SB590885 – DMXAA chemical

The identiﬁcation of potent and selective imidazole-based inhibitors of B-Raf kinase
Andrew K. Takle,a,* Murray J. B. Brown,b Susannah Davies,a David K. Dean,a Gerraint Francis,a Alessandra Gaiba,a Alex W. Hird,a Frank D. King,a Peter J. Lovell,a Antoinette Naylor,a Alastair D. Reith,c Jon G. Steadmana and David M. Wilsona
aDepartment of Medicinal Chemistry, Neurology and GI Centre of Excellence for Drug Discovery, GlaxoSmithKline Pharmaceuticals, New Frontiers Science Park, Third Avenue, Harlow, Essex CM19 5AW, UK
bDepartment of Screening and Compound Proﬁling, GlaxoSmithKline Pharmaceuticals, Medicines R&D Centre,
Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
cDepartment of High Throughput Biology, GlaxoSmithKline Pharmaceuticals, Medicines R&D Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
Received 28 July 2005; revised 23 September 2005; accepted 28 September 2005
Available online 2 November 2005

Abstract—A novel triarylimidazole derivative, SB-590885 (33), bearing a 2,3-dihydro-1H-inden-1-one oxime substituent has been identiﬁed as a potent and extremely selective inhibitor of B-Raf kinase.
© 2005 Elsevier Ltd. All rights reserved.

Mitogen-activated protein (MAP) kinases are a family of serine/threonine protein kinases that participate in signal transduction pathways controlling numerous intra-cellular events.1 MAP kinases are regulated by phosphorylation cascades whereby activation of an up- stream kinase leads to phosphorylation of a downstream substrate which itself has protein kinase activity. Typi-

focal cerebral ischemia in in vivo rodent models of stroke.4 Furthermore, inhibition of the cascade at the MEK level has proved to be neuroprotective leading to a signiﬁcant reduction in infarct volume in such animal models.5

cally in such MAP kinase cascades, three protein kinases H N N
are sequentially activated in response to appropriate N
extracellular stimuli and allow for signal ampliﬁcation N
at each step of the cascade. F OH

The RAF-MEK-ERK MAP kinase cascade appears to be intimately involved in the regulation of cell cycle pro- gression and apoptosis. Indeed, activating mutations in B-Raf, one of the Raf family members, are reported to be present in 66% of malignant melanomas.2 Disruption of this signaling cascade could thus oﬀer a novel ap- proach for cancer chemotherapy.3 Conversely, increased activation of ERK1/2 has been reported in a number of in vitro models of neuronal cell death and following

Keywords: B-Raf; Imidazole.
* Corresponding author. Tel.: +44 (0) 1279 627686; fax: +44 (0) 1279
622260; e-mail: [email protected]

(1) B-Raf IC50 900nM (2) B-Raf IC50 10nM

We sought to identify inhibitors of B-Raf, the likely ma- jor Raf isoform in the central nervous system, to assess their potential as neuroprotective agents in the treat- ment of stroke.

Screening of the SmithKline Beecham compound bank identiﬁed the tri-substituted imidazole (1) as a sub- micromolar inhibitor of B-Raf. Concurrently we became aware of a poster publication from Merck disclosing a related tri-aryl imidazole 2 (L-779,450) as a highly potent low nanomolar inhibitor of Raf.6 Compound (2)

A. K. Takle et al. / Bioorg. Med. Chem. Lett. 16 (2006) 378–381 379

was reported to demonstrate good selectivity over other kinases tested, with the exception of p38 where the selec- tivity was approximately 30-fold. Unfortunately 2 is

Table 2. B-Raf activity of C5 modiﬁed imidazoles
H
N

R
poorly soluble in aqueous systems ( 0.005 mg/ml in N
pH 5 buﬀer) thus precluding its use as an in vivo tool
Cl
where administration via iv infusion is desirable. We OH
sought to explore the SAR of the imidazole leads whilst

being mindful of the selectivity and developability issues described above.

Our studies began with the synthesis of the hybrid mol- ecule 3 which displayed comparable activity to 2. Further investigation demonstrated that the imidazole C2 position was tolerant of a variety of substituents (Table 1). Heteroaryl derivatives such as 4 and 5 showed comparable potency with 2, whereas the unsubstituted derivative 6 only showed a modest reduction in potency. Introduction of a basic amine substituent on either the C2 aromatic or tert-butyl groups was well tolerated (e.g., compounds 7 and 8) as was the carboxamide linked derivative 9. Such derivatives lead to enhanced aqueous solubility but also provide a handle for further substitution (e.g., 10, 11) and modulation of the physi- cochemical properties of the series.

