Discovery of 4-alkylamino-7-aryl-3-cyanoquinoline LRRK2 kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2013.02.041

Mutations in leucine-rich repeat kinase 2 (LRRK2) are associated with familial Parkinson's disease (PD). The kinase activity of this complex protein is increased by pathogenic mutations. Inhibition of LRRK2 kinase activity has therefore emerged as a promi

Full text of DOI:10.1016/j.bmcl.2013.02.041

Bioorganic & Medicinal Chemistry Letters 23 (2013) 1974–1977
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of 4-alkylamino-7-aryl-3-cyanoquinoline LRRK2 kinase
inhibitors
b
a
c
c
Albert W. Garofalo a, , Marc Adler , Danielle L. Aubele , Elizabeth F. Brigham , David Chian ,
Maurizio Franzini a, Erich Goldbach d, Grace T. Kwong c, Ruth Motter c, Gary D. Probst a, Kevin P. Quinn d,
Lany Ruslim e, Hing L. Sham a, Danny Tam e, Pearl Tanaka c, Anh P. Truong a, Xiaocong M. Ye a, Zhao Ren e
⇑
a Department of Medicinal Chemistry, Elan Pharmaceuticals, 180 Oyster Point Blvd., South San Francisco, CA 94080, United States
b Department of Molecular Design, Elan Pharmaceuticals, 180 Oyster Point Blvd., South San Francisco, CA 94080, United States
c Department of Pharmacology, Elan Pharmaceuticals, 180 Oyster Point Blvd., South San Francisco, CA 94080, United States
d Department of Drug Metabolism and Pharmacokinetics, 180 Oyster Point Blvd., South San Francisco, CA 94080, United States
e Department of Assay Development, Elan Pharmaceuticals, 180 Oyster Point Blvd., South San Francisco, CA 94080, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Mutations in leucine-rich repeat kinase 2 (LRRK2) are associated with familial Parkinson’s disease (PD).
The kinase activity of this complex protein is increased by pathogenic mutations. Inhibition of LRRK2
kinase activity has therefore emerged as a promising approach for the treatment of PD. Herein we report
our findings on a series of 4-alkylamino-7-aryl-3-cyanoquinolines that exhibit kinase inhibitory activity
against both wild type and G2019S mutant LRRK2. Activity was determined in both biochemical and cel-
lular assays. Compound 14 was further evaluated in an in vivo pharmacodynamic study and found to sig-
nificantly inhibit Ser935 phosphorylation after oral dosing.
Received 14 November 2012
Revised 2 February 2013
Accepted 7 February 2013
Available online 15 February 2013
Keywords:
Lrrk2
Leucine-rich repeat kinase 2
Parkinson’s disease
Ó 2013 Elsevier Ltd. All rights reserved.
Parkinson’s disease (PD) is a progressive neurological move-
ment disorder affecting 1–2% of the population over the age of
60 and up to 4% over the age of 80.1,2 In the United States alone
it is estimated that there are about 50,000 new cases reported
annually and this number is expected to grow as life expectancy
increases.3 Current standards of care have serious limitations
and/or drawbacks and largely address only symptoms.4 Because
of these shortcomings there is a clear need for improved therapies.
While most occurrences of PD are sporadic, the discovery of
causal mutations in familial forms of PD has opened numerous
avenues of investigation. Mutations in leucine-rich repeat kinase
2 (LRRK2) are known to cause autosomal-dominant parkinsonism.5
There are at least five pathogenic mutations in LRRK2 that have
been identified and, of those, the G2019S mutation is the most pre-
valent amino acid substitution.6 This mutation is found in the
kinase domain of LRRK2 and is reported to cause an increase in
kinase activity.7–9 This enhanced activity suggests that small mol-
ecule LRRK2 kinase inhibitors may lead to new PD therapeutics.
