Discovery of selective and nonpeptidic cathepsin S inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.06.009

Nonpeptidic, selective, and potent cathepsin S inhibitors were derived from an in-house pyrrolopyrimidine cathepsin K inhibitor by modification of the P2 and P3 moieties. The pyrrolopyrimidine-based inhibitors show nanomolar inhibition of cathepsin S with

Full text of DOI:10.1016/j.bmcl.2008.06.009

Bioorganic & Medicinal Chemistry Letters 18 (2008) 3959–3962
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of selective and nonpeptidic cathepsin S inhibitors
a
a
a
a
a
Osamu Irie a, , Takeru Ehara , Atsuko Iwasaki , Fumiaki Yokokawa , Junichi Sakaki , Hajime Hirao ,
Takanori Kanazawa a, Naoki Teno a, Miyuki Horiuchi a, Ichiro Umemura a, Hiroki Gunji a, Keiichi Masuya a,
Yuko Hitomi a, Genji Iwasaki a, Kazuhiko Nonomura a, Keiko Tanabe a, Hiroaki Fukaya a, Takatoshi Kosaka a,
Christopher R. Snell b, Allan Hallett b
*
a Novartis Institutes for BioMedical Research, Ohkubo 8, Tsukuba, Ibaraki 300-2611, Japan
b Novartis Institutes for BioMedical Research, 5 Gower Place, London WC1E 6BS, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Nonpeptidic, selective, and potent cathepsin S inhibitors were derived from an in-house pyrrolopyrimi-
dine cathepsin K inhibitor by modification of the P2 and P3 moieties. The pyrrolopyrimidine-based inhib-
itors show nanomolar inhibition of cathepsin S with over 100-fold selectivity against other cysteine
proteases, including cathepsin K and L. Some of the inhibitors showed cellular activities in mouse spleno-
cytes as well as oral bioavailabilities in rats.
Received 9 May 2008
Revised 3 June 2008
Accepted 5 June 2008
Available online 10 June 2008
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Cathepsin S inhibitor
Pyrrolopyrimidine
Nitrile
Nonpeptidic
Cysteine protease
Cathepsin S (Cat S) is a lysosomal cysteine protease belonging to
the papain superfamily, which is expressed in spleen, antigen pre-
senting cells, such as dendritic cells, B cells, and macrophages.1 The
major role of Cat S is the processing of the major histocompatibility
complex (MHC) class II associated invariant chain, which is essen-
tial for the normal functioning of the immune system. Cat S is thus
an attractive therapeutic target for the treatment of autoimmune
disorders. It is also reported that Cat S is implicated in various dis-
eases such as cancer, Alzheimer’s disease, and neuropathic pain.2
Other cysteine proteases, Cat K and L, play a significant role in
numerous important physiological and pathological processes,
such as bone resorption, cancer progression, and atherosclerosis.3
Herein, we describe the discovery of novel selective Cat S inhibi-
tors, which should be safer therapeutic agents than nonselective
inhibitors by avoiding off-target side effects.4,5
The X-ray crystal structures of human Cat S, K, and L in complex
with inhibitors have been disclosed.6–8 The high sequence homol-
ogy among Cat S, K, and L reflects the similarity in 3D structures of
the proteins. Although the design of specific Cat S inhibitors repre-
sents a considerable challenge, there are some differences in the
subsite functionalities among the cathepsins. The S2 subsite in
Cat S has Gly137 and Gly165, while the corresponding regions in
the S2 subsites comprise of Ala134/Ala163 in Cat K and Ala135/
Gly164 in Cat L (Fig. 1). This means that the S2 subsite of Cat S ac-
cepts slightly bulkier groups than that of Cat K and L. The S3 sub-
site in the Cat S enzyme has Phe70 and Lys64, which exhibit
distinctive electrostatic features compared to Tyr67/Asp61 in Cat
K and Leu69/Glu63 in Cat L, respectively. The positively charged
ammonium group of Lys64 in the S3 subsite of Cat S could make
a favorable electrostatic interaction with the negatively charged
functional group on the P3 moiety of ligand, such as a carboxylate
group, although this functional group should have electrostatic
repulsion with the negatively charged side chains of Asp61 and
Glu63 in the S3 subsites of Cat K and L, respectively. Those struc-
tural features in both subsites gave us a clue to the discovery of
Cat S selective compounds against Cat K and L.
