Characterization and optimization of selective, nonpeptidic inhibitors of cathepsin S with an unprecedented binding mode

Article abstract of DOI:10.1021/jm070111+

The substrate activity screening (SAS) method, a substrate-based fragment identification and optimization method for the development of enzyme inhibitors, was previously applied to cathepsin S to obtain a novel (2-arylphenoxy) acetaldehyde inhibitor, 2, w

Full text of DOI:10.1021/jm070111+

J. Med. Chem. 2007, 50, 2693-2699
2693
Characterization and Optimization of Selective, Nonpeptidic Inhibitors of Cathepsin S with an
Unprecedented Binding Mode†
Hiroaki Inagaki,‡ Hiroyuki Tsuruoka,‡ Michael Hornsby,§ Scott A. Lesley,§ Glen Spraggon,*,§ and Jonathan A. Ellman*,‡
Department of Chemistry, UniVersity of California, Berkeley, California 94720, and Genomics Institute of the NoVartis Research Foundation,
10675 John Jay Hopkins DriVe, San Diego, California 92121
ReceiVed January 27, 2007
The substrate activity screening (SAS) method, a substrate-based fragment identification and optimization
method for the development of enzyme inhibitors, was previously applied to cathepsin S to obtain a novel
(2-arylphenoxy)acetaldehyde inhibitor, 2, with a 0.49 µM Ki value (Wood, W. J. L.; Patterson, A. W.;
Tsuruoka, H.; Jain, R. K.; Ellman, J. A. J. Am. Chem. Soc. 2005, 127, 15521-15527). In this paper we
disclose the X-ray structure of a complex between cathepsin S and inhibitor 2 which reveals an unprecedented
binding mode. On the basis of this structure, additional 2-biaryloxy substrates with greatly increased cleavage
efficiency were designed. Conversion of the optimized substrates to the corresponding aldehyde inhibitors
yielded a low molecular weight (304 Daltons) and potent (9.6 nM) cathepsin S inhibitor that showed from
100- to >1000-fold selectivity relative to cathepsins B, L, and K.
Introduction
Recently, we disclosed a new method called substrate
activity screening (SAS) for the substrate-based identification
of novel fragments that bind to serine or cysteine protease
targets.1,2 This method further enables the subsequent conversion
of identified binding fragments to novel and low molecular
weight mechanism-based protease inhibitors. In the first ap-
plication of the SAS method,1 two distinct nonpeptidic aldehyde
inhibitors (Figure 1) were identified to the lysosomal cysteine
protease cathepsin S, which is expressed in antigen-presenting
cells such as macrophages, dendritic cells, and B cells. Cathepsin
S degrades the MHCa class II-associated invariant chain that is
required for productive loading of antigen onto the MHC class
II complex, and therefore, inhibition of cathepsin S may be
useful for attenuating antigen presentation in the treatment of
autoimmune diseases.3
Figure 1. Cathepsin S inhibitors identified by the SAS method.
Chemistry
Central to the synthesis of analogues of 2 was the preparation
of the appropriately substituted arylboronic acids that were
needed for Suzuki cross-coupling reactions to prepare the biaryl
frameworks. Accordingly, ortholithiation of 1,2-difluorobenzene
with n-butyllithium in TMEDA/THF followed by addition of
the requisite alkyl iodide provided the orthoalkylated difluo-
robenzenes 3 with high levels of regiocontrol (Scheme 1).5 A
second fluorine-directed ortholithiation followed by reaction
with B(OMe)3 and aqueous workup then provided the desired
arylboronic acids 4.
More recently, we reported on further optimization of the
potency and selectivity of inhibitor 1.4 Here we report that the
crystal structure of cathepsin S in complex with 2 establishes
that this novel inhibitor, which lacks nitrogen functionality or
chiral centers, has an unprecedented binding mode. Moreover,
this structural information has provided the basis for the
development of a low molecular weight (304 Daltons) inhibitor
with a low nanomolar Ki value and from 100- to >1000-fold
selectivity relative to the closely related enzymes cathepsins B,
L, and K.
For the synthesis of the 2-arylphenoxy substrates, two
approaches were taken.
In the first approach (Scheme 2), Suzuki cross-coupling
reactions were performed between arylboronic acids 4a and 4b
and 2-bromophenol to provide the 2-arylphenols 5a and 5b. For
the synthesis of 5f (Scheme 2), Suzuki cross-coupling was
performed with 4 (R ) OMe), which was prepared by literature
methods.6 The support-bound 7-(chloroacetamido)coumarin 61
was then alkylated with the phenols 5 mediated by the strong
base 2-(tert-butylimino)-2-(diethylamino)-1,3-dimethylperhydro-
1,3,2-diazaphosphorine (BEMP) to provide the support-bound
substrates 7. Cleavage from the support with trifluoroacetic acid
in CH2Cl2 then provided the desired substrates 8.
The second method for preparing 2-arylphenoxy substrates
8 was carried out in solution (Scheme 3). Suzuki cross-coupling
of 4 with commercially available 9 provided 10, which either
could be taken on to substrate 8 (Scheme 3) or could be readily
converted to the corresponding aldehyde inhibitor (vide infra).
Saponification of 10 followed by coupling with 7-amino-4-
methylcoumarin acetic acid ethyl ester provided 11, which upon
saponification provided the desired substrate 8. The synthesis
† Data for the X-ray crystal structures have been deposited in the Protein
Data Bank, PDB ID 2OP3.
* To whom correspondence should be addressed. (G.S.) Phone:
(858) 812-1567. Fax: (858) 812-1746. E-mail: gspraggon@gnf.org.
(J.A.E.) Phone: (510) 642-4488. Fax: (510) 642-8369. E-mail: jellman@
uclink.berkeley.edu.
‡ University of California.
§ Genomics Institute of the Novartis Research Foundation.
a Abbreviations: AMC, 7-amino-4-(carbamoylmethyl)coumarin; AMCA,
7-amino-4-methylcoumarin acetic acid; E-64, L-trans-epoxysuccinyl-Leu-
4-guanidinobutylamide; E-64c, L-trans-epoxysuccinyl-Leu-3-methylbuty-
lamide; DMPU, 1,3-dimethyl-3,4,5,6-tetrahydro-2(1H)-pyrimidinone; MHC,
major histocompatibility complex; HATU, O-(7-azabenzotriazol-1-yl)-
N,N,N′,N′-tetramethyluronium hexafluorophosphate; HPFC, high-perfor-
mance flash chromatography; TMEDA, N,N,N′,N′-tetramethylethylenedi-
amine.
10.1021/jm070111+ CCC: $37.00 © 2007 American Chemical Society
Published on Web 05/01/2007

2694 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Scheme 1. Synthesis of Arylboronic Acid Derivativesa
Inagaki et al.
Scheme 3. Synthesis of Substrates in Solutiona
a Reagents: (a) n-BuLi, TMEDA/THF, RI, DMPU; (b) n-BuLi, TMEDA/
THF, B(OMe)3, aqueous HCl.
Scheme 2. Synthesis of Substrates on Solid Supportsa
a Reagents: (a) Pd2(dba)3-CHCl3, 2-(dicyclohexylphosphino)-2′,6′-
dimethoxybiphenyl, K3PO4, dioxane; (b) aqueous NaOH, MeOH/THF; (c)
7-amino-4-methylcoumarin acetic acid ethyl ester, HATU, collidine, DMF;
(d) aqueous NaOH, MeOH/THF.
