Identification of dipeptidyl nitriles as potent and selective inhibitors of cathepsin B through structure-based drug design

Article abstract of DOI:10.1021/jm010206q

Cathepsin B is a member of the papain superfamily of cysteine proteases and has been implicated in the pathology of numerous diseases, including arthritis and cancer. As part of an effort to identify potent, reversible inhibitors of this protease, we examined a series of dipeptidyl nitriles, starting with the previously reported Cbz-Phe-NH-CH₂CN (19, IC₅₀ = 62 μM). High-resolution X-ray crystallographic data and molecular modeling were used to optimize the P₁, P₂, and P₃ substituents of this template. Cathepsin B is unique in its class in that it contains a carboxylate recognition site in the S₂′ pocket of the active site. Inhibitor potency and selectivity were enhanced by tethering a carboxylate functionality from the carbon α to the nitrile to interact with this region of the enzyme. This resulted in the identification of compound 10, a 7 nM inhibitor of cathepsin B, with excellent selectivity over other cysteine cathepsins.

Full text of DOI:10.1021/jm010206q

4524
J . Med. Chem. 2001, 44, 4524-4534
Id en tifica tion of Dip ep tid yl Nitr iles a s P oten t a n d Selective In h ibitor s of
Ca th ep sin B th r ou gh Str u ctu r e-Ba sed Dr u g Design
Paul D. Greenspan,* Kirk L. Clark, Ruben A. Tommasi, Scott D. Cowen, Leslie W. McQuire, David L. Farley,
J ohn H. van Duzer, Ronald L. Goldberg, Huanghai Zhou, Zhengming Du, J ohn J . Fitt, David E. Coppa,
Zheng Fang, William Macchia, Lijuan Zhu, Michael P. Capparelli, Robert Goldstein, Andrew M. Wigg,
J ohn R. Doughty, Regine S. Bohacek, and Ania K. Knap
Arthritis and Bone Metabolism Research, Novartis Pharmaceuticals Corporation, 556 Morris Avenue,
Summit, New J ersey 07901
Received May 9, 2001
Cathepsin B is a member of the papain superfamily of cysteine proteases and has been
implicated in the pathology of numerous diseases, including arthritis and cancer. As part of
an effort to identify potent, reversible inhibitors of this protease, we examined a series of
dipeptidyl nitriles, starting with the previously reported Cbz-Phe-NH-CH2CN (19, IC50
)
62 µM). High-resolution X-ray crystallographic data and molecular modeling were used to
optimize the P1, P2, and P3 substituents of this template. Cathepsin B is unique in its class in
that it contains a carboxylate recognition site in the S2′ pocket of the active site. Inhibitor
potency and selectivity were enhanced by tethering a carboxylate functionality from the carbon
R to the nitrile to interact with this region of the enzyme. This resulted in the identification of
compound 10, a 7 nM inhibitor of cathepsin B, with excellent selectivity over other cysteine
cathepsins.
In tr od u ction
Cathepsin B (cat B) is a lysosomal cysteine protease,
which belongs to the papain superfamily.1 It is an
ubiquitously expressed protein (in contrast to its more
highly regulated relatives, such as cat S or cat K)2 and
has long been thought to function primarily in normal
intracellular protein turnover.3 However, more recent
studies have implicated cat B in the pathology of a
number of important human diseases, including cancer4
and neurodegenerative disorders.5 Cat B has been
shown to be upregulated in patients with rheumatoid
arthritis,6 and components of the extracellular matrix
are known to be substrates for this enzyme.7 Unselective
cysteine cathepsin inhibitors have previously been
reported to be effective in animal models of rheumatoid
arthritis.8
The majority of cysteine protease inhibitors are pep-
tidic or peptidomimetic compounds in which the hydro-
lyzable amide is replaced by an electrophilic function-
ality. In this way, the catalytic thiol of the enzyme reacts
with the inhibitor to form a covalent complex.9 Until
recently, most potent cysteine protease inhibitors were
irreversible inhibitors, in which the electrophilic “war-
head” alkylates the enzyme, through nucleophilic dis-
placement or conjugate addition. Examples of this
approach include fluoromethyl ketones, acyloxymethyl
ketones, vinyl sulfones, or epoxysuccinates.10 Alterna-
tively, potent, reversible inhibition can be achieved
through highly electrophilic warheads such as aldehydes
or R-ketoamides, which form reversible, covalent bonds
to the active site thiol.10 Recently, researchers at Smith-
Kline Beecham have reported hydrazides11 and bis-R-
amidoketones12 as potent inhibitors of cat K, resulting
F igu r e 1. Mechanism of nitrile inhibition of cysteine pro-
teases.
from elegant structure-based optimization. These com-
pounds bind in a manner similar to that of aldehydes
and ketoamides.
