Synthesis and evaluation of arylaminoethyl amides as noncovalent inhibitors of cathepsin S. Part 3: Heterocyclic P3

Article abstract of DOI:10.1016/j.bmcl.2005.12.095

A series of N_α-2-benzoxazolyl-α-amino acid-(arylaminoethyl)amides were identified as potent, selective, and noncovalent inhibitors of cathepsin S. Structure-activity relationships including strategies for modulating the selectivities among cathepsins S, K, and L, and in vivo pharmacokinetics are discussed. A X-ray structure of compound 3 bound to the active site of cathepsin S is also reported.

Full text of DOI:10.1016/j.bmcl.2005.12.095

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1975–1980
Synthesis and evaluation of arylaminoethyl amides as
noncovalent inhibitors of cathepsin S. Part 3: Heterocyclic P3
David C. Tully,* Hong Liu, Phil B. Alper, Arnab K. Chatterjee, Robert Epple,
Michael J. Roberts, Jennifer A. Williams, KhanhLinh T. Nguyen,
David H. Woodmansee, Christine Tumanut, Jun Li, Glen Spraggon,
Jonathan Chang, Tove Tuntland, Jennifer L. Harris and Donald S. Karanewsky
Genomics Institute of the Novartis Research Foundation, 10675 John J. Hopkins Dr., San Diego, CA 92121, USA
Received 15 November 2005; revised 19 December 2005; accepted 20 December 2005
Available online 30 January 2006
Abstract—A series of Na-2-benzoxazolyl-a-amino acid-(arylaminoethyl)amides were identified as potent, selective, and noncovalent
inhibitors of cathepsin S. Structure–activity relationships including strategies for modulating the selectivities among cathepsins S, K,
and L, and in vivo pharmacokinetics are discussed. A X-ray structure of compound 3 bound to the active site of cathepsin S is also
reported.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Cathepsin S is a lysosomal cysteine protease and a mem-
ber of the papain family, and it is expressed primarily in
antigen presenting cell. It has been shown to play a key
role in antigen presentation through the targeted degra-
dation of the invariant chain (li) that is associated with
the major histocompatibility class II complex (MHC
II). The primary role of the invariant chain is to block
the MHC II binding groove, and proteolytic removal
is required prior to productive antigen loading on the
MHC II complex.1 Cathepsin S knockout mice display
significant impairment of invariant chain degradation
in antigen presenting cells and consequently exhibit a
clear resistance to the development of collagen-induced
arthritis and autoimmune myasthenia gravis in compar-
ison to the wild type mice.2 These data point to cathep-
sin S as an attractive therapeutic target for the
modulation and regulation of immune hyperresponsive-
ness, such as autoimmune diseases, asthma, multiple
sclerosis, and rheumatoid arthritis.3
tively nonselective inhibitor of this class of proteases, 1
was also pulled out of a high-throughput screen of our
compound collection for inhibitors of cathepsin S. As
we have reported previously,5 early hit-to-lead optimiza-
tion on this scaffold produced compound 2 as a potent
O
H
H
N
N
CatS Ki = 0.062 µM
CatK Ki = 0.043 µM
CatL Ki = 0.007 µM
CatB Ki > 30 µM
Me
N
H
O
OMe
1
O
H
H
N
N
CatS Ki = 0.003 µM
CatK Ki > 65 µM
CatL Ki > 17 µM
CatB Ki > 100 µM
CatF Ki > 20 µM
CatV Ki > 20 µM
O
N
H
O
F3C
OMe
2
P1
R1
P3
O
O
N
H
H
Compound 1 (Fig. 1) was originally reported by Alt-
mann et al. as a noncovalent inhibitor of the highly
homologous cysteine protease cathepsin K.4 As a rela-
N
N
N
CatS Ki = 0.029 µM
CatK Ki > 100 µM
CatL Ki > 10 µM
H
OMe
CatB Ki > 100 µM
Keywords: Cathepsin S; Cathepsin; Cysteine protease inhibitor; Non-
covalent inhibitor; Peptidomimetics.