Scheme 4. Synthesis of Aldehyde Inhibitorsa
a Reagents: (a) 2-bromophenol, Pd(PPh3)4, Na2CO3, benzene/MeOH/
H2O; (b) BEMP, THF; (c) trifluoroacetic acid, CH2Cl2.
a Reagents and conditions: (a) DIBAL-H, toluene, -78 °C.
Table 1. Data Collection and Refinement Statistics for Cathepsin S
Data Collection and Processing
of the aldehyde inhibitors 12 could readily be achieved by
diisobutylaluminum hydride (DIBAL-H) reduction of the late-
stage intermediate 10 (Scheme 4).
space group
a ) b, Å
c, Å
P4122
85.384
151.2
Results and Discussion
wavelength, Å
resolution, Å
outer shell diam, Å
1.0
1.6
1.6-1.66
0.082
Proteases uniformly recognize their substrates as â strands
in their active sites.7 Consequently, peptide backbone hydrogen-
bonding is featured prominently in substrate and inhibitor
binding interactions. In contrast, the novel inhibitor 2 lacks
amide NH groups or alternative hydrogen bond donors. More-
over, H-bond-accepting interactions typically provided by the
amide carbonyls or heterocyclic isosteres of protease substrates
or inhibitors are also absent. The structure of 2 therefore suggests
that its binding interactions with cathepsin S would be quite
distinct from that typically observed for peptidic or peptidomi-
metic inhibitors.
Rmerge, % (outer shell)
I/sigI (σ) (outer shell)
no. of unique reflns (obsd)
completeness, % (outer shell)
45.6 (5.0)
74679 (994427)
100.0 (100.0)
Refinement
Rfactora (Rfreeb), %
no. of protein atoms
no. of water atoms
no. of hetero atoms
rmsd(bonds), Å
0.16 (0.18)
3906
475
76
0.012
1.310
13.328
rmsd(angles), deg
mean B factor, Å2
A crystal structure of 2 in complex with cathepsin S was
therefore obtained to determine its mode of binding. Crystal-
lographic statistics are summarized in Table 1. Indeed, as
illustrated in the crystal structure (Figure 2), inhibitor 2 has a
very distinctive mode of binding. While the aldehyde pharma-
cophore shows the expected tetrahedral adduct with the active
site cysteine, the inhibitor does not make any interactions with
the S1 and S3 pockets, which typically accommodate the P1
a Rfactor ) ∑|Ii - Ii ||/∑|Ii|, where Ii is the scaled intensity of the ith
b
measurement and Ii is the mean intensity for that reflection. Rfree is the
same as Rcryst, but for 5.0% of the total reflections chosen at random and
omitted from refinement.
and P3 side chains of peptidic inhibitors. However, the biaryl
functionality does make multiple hydrophobic contacts with the
three fluorine substituents buried within a hydrophobic cavity,

Unprecedented Inhibitors of Cathepsin S by SAS
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2695
Table 2. Cathepsin S-Mediated Substrate Cleavage Efficiency
k
cat/KM × 103
entry
compd
R
(M-1 s-1
)
1
2
3
4
5
6
7
13
8f
F
OMe
Et
n-Pr
n-Bu
i-Bu
27.0 ( 0.4
2.70 ( 0.04
25.9 ( 0.2
145 ( 1
8c
8d
8a
8e
8b
85 ( 3
288 ( 30
15.7 ( 0.2
CH2Chx
Figure 2. Structure of cathepsin S in complex with inhibitor 2. Protein
carbon atoms are colored yellow, while those from the inhibitor are
shaded cyan. Nitrogens are in blue, oxygens in red, and fluorines in
green. (a) Surface representation of the cathepsin S active site cleft
showing the position of inhibitor 2 in relation to the specificity subsites
S1, S2, and S3. (b) Stick representations of inhibitor 2 in the active
site. Weak electrostatic interactions between fluorines and main chain
atoms under 4.0 Å are represented as dotted lines. (c) Stick representa-
tions of inhibitor 2 in the active site oriented toward the hydrophobic
S2 pocket. Potential hydrophobic interactions between fluorines and
carbon atoms under 5.0 Å are represented as dotted lines. The figure
was produced by pymol (www.pymol.org).
Table 3. Inhibition of Cathepsins B, K, L, and S by Aldehyde
Inhibitors
Ki (nM)
compd
R
Cat S
Cat L
Cat K
Cat B
defined by residues Phe 70, Met 71, Val 162, and Val 135-
Gly 137 (Figure 2c). These contacts are consistent with
previously reported substrate and inhibitor SAR, indicating the
importance of the binding interactions provided by the fluorine
substituents: removal of the 4-fluoro substituent from 2 reduces
Ki by 2.5-fold and replacement of all three fluoro substituents
reduces Ki by 80-fold.1 The 2- and 3-fluoro groups are likely
to contribute binding interactions and specificity via weak
hydrogen bond interactions with the main chain amides of Gly
165 and Gly 69, respectively (Figure 2b).
Importantly, the structure suggests numerous opportunities
for enhancing the potency of inhibitor 2 by extending into the
S1, S2, and S3 pockets. Designing inhibitors that extend into
the S2 pocket provides a particularly attractive means to achieve
selectivity relative to cathepsin B, and more significantly
cathepsins L and K, which have particularly high homology
with cathepsin S, because Gly 137 contributes to the wall of
the S2 pocket, while for cathepsins B, L, and K the larger Ala
residue is located at this site.
As previously reported in the development of inhibitors 1
and 2,1 and in the further optimization of 1,4 outstanding
correlation between substrate cleavage efficiency and inhibitory
activity was observed. Substrate analogues were therefore first
prepared to evaluate the effect of introducing substituents that
extend into the S2 pocket (Table 2). Replacing the 4-fluoro
substituent with alkyl or ether substituents would provide an
effective means to append functionality that might interact with
the hydrophobic S2 pocket. Two analogues, 8f and 8c, with
methoxy and ethyl groups at the 4-position, respectively (entries
2 and 3), were first prepared and evaluated as substrates to
identify whether an ether or alkyl group would provide the most
effective linkage strategy. While the methoxy analogue 8f
showed a 10-fold reduction in substrate cleavage efficiency
(entry 2) relative to that of the benchmark substrate 13 (entry
1), the ethyl derivative 8c was a substrate comparable in
efficiency to 13 (entry 3). These results are consistent with the
hydrophobic contacts made by the 4-fluoro group in the
2
12d
12e
fluoro
n-Pr
i-Bu
490 ( 50
1540 ( 320 1510 ( 310 >15000
48.7 ( 2.0 5190 ( 430 1280 ( 170 >15000
9.6 ( 2.4
>15000
1240 ( 280 >15000
cocrystal structure of 2 (Figure 2). Because 8c was a more
efficient substrate than 8f, an alkyl linkage was employed for
all additional analogues designed to extend into the S2 pocket.
As shown in entry 4, compound 8d with R ) n-propyl was
cleaved 5-fold more efficiently than compound 13, while 8a
with R ) n-butyl was cleaved only 3-fold more efficiently (entry
5). Introducing branching to provide 8e with R ) isobutyl (entry
6) provided the greatest enhancement in cleavage efficiency
(>10-fold) relative to that of 13. However, branched substrate
8b incorporating a cyclohexylmethyl substituent showed reduced
activity and serves to define the size of the pocket (entry 7).