This paper will describe our initial efforts toward
identification of potent peptidyl nitrile inhibitors of cat
B. Representatives of this class of compounds have long
been known to inhibit cysteine proteases.10 Nuclear
magnetic resonance (NMR) studies indicate that this
inhibition occurs through the formation of a thioimidate
intermediate (Figure 1) and that this process is fully
reversible.13 Mutation studies have suggested that this
intermediate is stabilized by a neighboring Gln resi-
due,14 which, along with the backbone NH of the
catalytic cysteine residue, forms the putative “oxyanion
hole” of the enzyme.15 However, there are no reports to
date of highly potent nitrile inhibitors of cysteine
proteases (IC50 or Ki , 1 µM) and very few of inhibition
of a cysteine cathepsin by this class of inhibitor.16-18 It
was apparent, therefore, that substantial optimization
of the template would be necessary to provide adequate
potency.
Ch em istr y
All of the compounds discussed are based on a
dipeptidyl nitrile scaffold. Compounds with no substitu-
ent R to the nitrile, such as 3, were prepared using the
procedure shown in Scheme 1. A Boc-protected amino
acid 2 was coupled to aminoacetonitrile under standard
conditions. After amine deprotection, acylation with the
* To whom correspondence should be addressed. Tel: (908)277-4775.
Fax: (908)277-4885. E-mail: paul.greenspan@pharma.novartis.com.
10.1021/jm010206q CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/14/2001

Dipeptidyl Nitriles as Potent and Selective Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26 4525
Sch em e 1. Representative Synthesis of Dipeptidyl
Nitriles with No R-Substituenta
To synthesize arylpropyl-tethered nitriles such as 14,
the diprotected propargyl glycine 11 was first prepared
as previously reported21 (Scheme 4). The terminal
acetylene was then coupled with methyl 3-bromoben-
zoate and subjected to alcalase hydrolysis to yield the
(S)-enantiomer of acid 12, in >97% ee. The free acid was
then converted to its corresponding primary amide and
deprotected, yielding 13. This intermediate was carried
on as before to yield, after ester deprotection,22 com-
pound 14.
Aryl ether-substituted amino nitrile 17 was prepared
from methyl 3-hydroxybenzoate 15 via O’Donnell type
alkylation23 of iodoether 16. Racemic 17 was employed
in this case, yielding 18 as a 1:1 mixture of isomers at
the P1 position (Scheme 5).
Resu lts a n d Discu ssion
a
(a) 3-Methylbenzyl bromide, NaH, DMF, room temperature,
As a starting point, we prepared compound 19 (Table
1), which was previously reported to be a sub-micromo-
lar inhibitor of bovine cat B.16 In our hands, against the
recombinant human enzyme, the compound was found
to be much less potent, with an IC50 of 62 µM. To
generate a molecular model of inhibitor binding, we
employed the coordinates of the structure of recombi-
nant rat cat B bound to the irreversible inhibitor Z-Arg-
Ser(OBn)-CMK.24 We were confident that the use of the
rat enzyme would not compromise our modeling work,
since the active site contained only two point muta-
tions: Ser175Thr (at the base of the S2 pocket, remote
to any simple aromatic P2 substituent) and Glu122Gly
(in a highly disordered portion of the occluding loop). It
was felt that the peptidyl nitrile inhibitor would bind
in an orientation similar to this chloromethyl ketone,
since they were both covalently linked peptidic inhibi-
tors. Because it had previously been established that
peptidyl nitriles bind through formation of a thioimidate
ester intermediate,13 we incorporated this linkage into
our enzyme-inhibitor model. Using a subset of the
enzyme (including residues ∼12 Å from the center of
the binding site), the model was subjected to Monte
Carlo docking using the McDock algorithm of the QXP
molecular modeling suite.25 During minimizations, the
side chains of Gln 23, Cys 29, Tyr 75, Glu 122, Glu 245,
and His 199 were allowed to be flexible, and the covalent
bond from Cys 29 to the inhibitor was maintained. The
predicted lowest energy conformation, depicted in Fig-
ure 2, was consistent with the mode of binding observed
in the crystal structure of rat cat B bound to Z-Arg-Ser-
(OBn)-CMK.24 The S2 and S3 subsites are occupied by
the Phe and Cbz groups, respectively, and hydrogen-
bonding interactions are formed from the P1 and P2
amide NHs and from the P2 carbonyl to the enzyme. If
one assumes that this is the correct binding mode, then
close inspection clearly reveals several structural as-
pects that can be exploited to improve potency in this
template. The most obvious of these are the suboptimal
occupancy of the S2 and S3 pockets of the enzyme and
complete absence of binding into the S′ side of the active
site.