P1'
3 R1 = H
P2
*
Corresponding author. Tel.: +1 8588121559; fax: +1 8588121648;
e-mail: dtully@gnf.org
Figure 1.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.12.095

1976
D. C. Tully et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1975–1980
(Ki = 3 nM) inhibitor of cathepsin S with greater than
5000-fold selectivity over cathepsins K, L, B, F, and V.
Lineweaver–Burk analysis confirmed that compounds
from this series are purely competitive inhibitors.5,6 In
addition, the nature of inhibition of arylaminoethyl
amides was shown to be fully reversible as enzymatic
activity is restored following dilution and dialysis of
the enzyme–inhibitor complex.4–7
10 mg/kg oral dose. The poor oral bioavailability is pre-
sumably due to rapid first pass metabolism since 3
exhibits poor in vitro metabolic stability, with <5% of
parent compound remaining after incubation for
30 min in rat liver microsomes. Lead optimization of
this series was then aimed primarily at improving the
oral bioavailability and in vivo PK by exploring the
SAR of the P3 heterocycle, the P1 side chain, and the
aniline moiety at the prime side.
Compound 2, however, suffered a short half-life and ra-
pid clearance when administered intravenously in rats.
Lead optimization efforts were initiated with the intent
to decrease the peptidic nature of compound 2 and im-
prove the pharmacokinetic (PK) and physicochemical
properties of this series. Structure analysis suggested
that the benzamide carbonyl in compounds 1 and 2 does
not play a direct role in binding to cathepsin S. Thus, in
order to reduce the peptidic character of 2 we envisioned
that the benzamide P3 could be replaced with an NH-ar-
yl/heteroaryl moiety in which the analogous hydrogen
bond is formed with Gly69 as in 2, and the NH-linked
aryl/heteroaryl moiety would fill the shallow S3 pocket
formed by Gly62 and Phe70 (Fig. 2).5,8
The synthesis of analogs lacking P1 substitution on the
diamine side chain is depicted in Scheme 1. Substituted
2-chlorobenzoxazoles were prepared from the appropri-
ate ortho-aminophenol.9 N-Arylation of cyclohexylala-
nine methyl ester with the appropriate 2-
chlorobenzoxazole 5 followed by saponification led to
the N-aryl amino acid 6. Condensation of 6 with either
N-(4-methoxyphenyl)- or N-(4-fluorophenyl)-ethylene
diamine4–6 afforded compounds 3 and 7a–g, respective-
ly, in 25–50% overall yield.
Substitution on the benzoxazole ring is fairly well toler-
ated as shown in Table 1. Replacement of the undesir-
able para-methoxyaniline group10 with 4-fluoroaniline
decreased the potency by about threefold (3–7a). The
addition of a halogen to the 5-position of the benzoxaz-
ole resulted in a twofold loss in activity (7b and e),
whereas substitutions at the 6-position of the benzoxaz-
ole were roughly equipotent (7d) or were slightly im-
proved in their potency toward cathepsin S (7c and g),
while still remaining selective over cathepsins K and L.
The most significant improvement in potency was found
with the 7-chlorobenzoxazole 7f, which gave a nearly
fivefold boost in affinity to cathepsin S over 7a.
Although highly selective over cathepsin K, the selectiv-
ity of 7f over cathepsin L was reduced to about 30-fold.
As shown in Table 1, our initial attempts at replacing
the biaryl amide P3 subunit resulted in compound 3,
in which the biaryl amide was successfully replaced with
a benzoxazole. The affinity toward cathepsin S was
diminished about 10-fold versus compound 2, neverthe-
less this modification eliminated an amide and decreased
the molecular weight while simultaneously preserving
the selectivity over cathepsins K and L. Furthermore,
this modification of the P3 subunit had a dramatic effect
on the in vivo pharmacokinetic parameters of this series.