The aldehyde inhibitors corresponding to the two substrates
with the highest cleavage efficiency were prepared, and their
inhibitory activities were determined (Table 3). Aldehyde 12d
showed 10-fold greater potency than inhibitor 2, and aldehyde
12e, which corresponds to the best substrate, was 50-fold more
potent with a 9.6 nM Ki value. On the basis of the crystal
structure, the propyl and isobutyl substituents of 12d and 12e,
respectively, would be expected to interact with Gly 137 in
cathepsin S. Because Ala is present at the corresponding site in
cathepsins B, L, and K, an increase in selectivity could be
anticipated. Indeed, while 2 was only marginally selective for
cathepsin S over cathepsin K and cathepsin L, replacing the
4-fluoro group with the alkyl side chains resulted in dramatic
improvements in selectivity, with inhibitor 12e showing >100-
fold selectivity for cathepsin K and >1000-fold selectivity for
cathepsins B and L (Table 3).
In conclusion, the structure of novel, nonpeptidic inhibitor 2
in complex with cathepsin S has been reported and shows a
binding mode that is distinct from that observed for peptidic
and peptidomimetic inhibitors. The structural information
guided the design of aldehyde inhibitor 12e, which is both potent
(Ki ) 9.6 nM) and selective (from 100- to >1000-fold selectivity

2696 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Inagaki et al.
2.98 g (19 mmol) of product as a colorless oil (yield 95%): 1H
NMR (400 MHz) δ 0.95 (t, J ) 7.4, 3H), 1.64 (sextet, J ) 7.5,
2H), 2.64 (dt, J ) 1.3, 7.0, 2H), 6.90-7.01 (m, 3H); GC/MS (EI)
m/z 156 (M+).
2,3-Difluoro-1-(2-methylpropyl)benzene (3e). The general pro-
cedure was followed using 0.99 mL (10 mmol) of difluorobenzene
to give 0.49 g (2.9 mmol) of product as a colorless oil (yield
29%): 1H NMR (400 MHz) δ 0.92 (d, J ) 6.7, 6H), 1.91 (m, 1H),
2.54 (dd, J ) 1.5, 7.2, 2H), 6.86-7.03 (m, 3H); GC/MS (EI) m/z
170 (M+).
relative to cathepsins K, L, and B). The structure of 2 in complex
with cathepsin S suggests additional opportunities for enhancing
potency and selectivity. The S3 pocket, which is unoccupied,
provides a clear opportunity for introducing additional binding
interactions. Moreover, the low molecular weight of 12e enables
the aldehyde pharmacophore to be replaced by more metaboli-
cally stable covalent8 or noncovalent9,10 pharmacophores that
extend into the prime binding pockets, a strategy that has proven
to be successful in recent cathepsin S and K inhibitor develop-
ment efforts.
Synthesis of 4-Alkyl-2,3-difluorobenzeneboronic Acids 4a-
e. 4-(Cyclohexylmethyl)-2,3-difluorobenzeneboronic Acid (4b).
To a solution of 3b (1.41 g, 6.7 mmol) and N,N,N′,N′-tetrametyl-
ethylenediamine (TMEDA; 1.01 mL, 6.7 mmol) in THF (20 mL)
was added a solution of n-butyllithium in hexane (2.5 M, 3.5 mL,
8.8 mmol) dropwise over 5 min at -78 °C under a nitrogen
atmosphere. After the mixture was stirred for 1 h at -78 °C,
trimethyl borate (0.97 mL, 8.8 mmol) was added dropwise over 5
min, and then the mixture was stirred for 1.5 h at -78 °C. After
the reaction was quenched by adding water (5 mL) at -78 °C, the
mixture was allowed to warm to ambient temperature, and then 2
N HCl(aq) (25 mL) was added. The mixture was stirred overnight
and extracted with EtOAc. The extracts were washed with brine,
dried with anhydrous Na2SO4, filtered, and concentrated in vacuo.
Purification using silica gel column chromatography (DCM/MeOH
) 50/1 f 20/1 f 10/1) gave the title compound (1.12 g, 4.4 mmol,
66%) as a colorless solid: 1H NMR (400 MHz) δ 0.91-1.05 (m,
2H), 1.10-1.26 (m, 4H), 1.62-1.74 (m, 5H), 2.55 (m, 2H), 5.05
(d, J ) 5.6, 1H), 6.89-6.97 (m, 1H), 7.35-7.46 (m, 1H).
4-n-Butyl-2,3-difluorobenzeneboronic Acid (4a). Using the
same procedure as for 4b, 3a (1.70 g, 10 mmol) was converted to
109 mg (0.51 mmol) of the title compound as a colorless solid
(yield 5%): 1H NMR (400 MHz) δ 0.91-0.97 (m, 3H), 1.31-
1.49 (m, 2H), 1.57-1.69 (m, 2H), 2.65-2.75 (m, 2H), 5.03 (d, J
) 5.5, 1H), 6.93-7.09 (m, 1H), 7.37-7.48 (m, 1H).
4-Ethyl-2,3-difluorobenzeneboronic Acid (4c). Using the same
procedure as for 4b, 3c (350 mg, 2.24 mmol) was converted to
0.28 g (1.51 mmol) of the title compound as a colorless solid (yield
67%): 1H NMR (300 MHz) δ 1.21-1.33 (m, 3H), 2.70-2.82 (m,
2H), 5.00 (br s, 1H), 7.00-7.09 (m, 1H), 7.47 (m, 0.5H), 7.78 (m,
0.5H).
2,3-Difluoro-4-n-propylbenzeneboronic Acid (4d). Using the
same procedure as for 4b, 3d (2.98 g, 19.1 mmol) was converted
to 2.34 g (11.7 mmol) of the title compound as a colorless solid
(yield 61%): 1H NMR (400 MHz) δ 0.90-1.01 (m, 3H), 1.61-
1.74 (m, 2H), 2.62-2.77 (m, 2H), 4.98 (br s, 0.4H), 6.94-7.08
(m, 1H), 7.36-7.47 (m, 0.4H), 7.76 (m, 0.6H).
Experimental Section
General Synthesis Methods. Unless otherwise noted, all
reagents were obtained from commercial suppliers and used without
purification. Tetrahydrofuran (THF), dioxane, CH2Cl2, ether, and
toluene were passed through a column of activated alumina (type
A2, 12 × 32, Purify Co.) under nitrogen pressure immediately prior
to use. TMEDA and DMPU were distilled under N2, and i-Pr2EtN,
pyridine, and CH3OH were distilled under N2 over CaH2 im-
mediately prior to use. Low amine content DMF was purchased
from Acros, Wang resin and Sieber amide resin were purchased
from Novabiochem (San Diego, CA), and O-(7-azabenzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HATU)
was purchased from PerSeptive Biosystems (Foster City, CA).
Fmoc-protected 7-amino-4-methylcoumarin acetic acid (Fmoc-
AMCA) was synthesized according to a method analogous to the
synthesis of 7-amino-4-(carbamoylmethyl)coumarin (AMC).11 Com-
pounds 2, 13, and Fmoc-AMCA-Wang resin were synthesized as
previously reported.1 All solution-phase reactions were carried out
in flame-dried glassware under an inert N2 atmosphere. Normal-
phase HPFC purification was carried out on a Biotage SP1
instrument (Charlottesville, VA) equipped with a Biotage Si flash
column. Reversed-phase HPLC analysis and purification were
conducted with an Agilent 1100 series instrument. Solid-phase
reactions were conducted in polypropylene cartridges equipped with
70 mm PE frits (Applied Separations, Allentown, PA) and Teflon
1
stopcocks and rocked on an orbital shaker. H NMR spectra were
obtained with Bruker AV-300 and AV-400 spectrometers. Unless
otherwise specified, all spectra were obtained in CDCl3, and
chemical shifts are reported in parts per million relative to the peak
for internal CHCl3. Coupling constants are reported in hertz. Mass
spectrometry analyses including high-resolution mass spectrometry
analyses were performed by the University of California at Berkeley
Micro Analysis and Mass Spectrometry Facilities. GC/MS analyses
were performed with an Agilent Technologies 6890N GC system
with an HP-5MS column coupled with an Agilent 5973 mass-
selective detector.