Recently, we reported a novel and potent nonpeptidic pyrrolo-
pyrimidine Cat K inhibitor 19 (Fig. 2). Our strategy for the discovery
of selective Cat S inhibitors was to switch selectivity preference to
Cat S starting with compound 1, which exhibits a modest potency
against Cat S (IC50 = 460 nM) and good synthetic feasibility for the
modification of both P2 and P3 parts. Initial efforts to improve po-
tency and selectivity were focused on finding an optimal size for
the P2 functionalities of compound 1 by replacing the neopentyl
group.
Our parallel synthetic approach to explore SAR on the P2 part is
shown in Scheme 1. The key intermediate 4 was prepared from
commercially available 4-amino-5-bromo-2-chloropyrimidine 2.
Treatment of 2 with NaCN and a catalytic amount of DABCO in
* Corresponding author. Tel.: +81 29 865 2384; fax: +81 29 865 2308.
E-mail address: osamu.irie@novartis.com (O. Irie).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.06.009

3960
O. Irie et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3959–3962
Figure 1. Compound 1 docked with human cathepsin S (Magenta), K (Green)—left, and L (Violet)—right showing 3D arrangement of subsite side chains.
Table 1
Cat S IC50 = 460 nM
Inhibition of human Cat K, L, and S by compounds 1 and 5a–l—initial optimization of
P2
Cl
Cat K IC50 = <1 nM
Cat L IC 50 = 96 nM
Cl
N
P3
N
N
N
N
P2
N
R
N
N
1
a
Compound
R
IC50 (nM)
Figure 2. Nonpeptidic cathepsin K inhibitor.
Cat S
Cat K
Cat L
96
18
>1000
370
1
ꢀCH2tBu
460
<1
<1
100
6
220
ꢀðCH2Þ tBu
2
320
110
21
Br
Br
5a
5b
5c
5d
5e
5f
5g
5h
5i
a
b
ꢀðCH2Þ2tBu
N
N
3
-CH2-Cyclohexyl
H2N
N
H2N
Cl
N
Cl
N
-(CH2)2-Cyclohexyl
-(CH2)3-Cyclohexyl
-(CH2)2-Cyclopentyl
-(CH2)2-Cycloheptyl
-(CH2)2-Cyclooctyl
-(CH2)2-Phenyl
320
2
3
150
7
>1000
17
>1000
260
Cl
33
99
64
>1000
>1000
110
840
>1000
23
c
N
N
N
R
N
N
5j
5k
-(CH2)2(p-Cl)Phenyl
-(CH2)2(m-Cl)Phenyl
19
190
>1000
>1000
86
40
N
H
N
N
4
5
5l
>1000
>1000
>1000
CH2CH2
N
Scheme 1. Reagents and conditions: (a) NaCN, DABCO, DMSO–H2O, 60 °C, 4 h, 83%;
(b) prop-2-ynyl-p-chlorobenzene, Pd(PPh3)2Cl2, CuI, Et3N, DMF, 80 °C, 7 h, 40%;
(c) R-X (X = Cl, Br, or I), K2CO3, DMF, rt, 5 h, 65–80%, or R-OH, DEAD, PPh3, THF, rt,
10 h, 45–60%.