Upon intravenous administration (3 mg/kg) the half-life
of 3 was nearly 5 h with a low clearance and moderate
volume of distribution (Table 4). However, compound
3 had effectively no plasma exposure following a
The 6- and 7-chlorobenzoxazoles were chosen as the
preferred P3 subunits for further exploration into the
nature of the P1 side-chain and aniline moieties. The
synthesis of diamines with a P1 side chain is shown in
Scheme 2. The appropriate Cbz-protected amino acid
8 was first reduced to the alcohol and then oxidized to
the aldehyde to give 9. Reductive amination with the
desired aniline or indoline afforded the N-aryl diamine
10. Hydrogenolysis of the Cbz-group, followed by
amide coupling to the N-aryl-cyclohexylalanine 6, affor-
ded both the 7-chlorobenzoxazole series 12a–g and the
6-chlorobenzoxazole series 13a–i in 20–40% overall yield
from amino acid 8.
We have shown previously that incorporating an alkyl
group onto the P1 side chain boosts the potency of com-
pounds from this series over the unsubstituted ethylene-
diamine analogs.11 Therefore, the diversification of the
N-aryl moiety, which can have dramatic influence over
the cathepsin S affinity selectivity over the related en-
zymes, was initiated with a (S)-methyl group in the P1.
The 4-methoxyaniline 12a (Table 2) is the most potent
analog in the series (Ki = 0.001 lM), but it also showed
a proportional increase in the affinity toward cathepsins
K and L, leaving only a narrow window of selectivity
over cathepsin L. The potency of the 7-chlorobenzoxaz-
˚
Figure 2. Crystal structure of cathepsin S with compound 3 at 1.8 A
resolution (RCSB PDB ID: 2F1G). Protein atoms are colored with
yellow carbons, red oxygens, and blue nitrogens. Compound atoms are
colored with cyan carbons, red oxygens, and blue nitrogens, while those
atoms ill defined with no visible electron density (P0 aniline moiety) are
labeled with the corresponding lighter shades. Potential hydrogen bonds
are represented by dotted black lines. Figure generated using PyMOL.

D. C. Tully et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1975–1980
1977
Table 1. Inhibition of cathepsins S, K, and L: influence of substitution on P3 benzoxazolea
O
H
H
R3
N
N
N
H
Y
Compound
R3
Y
CatS Ki (lM)
CatK Ki (lM)
CatL Ki (lM)
O
N
O
N
3
OMe
F
0.029
>30.0
>10.0
7a
7b
7c
7d
7e
7f
0.094
0.197
0.043
0.093
0.173
0.019
0.043
>30.0
>30.0
>30.0
>20.0
>20.0
4.56
>10.0
O
N
F
8.37
F
F
O
N
F
9.04
7.46
Cl
O
F
N
O
F
11.8
N
Cl
Cl
O
N
F
0.541
1.73
MeO2C
O
N
7g
F
>10.0
a Details of the assay conditions can be found in Supplementary material.
OH
solubility but led to an order of magnitude decrease in
cathepsin S activity and contributed very little to any
improvements in selectivity over cathepsins K and L.
Replacement of the aniline with 5-fluoroindoline (12d)
showed a significant improvement in the selectivity over
both cathepsins K and L despite the somewhat dimin-
ished cathepsin S activity over the methoxyaniline.
More importantly, the introduction of the indoline moi-
ety gave a dramatic improvement in the pharmacokinet-
ics of this series, as compound 12d had a 24% oral
O
N
b
a
SH
X
X
X
NH2
4
O
O
H
N
X
O
N
OH
c, d
N
Cl
6
5
O
O
N
H
H
N
N
X
N
e
H
R1
R1
c
R1
a,b
Y
Cbz
H
Cbz
NAr
NAr
Cbz
OH
N
N
3
Y = OMe
N
H
H
H
7a-g Y = F
O
O
10
9
8
Scheme 1. Reagents and conditions: (a) KSCSOEt, EtOH, reflux; (b)
PCl5, POCl3, 100 ꢁC; (c) 3-cyclohexyl-L-alanine methyl ester hydro-
chloride, DIEA, DMF, 0 ꢁC to rt, 70–85%; (d) LiOH, H2O, dioxane,
95–99%; (e) N1-(4-fluorophenyl)-ethylenediamine, DIEA, HATU,
CH2Cl2, 75–85%.