Acylation of the C2 benzylamine 7 was also of value in the generation of a ﬂuorescent derivative 12 (SB-477790) which proved to be suitable for the development of a B-Raf ﬂuorescent ligand binding assay (FP). The two B-Raf assay formats showed a good correlation in rank order potency but, due to increased throughput, the FP format was used to generate all subsequent SAR.

Compound R R2 B-Raf (FP)
Kd nM
2 H 4-Pyridyl 2.4
13 H Phenyl >200
14 H 3-Pyridyl >200
15 H 4-Pyrimidinyl 14
16 H 4-Pyridazinyl 87
17 CN 2-Amino, 4-pyrimidinyl 5.7
Our investigation of the imidazole C5 substituent showed trends similar to those reported in the p38 ki- nase area (Table 2).8 Replacement of the 4-pyridyl C5 substituent of 2 by phenyl- or 3-pyridyl groups (13 and 14, respectively) produced compounds that were essen- tially devoid of activity in the FP assay, thus implying that an appropriately positioned H-bond acceptor at C5 is essential for activity. This is consistent with the well-established binding mode of such imidazole-based kinase inhibitors where the C5 substituent forms an H- bond to the ‘hinge’ region of the protein.8 As with p38, B-Raf activity can be retained if the C5 pyridine group is replaced by an alternative H-bond acceptor such as in the pyrimidinyl- and pyridazinyl-derivatives 15 and 16, albeit with a 5- to 30-fold reduction in potency. The potency drop can be mitigated through the intro- duction of a 2-amino-4-pyrimidinyl substituent as in 17, which is postulated to pick up a second H-bonding interaction with the hinge region of the protein.

2
Table 3. B-Raf activity of C4 modiﬁed imidazolesa,b

Cl OH

H
H2N N
SB-477790 (12) N
R3

Table 1. B-Raf activity of C2 modiﬁed imidazoles7

H N

Compound R3 B-Raf (FP)
Kd nM

R1
N

Cl
OH

Compound R1

B-Raf IC50
nM

B-Raf (FP)
Kd nM

a Data from the B-Raf kinase assay is shown in parentheses (see Ref. 7).
b Oxime derivatives isolated as mixtures of E and Z isomers.
c Data derived from a variant of the FP binding assay and hence is not directly comparable.

380 A. K. Takle et al. / Bioorg. Med. Chem. Lett. 16 (2006) 378–381

Early investigation of the imidazole C4 position SAR demonstrated that the 3-hydroxy phenyl substituent was key for potent B-Raf activity (Table 3). The mono-substituted 3-hydroxy derivative 18 showed comparable aﬃnity to 2, whereas the 4-hydroxy isomer
19 and 3-methoxy derivative 20 showed signiﬁcantly reduced B-Raf aﬃnity. These data suggested that the 3-hydroxy group was acting as an H-bond donor and so we embarked upon a search to identify alterna- tive groups to perform this role. We were encouraged to ﬁnd that the 4-hydroxymethyl derivative 21 showed modest aﬃnity in the FP assay (Kd 71 nM), which was in contrast to the regioisomeric 3-hydroxymethyl deriv- ative 22. Further substitution of the C4 phenyl ring was tolerated as exempliﬁed by the 3-hydroxy- and 3-chloro-4-hydroxymethyl derivatives 23 and 24 but this had only a modest eﬀect on potency. Transforma- tion of 21 into the oxime derivative 25 however pro- duced a 15-fold increase in potency (Kd 5 nM). Once again, substitution with hydroxyl- or chloro-substitu- ents in the 3-position was tolerated (26 and 27) although the hydroxyl derivative 26 showed a 5-fold drop in potency compared to 25. Further exploration of the oxime unit demonstrated that substitution on the oxime carbon atom was also tolerated in the form of the ketoxime 28 (Kd 11 nM), although the amidox-

Table 4. B-Raf activity of cyclic oximes

H
N

ime derivative 29 showed a 10-fold drop in potency relative to 25.

