(1H-Imidazo[4,5-c]pyridin-2-yl)-1,2,5-oxadiazol-3-ylamine derivatives: A novel class of potent MSK-1-inhibitors

Article abstract of DOI:10.1016/j.bmcl.2005.05.021

A novel series of imidazo[4,5-c]pyridines bearing a 1,2,5-oxadiazol-3- ylamine functionality has been developed. These are potent inhibitors of mitogen and stress-activated protein kinase-1.

Full text of DOI:10.1016/j.bmcl.2005.05.021

Bioorganic & Medicinal Chemistry Letters 15 (2005) 3402–3406
(1H-Imidazo[4,5-c]pyridin-2-yl)-1,2,5-oxadiazol-3-ylamine
derivatives: A novel class of potent MSK-1-inhibitors
Mark J. Bamford,a,* Michael J. Alberti,b Nicholas Bailey,a Susannah Davies,a
David K. Dean,a Alessandra Gaiba,a Stephen Garland,c John D. Harling,d David K. Jung,b
Terence A. Panchal,a Christopher A. Parr,a Jon G. Steadman,a Andrew K. Takle,a
James T. Townsend,a David M. Wilsona and Jason Witheringtona
aGlaxoSmithKline Pharmaceuticals, Neurology and GI Centre of Excellence for Drug Discovery, New Frontiers Science Park,
Third Avenue, Harlow, Essex CM19 5AW, UK
bHigh Throughput Chemistry, Research Triangle Park, Five Moore Drive, RTP, NC 27709, USA
cGlaxoSmithKline Pharmaceuticals, Computational and Structural Sciences, New Frontiers Science Park,
Third Avenue, Harlow, Essex CM19 5AW, UK
dCardiovascular and Urology CEDD, Medicines Research Centre, Gunnels Wood Road, Stevenage,
Hertfordshire SG1 2NY, UK
Received 12 January 2005; accepted 9 May 2005
Available online 9 June 2005
Abstract—A novel series of imidazo[4,5-c]pyridines bearing a 1,2,5-oxadiazol-3-ylamine functionality has been developed. These are
potent inhibitors of mitogen and stress-activated protein kinase-1.
Ó 2005 Elsevier Ltd. All rights reserved.
Mitogen and stress-activated protein kinases 1 (MSK-1)
and MSK-2 (also named RSKB or RLPK) are localised
to the nucleus and are activated by both mitogen (ERK)
and stress-activated (p38) protein kinases to phosphory-
late the transcription factors CREB and histone H3.1,2
In macrophages and monocytes, MSK-1 is involved in
CREB-mediated transcriptional regulation of IL-1b
and COX2 in response to bacterial lipopolysaccharide.
MSK-1 and -2 were also shown to be required for the
stress-induced phosphorylation of both CREB and
ATF1, and of histone H3 and HMG-14 in fibroblasts.3,4
MSK-1 is also involved in the transcriptional activation
of the NF-jB p65 subunit.5 Thus, MSK-1 and -2 may
have a critical role in linking cellular signalling pathways
to chromatin modification and modulation of transcrip-
tion factor complexes and have potential therapeutic
utility. Indeed, it has been demonstrated that MSK-1
mediates excitotoxicity-induced death of hippocampal
neurones6 and that inhibitors of MSK-1 and -2 may,
therefore, be of use in the treatment of diseases involv-
ing ischaemic injury.7 Since inhibitors of the p38 path-
way8 and the ERK pathway9,10 are reported to be
neuroprotectants, MSK-1 and -2 inhibitors are also of
interest in this regard.
To date only broad spectrum kinase inhibitors H89
(IC50 120 nM) and Ro318220 (IC50 8 nM) have been re-
ported as MSK-1 inhibitors.11 We recently conducted a
screening programme with the aim of providing a novel
MSK-1 inhibitor template. The 3-amino-1,2,5-oxadia-
zole compound 1 was identified as a tractable starting
point for optimisation (MSK-1, IC50 140 nM).
To aid the prioritisation of molecules for synthesis, pro-
tein docking studies were carried out. Although the
X-ray crystal structure of the N-terminal domain of
the inactive conformation of MSK-1 has been solved,
this structure is likely to differ significantly from its ac-
tive conformation.13 Therefore, a computational homol-
ogy model of active MSK-1 was built, based on the
known12 crystal structure of the closely related protein
kinase A (PKA) complexed with H89 (an equipotent
inhibitor of PKA, IC50 135 nM).11 Compound 1 was
docked into the modelled ATP-binding site using the
Keywords: MSK-1; Mitogen and stress-activated protein kinase;
Imidazo[4,5-c]pyridine.
*
Corresponding author. Tel.: +44 1279 875078; fax: +44 1279
622260; e-mail: Mark_J_Bamford@GSK.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.05.021