R1
R1
O
O
N
d
H
e
N
X
NAr
N
H2N
H
11
12a-g
13a-h
ole P3 subunit remained attractive, however, and we
therefore sought to improve the selectivity profile by
optimization of the aniline moiety at the prime side.
Scheme 2. Reagents and conditions: (a) i—iso-BuOCOCl, Et3N, THF;
ii—NaBH4, H2O, 60–85%; (b) Dess–Martin periodinane, CH2Cl2; (c)
aniline or indoline, NaB(CN)H3, AcOH, MeOH, 60–75% combined
yield steps (b) and (c); (d) H2, Pd/C, MeOH, 90–95%; (e) 6, HATU,
DIEA, CH2Cl2, 75–85%.
Electron-withdrawing groups on the aniline ring are
also tolerated. Methylsulfones 12b and c improved the

1978
D. C. Tully et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1975–1980
Table 2. Inhibition of cathepsins S, K, and L: variation of aniline moiety
Cl
O
Me
O
H
N
N
X
N
H
Compound
X
CatS Ki (lM)
CatK Ki (lM)
CatL Ki (lM)
H
N
12a
0.001
0.286
0.024
OMe
H
N
12b
12c
12d
12e
0.011
0.026
0.015
0.015
1.05
2.13
0.108
0.199
0.849
4.67
SO2Me
SO2Me
H
N
N
N
14.7
F
F
>100
12f
0.003
12.3
3.71
0.123
N
N
F
F
O
12g
>100
>30
bioavailability in rats. However, 12d also had a rather
high clearance (33.1 mL/min/kg) (Table 4).
drops off considerably. However, the addition of a
methyl group at P1 (13b) leads to a tenfold improvement
in the Ki over 13a. Yet another fivefold increase in activ-
ity is gained by further extending the side chain into P1
with a benzyloxy group as in 13c, while simultaneously
preserving a 300-fold selectivity over cathepsin L. Addi-
tion of a chlorine atom to the 6-position of the benzox-
azole ring (13d) resulted in a tenfold reduction in activity
With the 5-fluoroindoline in place, we sought to explore
the effects of substitution on the P1 side chain, and
determine what effects, if any, such substitution had
on selectivity and PK of this series. In the case of unsub-
stituted benzoxazoles in P3 (13a, Table 3), the potency
Table 3. Inhibition of cathepsins S, K, and L—exploration of P1 side chain
X
O
R
N
F
O
N
H
N
N
H
Compound
X
R1
CatS Ki (lM)
CatK Ki (lM)
CatL Ki (lM)
13a
13b
13c
13d
13e
13f
13g
13h
13i
H
–H
–CH3
0.222
0.026
0.005
0.151
0.021
0.012
0.027
0.071
0.730
>30
>10
>10
2.83
1.49
9.67
0.369
>10
H
H
–CH2OCH2Ph
–CH3
16.7
>30
Cl
Cl
H
–CH2CH2SO2Me
–CH2CO2H
–CH2CO2H
–CH2CO2Et
–CH2CO2-t-Bu
4.87
>20
Cl
Cl
Cl
>100
>100
>100
>70
>50
>100

D. C. Tully et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1975–1980
1979
versus the 7-chlorobenzoxazole 12d for both cathepsins
S and L. Incorporation of a more polar P1 moiety, such
as the sulfone in compound 13e which bears a side chain
derived from methionine, returns the cathepsin S activi-
ty, but not without the emergence of moderate cathepsin
L activity. On the other hand, increasing the polarity
further, as in the aspartate derived side chains 13f–h,
afforded potent cathepsin S inhibitors while completely
dialing out cathepsins K and L activity.11,12 The posi-
tively charged guanidinium group in the S1 pocket ema-
nating from Arg141 represented a distinct feature of
cathepsin S that may have implications for the specific-
ity of 13f–h (Fig. 2). The analogous S1 region is occu-
pied by Asn158 in cathepsin K and Asp162 in
cathepsin L.13
selectivity of 12f remained a robust 1200-fold over
cathepsin K, but narrowed somewhat over cathepsin L
(41·). More importantly, the gem-dimethyl substitution
at the indoline 3-position considerably improved the
PK, as 12f has a long in vivo half-life in rats and low clear-
ance. Compound 12f is also orally bioavailable
(F = 15%), achieving a Cmax of over 1.2 lM with a half-
life of over 5 h following oral administration. Interesting-
ly, the 2-indolone derivative 12g is virtually inactive to-
ward all of the cathepsins assayed, suggesting that the
indoline nitrogen lone pair plays an essential role in the
binding to the enzyme.15
The X-ray co-crystal structure of compound 3 bound to
the active site of cathepsin S is shown in Figure 2. Re-
combinant human cathepsin S was crystallized in the
presence of 2-mercaptopyridine, which was added prior
to crystallization to prevent oxidation of the catalytic
cysteine (Cys25) and to aid in the crystallization itself.