then AcOH, reflux, then alcalase 2.4 L; (b) aminoacetonitrile‚HCl,
EDCI, HOBt, NMM, CH2Cl2; (c) HCOOH, then aqueous NaHCO3;
(d) Ph2COCl, NMM, CH2Cl2.
Sch em e 2. Synthesis of N-Substituted Nitrilea
a
(a) NaH, MeI, DMF; (b) HCl, EtOAc; (c) diphenylacetyl
chloride, NMM, CH2Cl2; (d) LiOH, THF, water; (e) aminoaceto-
nitrile‚HCl, EDCI, HOBt, NMM, CH2Cl2.
appropriate capping group yielded the desired product
3. Racemic Boc-protected amino acids that were not
commercially available were prepared using standard
aminomalonate alkylation chemistry, as shown in
Scheme 1.19 Where optically active amino acids were
required, the racemic Boc amino esters were resolved
via enzymatic hydrolysis20 to produce the (S)-enanti-
omer as the free acid, in >98% ee. In these cases, the
alkylation, malonate decarboxylation, and subsequent
enzymatic hydrolysis were all performed in one pot,
resulting in very efficient generation of the requisite
amino acids.
The N-alkylated nitrile 5 was prepared via the race-
mic ethyl ester 4 (Scheme 2). After N-methylation and
cleavage of the tert-butylcarbamate, the amino ester was
acylated, hydrolyzed, and coupled with aminoacetoni-
trile to yield 5.
In other cases, it was necessary to first prepare a
substituted aminonitrile fragment before assembling the
final product. Benzyloxymethyl-substituted amino ni-
triles, exemplified by compound 9, were prepared through
alkylation of the dianion of Boc L-serine with an ap-
propriately substituted benzylic halide, such as 7 (Scheme
3). The resulting amino acid was converted to primary
amide 8, which was subsequently dehydrated and
deprotected to yield 9. This aminonitrile was then
subjected to a sequence similar to that described previ-
ously. Pd(0)-catalyzed deprotection of the allyl ester was
accomplished in the final step to yield 10.
To optimize this inhibitor, we decided to retain the
peptidic template so as to preserve the key backbone
H-bond interactions with the enzyme. We then chose
to take a modular approach toward optimization, in
which substituents would be modified iteratively, based

4526 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26
Greenspan et al.
Sch em e 3. Preparation of Aminonitriles Containing a Benzyloxymethyl Substituenta
a
i
(a) K2CO3, acetone, reflux; (b) NaI, acetone; (c) Boc-serine, 2 NaH, DMF, 0 °C; (d) ClCO2 Bu, NMM, CH2Cl2, -15 °C, then NH3, room
temperature; (e) (COCl)2, DMF, pyridine; (f) HCOOH, then aqueous NaHCO3; (g) 2, EDCI, HOBt, NMM, CH2Cl2; (h) HCOOH, then
aqueous NaHCO3; (i) 2,4-difluorobenzoyl chloride, NMM, CH2Cl2; (j) morpholine, Pd(PPh3)4, THF.