M. J. Bamford et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3402–3406
3403
program GOLD and key features of the docking are
shown in Figure 1.
ligand indicates virtual co-planarity of the two hetero-
aryl rings, with possible intramolecular hydrogen
bonding between the amino group and the benzimid-
azole ring nitrogen. The N1-ethyl group is suggested
to bind into a lipophilic pocket constituted by Val53,
Leu55, Phe348 and Leu184. Of particular interest to
our initial chemical strategy, the protonated amine of
conserved Lys81 is predicted to be in close proximity
to C5. Thus, the first targeted analogues (Table 1)
were those with substituents able to act as an H-bond
This model suggests that the 3-amino-1,2,5-oxadiazole
fulfils a key bifurcated hydrogen bonding interaction
with the protein amide backbone of residues Asp132
and Ile134 in the conserved ÔhingeÕ region. The model
suggests restricted space around the 3-amino-1,2,5-
oxadiazole due to ÔgatekeeperÕ residue Leu131 adjacent
to the amine functionality. Energy minimisation of the
Figure 1. GOLD docking of compound 1 into MSK-1 homology model.
Table 1. Inhibition of hMSK-1 by 5- and 6-substituted benzimidazole analogues
Compound
R1
R2
X
Y
MSK-1 (IC50, nM)
1
2
3
4
5
6
7
8
9
10
H
H
H
H
H
H
H
F
C
C
C
C
C
C
C
C
N
C
C
C
C
C
C
C
C
C
C
N
140
1200
4390
974
38
OMe
COCH3
CH2OH
F
OH
H
17
2120
258
3
H
—
OH
H
—
H
1540

3404
M. J. Bamford et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3402–3406
acceptor in the region of C5, aimed at interaction with
Lys81.
data are consistent not only with a very specific H-bond-
ing interaction with the protein as proposed in docking
studies (Fig. 1) but also with the proposed role of the
exocyclic amino group in intramolecular H-bonding.
Whilst the 5-fold greater potency of pyrazine 19 over
pyridine 16 may reflect differences in basicity, it is also
reasonable to suggest that the CH group of the latter
(X = CH) may have a detrimental steric effect on the
proposed coplanarity of the pyridyl and imidazopyri-
dine rings. Indeed, replacement of the N5 ring nitrogen
in oxadiazole 9 with a methyl-bearing carbon in com-
pound 20 completely ablates all MSK-1 activity. Unfor-
tunately the desmethyl 3-amino-isoxazole analogue of
20 proved synthetically elusive. This steric effect may
also be relevant to the 3-aminopyrazole derivative 21
which is 18-fold less potent than 9, although this com-
pound is also capable of existing in an alternative NH-
tautomeric form. N-substitution of 21 to give 22 gives
a further 30-fold reduction in activity, presumably due
to methyl steric clash with the protein backbone adja-
cent to the H-bonding interaction. The importance of
the 2-amino group in the amino-oxadiazole is further
borne out by inactive compounds 23, in which the amine
is acylated so removing one H-bond donor and intro-
ducing steric clashes, and chloro compound 24 which
has no H-bond donor and reduced H-bond acceptor
capability. The corresponding des-amino compound
proved synthetically elusive.
Whilst the more bulky substituents in compounds 2–4
gave a reduction in potency relative to 1, the smaller flu-
oro and hydroxy functionality in compounds 5 and 6
gave significant enhancements to potency. Furthermore,
we were extremely encouraged to find that the 5-aza-
derivative 9 provided a 50-fold increase in potency over
1. In contrast, when the same fluoro or hydroxy func-
tionality was positioned at C6 to give compounds 7
and 8, respectively, no improvement in potency was ob-
served relative to unsubstituted 1, and indeed fluoro
compound 7 is significantly less potent than 1. This
reduction in potency was even more marked in the 6-
aza-analogue 10. These results are consistent with our
original theory that H-bonding substituents at C5 would
be favoured. Furthermore, the introduction of electron
rich groups at C6 is likely to be disfavoured due to the
proposed (Fig. 1) proximity of Asp195.
Having established the imidazopyridine compound 9 as
a potent baseline compound, we next investigated the
requirements of the 3-amino-1,2,5-oxadiazole moiety
(Table 2).
Simple replacement of this moiety with a phenyl (11) or
3-furyl (12) group results in loss of all activity. There is
also no activity in 2-, 4- and 3-pyridyl analogues 13–15.
However, the 2-amino, 3-pyridyl analogue 16 shows
comparable potency to the original lead 1, albeit still
35-fold less potent than direct comparator 9 despite hav-
ing functionality capable of fulfilling the key bifurcated
H-bond. Replacement of the 2-amino group in 16 with
an hydroxyl group to give 17 or movement of the amino
group to the 6-position as in 18 ablates all activity. The
most effective 3-amino-1,2,5-oxadiazole replacement yet
identified is the 2-aminopyrazine compound 19. These
Thus, the 3-amino-1,2,5-oxadiazole group and the
5-aza-modification are key to potency of the benzimid-
azole template. We next turned our attention to modifi-
cation of the N1-substituent (Table 3).
The unsubstituted and methyl substituted analogues 25
and 26, respectively, are approximately 650 and 40 times
less potent than the parent ethyl derivative, supporting
the proposed crucial lipophilic interaction in this region
Table 2. Inhibition of hMSK-1 by selected amino-1,2,5-oxadiazol-modified analogues of compound 1
Compound
X
Y
R
MSK-1 (IC50, nM)
11
12
13
14
—
—
—
—
—
—
—
—
Ph
>10,000
>10,000
>10,000
>10,000
3-Furyl
2-Pyridyl
4-Pyridyl
15
16
17
18
19
CH
CH
CH
CH
N
H
>10,000
107
—
—
—
—
2-NH2
2-OH
6-NH2
2-NH2
>10,000
>10,000
20
9
N
O
NH2
NH2
NH2
NH2
NHAc
Cl
3
>10,000
55
20
21
22
23
24
CMe
CH
CH
N
O
NH
NMe
O
1520
5097
N
O
>10,000