a
Inhibition profiles were determined by a fluorometric assay with recombinant
human Cat K, L, and S, employing Z-Phe-Arg-AMC (Cat K and L) and Z-Leu-Leu-Arg-
AMC (Cat S) as synthetic substrates.9
DMSO–H2O, followed by a Castro-Stephens type reaction using cat-
alytic amounts of Pd(PPh3)2Cl2 and CuI in DMF, afforded pyrrolo-
pyrimidine 4. The Mitsunobu coupling reaction of alcohols or
treatment of alkyl halides with the key intermediate 4 provided
compounds 5a–l bearing a variety of P2 parts, in order to allow
the identification of the optimum P2 part for the S2 subsite in
Cat S.
moiety showed promising potency against Cat S and selectivity
over Cat K. Substitution of a polar functional group on the P2 part
was not tolerated due to the hydrophobic properties of the S2
susbsites in Cat S, K, and L (5d vs 5l). This SAR study for the P2 opti-
mization led to the conclusion that the cyclohexylethyl group was
suitable for the P2 moiety on pyrrolopyrimidine because of an
excellent balance between Cat S potency (5d; IC50 = 21 nM) and
selectivity against Cat K and L (over 10-fold selectivity), while
the cycloheptylethyl 5g was also promising P2 moiety. We next
set out to explore the optimization of the P3 part on the pyrrolo-
pyrimidine scaffold to improve selectivity against Cat K and L.
A key intermediate, bromide 10, was prepared following the
previously described synthetic route for the parallel synthesis ap-
proach7 to optimize the P3 moiety (Scheme 2). Suzuki-coupling
reactions between the bromide 10 and commercially available aryl
boronic acids with a catalytic amount of Pd(dppf)Cl2 in THF gave
corresponding 6-arylmethyl pyrrolopyrimidines 11a, b. Treatment
of commercially available anilines, thiol, or phenol derivatives with
bromide 10 under basic conditions provided compounds 11d–q
having a variety of P3 moieties.10
Insertion of a methylene group into the P2 part of compound 1
dramatically enhanced the inhibitory activity against Cat S because
the S2 subsite in Cat S possesses a deeper pocket than that of Cat K
and L (Table 1, 1 vs 5a, 5c vs 5d). Extension of one more methylene
group into the P2 subsite of 5a significantly decreased Cat S affinity
(5a vs 5b, 5d vs 5e). We expected the structural plasticity of the S2
subsite in Cat S, because the Phe211 near the bottom of the S2
pocket is known to move its position to adjust to either a small
t
or a bulky P2 side chain of ligands.6 Replacement of the Bu group
on 5a with a bulkier cycloalkane group improved selectivity to-
ward Cat S (5a vs 5d, 5f–h). Introduction of phenyl derivatives with
an ethyl spacer led to loss of selectivity against Cat L (5d vs 5i–k),
although compound 5j having a p-chlorophenethyl group on the P2

O. Irie et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3959–3962
3961
Table 2
Br
a
Br
Cl
b
N
Inhibition of human Cat S, K, and L by compounds 11a–q—optimization of P3
N
N
N
R
N
Cl
N
N
N
H
N
N
6
7
N
HO
HO
N
c
N
N
a
Compound
R
IC50 (nM)
N
N
N
H
N
Cat S
Cat K
Cat L
11a
11b
11c11
11d
11e
11f
4-MeO-Phenyl-
1-Naphtyl-
Cyclohexyl-
Phenyl-N(H)-
Phenyl-N(Me)-
Phenyl-O-
110
440
140
24
52
10
280
—
—
>1000
>1000
>1000
250
>1000
—
9
8
>1000
>1000
>1000
>1000
790
Br
R
N
N
e
d
N
N
N
N
N
N
11g
11h
2-Pyridyl-O-
4-Pyridyl-S-
11
6
10
11a-q
480
>1000
HO
O
Scheme 2. Reagents and conditions: (a) i—cyclohexylethylamine, Et3N, MeOH, 0 °C
to rt, 10 h, 76%, ii—NaCN, DABCO, DMSO–H2O, 60 °C, 10 h, 83%; (b) propargyl
alcohol, Pd(PPh3)2Cl2, CuI, Et3N, DMF, 80 °C, 2 h, 96%; (c) DBU, DMF, 100 °C, 2 h,
56%; (d) CBr4, Ph3P, CH2Cl2, 0 °C, 0.5 h, rt, 3 h, 78%; (e) ArB(OH)2, Pd(dppf)Cl2-
CH2Cl2, Cs2CO3, THF, 60 °C, 1 h, 46–77% for 11a, b or ArXH, (X = NH, NMe, S, O),
K2CO3, DMF, 0 °C to rt, 15–88% for 11c–q.