Cyclization between the 3 and 4 positions of the C4 substi- tuent to form the 2,3-dihydroindene and 3,4-dihydro- naphthalenone derivatives (30–32) also provided some interesting SAR ﬁndings (Table 4). Whilst the racemic 1-hydroxy 2,3-dihydroindene 30 showed weak aﬃnity, the oxime 31 (Kd 1.3 nM) was extremely potent with activ- ity at least equal to that of the early lead 2. The ring expanded 3,4-dihydro-1(2H)-naphthalenone oxime 32 (Kd 17 nM) also showed signiﬁcant activity, although this was 10-fold lower than that of 31. SB-590885 (33, Kd
0.3 nM), the C2 (4-dimethylaminoethoxy) phenyl ana- logue of 31, was prepared to improve the aqueous solubil- ity (>1 mg/ml in pH 5 buﬀer) but also showed enhanced potency, being some 8-fold more potent than the early lead 2.

NMe2
O

HO
SB-590885 (33)
B-Raf Kd 0.3nM

The selectivity proﬁles of SB-590885 (33) and 2 were as- sessed against a panel of 21 protein kinases (Table 5). The data conﬁrmed the generally good selectivity proﬁle

X Table 6. Selectivity of SB-590885 (33) and (2) versus p38a, GSK3b and lck

Compound X n B-Raf (FP) Kd nM Compound Fold selectivity versus B-Raf

30 H, OH (R/S) 1 >200 p38a GSK3b lck
31 HON 1 1.3 2 7 30 70
32 HON 2 17 33 >1000 >1000 >1000

Table 5. Selectivity proﬁling of SB-590885 (33) and (2)a

Com

AMPK Chk1 CK2 GSK JNK1 lck MAPK2 MAPK MEK1 MSK1 p70

Phos. PKA PKB PKC PRAK ROCK- p38 p38 p38 p38d SGK

pound 3b

AP-K2

S6K K a a

II a b c

2 12 1 —3 83 0 88 —3 —1 5 7 26 17 12 44 2 19 3 95 81 5 —15 2
33 14 9 8 25 19 39 11 9 14 38 13 3 —2 8 10 18 26 46 9 10 10 17

a Values are %inhibition at 10 lM drug concentration in kinase activity assays in the presence of 100 lM ATP (see Ref. 9 for details).

OTBDMS

OH
iii, iv v

H
vi N

NMe2

(37) R

i (34) X = O; R = Br
(35) X = NOMe; R = Br
(36) X = NOMe; R = CHO

MeON

(38)

MeON

(39)

X
vii

(40) X = NOMe
(41) X = O
(33) X = NOH

viii

Scheme 1. Reagents and conditions: (i) MeONH2 HCl, pyridine, ethanol (97%); (ii) nBuLi, DMF, THF —60 °C (65%); (iii) LDA, THF —60 °C, add 36; (iv) TBAF, THF (64% over two steps); (v) DMSO, oxalyl chloride, triethylamine dichloromethane 60 C to room temperature (94%); (vi) 4-(2-dimethylaminoethoxy)benzaldehyde, ammonium acetate, acetic acid 100 °C (30%); (vii) 5 M HCl, acetone, dioxane 100 °C (56%); (viii) 50% aqs
NH2OH, ethanol, reﬂux (100%).

A. K. Takle et al. / Bioorg. Med. Chem. Lett. 16 (2006) 378–381 381

of 2 but also identiﬁed some activity against GSK3b and lck. SB-590885, on the other hand, appeared to be de- void of signiﬁcant activity against this panel of enzymes. More in-depth proﬁling of both compounds against p38a, GSK3b, and lck in the ﬂuorescent binding assay format conﬁrmed the enhanced selectivity proﬁle for SB-590885 (Table 6).

Triaryl imidazoles, such as SB-590885 (33), were pre- pared as outlined in Scheme 1.10,11 Thus, commercially available 5-bromo-2,3-dihydro-1H-inden-1-one 34 was converted to the O-methyl oxime 35 and then formylat- ed by treatment with n-butyl lithium followed by the addition of DMF, to aﬀord 36. Reaction of 36 with the anion derived from the 4-hydroxymethylpyridine derivative 37,12 followed by desilylation, aﬀorded the ra- cemic 1,2-diol 38. This was then oxidized to the dione 39 under Swern conditions and converted to the imidazole 40 in modest yield, by treatment with 4-(2-dimethylami- noethoxy) benzaldehyde and ammonium acetate in ace- tic acid at 100 °C. Hydrolysis of the O-methyl oxime, by treatment with 5 M hydrochloric acid and acetone in dioxane, formed the ketone 41 which was converted to 33 by treatment with aqueous hydroxylamine.