Sch em e 4. Preparation of Arylpropyl-Substituted
Sch em e 5. Preparation of Aryloxyethyl Subsituted
Aminonitrilesa
Aminonitrilesa
a
(a) Methyl 3-bromobenzoate, CuI, (Ph3P)2PdCl2, TEA; (b)
i
alcalase 2.4 L, MeCN, 0.2 M NaHCO3; (c) ClCO2 Bu, NMM,
CH2Cl2, -15 °C, then NH3, room temperature; (d) H2, 10% Pd/C,
EtOH, 1 atm; (e) HCl, EtOAc; (f) 1, EDCI, HOBt, NMM, CH2Cl2;
(g) (c) (COCl)2, DMF, pyridine; (h) HCOOH, then aqueous NaH-
CO3; (i) 2,4-difluorobenzoyl chloride, NMM, CH2Cl2; (j) potassium
trimethylsilanoate, CH2Cl2.
a
(a) 1,2-Dibromoethane, K2CO3, acetone, reflux; (b) NaI, ace-
tone, reflux; (c) N-(diphenylmethylene)aminoacetonitrile, NaN-
(TMS)2, THF, -78 °C to room temperature; (d) 1 N HCl, Et2O; (e)
1, EDCI, HOBt, NMM, CH2Cl2; (f) HCOOH, then aqueous NaH-
CO3; (g) diphenylacetyl chloride, NMM, CH2Cl2; (h) potassium
trimethylsilanoate, THF.
on structural information. Starting with the P3 acyl
substituent (Table 1), it was observed that the CBz
phenyl group of compound 19 was tethered too far from
the P2-P3 amide to interact optimally with Tyr 75,
which is the most prominent feature in the S3 pocket.
Removal of the oxygen from the CBz group to reduce
the phenyl tether length resulted in a modest increase
in potency (20). However, a ∼100-fold improvement was
observed upon replacing the phenylacetyl group with a
diphenylacetyl to further occupy this hydrophobic pocket,
thus producing a 0.5 µM cat B inhibitor (21).
attention was turned to the P2 position. Examination
of the aromatic ring within the S2 binding site suggested
that there was room for an additional hydrophobic
moiety attached to the phenyl group. Because a sub-
stituent at the meta position appeared to be most
favored in the model, several 3-substituted phenylala-
nine-containing compounds were prepared, initially in
racemic form. 3-Iodo substitution led to a 4-fold en-
hancement in potency (22). Consistent with our model-
ing results, the 4-iodo compound was found to be
inactive (23). A survey of other small hydrophobic
groups indicated that the methyl substituent was
Before undertaking a substantial characterization of
the P3 substituent structure-activity relationship (SAR),

Dipeptidyl Nitriles as Potent and Selective Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26 4527
Ta ble 1. Dipeptidic Nitriles Containing No P1 Substituent
cmpd
19 PhCH2O
P3
P2
IC50 (nM)
Ph
Ph
Ph
62 000b
34 000b
496 ( 108
121 ( 38
>1000b
87 ( 26
130 ( 34
136 ( 26
45 ( 7.6
1900b
20 PhCH2
21 Ph2CH
22 Ph2CH
23 Ph2CH
24 Ph2CH
25 PH2CH
26 Ph2CH
3-iodophenyla
4-iodophenyl
3-methylphenyla
3-chlorophenyla
3-ethylphenyla
3-methylphenyl
3-methylphenyla
3-methylphenyl
3-methylphenyl
3-methylphenyl
3-methylphenyl
3
Ph2CH
27 Ph
28 n-Bu
29 4-chlorophenyl
30 2,4-difluorophenyl
31 3,4-dichlorophenyl
3000b
F igu r e 3. Overview of cat B-3 complex X-ray structure.
308 ( 35
591b
31.3 ( 1.3
72b
11.9 ( 2.3
32 2-fluoro-4-chlorophenyl 3-methylphenyl
34 Ph2CH
35 Ph2CH
36 Ph2CH
37 Ph2CH
3,5-dimethylphenyla
3-(NH2CH2)-5-Me-Pha 358 ( 8.5
3-amidinoPha
>1000b
652 ( 28
3-(5-methylpyridyl)a
a
b
The compound is racemic. n ) 1.