11i
11j
16
6
>2000
>5000
>2000
>5000
O
HO
O
O
O
H2N
Substitution of the P3 aromatic ring resulted in a loss of inhib-
itory activity to Cat S (Table 1, 5d vs Table 2; 11a, b). Saturation of
the P3 phenyl ring decreased the Cat S potency (5d vs 11c), due in
11k
7
390
>1000
O
part to the loss of p–p interaction between the phenyl ring on the
O
pyrrolopyrimidine and Phe70 in the S3 subsite of Cat S. Insertion of
a polar hetero atom, N, O, or S, between the P3 part and the pyrrol-
opyrimidine scaffold for changing the distance of the P3 aromatic
ring from the pyrrolopyrimidine scaffold enhanced the Cat S affin-
ity along with improved selectivity over Cat K and L (5d vs 11d, f,
h), presumably due to the structural feature unique to the Cat S en-
zyme owing to the Phe70 and Lys64 in the S3 subsite. The elec-
tronic requirement in the P3 moieties was investigated next.
Replacement of the phenyl ring with a pyridyl ring caused a loss
of selectivity against Cat K and L (11f vs 11g). In order to give addi-
tional attractive interaction between the positively charged Lys64
in the S3 subsite of Cat S and negatively charged functional group,
a carboxylic acid was introduced to the P3 phenyl group, which
showed excellent potency (11j; IC50 = 6 nM) and high selectivity
(>450-fold) against Cat K and L.
11l
2
8
280
>1000
>1000
O
H2N
O
11m
>1000
N
O
H
O
11n
11o
11p
12
9
>1500
>1500
>1000
>1000
>4000
>4000
>1000
>1000
N
O
H
N
N
O
O
O
O
N
N
O
O
15
13
While compound 11j had both excellent potency and selectivity
for Cat S, it was inactive at 10 l
M in a secondary cellular assay12
11q
measuring MHC II associated invariant chain degradation in mouse
a
Inhibition profiles were determined by a fluorometric assay with recombinant
human Cat K, L, and S, employing Z-Phe-Arg-AMC (Cat K and L) and Z-Leu-Leu-Arg-
AMC (Cat S) as synthetic substrates.9
splenocytes. Another compound 11i (IC50 = 16 nM) with a carbox-
ylic acid also showed no cellular activity at 10
pound 5a and LHVS12 as a positive control, demonstrated high
sensitivity of cellular activity at 1 M. From these results, it was
lM, although com-
l
anticipated that a carboxyl group would decrease permeability
into cells. We thus turned to improving the cellular potency.
Replacement of the carboxyl group with amide was well toler-
ated by Cat S (11i vs 11l). Extension of the P3 part on 11l into the
deep S3 subsite of Cat S resulted in high affinity (IC50; 9–15 nM) to
Cat S with excellent (ꢁ100-fold) selectivity against Cat K and L
(11m–q). Interestingly, compounds 11o and 11q showed cellular
Table 3
Pharmacokinetic parameters of 11l and 11o–q in male Sprague–Dawley rats (11l; iv
1.6 mg/kg; po 4.0 mg/kg, 11o–q; iv 1 mg/kg; po 3 mg/kg), where values are means of
n = 3
poAUCa (nM h)
a
Compound
Cl (L/h/kg)
ivt1/2 (h)
F (%)
poCmax (nM)
11l
5.0
4.0
5.3
2.5
0.7
1.2
1.3
1.4
21
47
25
38
25
34
17
43
108
242
103
346
11o
11p
11q
activities at 1 lM in mouse splenocytes.