We employed a kinase inhibitor-focused screen that identified
quinoline amides as hits of which the most potent was compound
1 (Table 1). This screen used an HTRF (homogeneous time-resolved
fluorescence) assay measuring the inhibition of LRRKtide
Table 1
Initial screening hit and amide replacements
NH
5
8
4
3
R
6
2
N
1
7
MeO2S
Compound
R
LRRK2 (wt) LRRK2 (G2019S)
IC50 M) IC50 M)
(l
(l
1
2
3
4
5
6
CONH2
0.039
0.066
1.747
0.041
0.071
2.893
88.597
2.229
0.028
Acetyl
Thiazol-2-yl
4,5-Dihydro-1H-imidazol-2-yl 91.856
1-Methyl-1H-pyrazol-4-yl
2.020
0.031
CN
phosphorylation.10 Compounds of this chemotype are known
inhibitors of CSF1R as well as PDE4 and we were therefore inter-
ested in modifying this hit scaffold.11,12 We also observed that
these quinoline amides have poor CNS penetrance (B/P = 0.1 for
⇑
Corresponding author.
E-mail address: albertwgarofalo@gmail.com (A.W. Garofalo).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.02.041

A. W. Garofalo et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1974–1977
1975
Figure 1. A model of the complex between compound 6 and LRRK2.
NC
CO2Et
b
a
CO2Et
CN
+
I
N
H
I
NH2
EtO
9
8
7
Cl
OH
CN
CN
c
d
I
N
I
N
11
10
NH
N
NH
N
O
B
CN
CN
e
O
+
N
N
I
N
N
14
13
12
Scheme 1. Reagents and conditions: (a) neat, 120 °C (98%); (b) Dowtherm, reflux; (c) POCl3, reflux (43%, 2 steps); (d) isopropylamine, DMA, 100 °C (76%); (e) Pd(PPh3)4,
K2CO3, dioxane, H2O (67%).
1) and significant P-gp mediated efflux (in vitro efflux ratio = 6.2
for 1). We ascribed this poor uptake and high efflux to the presence
of the amide substituent which led to the search for suitable
replacements as listed in Table 1. A twofold loss in potency was
the result of replacing the amide with an acetyl group (compound
2) and much larger decreases in activity were observed with all
heterocyclic replacements examined (e.g., 3, 4, and 5). However,
incorporation of a nitrile at C3 afforded a compound 6, that was
at least equipotent to the screening hit.
We developed a homology model for LRRK2 based on the crystal
structure of mixed-lineage kinase MLK1 (pdb code: 3DTC).13,14
Compound 6 was docked into our model and showed a hinge bind-
ing interaction between the quinoline nitrogen and the A1950
amide (Fig. 1). In this model the methylphenylsulfone is oriented
toward the solvent channel and the isopropylamine occupies space
between V1893 on the P-loop and L2001 in the C-terminal domain.
Encouraged by the activity of 6 we set about making a number of
3-cyanoquinolines. Taking into consideration the binding mode
from our model, we opted to make a set of compounds that were
diversified at C4 and C7. These compounds were generally prepared
as depicted in Scheme 1. Addition of 3-iodoaniline (7) to ethyl 2-
cyano-3-ethoxyacrylate (8) afforded cyanoacrylate 9. Thermal
cyclization followed by chlorination using POCl3 gave the
4-chloro-7-iodoquinoline 11. Treatment of 11 with an appropriate
amine in DMA followed by Suzuki coupling to C7 provided the
desired analogs.
Compound potency was measured in both biochemical and cel-
lular assays and results are reported in Table 2. The in vitro poten-
cies of compounds maintaining the C4 isopropylamino group and
varying substitution at C7 (6, 14–22) were not greatly different.
However, comparisons with the C7 unsubstituted quinoline, 22,
indicate a clear potency advantage for compounds with substitu-
tion at this position. Several analogs with tertiary amine substitu-
tion at C4 were prepared (23–26). Compounds 23–25 exhibited
slightly improved potency over the screening hit, 6, whereas 26
containing a slightly larger, more lipophilic 3° amine, was approx-
imately equipotent. Overall, the data in Table 2 show an increase
in biochemical potency with increasing amine basicity; the most
potent compounds as measured in the biochemical assays also
contain the most basic amine substitution at C4. We have previ-
ously reported on a series of cinnoline amides that demonstrated
improved potency by changing the C4 isopropylamine to (R)-1-
cyclopropylethylamine,15 however, the same benefit was not
realized in this series (cf. 6 and 31, 32).