F igu r e 4. X-ray structure of compound 3 in cat B active site,
overlaid with modeling prediction generated from rat cat B
X-ray structure.
of 15.9% at a resolution of 1.9 Å. The asymmetric unit
in this crystal form contains three independent enzyme-
inhibitor complexes, which are nearly identical with
respect to the active site and inhibitor. Two of these
complexes contacted each other near their active sites
in a pseudo-C2 symmetric fashion. This proximity raised
some concern that crystal packing could be influencing
the bound inhibitor conformation (Figure 3). Fortu-
nately, however, the third enzyme in the asymmetric
unit was fully solvent exposed and served to confirm
the validity of the observed binding orientations. Figure
4 displays the experimental enzyme-inhibitor complex
overlaid with the lowest energy structure predicted
using molecular modeling. It is evident that there was
excellent agreement between the results of molecular
modeling and the results of X-ray crystallography, thus
giving us confidence that we could use our model in a
predictive manner. The crystal structure provided us
with the opportunity to critically examine the ligand’s
covalent bond to the active site cysteine as well as its
nonbonded interactions with the enzyme.
F igu r e 2. Compound 19 modeled into rat cat B active site.
marginally better than the others (24-26). Preparation
of the (S)-isomer of this compound identified 3 as a 45
nM inhibitor of cat B.
At this point, the P3 SAR was further explored. An
unsubstituted benzamide (27, a racemate) showed only
weak activity in comparison to 3, as did an alkyl
substituent (28), again presumably due to inadequate
interaction with Tyr 75. However, incorporation of a
para-chloro substituent (29) led to a substantial increase
in potency over 27. Additional incorporation of a 3-chloro
or 2-fluoro (31 and 32, respectively) led to compounds
with potencies similar to or better than 3, further
highlighting the importance of an appropriate hydro-
phobic aromatic substituent in this position.
Cocrystallization experiments were then carried out
with recombinant human cat B and compound 3. Dif-
fraction data were collected at cryogenic temperatures,
and the crystal structure was fully refined to an R factor

4528 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26
Greenspan et al.
One of the most intriguing observations was that the
positioning of the 3-methyl group on the P2 phenylala-
nine within the cat B active site was ambiguous, since
two of the complexes in the unit cell showed methyl
group density on opposite sides of the phenyl ring, and
the third was not definitive. This led us to prepare the
dimethylated compound 34 (racemate), which was found
to be 4-fold more potent than 3.
The S2 pocket of cat B also contains a glutamate
residue (Glu 245), which has previously been demon-
strated to provide the enzyme with the ability to
recognize substrates with positively charged P2 resi-
dues.26 The crystal structure shows clearly that this
carboxylate residue is able to rotate out of the way of
an aromatic substituent in order to avoid an unfavorable
interaction. We were hopeful that we could capitalize
on this structural feature by incorporating positive
charge into the P2 substituent. Unfortunately, we were
unsuccessful in this endeavor, as inhibitors with an
aminomethyl substituent (35), an amidino group (36),
or a 3-pyridylalanyl P2 moiety (37) were all less potent
than the 3-methylphenylalanyl. There is currently no
satisfactory explanation for the failure of this approach,
although it does serve to highlight the sometimes
empirical nature of structure-based drug design.
F igu r e 5. Electron density map of cat B-3 structure il-
lustrating the mode of interaction of the nitrile functionality
with the catalytic site of the enzyme. Fo-Fc electron density
map with inhibitor and Cys 29’s CB and SG atoms omitted
from Fc. Inset panel has been rotated to better show the P2
and P3 ring systems.
The inhibitors thus far discussed bind solely into the
S2 and S3 pockets of the enzyme active site, assuming
that the observed binding mode for inhibitor 3 is
unchanged throughout the series. If this is correct, then
the large S2′ portion of the enzyme is completely vacant.
This site is very important when designing inhibitors
of cat B, since this enzyme has an S′ pocket that is
unique among papain superfamily proteases. Other
cathepsins, such as cat L, S, and K, have relatively open
S′ regions, whereas in cat B this region is blocked by a
19-residue insertion.27 His 110 and His 111 in this
insertion have previously been shown to bind carboxy-
late groups,28 thus providing cat B with its unusual
C-terminal dipeptidyl exopeptidase activity.29 We sought
to exploit this binding pocket to further improve the
potency of the inhibitors.