The cathepsin S crystals were soaked with compound
3 for 24 h in the presence of 4 mM DTT (also present
4 h prior to addition of 3) to remove the 2-mercapto-
pyridine covalently bound to Cys25. The presence of
the DTT reduced the occupancy of the 2-mercaptopyri-
dine in the S1-prime site but did not eliminate it com-
pletely. As a result, the methoxyaniline group was
completely disordered such that virtually no electron
density was visible, and therefore was modeled into
the S1-prime site. The lack of electron density for the
methoxyaniline could have been due either to the resid-
ual 2-mercaptopyridine in the S1-prime site or to the
intrinsic disorder within this sight. The two carbons of
the ethylenediamine chain are completely visible howev-
The carboxylic acid derivative 13g is not orally bioavail-
able, and it has a relatively short half-life and moderate
clearance when administered intravenously. When its eth-
yl ester analog 13h is administered orally at 10 mg/kg, it is
not detected in the plasma, however its carboxylic acid
metabolite 13g is detected with a peak plasma concen-
tration of over 1 lM, providing a 22% bioavailability
of 13g as its ethyl ester prodrug (Table 4). Interesting-
ly, increasing the steric bulk of the side chain to the
tert-butyl ester, as in compound 13i, led to a significant
drop in activity. This is in contrast to what was ob-
served with 13c, where the relatively large benzyloxym-
ethyl side chain drove the cathepsin S Ki down to
5 nM, suggesting a preference for aromatics over bulky
aliphatics in the S1 pocket.14
Metabolic ID studies on indolines 12d and 13a–h showed
that one of the primary modes of metabolism is a hydrox-
ylation/dehydration on the aminoethylindoline moiety,
suggesting a possible explanation for the relatively high
in vivo clearance and short half-lives (Table 4). Attempts
to block this oxidative metabolism were made by intro-
ducing substituents onto the 2- and 3-positions on the
indoline ring. Geminal dimethyl substitution at the 2-po-
sition of the indoline ring (12e, Table 2) led to a major
boost in selectivity over cathepsin L to greater than
300-fold, while completely preserving cathepsin S activity
(Ki = 0.015 lM). Although the half-life (t1/2 = 4 h) im-
proved significantly over 12d, the AUC and Cmax were
not improved and as such did not warrant further iv pro-
filing (Table 4). Moving this gem-dimethyl group over to
the 3-position on the indoline ring (12f) resulted in a 5-
fold gain in cathepsin S affinity (Ki = 0.003 lM). The
˚
er, and are situated 4.8 and 3.5 A from the sulfur atom
of Cys25. As expected, the cyclohexyl P2 group fills the
enzyme S2 pocket, and the benzoxazole sits squarely in
the S3 pocket, which is formed by Lys64, Gly62, Gly69,
and Phe70 through induced fit. The hydrogen bonds
formed between benzoxazole 3 and the enzyme are rep-
resented by the dotted lines in Figure 2. The N–H
groups of both the cyclohexylalanine and the ethylenedi-
amine amide each donates a putative hydrogen bond to
the protein backbone carbonyls of Gly69 and Asn163,
respectively, while the cyclohexylalanine carbonyl ac-
cepts a hydrogen bond from N–H of Gly69.