1976
A. W. Garofalo et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1974–1977
Table 2
Inhibitory activity against wild type (wt) and mutant G2019S LRRK2
A
N
CN
R
Compound
R
A
LRRK2 (wt)
IC50 M)
LRRK2 (G2019S)
IC50 M)
LRRK2 (G2019S)
EC50 M)
(
l
(
l
(l
6
4-(Methylsulfonyl)phenyl
1-Methyl-1H-pyrazol-4-yl
4-(N-Methylsulfamoyl)phenyl
4-(Morpholinosulfonyl)phenyl
2-Methylthiazol-5-yl
4-(Pyrrolidin-1-ylsulfonyl)phenyl
4-(N,N-Dimethylsulfamoyl)phenyl
6-(Methylsulfonyl)pyridin-3-yl
5-Chlorothiophen-2-yl
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Isopropylamino
Azetidin-1-yl
Azetidin-1-yl
Dimethylamino
Piperidin-1-yl
4-Methylpiperazin-1-yl
Methyl(phenyl)amino
Imidazol-1-yl
0.032
0.007
0.009
0.014
0.023
0.012
0.037
0.064
0.109
0.103
0.010
0.016
0.020
0.052
0.281
0.450
0.682
2.333
0.046
0.048
0.028
0.005
0.006
0.010
0.016
0.019
0.028
0.048
0.065
0.288
0.013
0.014
0.027
0.059
0.120
0.278
0.498
1.177
0.039
0.027
0.435
0.140
0.226
0.416
0.545
—
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
—
1.599
2.621
—
0.504
—
1.134
1.599
—
—
—
—
—
H
1-Methyl-1H-pyrazol-4-yl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(Methylsulfonyl)phenyl
4-(2-Oxooxazolidin-3-yl)phenyl
N-Methylacetamido
(R)-Cyclopropylethylamino
(R)-Cyclopropylethylamino
0.731
Biochemical enzyme activity was determined using GST-LRRK2(970-2527), GST-(G2019S)LRRK2(970-2527), and LRRKtide in the presence of 100
was determined using HEK293 cells stably transfected with LRRK2 (G2019S).
lM ATP. Cellular activity
Our cellular assay monitored Ser935 phosphorylation which is
known to depend on LRRK2 kinase activity.16 Good correlation be-
tween in vitro and cellular measures was observed. A correlation
analysis of LRRK2 (G2019S) biochemical IC50’s with cellular EC50’s
showed R2 = 0.88. The most potent compound was 14 which regis-
tered a 140 nM EC50
We profiled compound 6, 14, and 32 against a panel of 40 ki-
nases.17 At a concentration of 1
M, compound 14 inhibited the ki-
nase activities of AURA, FLT3, PLK2 and SYK at greater than 80%.
However, at 1 and 10 M, respectively, neither compounds 6 nor
.
l
l
32 inhibited any kinases in the panel to such an extent. The highest
inhibition observed with these compounds was against SYK (71%
and 80%). These three compounds differ most greatly in their sub-
stitution at C7 and this observation suggests that inhibitor interac-
tions with residues near the solvent channel might be exploited for
improving LRRK2 selectivity.
We assessed the CNS exposure of compounds 6, 14, and 32 in
both Mdr1 a/b (À/À) and P-gp competent wild-type FVB mice to
obtain a measure of P-gp efflux and brain uptake. CNS penetration
was observed at 5 min following a 1 mg/kg IV dose. Penetration
was assessed from the brain to plasma concentration ratios (Kp)
in the wild-type mice and was determined to be 0.7, 1.3, and 0.1
for compounds 6, 14, and 32 respectively. No significant P-gp efflux
was observed and this was consistent with values obtained in vitro
from MDR1-MDCK transfected cells.