Synthesis of 1-Alkyl-2,3-difluorobenzenes 3a-e. According to
the procedure reported by Reiffenrath,4 1,2-difluorobenzene was
alkylated with the appropriate alkyl iodide to obtain the title
compounds.
1-n-Butyl-2,3-difluorobenzene (3a). The general procedure was
followed using 0.99 mL (10 mmol) of difluorobenzene to give 1.15
g (6.8 mmol) of product as a colorless oil (yield 68%): 1H NMR
(400 MHz) δ 0.93 (t, J ) 7.3, 3H), 1.37 (sextet, J ) 7.4, 2H), 1.59
(quintet, J ) 7.7, 2H), 2.66 (t, J ) 7.7, 2H), 6.92-7.02 (m, 3H);
MS (EI) m/z 170 (M+).
2,3-Difluoro-4-(2-methylpropyl)benzeneboronic Acid (4e). Us-
ing the same procedure as for 4b, 3e (592 mg, 3.48 mmol) was
converted to 490 mg (2.29 mmol) of the title compound as a
colorless solid (yield 66%): 1H NMR (300 MHz) δ 0.90-0.98 (m,
6H), 1.85-1.99 (m, 1H), 2.52-2.58 (m, 2H), 4.96 (d, J ) 5.7,
0.6H), 6.90-7.02 (m, 1H), 7.39-7.48 (m, 1H).
Synthesis of 2-(4-Alkyl-2,3-difluorophenyl)phenols 5a, 5b, and
5f. 2-(4-n-Butyl-2,3-difluorophenyl)phenol (5a). A mixture of 4a
(105 mg, 0.49 mmol), 2-bromophenol (57 µL, 0.49 mmol), Pd-
(Ph3P)4 (11 mg, 0.010 mmol), sodium carbonate (208 mg, 2.0
mmol), benzene (4 mL), MeOH (0.8 mL), and water (0.2 mL) was
heated at 80 °C with stirring for 15 h under a nitrogen atmosphere.
The resultant mixture was partitioned between EtOAc (50 mL) and
1 N HCl(aq) (10 mL), and the organic layer was isolated. The
organic layer was washed with water (10 mL) and brine (10 mL),
dried with anhydrous Na2SO4, filtered, and concentrated in vacuo.
Purification by HPFC (hexane/EtOAc ) 100/0 f 88/12) gave the
title compound (95 mg, 0.36 mmol, 74%) as a colorless oil: 1H
NMR (400 MHz) δ 0.97 (t, J ) 7.3, 3H), 1.41 (sextet, J ) 7.4,
2H), 1.65 (quintet, J ) 7.6, 2H), 2.72 (t, J ) 7.7, 2H), 4.96 (br s,
1H), 6.98-7.07 (m, 4H), 7.21-7.33 (m, 2H); MS (EI) m/z 262
(M+).
1-(Cyclohexylmethyl)-2,3-difluorobenzene (3b). The general
procedure was followed using 0.27 mL (2.7 mmol) of difluoroben-
zene to give 205 mg (0.98 mmol) of product as a colorless oil (yield
36%): 1H NMR (400 MHz) δ 0.90-1.03 (m, 2H), 1.08-1.22 (m,
4H), 1.60-1.75 (m, 5H), 2.54 (d, J ) 6.7, 2H), 6.88-7.00 (m,
3H); MS (EI) m/z 210 (M+).
1-Ethyl-2,3-difluorobenzene (3c). The general procedure was
followed using 1.97 mL (20 mmol) of difluorobenzene to give 350
mg (2.2 mmol) of product as a colorless oil (yield 11%): 1H NMR
(300 MHz) δ 1.24 (t, J ) 7.6, 3H), 2.70 (dq, J ) 1.2, 6.4, 2H),
6.92-7.02 (m, 3H); GC/MS (EI) m/z 142 (M+).
2,3-Difluoro-1-n-propylbenzene (3d). The general procedure
was followed using 1.97 mL (20 mmol) of difluorobenzene to give
2-[4-(Cyclohexylmethyl)-2,3-difluorophenyl]phenol (5b). Us-
ing the same procedure as for 5a, 4b (178 mg, 0.70 mmol) was

Unprecedented Inhibitors of Cathepsin S by SAS
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2697
converted to 173 mg (0.57 mmol) of the title compound as a
colorless gum (yield 82%): 1H NMR (400 MHz) δ 0.96-1.06 (m,
2H), 1.15-1.28 (m, 3H), 1.60-1.74 (m, 6H), 2.59 (d, J ) 7.1,
2H), 5.04 (br s, 1H), 6.98-7.06 (m, 4H), 7.23-7.33 (m, 2H); MS
(EI) m/z 302 (M+).
2-(2,3-Difluoro-4-methoxyphenyl)phenol (5f). Using the same
procedure as for 5a, 4f (120 mg, 0.64 mmol) was converted to 112
mg (0.47 mmol) of the title compound as a colorless oil (yield 74%);
1H NMR (400 MHz) δ 3.96 (s, 3H), 4.93 (s, 1H), 6.85 (m, 1H),
6.97-7.03 (m, 2H), 7.08 (m, 1H), 7.22 (d, J ) 7.4, 1H), 7.30 (m,
1H); MS (EI) m/z 236 (M+).
1H), 6.97 (t, J ) 7.4, 1H), 7.06-7.09 (m, 2H), 7.29-7.42 (m, 2H);
MS (FAB) m/z 550 (M + H+).
Ethyl [[2′,3′-Difluoro-4′-(2-methylpropyl)biphenyl-2-yl]oxy]-
acetate (10e). Using the same procedure as for 10c, 4e (127 mg,
0.59 mmol) was converted to 101 mg (0.29 mmol) of the title
compound as a pale yellow oil (yield 98%): 1H NMR (400 MHz)
δ 0.96 (d, J ) 6.7, 6H), 1.26 (t, J ) 7.1, 3H), 1.95 (m, 1H), 2.57
(d, J ) 7.2, 2H), 4.22 (q, J ) 7.1, 2H), 4.62 (s, 2H), 6.87 (d, J )
8.2, 1H), 6.93 (t, J ) 7.3, 1H), 7.06-7.12 (m, 2H), 7.30-7.36 (m,
2H); MS (FAB) m/z 564 (M + H+).
Ethyl (7-Amino-4-methyl-2-oxo-2H-chromen-3-yl)acetate
(AMCA Ethyl Ester). A mixture of [7-(9-fluorenylmethoxycar-
bonylamino)-4-methyl-2-oxo-2H-chromen-3-yl]acetic acid (Fmoc-
AMCA; 4.56 g, 10.0 mmol), potassium carbonate (1.52 g, 11.0
mmol), iodoethane (1.20 mL, 15.0 mmol), and DMF (30 mL) was
stirred at ambient temperature for 6 h. The resultant mixture was
poured into 1 N HCl(aq) (50 mL) and was extracted with DCM
(200 mL + 2 × 50 mL). The combined organic layers were washed
with water (2 × 50 mL) and brine (50 mL), and after dilution with
DCM/MeOH ) 9/1 solution (500 mL), the organic solution was
dried with anhydrous Na2SO4, filtered, and concentrated in vacuo.