In summary, our early lead optimization from com-
pound 2 led to a novel series of potent cathepsin S inhib-
itors lacking an electrophilic, covalent warhead.16 By
Table 4. Pharmacokinetics of selected analogsa
Compound
Single iv dose (3 mg/kg)
Single po dose (10 mg/kg)
CL (mL/min/kg)
Vss (L/kg)
t1/2 (h)
AUC (min lg/mL)
Cmax (nM)
t1/2 (h)
F (%)
3
11.4
33.1
23.2
—
1.63
2.57
0.74
—
4.8
2.3
1.0
—
0.9
65
12
703
5.3
1.1
1.0
4.1
5.1
<1
24
12d
13g/h
12e
12f
95
54
1100
533
22b
—
10.1
0.90
8.3
144
1261
15
a Pharmacokinetic parameters of single iv dose (3 mg/kg) and po dose (10 mg/kg) in male Wistar rats. All in vivo pharmacokinetic values are means
of three experiments.
b PK parameters shown are for IV administration of carboxylic acid 13g and oral administration of ethyl ester prodrug 13h, and the bioavailability of
13g from prodrug 13h is calculated using iv AUC of 13g.

1980
D. C. Tully et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1975–1980
L. W.; Tommasi, R. A.; Farley, D. L.; Quadros, E.;
Coppa, D. E.; Du, Z.; Fang, Z.; Zhou, H.; Doughty, J.;
Toscano, K. T.; Wigg, A. M.; Zhou, S. Bioorg. Med.
Chem. Lett. 2003, 13, 4121.
replacing the P3 amide of compound 2 with a heterocy-
cle, the in vivo PK of this series was dramatically im-
proved. Furthermore, optimization of the P1 aniline
moiety to a more drug-like indoline group10 has led to
potent, orally bioavailable cathepsin S inhibitors, such
as 12f, with much improved pharmacokinetics in rats
over the early lead compounds 2 and 3.
9. Haviv, F.; Ratajczyk, J. D.; DeNet, R. W.; Kerdesky, F.
A.; Walters, R. L.; Schmidt, S. P.; Holms, J. H.; Young, P.
R.; Carter, G. W. J. Med. Chem. 1988, 31, 1719.
10. The potential safety liabilities of the methoxyaniline
moiety as in compounds 1–3 have been well documented:
Park, B. K.; Kitteringham, N. R.; Maggs, J. L.; Pirmoha-
med, M.; Williams, D. P. Annu. Rev. Pharmacol. Toxicol.
2005, 45, 177.
Acknowledgments
11. Alper, P. B.; Liu, H.; Chatterjee, A. K.; Nguyen, K. T.;
Tully, D. C.; Tumanut, C.; Li, J.; Harris, J. L.; Tuntland,
T.; Chang, J.; Gordon, P.; Hollenbeck, T.; Karanewsky,
D. S. Bioorg. Med. Chem. Lett., 2006, in press.
12. Carboxylic acids 13f and g were obtained by deprotection
of the corresponding tert-butyl esters with TFA (50%) in
CH2Cl2 as previously described in Ref. 11. Ethyl ester 13h
was then prepared by heating 13g in ethanol:triethylor-
thoformate (2:1) with catalytic p-toluenesulfonic acid.
13. 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.
14. A similar SAR trend has been observed with a series of
dipeptide nitrile inhibitors of cathepsin S: Ward, Y. D.;
Thomson, D. S.; Frye, L. L.; Cywin, C. L.; Morwick, T.;
Emmanuel, M. J.; Zindell, R.; McNeil, D.; Bekkali, Y.;
Giradot, M.; Hrapchak, M.; DeTuri, M.; Crane, K.;
White, D.; Pav, S.; Wang, Y.; Hao, M.-H.; Grygon, C. A.;
Labadia, M. E.; Freeman, D. M.; Davidson, W.; Hopkins,
J. L.; Brown, M. L.; Spero, D. M. J. Med. Chem. 2002, 45,
5471.