Lead compound 14 was further evaluated in an in vivo pharma-
codynamic study. BAC transgenic mice (n = 5 per group) expressing
mouse LRRK2 with the human G2019S mutation were adminis-
tered a single oral dose of compound 14 at 30 and 100 mg/
kg.18,19 Brain samples were collected at 3 h post dose and analyzed
by sandwich ELISA to determine p-Ser935 and total LRRK2 lev-
els.20,21 Figure 2 shows that treatment with LRRK2 inhibitor 14 re-
sulted in a significant reduction of the percent of phosphorylation
at Ser935 at 30 and 100 mg/kg (30% and 44% reduction respec-
tively, ⁄p <0.05 and ⁄⁄⁄p <0.001). Consistent with our observations
Figure 2. In vivo G2019S LRRK2 transgenic mouse pharmacodynamic study results.
Transgenic mice were administered a single oral dose of compound 14 and brains
were collected 3 h post dose for analysis. The p-Ser935 levels were determined by
ELISA and normalized to total LRRK2. Data are expressed as the mean ratio of
p-Ser935 to total LRRK2 normalized to values observed in the vehicle control group.
Error bars represent SEM.
in FVB mice, we observed moderate CNS exposure of compound
14 in these transgenic animals. Brain to plasma concentration ra-
tios were determined to be 1.2 at 30 mg/kg and 2.1 at 100 mg/kg.22
In summary, we have developed a series of 4-alkylamino-7-
aryl-3-cyanoquinolines that inhibit LRRK2 kinase activity as mea-
sured in biochemical and cellular assays. Several compounds in
this series were found to have high selectivity for LRRK2 inhibition

A. W. Garofalo et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1974–1977
1977
D. W.; Kesavapany, S.; Singleton, A.; Lees, A.; Harvey, R. J.; Harvey, K.; Cookson,
M. R. Neurobiol. Dis. 2006, 23, 329.
9. Jaleel, M.; Nichols, R. J.; Deak, M.; Campbell, D. G.; Gillardon, F.; Knebel, A.;
Alessi, D. R. Biochem. J. 2007, 405, 307.
10. Detection was done using biotinylated LRRKtide, Eu(K)-Ab (p-Moesin)/SA-
XL665 measured at 620 and 665 nm. GST-LRRK2 (aa970-2527) proteins
containing either the wt sequence or a G2019S mutation were purchased
from Life Technologies (Carlsbad, CA), and stored at À80 °C in small aliquots
until use.
as measured in a panel of representative kinases. Compound 14
was dosed orally in an in vivo mouse model and found to signifi-
cantly inhibit Ser935 phosphorylation at doses of 30 and 100 mg/
kg in brain.
Acknowledgments
11. Scott, D. A.; Dakin, L. A.; Del Valle, D. J.; Diebold, R. B.; Drew, L.; Gero, T. W.;
Ogoe, C. A.; Omer, C. A.; Repik, G.; Thakur, K.; Ye, Q.; Zheng, X. Bioorg. Med.
Chem. Lett. 2011, 21, 1382.
12. Lunniss, C. J.; Cooper, A. W. J.; Eldred, C. D.; Kranz, M.; Lindvall, M.; Lucas, F. S.;
Neu, M.; Preston, A. G. S.; Ranshaw, L. E.; Redgrave, A. J.; Robinson, J. E.; Shipley,
T. J.; Solanke, Y. E.; Somers, D. O.; Wiseman, J. O. Bioorg. Med. Chem. Lett. 2009,
19, 1380.
13. Hudkins, R. L.; Diebold, J. L.; Tao, M.; Josef, K. A.; Park, C. H.; Angeles, T. S.;
Aimone, L. D.; Husten, J.; Ator, M. A.; Meyer, S. L.; Holskin, B. P.; Durkin, J. T.;
Fedorov, A. A.; Fedorov, E. V.; Almo, S. C.; Mathiasen, J. R.; Bozyczko-Coyne, D.;
Saporito, M. S.; Scott, R. W.; Mallamo, J. P. J. Med. Chem. 2008, 51, 5680.
14. MLK1 (3DTC) was selected as the template for the homology model because its
sequence was the closest human kinase domain in the pdb (BLAST
score = 1eÀ25, with 31% identities). However, several features of the 3DTC
structure indicate that it is an inactive conformation. The P-loop in 3DTC
undergoes a ligand induced conformational change and the DFG sequence is
partially disordered. The corresponding parts in our homology models were
rebuilt using the structure of homologous kinases that were in the fully
activated state.