F igu r e 6. N-substituted dipeptidyl nitriles (both inactive
against cat B at 10 µM).
The high-resolution crystal structure demonstrated
conclusively that the nitrile inhibitor binds through
formation of a thioimidate ester with Cys 29 of the cat
B active site. The thioimidate nitrogen was clearly
planar and sp2 hybridized. In addition, this structure
confirms the role of the backbone NH of Cys 29 and the
side-chain amide of Gln 23 in stabilization of the
intermediate (Figure 5) with hydrogen bond distances
of 3.1 and 3.0 Å, respectively.
A limitation of a nitrile-based inhibitor template is
the fact that the nitrile is a terminating functionality,
which, unlike a ketone or hydrazide, for example, only
allows for elaboration in one direction. However, ex-
amination of the enzyme-inhibitor crystal structure led
us to hypothesize that the histidine region of S′ could
be accessed via P1 substitution. Previous reports have
indicated that O-benzyl serine is tolerated in the P1
position of cat B inhibitors.30 In our series, incorporation
of this substituent led to a 5-fold enhancement in
binding (38, Table 2), as compared to compound 3.
Modeling of compound 38 into the cat B active site
shows that the phenyl group is likely to be positioned
into a shallow hydrophobic pocket. This prediction is
consistent with the X-ray structure of the chloromethyl
ketone discussed previously.24 In this orientation, it
appeared that placement of a carboxylate in the meta-
position of the pendant aromatic ring would allow for a
salt-bridging interaction with one of the histidines of
the occluding loop (Figure 7). Indeed, this carboxylated
compound (39) was found to be highly potent, with an
IC50 of 1.8 nM. The importance of the placement of the
carboxylate was confirmed when the para-carboxylate
The backbone hydrogen-bonding pattern predicted
through modeling was observed in the crystal structure.
This result was further verified by the observation that
compounds methylated at either NH had no measurable
cat B activity (5 and 33, Figure 6).
As anticipated, one phenyl group of the P3 diphenyl-
acetyl moiety appears to form a favorable edge-to-face
interaction with Tyr-75, and its position is well-defined
by the electron density in the crystal structure. The
weak electron density for the second P3 aromatic ring
indicates that its orientation is not well-defined (Figure
5); however, the SAR data indicate that this group is
essential for high potency. The high potency seen with
other hydrophobic substituents at P3 (compounds 31 and
32) indicates that this result may simply be due to
increased desolvation energy.

Dipeptidyl Nitriles as Potent and Selective Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26 4529
Ta ble 3. Selectivity of Representative Nitriles
Ta ble 2. Dipeptidic Nitriles with P1 Substitution
cmpd
cat B IC50 (nM)
cat L IC50 (nM)
cat S IC50 (nM)
20
21
3
39
10
34 000a
170b
226a
496 ( 108
45 ( 7.6
1.8 ( 0.75
6.8 ( 0.6
31 ( 1.6
6 ( 3.6
20 ( 7.6
554b
48 ( 4.5
88b
46 ( 5.0
937b
a
b
n ) 1. n ) 2.
We discovered that with the 3-carboxy-O-benzylserine
P1 substituent in place, dramatic potency enhancement
was observed regardless of the P3 subsituent. In fact,
substituents that afforded little potency in compounds
with only hydrogen at P1 were now nanomolar inhibitors
(compare 27 with 48, 28 with 49, and 30 with 10). As
before, the benzyl ether tether was found to be optimal
(10 and 14), and the carboxylate functionality was
shown to improve potency (50). With the more potent
2-fluoro-4-chlorobenzoyl group at P3, we again saw
potency improvement upon incorporation of the O-
benzylserine P1 substituent and then upon meta-car-
boxylation (28, 46, and 47).
