N. A. Powell et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1046–1050
1049
Figure 2. X-ray crystal structure of 29 (yellow stick) bound in the ATP active site of Sky (pdb ID code: 4FF8).10 The glycine-rich loop has been removed for clarity.
ing the Ala571 pocket is key to achieving selectivity for Sky versus
a representative selectivity panel.8 As expected from the observed
binding mode in the crystal structure, these compounds also inhib-
ited off-target kinases as illustrated by analog 18 which exhibited
potent inhibition of Abl, FGFR, and Lck, as well as other members of
the HGFR family, Axl, and Mer (Table 2).11
To improve the kinase selectivity, we hypothesized that a short-
er and less conformationally constrained sidechain at the pyrimi-
dine 2-position would better fill the Ala571 subpocket.
Accordingly, we investigated the SAR of a series of 2-benzylamino
analogs (Table 3). The unsubstituted benzyl analog 21 exhibited
2.6-fold reduced inhibitory activity compared to the phenethyl
analog 8. Substitution at the para-position resulted in further
reduction in activity (analogs 22–24), indicating that steric con-
straints were overriding any electronic or lipophilic substituent
properties. The 3-Cl analog 25 exhibited improved Sky inhibition
that the primary metabolite was oxidative hydroxylation at the
-carbonyl C atom in the butyrolactam ring. To block this route
of metabolism, we converted the butyrolactam ring to an oxazoli-
din-2-one ring, which exhibited potent Sky inhibition activity (32,
IC50 = 0.070 lM), but no improvement in HLM stability.
In conclusion, we have discovered a series of Sky inhibitors
based on a 2,4-diaminopyrimidine-5-carboxamide template as po-
tential anti-platelet therapy. 2-Phenethylamino analogs with
hydrogen-bond accepting groups or phenols at the para- or meta-
positions of the phenyl ring exhibited potent inhibition of Sky,
but poor kinase selectivity. 2-Benzylamino analogs with ortho- or
meta-substituents exhibited good inhibition activity and excellent
kinase selectivity. These analogs suffer from poor HLM stability. Ef-
forts to improve the HLM stability of this series will be the subject
of the following article.
a
activity (IC50 = 0.85 lM), while the 3-CF3 and 3-CH3 substituents
were detrimental to activity (26 and 27, respectively). Electron-
Supplementary data
withdrawing groups were also not tolerated at the ortho-position
Supplementary data associated with this article can be found, in
(28, IC50 = 1.58
l
M), while the electron-donating 2-OCH3 group
sharp improvement in Sky inhibition (29,
M). Further extension of the ortho-substituent to a 2-
exhibited
IC50 = 0.40
a
l
OEt group resulted in further improved inhibition activity (30,
IC50 = 0.215 M). The combination of an ortho-Cl and meta-Cl sub-
stituents into a 2,5-disubstitution pattern also resulted in im-
l
References and notes
1. American Heart Association. Heart Disease and Stroke Statistics—2007 Update.
2. Cattaneo, M. Arterioscler. Thromb. Vasc. Biol. 2004, 24, 1980.
3. Saller, F.; Burnier, L.; Schapira, M.; Angelillo-Scherrer, A. Blood Cells Mol. Dis.
2006, 36, 373.
4. Angelillo-Scherrer, A.; de Frutos, P. G.; Aparicio, C.; Melis, E.; Savi, P.; Lupu, F.;
Arnout, J.; Dewerchin, M.; Hoylaerts, M. F.; Herbert, J.-M.; Collen, D.; Dahlbäck,
B.; Carmeliet, P. Nat. Med. 2001, 7, 215.
5. Gould, W. R.; Baxi, S. M.; Schroeder, R.; Peng, Y. W.; Leadley, R. J.; Peterson, J. T.;
Perrin, L. A. J. Thromb. Haemost. 2005, 3, 733.
6. Nagata, K.; Ohashi, K.; Takano, T.; Arita, H.; Zong, C.; Hanafusa, H.; Mizuno, K. J.
Biol. Chem. 1996, 271, 30022.
7. Angelillo-Scherrer, A.; Burnier, L.; Flores, N.; Savi, P.; DeMol, M.; Schaeffer, P.;
Herbert, J.-M.; Lemke, G.; Goff, S. P.; Matsushima, G. K.; Earp, H. S.; Vesin, C.;
Hoylaerts, M. F.; Plaisance, S.; Collen, D.; Conway, E. M.; Wehrle-Haller, B.;
Carmeliet, P. J. Clin. Invest. 2005, 115, 237.
8. Powell, N. A.; Kohrt, J. T.; Filipski, K. J.; Kaufman, M.; Sheehan, D.; Edmunds,
Jeremy J.; Delaney, A.; Wang, Y.; Bourbonais, F.; Lee, D.-Y.; Schwende, F.; Sun,
proved activity relative to the monosubstituted chlorobenzyl
analogs (31, IC50 = 0.355 lM). A crystal structure of analog 29
was obtained that confirmed our hypothesis that a less conforma-
tionally mobile benzyl sidechain would fill the Ala571 subpocket
(Fig. 2).8 As expected from this binding mode, analogs 29 and 31
exhibited excellent kinase selectivity against a panel of 20 & 28
kinases, respectively (Table 2).
All compounds examined thus far exhibited low metabolic sta-
bility when incubated with human liver microsomes. While the
aromatic ring in the 2-benzylamine sidechain represented a poten-
tial site of metabolism, the presence of electron-withdrawing or
electron-donating substituents had no affect on the HLM stability.
Identification of the metabolites by LC–MS–MS methods indicated