Selected compounds, 11l, o, p, were evaluated for pharmacoki-
netic properties in cassette dosing experiments13 with male Spra-
gue–Dawley rats (Table 3). All compounds had acceptable oral
bioavailabilities of greater than 21% and moderate to high plasma
clearances of 2.5–5.3 L/h/kg and moderate terminal elimination
half-lives of 0.7–1.4 h after intravenous administration.
In summary, potent and selective nonpeptidic Cat S inhibitors
were discovered. The novel pyrrolopyrimidine scaffold 1 derived
from the Cat K inhibitors was modified by stepwise optimization
a
Dose-normalized to 1 mg/kg.
utilizing key structural and electrostatic features of the S2 and S3
subsites of the Cat S enzyme. Some of the selective Cat S inhibitors
identified had potent cellular efficacies in mouse splenocytes and
modest PK profiles in rats. The biological activities in rodents by

3962
O. Irie et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3959–3962
Tuntland, T.; Chang, J.; Gordon, P.; Hollenbeck, T.; Tumanut, C.; Li, J.; Harris, J. L.
their Cat
S specific inhibition will be described in future
Bioorg. Med. Chem. Lett. 2007, 17, 2899; Bekkali, Y.; Thomson, D. S.; Betageri, R.;
Emmanuel, M. J.; Hao, M.-H.; Hickey, E.; Liu, W.; Patel, U.; Ward, Y. D.; Young, E.
R. R.; Nelson, R.; Kukulka, A.; Brown, M. L.; Crane, K.; White, D.; Freeman, D. M.;
Labadia, M. E.; Wildeson, J.; Spero, D. M. Bioorg. Med. Chem. Lett. 2007, 17, 2465;
For a recent nonpeptidic cathespin S inhibitors, see: (b) Wei, J.; Pio, B. A.; Cai,
H.; M eduna, S. P.; Sun, S.; Gu, Y.; Jiang, W.; Thurmond, R. L.; Karlsson, L.;
Edwards, J. P. Bioorg. Med. Chem. Lett. 2007, 17, 5525; Inagaki, H.; Tsuruoka, H.;
Hornsby, M.; Lesley, S. A.; Spraggon, G.; Ellman, J. A. J. Med. Chem. 2007, 50,
2693.
publications.
Acknowledgments
We thank Michie Kobayashi, Tomoko Ohkubo, Andrew
McBryde, and Caroline Huntley for excellent technical assistance.
The authors are grateful to Pamposh Ganju, Shinichi Koizumi,
and Terance W. Hart for valuable discussions. We also acknowl-
edge professor Dr. Masakatsu Shibasaki (The University of Tokyo)
for fruitful discussions regarding palladium chemistry.
6. Pauly, T. A.; Sulea, T.; Ammirati, M.; Sivaraman, J.; Danley, D. E.; Griffor, M. C.;
Kamath, A. V.; Wang, I.-K.; Laird, E. R.; Seddon, A. P.; Menard, R.; Cygler, M.;
Rath, V. L. Biochemistry 2003, 42, 3203.
7. Teno, N.; Miyake, T.; Ehara, T.; Irie, O.; Sakaki, J.; Ohmori, O.; Gunji, H.;
Matsuura, N.; Masuya, K.; Hitomi, Y.; Nonomura, K.; Horiuchi, M.; Gohda, K.;
Iwasaki, A.; Umemura, I.; Tada, S.; Kometani, M.; Iwasaki, G.; Cowan-Jacob, S.
W.; Missbach, M.; Lattmann, R.; Betschart, C. Bioorg. Med. Chem. Lett. 2007, 17,
6096.
References and notes
8. Chowdhury, S. F.; Sivaraman, J.; Wang, J.; Devanathan, G.; Lachance, P.; Qi, H.;
Ménard, R.; Lefebvre, J.; Konishi, Y.; Cygler, M.; Sulea, T.; Purisima, E. O. J. Med.