M. J. Bamford et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3402–3406
3405
Table 3. Inhibition of hMSK-1 by representative N1-substituted
analogues
the binding pocket shows a preference for small,
branched alkyl substituents. However, there is scope
for introduction of functionality capable of signifi-
cantly altering the overall properties of the molecule.
This bodes well for the future optimisation of this
template.
The benzimidazole derivatives, such as compound 5,
were prepared14,15 as outlined in Scheme 1. Thus, 4-flu-
orobenzene-1,2-diamine 32 was selectively mono-pro-
tected to give 33 by warming in di(t-butyl)dicarbonate
until homogeneous. This was then deprotonated with
potassium t-butoxide and treated with ethyl iodide to
give 34. The amino-1,2,5-oxadiazol synthon 36 was pre-
pared in moderate yield by treatment of 4-amino-6-
hydroxy-2-methyl-5-nitrosopyrimidine 35 in acetic acid
with lead tetra-acetate. Heating 34 and 36 together in
acetic acid then yielded the desired 5-fluoro-benzimid-
azole derivative 5 in moderate yield.
Compound
R
MSK-1 (IC50 nM)
9
Et
3
1960
126
1.8
3
25
26
27
28
29
30
31
H
Me
Cyclopropyl
Cyclohexyl
Ph
11
17
Piperidin-4-yl
Aminobut-4-yl
13
of the protein. Furthermore, the cyclopropyl derivative
27 and the cyclohexyl derivative 28 afford the greatest
potency, being at least as potent as 9. Lipophilic phenyl
derivative 29 also affords good potency as do the more
polar piperidin-4-yl and aminobutyl substituents in ana-
logues 30 and 31, respectively. The latter derivatives
thus afford the opportunity for manipulation of the
physicochemical properties of the template. Although
entropically less favourable, the aminobutyl chain may
be able to access a beneficial interaction such as with
Glu138 (Fig. 1).
The 5-aza-benzimidazole derivatives were prepared as
outlined in Scheme 2 and is representative of the
approach taken to this class of compound. Thus,
commercially
available
4-methoxy-3-nitropyridine
hydrochloride 37 was stirred with the appropriate pri-
mary amine in ethanol at reflux to afford good yields
of the substituted amino-nitropyridines, for example,
38. Catalytic hydrogenation then afforded the corre-
sponding diaminopyridine 39. Condensation with neat
ethylcyanoacetate at high temperature provided
acceptable yields of the cyanomethyl benzimidazole
intermediate 40. This was then cyclised in a moder-
Thus, the nature of the N1-substituent is also crucial to
the potency of the template. In keeping with the model,
Scheme 1. Preparation of substituted benzimidazole derivatives. Reagents and conditions: (a) di(t-butyl)dicarbonate, 50 °C (73%); (b) KOtBu, THF
then EtI, 21 °C (38%); (c) Pb(OAc)4, AcOH, 21 °C (32%); (d) AcOH, reflux (17%).
Scheme 2. Preparation of substituted aza-benzimidazole derivatives. Reagents and conditions: (a) ethylamine, ethanol, reflux (88%); (b) H2, 10% Pd/
C, 21 °C (94%); (c) ethylcyanoacetate, 190 °C (37%); (d) (i) NaNO2, methanol, HCl, 21 °C, (ii) NaOH, NH2OH, water, 90 °C (43%, two steps); (e)
2-aminopyrazine-3-carboxylic acid, polyphosphoric acid, 130–195 °C (6%).