Because our structure-based approach was directed
specifically at cat B, we were optimistic that we could
achieve selectivity over other cysteine proteases. Table
3 lists potency data obtained for representative nitriles
against cathepsins B, L, and S. Compound 20, a very
weak cat B inhibitor, was found to be relatively potent
against cat L and S. Incorporation of the diphenylacetyl
group at P3 (21) improved activity against all three
enzymes, although the effect was more pronounced for
cat B than the others. Subsequent optimization of the
P2 position (3) provided a compound that was modestly
selective for cat B over cat S but not cat L. Incorporation
of the tethered benzoate substituent at P1 yielded a
compound with selectivity for cat B over both cat L and
cat S (39). However, the most dramatic effect was
observed upon replacement of the diphenylacetyl sub-
stituent of compound 39 with a smaller group, such as
in compound 10. As discussed previously, this modifica-
tion has a modest effect on cat B potency. However, we
see a substantial loss of activity against both cat L and
cat S. This results in ∼100-fold selectivity for cat B over
the other enzymes.
cmpd
38 Ph2CH
P3
X
Y
Z
IC50 (nM)
O
O
O
O
O
CH2
H
10.2 ( 0.6
39 Ph2CH
40 Ph2CH
41 Ph2CH
42 Ph2CH
43 Ph2CH
18 Ph2CH
44 Ph2CH
45 Ph2CH
CH2 3-COOH 1.8 ( 0.75
CH2 4-COOH 30.7 ( 3
CH2 3-CH2OH 29b
CH2 3-COOMe 47b
CH2 CH2 3-COOH 18b
CH2
CtC
O
O
3-COOH 138b,c
3-COOH 5.6 ( 2.7
CH2 5-COOHa 5.0 ( 1.5
CH2
CH2 3-COOH 2.0 ( 0.1
CH2 3-COOH 9.4 ( 1.6
CH2 3-COOH 35.7 ( 7.2
CH2 3-COOH 6.8 ( 0.6
46 2-fluoro-4-chlorophenyl O
47 2-fluoro-4-chlorophenyl O
48 Ph
H
16.9 ( 0.4
O
O
O
49 n-Bu
10 2,4-difluorophenyl
14 2,4-difluorophenyl
50 2,4-difluorophenyl
CH2 CH2 3-COOH 110 ( 17
O
CH2
H
42, 43d
a
b
Substituent is 5-carboxy-2-furyl. n ) 1. c At P1, there is a 1:1
d
mixture of diastereomers. n ) 2.
These compounds were confirmed to be reversible
inhibitors of cat B. This was demonstrated by forming
a complex by preincubating the enzyme in the presence
of excess inhibitor (compounds 10 and 39) and measur-
ing enzyme activity after dialysis and dilution into the
assay mixture containing substrate. In both cases,
enzyme activity was fully restored.
F igu r e 7. X-ray structure of compound 35 modeled into cat
B active site.
isomer (40) was found to be roughly 15-fold less potent
than 39. The fact that 40 is 3-fold less potent than 38
may be explained by a decrease in desolvation energy
for compound 40, since the carboxylate functionality
should increase its water solubility. The requirement
for an acidic functionality in the meta-position was
confirmed by the reduced potency of the primary alcohol
and methyl ester (41 and 42). Additional tethers were
also examined but were found to be less optimal (18 and
43-44). The 2,5-disubstituted furan was found to be a
suitable phenyl isostere (45).
Con clu sion
The work described herein represents a successful
application of the principles of structure-based drug
design to the optimization of a series of peptidyl nitrile
inhibitors of cat B. Through systematic, iterative opti-
mization of the substituents at the P3, P2, and P1
positions, we were able to achieve a ∼1000-fold im-
provement in potency from our original lead compound.
High-quality X-ray crystallography of the complex of one
of our first nanomolar inhibitors (3) bound to the
enzyme confirmed the hypothesized mode of nitrile
inhibition and enabled further optimization of the
With our optimized P1 substituent in hand, we were
now able to reexamine the SAR of the P3 substituent.

4530 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26
Greenspan et al.
ethyl ether was added and stirring continued for 1 h. The
reaction mixture was extracted twice with ethyl ether. The
aqueous layer was cooled to 0 °C and acidified to pH 3 with 6
N HCl. The mixture was extracted with ethyl acetate (3 × 100
mL), which was washed with water and brine, dried (MgSO4),
and evaporated and chromatographed (silica gel, 5% MeOH,
CH2Cl2) to yield 2 (3.55 g, 63% yield). [R]D + 17.46° (CHCl3).