Then the residue was suspended in MeOH, and the insoluble solid
was collected by filtration to give as a crude product [7-(9-
fluorenylmethoxycarbonylamino)-4-methyl-2-oxo-2H-chromen-3-
yl]acetic acid ethyl ester (Fmoc-AMCA ethyl ester) as a colorless
powder: 1H NMR (400 MHz) δ 1.27 (t, J ) 7.1 Hz, 3H), 2.37 (s,
3H), 3.71 (s, 2H), 4.18 (q, J ) 7.1 Hz, 2H), 4.29 (t, J ) 6.0 Hz,
1H), 4.61 (d, J ) 6.3 Hz, 2H), 6.90 (br s, 0.9H), 7.33-7.45 (m,
6H), 7.53 (d, J ) 8.7 Hz, 1H), 7.63 (d, J ) 7.5 Hz, 2H), 7.80 (d,
J ) 7.6 Hz, 2H); MS (FAB) m/z 262 (M + H+).
Crude Fmoc-AMCA ethyl ester was suspended in DMF (50 mL),
and piperidine (5 mL) was added to the suspension at ambient
temperature. The mixture was stirred for 20 min at ambient
temperature and then concentrated in vacuo. The resultant residue
was crystallized in a hexane/DCM solution (90 mL/10 mL) and
collected by filtration to give the title compound (2.19 g, 8.34 mmol,
2 steps, 84%) as a pale yellow solid: 1H NMR (400 MHz) δ 1.26
(t, J ) 7.1 Hz, 3H), 2.33 (s, 3H), 3.68 (s, 2H), 4.16 (s, 2H), 4.17
(q, J ) 7.1, 2H), 6.54-6.58 (m, 2H), 7.38 (d, J ) 8.5 Hz, 1H);
MS (FAB) m/z 484 (M + H+).
Synthesis of [(4′-Alkyl-2′,3′-difluorobiphenyl-2-yl)oxy]acetyl-
AMCA Ethyl Esters 11c, 11d, and 11e. [(4′-Ethyl-2′,3′-difluo-
robiphenyl-2-yl)oxy]acetyl-AMCA Ethyl Ester (11c). NaOH(aq)
(1 N, 0.60 mL) was added to a mixture of 10c (61 mg, 0.190 mmol),
MeOH (0.6 mL), and THF (1.8 mL), and the mixture was stirred
for 6 h at ambient temperature. After 1 N HCl(aq) (2 mL) and
water (10 mL) were added, the resulting mixture was extracted with
EtOAc (50 mL + 10 mL). The combined extracts were dried with
anhydrous Na2SO4, filtered, and concentrated in vacuo to provide
[(4′-ethyl-2′,3′-difluorobiphenyl-2-yl)oxy]acetic acid. To a solution
of the crude [(4′-ethyl-2′,3′-difluorobiphenyl-2-yl)oxy]acetic acid,
AMCA ethyl ester (50 mg, 0.19 mmol), and 2,4,6-collidine (0.050
mL, 0.38 mmol) in DMF (1 mL) was added HATU (108 mg, 0.29
mmol) at ambient temperature, and the mixture was stirred for 17
h at the same temperature. After dilution with DCM (50 mL), the
resultant mixture was washed with 1 N HCl(aq) (10 mL), water
(10 mL), and brine (10 mL), then dried with anhydrous Na2SO4,
filtered, and concentrated in vacuo. The obtained crude product
was purified by silica gel column chromatography (DCM/EtOAc
) 20/1 f 10/1) to give the title compound (90 mg, 0.169 mmol,
89%) as a colorless foam: 1H NMR (400 MHz) δ 1.27 (t, J ) 7.1,
3H), 1.33 (t, J ) 7.6, 3H), 2.40 (s, 3H), 2.82 (q, J ) 7.5, 2H), 3.72
(s, 2H), 4.18 (q, J ) 7.2, 2H), 4.65 (s, 2H), 6.99 (d, J ) 8.4, 1H),
7.09-7.19 (m, 3H), 7.35-7.44 (m, 3H), 7.55-7.57 (m, 2H), 8.40
(br s, 1H); MS (FAB) m/z 536 (M + H+).
Synthesis of [2-(4-Alkyl-2,3-difluorophenyl)phenoxy]acetyl-
AMCAs 8a, 8b, and 8f. [2-(4-n-Butyl-2,3-difluorophenyl)-
phenoxy]acetyl-AMCA (8a). According to the reported procedure,1
5a (89 mg, 0.34 mmol) was reacted with resin-bound AMCA (6;
0.10 mmol). Cleavage from the resin followed by reversed-phase
HPLC purification yielded 22 mg (0.041 mmol) of the title
compound as a colorless solid (yield 41%): 1H NMR (400 MHz)
δ 0.95 (t, J ) 7.3, 3H), 1.43 (sextet, J ) 7.4, 2H), 1.67 (quintet, J
) 7.6, 2H), 2.43 (s, 3H), 2.77 (t, J ) 7.7, 2H), 3.78 (s, 2H), 4.65
(s, 2H), 6.99 (d, J ) 8.2, 1H), 7.09 (m, 2H), 7.16 (t, J ) 7.1, 1H),
7.35 (m, 2H), 7.44 (m, 1H), 7.57 (d, J ) 8.7, 1H), 7.64 (d, J )
2.1, 1H), 8.43 (d, J ) 3.1, 1H); HRMS (FAB+) m/z 536.1892
(MH+, C30H28F2NO6 requires 536.1885). The purity of the com-
pound was checked by RP-HPLC (see the Supporting Information
for MeOH/H2O/TFA and MeCN/H2O/TFA HPLC traces).
[2-[4-(Cyclohexylmethyl)-2,3-difluorophenyl]phenoxy]acetyl-
AMCA (8b). Using the same procedure as for 8a, 5b (91 mg, 0.30
mmol) was converted to 32 mg (0.056 mmol) of the title compound
as a colorless solid (yield 56%): 1H NMR (400 MHz) δ 1.00-
1.09 (m, 2H), 1.15-1.21 (m, 3H), 1.60-1.76 (m, 6H), 2.43 (s,
3H), 2.65 (d, J ) 6.7, 2H), 3.78 (s, 2H), 4.66 (s, 2H), 6.99 (d, J )
8.3, 1H), 7.06 (m, 2H), 7.17 (t, J ) 7.2, 1H), 7.35-7.37 (m, 2H),
7.43 (m, 1H), 7.58 (d, J ) 8.7, 1H), 7.67 (d, J ) 2.0, 1H), 8.45 (d,
J ) 3.3, 1H); HRMS (FAB+) m/z 576.2198 (MH+, C33H32F2NO6
requires 576.2198). The purity of the compound was checked by
RP-HPLC (see the Supporting Information for MeOH/H2O/TFA
and MeCN/H2O/TFA HPLC traces).
[2-(2,3-Difluoro-4-methoxyphenyl)phenoxy]acetyl-AMCA (8f).