In summary, SB-590885 (33), a potent and extremely selective inhibitor of B-Raf kinase, was identiﬁed follow- ing an evaluation of the SAR of a series of imidazole based leads. SB-590885 represents an excellent molecule with which to investigate the role of B-Raf in neurode- generative and other disease states.

Acknowledgments

The authors thank colleagues from the departments of Gene Expression and Protein Biochemistry for the pro- vision of human recombinant reagents and Screening and Compound Proﬁling for kinase activity data. We also acknowledge the Division of Signal Transduction Therapy, University of Dundee, for the generation of kinase selectivity data.

References and notes

1. Pearson, G.; Robinson, F.; Beers, T.; Xu, B.; Karandikar, M.; Berman, K.; Cobb, M. H. Endocr. Rev. 2001, 22, 153.
2. Davies, H. et al. Nature 2002, 417, 949.
3. (a) Chang, F.; Steelman, L. S.; Shelton, J. G.; Lee, J. T.; Navolanic, P. M.; Blalock, W. L.; Franklin, R.; McCu- brey, J. A. Int. J. Oncol. 2003, 22, 469; (b) Hilger, R. A.; Scheulen, M. E.; Strumberg, D. Onkologie 2002, 25, 511.
4. see: Irving, E. A.; Barone, F. C.; Reith, A. D.; Hading- ham, S. J.; Parsons, A. A. Mol. Brain Res. 2000, 77, 65, and references therein.
5. (a) Namura, S.; Iihara, K.; Takami, S.; Nagata, I.; Kicku- chi, H.; Matsushita, K.; Moskowitz, M. A.; Bonventre, J. V.; Alessandrini, A. Proc. Natl. Acad. Sci. U.S.A 2001, 98, 11569; (b) Wang, X.; Wang, H.; Xu, L.; Rozanski, D. J.; Sugawara, T.; Chan, P. H.; Trzaskos, J. M.; Feuerstein, G.
Z. J. Pharmacol. Exp. Ther. 2003, 304, 172.
6. Heimbrook, D. C.; Huber, H. E.; Stirdivant, S. M.; Patrick,
D. R.; Claremon, D.; Liverton, N.; Selnick, H.; Ahern, J.; Conroy, R.; Drakas, R.; Falconi, N.; Hancock, P.; Robin- son, R.; Smith, G.; Olif, A. American Association for Cancer Research, New Orleans, April 1998 poster #3793.
7. Compound activity is determined as either the ability of test compounds to inhibit B-Raf mediated phosphoryla- tion of kinase dead MEK (data expressed as IC50) or as a B-Raf-binding aﬃnity derived from the ﬂuorescent ligand displacement assay (FP, data expressed as Kd). For assay details, see Dean, D. K.; Takle, A. K.; Wilson, D. M. PCT Int. Appl. WO 02/24680.
8. Adams, J. L.; Badger, A. M.; Kumar, S.; Lee, J. C. Prog. Med. Chem. 2001, 38, 1.
9. Davies, S. P.; Reddy, H.; Caivano, M.; Cohen, P.
Biochem. J. 2000, 351, 95.
10. All novel compounds gave satisfactory 1H NMR and LC/ MS data in full agreement with their proposed structures.
11. SB-590885 was isolated as a 4 to 11:1 mixture of E and Z isomers. The conﬁguration of the major (E) isomer was conﬁrmed by NMR.
12. Gallagher, T. F.; Siebel, G. L.; Kassis, S.; Laydon, J. T.; Blumenthal, M. J.; Lee, J. C.; Lee, D.; Boehm, J. C.; Frier- Thompson, S. M.; Abt, J. W.; Soreson, M. E.; Smietana,
J. M.; Hall, R. F.; Garigipati, R. S.; Bender, P. E.; Erhard,
K. F.; Krog, A. J.; Hofmann, G. A.; Sheldrake, P. L.; McDonnell, P. C.; Kumar, S.; Young, P. R.; Adams, J. L. Bioorg. Med. Chem. 1997, 5, 49.SB590885