Chem. 2002, 45, 5321.
1. For a recent review of cathespin S, see: Gupta, S.; Singh, R. K.; Dastidar, S.; Ray,
A. Expert Opin. Ther. Targets 2008, 12, 291.
2. (a) Clark, A. K.; Yip, P. K.; Grist, J.; Gentry, C.; Staniland, A. A.; Marchand, F.;
Dehvari, M.; Wotherspoon, G.; Winter, J.; Ullah, J.; Bevan, S.; Malcangio, M.
Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 10655; (b) Barclay, J.; Clark, A. K.; Ganju,
P.; Gentry, C.; Patel, S.; Wotherspoon, G.; Buxton, F.; Song, C.; Ullah, J.; Winter,
J.; Fox, A.; Bevan, S.; Malcangio, M. Pain 2007, 130, 225.
3. For recent reviews of cysteine proteases, see: (a) Vasiljeva, O.; Reinheckel, T.;
Peters, C.; Turk, D.; Turk, V.; Turk, B. Curr. Pharm. Des. 2007, 13, 387;
(b) Palermo, C.; Joyce, J. A. Trends Pharmacol. Sci. 2007, 29, 22; (c) Maryanoff,
B. E.; Costanzo, M. J. Bioorg. Med. Chem. 2008, 16, 1562; (d) Lutgens, S. P. M.;
Cleutjens, K. B. J. M.; Daemen, M. J. A. P.; Heeneman, S. FASEB J. 2007, 21, 3029.
4. For a recent review of cathespin S inhibitors, see: Link, J. O.; Zipfel, S. Curr. Opin.
Drug Discov. Dev. 2006, 9, 471.
5. For a recent peptidic cathespin S inhibitors, see: (a) Gauthier, J. Y.; Black, W. C.;
Courchesne, I.; Cromlish, W.; Desmarais, S.; Houle, R.; Lamontagne, S.; Li, C. S.;
Massé, F.; McKay, D. J.; Ouellet, M.; Robichaud, J.; Truchon, J.-F.; Truong, V.-L.;
Wang, Q.; Percival, M. D. Bioorg. Med. Chem. Lett. 2007, 17, 4929; Chatterjee, A.
K.; Liu, H.; Tully, D. C.; Guo, J.; Epple, R.; Russo, R.; Williams, J.; Roberts, M.;
9. Altmann, E.; Aichholz, R.; Betschart, C.; Buhl, T.; Green, J.; Irie, O.; Teno, N.;
Lattmann, R.; Tintelnot-Blomley, M.; Missbach, M. J. Med. Chem. 2007, 50,
591.
10. New compounds have been characterized by 1H NMR and MS. The
experimental information is described in a patent application, see: Buxton, F.
P.; Ehara, T.; Ganju, P.; Hallett, A.; Irie, O.; Iwasaki, A.; Kanazawa, T.; Masuya,
K.; Nonomura, K.; Sakaki, J.; Snell, C. R.; Song, C.; Tanabe, K.; Teno, N;
Umemura, I.; Yokokawa, F. WO 2004069656, 2004.
11. Preparation of 11c; A Sonogashira reaction of 7 with prop-2-ynylcyclohexane
as an alkyne under the same condition as (b) followed by
a cyclization
condition (c) in Scheme 2, yielded 11c.
12. Riese, R. J.; Wolf, P. R.; Brömme, D.; Natkin, L. R.; Villadangos, J. A.; Ploegh, H. L.;
Chapman, H. A. Immunity 1996, 4, 357.
13. (a) Manitpisitkul, P.; White, R. E. Drug Discovery Today 2004, 9, 652; (b) Janser,
P.; Neumann, U.; Miltz, W.; Feifel, R.; Buhl, T. Bioorg. Med. Chem. Lett. 2006, 16,
2632.