NaOH; (c) H2, Pd–C; (d) RCO2H, EDC, HOBt; (e)
RSO2Cl, TEA, DCM.

3624
A. H. Bingham et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3622–3627
Table 1. Structure–activity relationship of substituents (R)
Structure
N
N
S
N
N
R
Number
R-group
IKK218
410
IKK118
2630
IKK1/IKK2
6
5a
N
N Me
O
O
S
5b
5c
5d
5e
440
24% at 10 lM
34% at 10 lM
14% at 10 lM
NT
>22
>14
>4
700
N
N
2500
3600
N
O
O
N
ND
N
N
N
5f
4000
6600
NT
NT
ND
ND
O
O
5g
O
H
N
9a
130
146
1
N
H
O
N
9b
9c
270
400
731
3
5
H
NH2
O
1910
N
NH
H
O
O
S
S
N
10a
10b
10c
N
H
230
4840
2990
NT
21
3
O
N
O
O
O
N
990
H
Cl
O
S
1500
ND
N
H
O
O
N
H
S
10d
10e
8900
8900
NT
NT
ND
ND
O
O
O
S
S
N
Cl
H
Cl

A. H. Bingham et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3622–3627
3625
was then reduced by hydrogenation to yield the anilino
compound 8. This was then reacted either with acids and
EDC to form amides 9 or sulfonyl chlorides in DCM
with triethylamine to yield sulfonamides 10.
Initially, the generality of our original hit sulfonamide
linker was investigated (Table 1) with analogues (5a–
5g), which had groups attached directly to the aryl ring.
The most active structures in this series were compounds
5a and 5b, with both the compounds having IKK2 activ-
ities in the 400 nM range. As neither of these com-
pounds had activities comparable to the starting lead
1, it appeared that a linker or spacer was required. This
concept was explored further with analogues which had
either an amide (9a–9c, Table 1) or a reverse sulfon-
amide (10a–10e, Table 1) as a replacement for the sul-
fonamide portion. In general, we found that an amide
linker could be used to yield some potent IKK2 inhibi-
tors especially when used in conjunction with a group
bearing a basic centre (cf 1 to 9a and 9b). Unfortunately,
in the case of these amides the IKK2/IKK1 selectivity
profile was less favourable, with a 5-fold selectivity for
IKK2 over IKK1 being the best we could achieve in
amide 9c (compared to 13-fold for our lead structure
1). In the case of the reverse sulfonamides their activities
The structure–activity relationship of all analogues was
explored against both the IKK1 and IKK2 enzymes.
H
Me2N
b
N
N
S
S
O
O
O
N
N
N
c, 4
R2
a
60-98%
R2
14
O
O
O
O
O
O
S
R2
S
SO2Cl
11
N
N
R1
13
12
R1
87-100%
R1
82-100%
Scheme 3. Synthesis of arylaminopyrimidine analogues. Reagents: (a)
RNH2, TEA, DCM; (b) DMF–DMA; (c) DMF, NaOH.
Table 2. Structure–activity relationship of sulfonamides
Structure
N
N
S
N
N
R
Number
R1
IKK218
64
IKK118
850
IKK1/IKK2
13
Cell16
5700
Cell/IKK2
89
PSA
104
O
N
O
S
1
NH
5a
9c
410
400
2630
1910
6
5
1270
3820
3.1
9.5
85
93
N
N Me
O
N
H
NH
O
O
N
H
S
N
10a
14a
230
250
4840
5000
21
20
17,800
289
77
135
88
O
N
O
N
O
O
S
1.2
S
O
N
S
14b
14c
380
410
6700
18
1810
1920
4.8
4.7
99
88
O
O
N
O
O
O
S
S
S
28% at 10 lM
>24
O
14d
14e
540
600
48% at 10 lM
30% at 10 lM
>18
>17
1500
2080
2.8
3.5
90
86
N
O
N