3406
M. J. Bamford et al. / Bioorg. Med. Chem. Lett. 15 (2005) 3402–3406
4. Soloaga, A.; Thomson, S.; Wiggin, G. R.; Rampersaud,
N.; Dyson, M. H.; Hazzalin, C. A.; Mahadevan, L. C.;
Arthur, J. S. C. EMBO J. 2003, 22, 2788.
5. Vermeulen, L.; De Wilde, G.; Van Damme, P.; Vanden
Berghe, W.; Haegeman, G. EMBO J. 2003, 22, 1313.
6. Hughes, J. P.; Staton, P. C.; Wilkinson, M. G.; Strijbos, P.
J. L. M.; Skaper, S. D.; Arthur, J. S. C.; Reith, A. D.
J. Neurochem. 2003, 86, 25.
ately yielding, two-step reaction sequence consisting of
nitrosylation with sodium nitrite in methanolic HCl,
followed by heating with a basic aqueous solution
of hydroxylamine to give the 3-amino-1,2,5-oxadiazol
derivative 9. Alternatively, diaminopyridine 39 was
added as a mixture with an appropriate aryl carbox-
ylic acid to hot polyphosphoric acid and further
heated to afford benzimidazole derivatives, for exam-
ple, 19.
7. Irving, E. A.; Bamford, M. J. Cereb. Blood Flow Metab.
2002, 22, 631.
8. Barone, F. C.; Irving, E. A.; Ray, A. M.; Lee, J. C.;
Kassis, S.; Kumar, S.; Badger, A. M.; Legos, J. J.;
Erhardt, J. A.; Ohlstein, E. H.; Hunter, A. J.; Harrison, D.
C.; Philpott, K.; Smith, B. R.; Adams, J. L.; Parsons, A.
A. Med. Res. Rev. 2001, 21, 129.
9. Gaiba, A.; Takle, A. K.; Wilson, D. M. WO Patent 66540,
2001.
10. Irving, E. A.; Ray, A. M.; Staton, P. C.; Davis, R. P.;
Hadingham, S. J.; Campbell, C. A.; Roberts, J.; Reith, A.
D.; Hunter, A. J.; Parsons, A. A. J. Cereb. Blood Flow
Metab. 2001, 21, S379.
11. Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P. Biochem
J. 2000, 351, 95.
12. Engh, R. A.; Girod, A.; Kinzel, V.; Huber, R.; Bosse-
meyer, D. J. Biol. Chem. 1996, 271, 26157.
13. Smith, K. J.; Carter, P. S.; Bridges, A.; Horrocks, P.;
Lewis, C.; Pettman, G.; Clarke, A.; Brown, M.; Hughes,
J.; Wilkinson, M.; Bax, B.; Reith, A. Structure 2004, 12,
1067.
A new MSK-1-inhibitory template has been identified
by high-throughput screening. An initial understanding
of the SAR has been established and potent analogues
developed using an MSK-1 protein homology model.
Further optimisation will be discussed in subsequent
papers.
Acknowledgments
The authors acknowledge the assistance of Glaxo-
SmithKline colleagues from the Departments of Gene
Expression and Protein Biochemistry for the recombi-
nant human enzyme, Computational and Structural Sci-
ences for protein homology model generation, and
Assay Development and Compound Profiling for assay
of the compounds.
14. MSK-1 was assayed at a concentration of 2 nM in 50 mM
HEPES buffer, pH 7.5, containing 2 lM biotinylated
peptide (biotin-GRPRTSSFAEG-OH); 20 lM ATP;
25 Bq/pmol ATP33; 10 mM MgCl2; 0.1 mM EDTA;
0.0025% Tween 20; 5 mM b-mercaptoethanol. After
incubation for 1 h at 20 °C, the reaction was stopped by
addition of 200 mM EDTA solution containing streptavi-
din-coated SPA beads (Amersham) to give a final 0.2 mg
of beads per assay well. The plates were shaken for 10 min
before centrifugation at 2500 rpm for 10 min. 33P incor-
poration was quantified by scintillation counting in a
Wallac Trilux.
References and notes
1. Deak, M.; Clifton, A. D.; Lucocq, L. M.; Alessi, D. R.
EMBO J. 1998, 17, 4426.
2. Thomson, S.; Clayton, A. L.; Hazzalin, C. A.; Rose, S.;
Barratt, M. J.; Mahadevan, L. C. EMBO J. 1999, 18,
4779.
3. Wiggin, G. R.; Soloaga, A.; Foster, J. M.; Murray-Tait,
V.; Cohen, P.; Arthur, J. S. C. Mol. Cell. Biol. 2002, 22,
2871.
15. All novel compounds gave satisfactory 1H NMR and LC/
MS data in full agreement with their proposed structures.