Using the same procedure as for 8a, 5f (56 mg, 0.24 mmol) was
converted to 21 mg (0.040 mmol) of the title compound as a
colorless solid (yield 40%): 1H NMR (400 MHz) δ 2.44 (s, 3H),
3.78 (s, 2H), 4.01 (s, 3H), 4.65 (s, 2H), 6.92 (m, 1H), 6.99 (d, J )
8.0, 1H), 7.10 (m, 1H), 7.16 (t, J ) 7.4, 1H), 7.34 (m, 1H), 7.38-
7.43 (m, 2H), 7.59-7.62 (m, 2H), 8.41 (br s, 1H); HRMS (FAB+)
m/z 510.1371 (MH+, C27H22F2NO7 requires 510.1364). The purity
of the compound was checked by RP-HPLC (see the Supporting
Information for MeOH/H2O/TFA and MeCN/H2O/TFA HPLC
traces).
Synthesis of Ethyl [(4′-Alkyl-2′,3′-difluorobiphenyl-2-yl)oxy]-
acetates 10c, 10d, and 10e. Ethyl [(4′-Ethyl-2′,3′-difluorobiphe-
nyl-2-yl)oxy]acetate (10c). A mixture of 4c (186 mg, 1.00 mmol),
ethyl (2-bromophenoxy)acetate (9; 130 mg, 0.50 mmol), Pd2(dba)3-
CHCl3 (16 mg, 0.015 mmol), 2-(dicyclohexylphosphino)-2′,6′-
dimethoxybiphenyl (11 mg, 0.031 mmol), potassium phosphate (318
mg, 1.50 mmol), and dioxane (3 mL) was heated at 110 °C with
stirring for 15 h under a a nitrogen atmosphere. The resultant
mixture was filtered through a short silica gel pad and concentrated
in vacuo. Purification by silica gel column chromatography (hexane/
EtOAc ) 20/1 f 10/1) provided the title compound (145 mg, 0.45
mmol, 91%) as a pale yellow oil: 1H NMR (400 MHz) δ 1.24-
1.57 (m, 6H), 2.74 (q, J ) 7.5, 2H), 4.23 (q, J ) 7.1, 2H), 4.63 (s,
2H), 6.87 (d, J ) 8.3, 1H), 6.99 (t, J ) 7.4, 1H), 7.06-7.12 (m,
2H), 7.30 (d, J ) 7.5, 1H), 7.34 (m, 1H); MS (FAB) m/z 536 (M
+ H+).
Ethyl [(2′,3′-Difluoro-4′-n-propylbiphenyl-2-yl)oxy]acetate (10d).
Using the same procedure as for 10c, 4d (130 mg, 0.50 mmol)
was converted to 164 mg (0.49 mmol) of the title compound as a
pale yellow oil (yield 98%): 1H NMR (400 MHz) δ 1.00 (t, J )
7.3, 3H), 1.26 (t, J ) 7.1, 3H), 1.69 (sextet, J ) 7.5, 2H), 2.67 (t,
J ) 7.6, 2H), 4.23 (q, J ) 7.1, 2H), 4.63 (s, 2H), 6.87 (d, J ) 8.2,
[(2′,3′-Difluoro-4′-n-propylbiphenyl-2-yl)oxy]acetyl-AMCA Eth-
yl Ester (11d). Using the same procedure as for 11c, 10d (117
mg, 0.35 mmol) was converted to 125 mg (0.23 mmol) of the title
compound as a colorless foam (yield 65%): 1H NMR (400 MHz)
δ 1.02 (t, J ) 7.3, 3H), 1.27 (t, J ) 7.1, 3H), 1.73 (sextet, J ) 7.5,
2H), 2.40 (s, 3H), 2.75 (t, J ) 7.6, 2H), 3.72 (s, 2H), 4.18 (q, J )

2698 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11
Inagaki et al.
7.1, 2H), 4.65 (s, 2H), 6.94 (d, J ) 8.2, 1H), 7.09 (m, 2H), 7.17 (t,
J ) 7.1, 1H), 7.34-7.42 (m, 2H), 7.44-7.46 (m, 1H), 7.55-7.57
(m, 2H), 8.40 (br s, 1H); MS (FAB) m/z 550 (M + H+).
[[2′,3′-Difluoro-4′-(2-methylpropyl)biphenyl-2-yl]oxy]acetyl-
AMCA Ethyl Ester (11e). Using the same procedure as for 11c,
10e (53 mg, 0.152 mmol) was converted to 75 mg (0.132 mmol)
of the title compound as a colorless foam (yield 87%): 1H NMR
(400 MHz) δ 0.98 (d, J ) 6.6, 6H), 1.27 (t, J ) 7.1, 3H), 2.01 (m,
1H), 2.40 (s, 3H), 2.65 (d, J ) 7.2, 2H), 3.72 (s, 2H), 4.18 (q, J )
7.1, 2H), 4.65 (s, 2H), 6.99 (d, J ) 8.2, 1H), 7.05-7.12 (m, 2H),
7.17 (t, J ) 7.5, 1H), 7.37 (m, 2H), 7.44 (m, 1H), 7.56 (d, J ) 8.8,
1H), 7.59 (d, J ) 2.0, 1H), 8.42 (br s, 1H); MS (FAB) m/z 564 (M
+ H+).
Synthesis of [2-(4-Alkyl-2,3-difluorophenyl)phenoxy]acetyl-
AMCAs 8c, 8d, and 8e. [2-(4-Ethyl-2,3-difluorophenyl)phenoxy]-
acetyl-AMCA (8c). NaOH(aq) (1 N, 0.50 mL) was added to a
mixture of 11c (88 mg, 0.165 mmol), MeOH (0.5 mL), and THF
(1.5 mL), and the resulting mixture was stirred for 2 h at ambient
temperature. After 1 N HCl(aq) (5 mL) and water (5 mL) were
added, the mixture was extracted with a DCM/MeOH (5/1) solution
(once with 50 mL and three times with 10 mL). The combined
extracts were dried with anhydrous Na2SO4, filtered, and concen-
trated in vacuo. The crude product was purified by reversed-phase
HPLC (MeCN/H2O/TFA) to give the pure title compound (50 mg,
0.098 mmol, 59%) as a colorless solid: 1H NMR (400 MHz) δ
1.33 (t, J ) 7.6, 3H), 2.43 (s, 3H), 2.82 (q, J ) 7.6, 2H), 3.77 (s,
2H), 4.65 (s, 2H), 6.99 (d, J ) 8.3, 1H), 7.11 (m, 2H), 7.17 (t, J )
7.6, 1H), 7.36 (m, 2H), 7.44 (m, 1H), 7.56-7.61 (m, 2H), 8.42 (br
s, 1H); HRMS (FAB+) m/z 508.1573 (MH+, C28H24F2NO6 requires
508.1572). The purity of the compound was checked by RP-HPLC
(see the Supporting Information for MeOH/H2O/TFA and MeCN/
H2O/TFA HPLC traces).
[2-(2,3-Difluoro-4-n-propylphenyl)phenoxy]acetyl-AMCA (8d).
Using the same procedure as for 8c, 11d (123 mg, 0.224 mmol)
was converted to 52 mg (0.100 mmol) of the title compound as a
colorless solid (yield 45%): 1H NMR (400 MHz) δ 1.02 (t, J )
7.3, 3H), 1.72 (sextet, J ) 7.5, 2H), 2.42 (s, 3H), 2.75 (t, J ) 7.5,
2H), 3.77 (s, 2H), 4.65 (s, 2H), 6.99 (d, J ) 8.2, 1H), 7.09 (m,
2H), 7.16 (t, J ) 7.4, 1H), 7.34-7.38 (m, 2H), 7.43 (m, 1H), 7.56
(d, J ) 8.7, 1H), 7.59 (d, J ) 2.0, 1H), 8.42 (br s, 1H); HRMS
(FAB+) m/z 522.1715 (MH+, C29H26F2NO6 requires 522.1728).