3626
A. H. Bingham et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3622–3627
tended to be lower than the original lead (10a, 230 nM
being the best observed), although it should be noted
that none of the groups posses the basic nitrogen which
seems to be preferred (although not essential) for activ-
ity. However, the most potent reverse sulfonamide 10a
did display reasonable selectivity (21-fold) for the de-
sired IKK2 isoform.
ing further analogues with optimized activity and cell-
penetration potential.19
In conclusion, a comprehensive strategy has been em-
ployed to explore the 4-phenyl substituents region of
our lead 1. Various novel analogues have been synthe-
sized using different synthetic routes. This has enabled
a number of alternative linkers to be accessed and their
activity explored with respect to the type of linker and
the actual substituents. The sulfonamide linker appears
to be optimal, with the original hit 1 having very good
activity in the IKK2 protein assay. This activity does
not translate well into the E-selectin reporter cell assay,
and one reason for this may be its high PSA. Compound
14a, however, despite having a lower IKK2 protein
assay result, showed very favourable activity in the
E-selectin reporter cell assay, and represented a consi-
derable improvement on the original lead molecule 1.
These improvements mean compound 14a is deemed
suitable for pharmacokinetic analysis.
Although removing and replacing the sulfonamide lin-
ker in our original lead compound yielded some com-
pounds with interesting selectivity profiles, no overall
improvement in IKK2 potency was achieved.
Attention was now turned to retaining the sulfonamide
linker in the same orientation as 1 and exploring the
piperazine ring. The sulfonamide series was synthesized
from commercially available starting materials (Scheme
3). Sulfonyl chloride 11 was reacted with a series of
amines to yield the sulfonamides 12 (82–100%). The
acetyl group was then elaborated to the enaminones 13
by refluxing in DMF–DMA (87–100%). This was then
reacted under basic conditions with 4 to yield the amino-
pyrimidines 14 (60–98%). If a Boc-protecting group was
present on R1R2NH this was removed using 10% triflu-
oroacetic acid in dichloromethane. The results from this
series are displayed in Table 2.
References and notes
1. Baldwin, A. S. Ann. Rev. Immunol. 1996, 14, 649.
2. Baeuerle, P. A.; Baltimore, D. Cell 1996, 87, 13.
3. Siebenlist, U.; Franzoso, G.; Brown, K. Annu. Rev. Cell
Biol. 1994, 10, 405.
The direct oxygen and sulfur analogues (14a and 14b)
were both found to be less active than 1. This again sug-
gested that a basic nitrogen was preferred for activity.
When the sulfide of 14a was oxidized to the sulfoxide
(data not shown) this also led to a loss of activity. The
need for a saturated ring was also examined using the di-
methyl analogue 14c, and although not more active than
the original hit, it was more active than the 5- and 6-
membered saturated ring 14d and 14e. When the free
piperazine NH of 1 was methylated it led to a large loss
of activity (data not shown), suggesting that there is lim-
ited space in this region of the protein.
4. DiDonato, J. A.; Hayakawa, M.; Rothwarf, D. M.; Zandi,
E.; Karin, M. Nature 1997, 388, 548.
5. Mercurio, F.; Zhu, H.; Murray, B. W.; Shevchenko, A.;
Bennett, B. L.; Li, J.; Young, D. B.; Barbosa, M.; Mann,
M. Science 1997, 278, 860.
6. Regnier, C. H.; Song, H. Y.; Gao, X.; Goeddel, D. V.;
Cao, Z.; Rothe, M. Cell 1997, 90, 373.
7. Woronicz, J. D.; Gao, X.; Cao, Z.; Rothe, M.; Goeddel,
D. V. Science 1997, 278, 866.
8. Hu, Y.; Baud, V.; Delhause, M.; Zhang, P.; Deerick, T.;
Ellisman, M.; Johnson, R.; Karin, M. Science 1999, 284,
316.
9. Li, Q.; Lu, Q.; Hwang, J. Y.; Buscher, D.; Lee, K. F.;
Izpisua-Belmonte, J. C.; Verma, I. M. Genes Dev. 1999, 13,
1322.
Having established that maintaining a sulfonamide lin-
ker group in the lower portion of the molecule was opti-
mal based on primary protein in vitro activity. The
original hit 1, some closely related compounds and rep-
resentative examples of other linkers (5a, 9c, 10a) were
evaluated further in the E-selectin reporter cell assay16
(Table 2). Despite showing a very encouraging protein
IKK2 activity, 1 demonstrated a significant drop-off in
activity when tested in a cellular system. Conversely
compound 14a which had a more modest protein
IKK2 activity, translated very favourably into the cellu-
lar system, with essentially no drop-off. The relative cell
penetrating abilities of the compounds may explain this.
A number of physicochemical factors are known to af-
fect cell penetration, with one being polar surface area
(PSA). Comparing the PSA of this small set of com-
10. Li, Q.; Van Antwerp, D.; Mercurio, F.; Lee, K. F.; Verma,
I. M. Science 1999, 284, 321.
11. Takeda, K.; Takeuchi, T.; Itami, S.; Adachi, O.; Kawai,
T.; sanjo, H.; Yoshikawa, K.; Terada, N.; Akira, S.
Science 1999, 284, 313.
12. Griffiths, R. J. Rheumatoid Arthritis 1997, 1, 473.
13. Bhagwat, S. S.; Brydon, L.; Bennett, Y. S.; Eoin, C.;
O’Leary, C.; Leisten, J.; Firestein, G. S.; Boyle, D. S.;
Dreano, M.; Anderson, D. W.; Grimshaw, C. E. The small
molecule IKK2 inhibitor SPC839 is efficacious in an
animal model of arthritis. Paper Presented at: ACR
Annual Scientific Meeting (San Francisco, CA, Arthritis
& Rheumatism), 2001.
14. Sors, A.; Jean-Louis, F.; Begue, E.; Parmentier, L.;
Dubertret, L.; Dreano, M.; Courtois, G.; Bachelez, H.;
Michel, L. Clin. Cancer. Res. 2008, 14, 901.
15. Merck Kga, A., Company Pipeline, 2007.
16. A549 cells were stably transfected with a segment of the E-
selectin promoter upstream of a luciferase reporter gene.
Cells activated with TNFa (R&D systems) and treated
with test compound for 6 h. Luminescence detected using
commercially available Luciferase reporter gene reagent
(Luclite, Perkin-Elmer). Values are the mean of at least
two experiments.
pounds it does appear that compounds with
a
PSA < 100 show very little drop-off when tested in the
cellular system, whereas a PSA > 100 in the case of 1
and 10a may lead to a dramatic drop-off. A larger set
of compounds would need to be tested to thoroughly
investigate this observation, although it is quite proba-
ble that PSA could be a good descriptor to use in design-

A. H. Bingham et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3622–3627
3627
17. Bingham, A. H.; Davenport, R. J.; Gowers, L.; Knight, R.
L.; Lowe, C.; Owen, D. A.; Parry, D. M.; Pitt, W. R.
Bioorg. Med. Chem. Lett. 2004, 14, 409.
18. Values are the mean of at least three experiments.
19. Ertl, P.; Rohde, B.; Selzer, P. J. Med. Chem. 2000, 43,
3714.