The authors acknowledge Mike Hornsby for protein
production, and Perry Gordon and Tom Hollenbeck
for bioanalytical support. The authors also thank Terry
Hart, Allan Hallett, Kirk Clark, and Raviraj Kulathila
(Novartis Institutes for Biomedical Research) for valu-
able discussions.
Supplementary material
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2005.12.095.
References and notes
1. (a) Thurmond, R. L.; Sun, S.; Karlsson, L.; Edwards, J. P.
Curr. Opin. Investig. Drugs 2005, 6, 473; (b) Leroy, V.;
Thurairatnam, S. Expert Opin. Ther. Patents 2004, 14, 301.
2. (b) Nakagawa, T. Y.; Brissette, W. H.; Lira, P. D.;
Griffiths, R. J.; Petrushova, N.; Stock, J.; McNeish, J. D.;
Eastman, S. E.; Howard, E. D.; Clarke, S. R. M.;
Rosloniec, E. F.; Elliott, E. A.; Rudensky, A. Y. Immunity
1999, 10, 207; (b) Yang, H.; Kala, M.; Scott, B. G.;
Goluszko, E.; Chapman, H. A.; Christadoss, P. J.
Immunol. 2005, 174, 1729.
3. Liu, W.; Spero, D. M. Drug News Prespect. 2004, 17, 357.
4. Altmann, E.; Green, J.; Tintelnot-Blomley, M. Bioorg.
Med. Chem. Lett. 2003, 13, 1997.
5. Liu, H.; Tully, D. C.; Epple, R.; Bursulaya, B.; Li, J.;
Harris, J. L.; Williams, J. A.; Russo, R.; Tumanut, C.;
Roberts, M. J.; Alper, P. B.; He, Y.; Karanewsky, D. S.
Bioorg. Med. Chem. Lett. 2005, 15, 4979.
6. Altmann, E.; Renaud, J.; Green, J.; Farley, D.; Cutting,
B.; Jahnke, W. J. Med. Chem. 2002, 45, 2352.
7. Experimental details of the enzymatic assays and data
from the dilution–dialysis experiments with compounds 2
and 12f are included in Supplementary material.
8. Similar approaches have been reported by others for
dipeptide nitrile inhibitors of cathepsins K and B: (a)
Robichaud, J.; Bayly, C.; Oballa, R.; Prasit, P.; Mellon,
C.; Falgueyret, J. P.; Percival, M. D.; Wesolowski, G.;
Rodan, S. B. Bioorg. Med. Chem. Lett. 2004, 14, 4291; (b)
Greenspan, P. D.; Clark, K. L.; Cowen, S. D.; McQuire,
15. Indolone derivative 12g was synthesized according to the
following scheme:
O
O
a,b
Boc
N
e
Boc
OH
N
N
F
H
H
O
c,d
N
12g
H2N
F
Reagents and conditions: (a) CBr4 (1.0 equiv), PPh3
(1.0 equiv), DCM (52%); (b) 5-fluoroisatin, NaH
(1.2 equiv), DMF, (36%); (c) H2NNH2 (50%) in EtOH,
reflux (89%); (d) TFA:DCM:H2O (50:45:5) (99%); (e) 6,
HATU, DIEA, DCM (77%).
16. During the course of this work, a series of nonpeptidic and
noncovalent inhibitors of cathepsin S have been reported
(a) Thurmond, R. L.; Sun, S.; Sehon, C. A.; Baker, S. M.;
Cai, H.; Gu, Y.; Jiang, W.; Riley, J. P.; Williams, K. N.;
Edwards, J. P.; Karlsson, L. J. Pharmacol. Exp. Ther.
2004, 308, 268; (b) Thurmond, R. L.; Beavers, M. P.; Cai,
H.; Meduna, S. P.; Gustin, D. L.; Sun, S.; Almond, H. J.;
Karlsson, L.; Edwards, J. P. J. Med. Chem. 2004, 47, 4799.