The purity of the compound was checked by RP-HPLC (see the
Supporting Information for MeOH/H2O/TFA and MeCN/H2O/TFA
HPLC traces).
[2-(4-Isobutyl-2,3-difluorophenyl)phenoxy]acetyl-AMCA (8e).
Using the same procedure as for 8c, 11e (72 mg, 0.128 mmol) was
converted to 45 mg (0.083 mmol) of the title compound as a
colorless solid (yield 65%): 1H NMR (400 MHz) δ 0.98 (d, J )
6.6, 6H), 1.99 (m, 1H), 2.43 (s, 3H), 2.64 (d, J ) 7.2, 2H), 3.77 (s,
2H), 4.65 (s, 2H), 6.99 (d, J ) 8.2, 1H), 7.07 (m, 2H), 7.17 (t, J )
7.5, 1H), 7.35-7.37 (m, 2H), 7.44 (m, 1H), 7.57 (d, J ) 8.7, 1H),
7.63 (m, 1H), 8.45 (br s, 1H); HRMS (FAB+) m/z 536.1877 (MH+,
C30H28F2NO6 requires 536.1885). The purity of the compound was
checked by RP-HPLC (see the Supporting Information for MeOH/
H2O/TFA and MeCN/H2O/TFA HPLC traces).
1.8H), 0.92 (t, J ) 7.4, 1.2H), 1.39-1.68 (m, 2H), 2.41-2.54 (m,
1.2H), 2.64 (t, J ) 7.7, 0.8H), 3.86 (d, J ) 5.0, 1.2H), 4.73 (s,
0.8H), 5.02 (t, J ) 5.2, 0.6H), 6.80-7.10 (m, 4H), 7.12-7.50 (m,
2H), 9.61 (s, 0.4H); HRMS (EI+) m/z 290.1117 (M+, C17H16F2O2
requires 290.1118). The purity of the compound was checked by
RP-HPLC (see the Supporting Information for MeOH/H2O/TFA
and MeCN/H2O/TFA HPLC traces).
[2-(4-Isobutyl-2,3-difluorophenyl)phenoxy]acetaldehyde (12e).
Using the same procedure as for 12d, 10e (67 mg, 0.193 mmol)
was converted to 22 mg (0.074 mmol) of the title compound (1/4
mixture of the aldehyde and its hydrate) as a colorless gum (yield
38%): 1H NMR (400 MHz, CD3CN/D2O ) 5/1) δ 0.80-0.90 (m,
4.8H), 0.90 (d, J ) 6.6, 1.2H), 1.72-1.90 (m, 1H), 2.32-2.42 (m,
1.6H), 2.55 (d, J ) 7.4, 0.4H), 3.86 (d, J ) 5.0, 1.6H), 4.72 (s,
0.4H), 5.01 (t, J ) 0.8H), 6.80-7.10 (m, 4H), 7.16-7.40 (m, 2H),
9.59 (s, 0.2H); HRMS (EI+) m/z 304.1275 (M+, C18H18F2O2
requires 304.1275). The purity of the compound was checked by
RP-HPLC (see the Supporting Information for MeOH/H2O/TFA
and MeCN/H2O/TFA HPLC traces).
General Assay Procedure. Cathepsins B, K, L, and S were
purchased from Calbiochem (San Diego, CA). Cbz-Leu-Arg-AMC,
Cbz-Phe-Arg-AMC, l-trans-epoxysuccinyl-Leu-4-guanidinobuty-
lamide (E-64), and l-trans-epoxysuccinyl-Leu-3-methylbutylamide
(E-64c) were purchased from Bachem (Torrance, CA). The pro-
teolytic cleavage of N-acylaminocoumarins by cathepsin S was
conducted in Dynatech Microfluor fluorescence 96-well microtiter
plates, and readings were taken on a Molecular Devices Spectra
Max Gemini XS instrument. The excitation wavelength was 370
nm and the emission wavelength was 455 nm with a cutoff of 435
nm for AMCA substrates; the excitation wavelength was 355 nm
and the emission wavelength was 450 nm for peptidyl-AMC
substrates. The assay buffer consisted of a 100 mM solution of pH
6.1 sodium phosphate buffer with 100 mM sodium chloride, 1 mM
DTT, 1 mM EDTA, and 0.001% Tween-20.
Assay Procedure for AMCA Substrates. Assays were con-
ducted at 37 °C in duplicate with and without the enzyme. In each
well were placed 38 µL of enzyme solution and 2 µL of a DMSO
substrate solution. Assays were performed at substrate concentra-
tions that were at minimum 6 times less than the KM for that
substrate, and assays were performed in duplicate. Relative
fluorescent units (RFUs) were measured at regular intervals over a
period of time (maximum 15 min). A plot of RFUs versus time
was made for each substrate with and without cathepsin S. The
slope of the plotted line gave Vmax/KM of each substrate for cathepsin
S. Using the RFUs per micromolar concentration for AMCA,
substrate concentration, and enzyme concentration determined by
E-64c titration for cathepsin S,12 kcat/KM was determined.
Assay Procedure for Inhibitors. The dissociation constants (Ki)
were determined from the IC50 values taken from plots of Vi/Vo
versus inhibitor concentration, where Vo is the velocity in the
absence of the inhibitor and Vi is the velocity with the inhibitor.
Because the substrate concentration is significantly below KM, the
IC50 values were converted to Ki by the equation Ki ) IC50 - Et/2,
where Et ) enzyme concentration.13 The cathepsin B concentration
in the assays was 2 nM, that of cathepsin K was 10 nM, that of
cathepsin L was 0.025 nM, and that of cathepsin S was 0.7 nM.
The following concentrations for the fluorogenic substrates were
used: (Cbz-Phe-Arg-AMC) 2.5 µM for cathepsins B and K, 0.3
µM for cathepsin L; (Cbz-Leu-Arg-AMC) 2.5 µM for cathepsin S.
Assays were conducted in duplicate with and without inhibitor at
five inhibitor concentrations to provide 10-90% enzyme inhibition.
In each well were placed 180 µL of enzyme solution and 10 µL of
a DMSO inhibitor solution. The resulting solutions were incubated
for 5 min at 37 °C, then 10 µL of the peptidyl-AMC substrate was
added, and generation of AMC was monitored over 5 min.
