4086
T. Troxler et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4085–4090
O
N
O
O
NH2
NH2
N
H
N
H
N
N
H
N
H
N
H
F
O
O
O
O
O
N
H
N
H
N
H
Sunitinib
1
2
Figure 1. Discovery of 1,4,6,7-tetrahydro-pyrrolo[3,2-c]pyridine containing LRRK2 inhibitors.
generation inhibitors like the 2,4-diaminopyrimidines LRRK2-IN-
147 or CZC-2514648 showed excellent potency and selectivity for
LRRK2, but lacked sufficient brain exposure to be used in rodent
models of PD. In contrast to this, the dual ALK/LRRK2 inhibitor
TAE684 has been shown to achieve significant mouse brain expo-
sure, but failed to inhibit LRRK2 phosphorylation in the brain after
oral dosing.36 More recently, further optimized 2,4-diaminopyrim-
idines with improved brain penetration have been disclosed by
several groups.30,40,41,44 Some of them were shown to significantly
reduce LRRK2 autophosphorylation in the brain of G2019S LRRK2
transgenic mice after oral dosing.30,44 In addition to these 2,4-dia-
minopyrimidines based compounds, cinnolines,42 triazolopyrida-
zines49 and 3-cyanoquinolines43 have been published that
potently inhibit wild-type and mutant LRRK2 in vitro.
Early on in our LRRK2 kinase inhibitors program, a medicinal
chemistry effort was initiated with two main goals: (i) identifica-
tion of a potent, reasonably selective LRRK2 kinase inhibitor with
an attachment point (preferably a primary or secondary amine) re-
mote from the kinase interaction part that allows for crosslinking
to a solid support for pulldown experiments,50 and (ii) identifica-
tion of a proprietary starting point for a derivation program with
the potential for high potency, selectivity and brain penetration.
For the first goal, we decided to start from the indolinone Sunitinib,
a broad-band kinase inhibitor that was found to also inhibit
LRRK251 (Fig. 1). In our biochemical assay, Sunitinib inhibited
the main goals of improving kinase selectivity and demonstrating
favorable PK properties, including brain penetration.
Derivatives 2–14 were prepared by condensation of an appro-
priately substituted indolinone (as exemplified by 17) with the
benzyl
protected
tetrahydropyrrolo-pyridine-2-carbaldehyde
building block 19, followed by debenzylation and introduction of
the final substituent at the tetrahydropyridine nitrogen (e.g., acetyl
for derivative 11, Scheme 1). 5-substituted indolinones were either
commercially available, or were prepared from the corresponding
indole by a bromination/hydrolysis sequence (e.g., 16–17, Scheme
1). Aldehyde building block 19 was prepared as described by
Voskressensky et al.52 from 1-benzyl-piperid-4-one oxime (KOH,
acetylene gas, DMSO), followed by selective formylation in position
2 of the pyrrole derivative 18 (POCl3, DMF/Et2O) (Scheme 1). Deriv-
ative 7 bearing a chloro substituent in position 3 of the pyrrole
moiety was prepared starting from 18 by de-benzylation (H2, Pd/
C), acetylation (AcCl, Et3N, DCM), formylation in position 2 (POCl3,
DMF/Et2O) and chlorination in position 3 (NCS, benzoylperoxide,
CCl4), followed by condensation with 5-methoxyindolinone (EtOH,
cat. piperidine, reflux). Compound 15 (Fig. 2) carrying a methyl
HO
N
N
H
O
LRRK2 kinase with an IC50 of 0.028
found to be highly unselective in a kinase selectivity panel (inhibi-
tion of 20 out of 54 kinases with an IC50 <1 M). A limited deriva-
lM, and, as expected, was
a
d, e
l
O
N
N
tion program around Sunitinib revealed that the ethylene linker at
the amide can be extended in length, and the tertiary amine at its
end be replaced with a primary amine without loss of activity in
the biochemical assay. In addition, replacement of the fluorine in
position 5 of the indolinone core with a methoxy considerably im-
proved kinase selectivity, while retaining full activity on LRRK2.
N
H
N
H
16
18
19
f
b, c
O
Hence, derivative 1 inhibited LRRK2 with an IC50 of 0.046
and blocked only 5 out of 36 other kinases in the selectivity panel
with an IC50 <1 M. This cross-linkable compound could success-
lM,
O
N
N
H
H
O
g
17
l
fully be coupled to sepharose solid support (see Supplementary
data), and turned out to be a highly versatile tool for pull-down
experiments.
N
In order to obtain proprietary starting points for a drug develop-
ment program, options were considered on how to morph the 2,4-
dimethyl-3-carboxamide substituted pyrrole into a novel, hitherto
unknown moiety. Docking studies of 1 in a LRRK2 homology model
(vide infra) suggested that forming an additional ring between the
2-methyl group and the amide nitrogen, with concomitant inver-
sion of the amide, should be tolerated by the kinase. Indeed,
4,5,6,7-tetrahydro-1H-pyrrolo[3,2-c]pyridine derivative 2, the cor-
N
H
O
O
N
H
20
h, i
11
responding analog of 1, showed an IC50 of 0.011 lM on LRRK2, ni-
Scheme 1. Synthesis of LRRK2 inhibitor 11. Reagents and conditions: (a) i-PrOH,
PPh3, DIAD, THF, rt, 16 h (51%); (b) Br2, DMF, 0°, 300 (quant.); (c) phosphoric acid,
MeOCH2CH2OH, 100°, 2 h (28%); (d) NH2OHꢀHClꢀK2CO3, EtOH, 80°, 1 h (98%); (e)
acetylene gas, KOH, DMSO, 90°, 6 h (36%); (f) POCl3, DMF, Et2O, rt, 1.5 h (71%); (g)
EtOH, cat. piperidine, 95°, 4 h (92%); (h) ammonium formate, Pd/C, MeOH, rflx, 2 h
(97%); (i) ClCOMe, DIEA, DCM, rt, 1 h (75%).
cely confirming this hypothesis. Although some of the kinase
selectivity of 1 was lost by incorporation of this new moiety (20
out of 32 other kinases inhibited with IC50 <1 lM), it was decided,
based on the high initial potency and the favorable IP position for
this novel moiety, to initiate a derivation program around 2 with