15. Garofalo, A. W.; Adler, M.; Aubele, D. L.; Bowers, S.; Franzini, M.; Goldbach, E.;
Lorentzen, C.; Neitz, R. J.; Probst, G. D.; Quinn, K. P.; Santiago, P.; Sham, H. L.;
Tam, D.; Truong, A. P.; Ye, X. M.; Ren, Z. Bioorg. Med. Chem. Lett. 2013, 23, 71.
16. Dzamko, N.; Deak, M.; Hentati, F.; Reith, A. D.; Prescott, A. R.; Alessi, D. R.;
Nichols, R. J. Biochem. J. 2010, 430, 405.
The authors thank Wayman Chan for performing difficult puri-
fications. We would like to acknowledge the work of Pui Seto,
Monica You, Jeanne Baker, and Kevin Tanaka for the production
and purification of antibodies used in our assays. We also want
to acknowledge Colin Lorentzen and Shawn Gauby for in vitro
DMPK support and Pam Santiago for in vivo DMPK support. We
also want to especially thank Dr. Takeshi Iwatsubo (University of
Tokyo) for the generous gift of the HEK293 cell line stably express-
ing LRRK2-G2019S.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.02.
041.
References and notes
1. Samii, A.; Nutt, J. G.; Ransom, B. R. Lancet 2004, 363, 1783.
2. de Lau, L. M. L.; Breteler, M. M. B. Lancet Neurol. 2006, 5, 525.
3. National Institute of Neurological Disorders and Stroke, National Institutes of
17. Compounds were profiled at Invitrogen in
Supplementary data).
a single-point assay (see
18. Li, X.; Patel, J. C.; Wang, J.; Avshalumov, M. V.; Nicholson, C.; Buxbaum, J. D.;
Elder, G. A.; Rice, M. E.; Yue, Z. J. Neurosci. 2010, 30, 1788.
Health.
http://www.ninds.nih.gov/disorders/parkinsons_disease/
parkinsons_disease_backgrounder.htm (accessed Feb 2013)
4. Rascol, O.; Payoux, P.; Ory, F.; Ferreira, J. J.; Brefel-Courbon, C.; Montastruc, J.-L.
Ann. Neurol. 2003, 53, S3.
5. Zimprich, A.; Biskup, S.; Leitner, P.; Lichtner, P.; Farrer, M.; Lincoln, S.;
Kachergus, J.; Hulihan, M.; Uitti, R. J.; Calne, D. B.; Stoessl, A. J.; Pfeiffer, R. F.;
Patenge, N.; Carbajal, I. C.; Vieregge, P.; Asmus, F.; Müller-Myhsok, B.; Dickson,
D. W.; Meitinger, T.; Strom, T. M.; Wszolek, Z. K.; Gasser, T. Neuron 2004, 44,
601.
19. Breeding pairs were licensed from the Mt. Sinai School of Medicine.
20. An antibody recognizing the N-terminal region of LRRK2 (Elan
Pharmaceuticals) was used for antigen capture, while another antibody
against either LRRK2 (Elan Pharmaceuticals) or pSer935 (Epitomics) was
used for detection. The data were normalized as a percentage of the vehicle-
treated controls.
21. Our ELISA was calibrated by comparing results for a cross-section of LRRK2
inhibitors against Western blots.
2 = 0.82. The correlation plot is included in the Supplementary data.
22. A table listing compound levels is included in the Supplementary data.
A linear correlation was observed with
6. Greggio, E.; Cookson, M. R. ASN Neuro 2009, 1, 13.
R
7. West, A. B.; Moore, D. J.; Biskup, S.; Bugayenko, A.; Smith, W. W.; Ross, C. A.;
Dawson, V. L.; Dawson, T. M. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 16842.
8. Greggio, E.; Jain, S.; Kingsbury, A.; Bandopadhyay, R.; Lewis, P.; Kaganovich, A.;
van der Brug, M. P.; Beilina, A.; Blackinton, J.; Thomas, K. J.; Ahmad, R.; Miller,