Crystallization and Data Collection. Cathepsin S was expressed
in the baculovirus system and its complex with inhibitor 2, produced
by adding 2 to the protein in a 2/1 molar ratio prior to concentration
of the protein to 15 mg/mL. A coarse screen of 480 crystallization
conditions was set up in low-profile 96-well Greiner plates at 4 °C
in a sitting drop format containing 250 nL of protein combined
Synthesis of [2-(4-Alkyl-2,3-difluorophenyl)phenoxy]acetal-
dehydes 12d and 12e. [2-(2,3-Difluoro-4-n-propylphenyl)phe-
noxy]acetaldehyde (12d). Diisobutylaluminum hydride solution
(1.0 M in toluene, 0.20 mL) was added dropwise over 5 min to
10d (44 mg, 0.130 mmol) in toluene (1.3 mL) at -78 °C, and the
mixture was stirred for 1 h at the same temperature. The mixture
was quenched with AcOH (0.12 mL), and then saturated aqueous
sodium tartrate solution (8 mL) was added to it. After being stirred
for 1.5 h at ambient temperature, the resultant mixture was extracted
with EtOAc (50 mL). The extracts were washed with brine (10
mL), dried with anhydrous Na2SO4, filtered, and concentrated in
vacuo. The crude product was purified by silica gel column
chromatography (hexane/EtOAc ) 2/1 f 1/1) and then by reversed-
phase HPLC (MeCN/H2O/TFA) to provide the title compound (12
mg, 31%, 2/3 mixture of the aldehyde and its hydrate) as a colorless
gum: 1H NMR (400 MHz, CD3CN/D2O ) 3/1) δ 0.75-0.90 (m,

Unprecedented Inhibitors of Cathepsin S by SAS
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 11 2699
(4) Patterson, A. W.; Wood, W. J. L.; Hornsby, M.; Lesley, S.; Spraggon,
G.; Ellman, J. A. Identification of selective, nonpeptidic nitrile
inhibitors of cathepsin S using the substrate activity screening method.
J. Med. Chem. 2006, 49, 6298-6307.
(5) Reiffenrath, V.; Krause, J. Process for the Preparation of 2,3-
Difluorobenzenes. German Patent Application DE 88-3807910
19880310, 1988.
(6) Dantlgraber, G.; Shen, D.; Diele, S.; Tschierske, C. Antiferroelectric
switchable mesophases of nonchiral bent-core liquid crystals contain-
ing fluorinated central cores. Chem. Mater. 2002, 14, 1149 -1158.
(7) Tyndall, J. D. A.; Nall, T.; Fairlie, D. P. Proteases universally
recognize beta strands in their active sites. Chem. ReV. 2005, 105,
973-1000.
(8) Palmer, J. T.; Hirschbein, B. L.; Cheung, H.; McCarter, J.; Janc, J.
W.; Yu, Z. W.; Wesolowskib, G. Keto-1,3,4-oxadiazoles as cathepsin
K inhibitors. Bioorg. Med. Chem. Lett. 2006, 16, 2909-2914.
(9) 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. Design and synthesis of arylaminoethyl
amides as noncovalent inhibitors of cathepsin S. Part 1. Bioorg. Med.
Chem. Lett. 2005, 15, 4979-4984.
(10) Tully, D. C.: Liu, H.; Chatterjee, A. K.; Alper, P. B.; Epple, R.;
Williams, J. A.; Roberts, M. J.; Woodmansee, D. H.; Masick, B. T.;
Tumanut, C.; Li, J.; Spraggon, G.; Hornsby, M.; Chang, J.; Tuntland,
T.; Hollenbeck, T.; Gordon, P.; Harris, J. L.; Karanewsky, D. S.
Synthesis and SAR of arylaminoethyl amides as noncovalent inhibi-
tors of cathepsin S: P3 cyclic ethers. Bioorg. Med. Chem. Lett. 2006,
16, 5112-5117.
(11) Maly, D. J.; Leonetti, F.; Backes, B. J.; Dauber, D. S.; Harris, J. L.;
Craik, C. S.; Ellman, J. A. Expedient solid-phase synthesis of
fluorogenic protease substrates using the 7-amino-4-carbamoylm-
ethylcoumarin (ACC) fluorophore. J. Org. Chem. 2002, 67, 910-
915.
(12) Barrett, A. J.; Kembhavi, A. A.; Brown, M. A.; Kirschke, H.; Knight,
C. G.; Tamai, M.; Hanada, K. L-trans-Epoxysuccinyl-leucylamido-
(4-guanidino)butane (E-64) and its analogs as inhibitors of cysteine
proteinases including cathepsins B, H, and L. Biochem. J. 1982, 201,
189-198.
(13) Cha, S.; Agarwal, R. P.; Parks, J. Tight-binding inhibitors-II: non-
steady state nature of inhibition of mild xanthine oxidase by
allopurinol and alloxanthine and of human erythrocytic adenosine
deaminase by coformycin. Biochem. Pharmacol. 1975, 24, 2187-
2197.
with an equal volume of crystallization solution. Small crystals (∼20
µm) appeared after 1-2 weeks in a solution of 0.1 M sodium
acetate, 20% Peg-8000, and 2.0 M LiSO4. Data were collected at
beamline 5.0.3 of the Advanced Light Source (ALS) in Berkeley,
CA. The crystal form belongs to space group P4122 and diffracted
to 1.6 Å resolution (dataset statistics are presented in Table 1). This
tetragonal crystal form was isomorphous with a previously deter-
mined crystal form.9,3 The structure was then iteratively refined
until convergence with Coot14 and Refmac5;15 all other crystal-
lographic manipulations were carried out with the CCP4 program
package.16 The refined structure exhibited excellent geometry with
no residues in disallowed regions of the Ramachandran plot (Table
1). Excellent agreement was observed between the conformations
of inhibitor 2 in the two molecules of the asymmetric unit although
the presence of an alternative conformation of Cys 25 of molecule
B and less defined electron density of the compound suggest that
inhibitor 2 was not at full occupancy within this monomer. The
structure was deposited with PDB code 2OP3.
Acknowledgment. J.A.E. gratefully acknowledges the NIH
(Grant GM54051) for support of this work. H.I. and H.T.
gratefully acknowledge the support of Daiichi Pharmaceuticals
and Sankyo. We also thank Andreas Kreusch, Mike Didonato,
and Christian Lee for help with data collection and Peter Schultz
for continued support and encouragement. The work in this
paper is based on experiments conducted at beamline 5.0.3 of
the Advanced Light Source (ALS). The ALS is supported by
the Director, Office of Science, Office of Basic Energy Sciences,
Material Sciences Division of the U.S. Department of Energy,
under Contract No. DE-AC03-76SF00098 at Lawrence Berkeley
National Laboratory. We thank all of the staff at these beamlines
for their continued support.
Supporting Information Available: Purity data for substrates
8a-8f and inhibitors 12a and 12b. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Wood, W. J. L.; Patterson, A. W.; Tsuruoka, H.; Jain, R. K.; Ellman,
J. A. Substrate activity screening: A fragment-based method for the
rapid identification of nonpeptidic protease inhibitors. J. Am. Chem.
Soc. 2005, 127, 15521-15527.
(2) Salisbury, C. M.; Ellman, J. A. Rapid identification of potent
nonpeptidic serine protease inhibitors. ChemBioChem 2006, 7, 1034-
1037.
(3) (a) Leroy, V.; Thurairatnam, S. Cathepsin S inhibitors. Expert Opin.
Ther. Pat. 2004, 14, 301-311. (b) Yasuda, Y.; Kaleta, J.; Broemme,
D. The role of cathepsins in osteoporosis and arthritis: Rationale
for the design of new therapeutics. AdV. Drug DeliVery ReV. 2005,
57, 973-993. (c) Leung Toung, R.; Zhao, Y.; Li, W.; Tam, T. F.;
Karimian, K.; Spino, M. Thiol proteases: inhibitors and potential
therapeutic targets. Curr. Med. Chem. 2006, 13, 547-581.
(14) Emsley, P.; Cowtan, K. Coot: Model-building tools for molecular
graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004, 60,
2126-2132.
(15) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method. Acta
Crystallogr. 1997, D53, 240-255.
(16) Collaborative computing project, number 4. The CCP4 suite: Pro-
grams for protein crystallography, version 3.1. Acta Crystallogr. 1994,
D50, 760-763.
JM070111+