1H NMR (300 MHz, CDCl3): δ 12.60 (1H, s); 6.90-7.20 (5H,
m); 4.03-4.12 (m, 1H); 2.96 (1H, dd, J ) 13.5, 4.5 Hz); 2.77
(1H, dd, J ) 13.5, 10.5 Hz); 2.26 (3H, s); 1.32 (9H, s).
template. The value of the molecular modeling tools
employed in this study was demonstrated by the excel-
lent agreement of modeling results with X-ray data. Of
particular note was the achievement of high potency and
selectivity through optimization of S′ site binding
interactions. A carboxylate group tethered from the P1
position was oriented to make a salt-bridge interaction
with the histidine-containing loop of the S2′ site. Even
in the absence of an X-ray structure of compound 39 or
one of its analogues bound to cat B, the SAR data
provides a high degree of confidence that our predicted
structural model was reflective of the actual inhibitor
binding mode. Additional support for this conclusion can
be found in the high selectivity observed for the inhibi-
tion of cat B over other cathepsins. In conclusion, we
utilized structural information to guide us in the
optimization of a high-micromolar lead compound (19)
into selective, low nanomolar cat B inhibitors. For
example, compound 10, while exhibiting low nanomolar
inhibition of cat B, was only weakly potent against
cathepsins L and S. Because of the absence of structural
data on cathepsins S and L, it is not possible to discuss
the structural basis of isozyme selectivity.
N-Cya n om eth yl-NR-(d ip h en yla cetyl)-3-m eth ylp h en yl-
a la n in a m id e (3). To a solution of 2, (5.0 g, 17.92 mmol),
HOBT (3.29 g, 21.5 mmol), and N-methylmorpholine (7.9 mL,
7.25 g, 71.7 mmol) in CH2Cl2 (100 mL) was added EDCI (5.15
g, 26.9 mmol) in one portion, and the solution was stirred at
room temperature for 16 h. The solution was then diluted with
CH2Cl2 (150 mL), washed with HCl (2 × 250 mL), saturated
aqueous NaHCO3 (250 mL), and brine (100 mL), dried (Mg-
SO4), and evaporated to yield 5.37 g of a white solid, which
was taken up in formic acid (64 mL) and stirred at room
temperature for 6 h. The formic acid was then evaporated
under high vacuum at room temperature. The residue was
dissolved in water (25 mL), basified with saturated aqueous
NaHCO3, and extracted with ethyl acetate (3 × 80 mL). The
combined organic layers were then washed with brine (50 mL)
and evaporated to yield N-cyanomethyl-NR-(tert-butoxycarbo-
nyl)-3-methylphenylalaninamide as a brown oil. (1.86 g, 48%
for 2 steps). 1H NMR (300 MHz, CDCl3): δ 7.90 (1H, br s);
7.22 (1H, t, 7.5 Hz); 6.96-7.14 (4H, m); 4.20 (2H, d, J ) 6.0
Hz); 3.67 (1H, dd, J ) 9.0, 4.1 Hz); 3.22 (1H, dd, J ) 14.0, 4.1
Hz); 2.71 (1H, dd, J ) 14.0, 9.0); 2.33 (3H, s). To a solution of
this material (0.21 g, 0.92 mmol) and N-methylmorpholine
(0.30 mL, 2.8 g, 2.76 mmol) in CH2Cl2 (10 mL) was added
diphenylacetyl chloride (0.20 g, 0.92 mmol) in one portion.
After it was stirred for 2 h, the solution was diluted with CH2-
Cl2 (40 mL) and then washed with 1 N HCl (50 mL) and brine
(50 mL) and then evaporated and chromatographed (silica,
50-75% ethyl acetate/hexane) to yield 3 as a white solid, mp
Exp er im en ta l Section
Deter m in a tion of In h ibition of Ca th ep sin B Activity.
Recombinant human cat B protein was expressed in baculovi-
rus and purified as described elsewhere.31 To a microtiter well
was added 100 uL of a 20 µM solution of inhibitor in assay
buffer (0.1 M, pH 5.8, phosphate buffer containing ethylene-
diaminetetraacetic acid (1.33 mM), dithiothreitol (2.7 mM), and
Brij (0.03%)) followed by 50 µL of a 6.4 mM solution of Z-Arg-
Arg-AMC substrate (Peptides International) in assay buffer.
After the solution was mixed, 50 uL of a 0.544 nM solution of
recombinant human cat B in assay buffer was added to the
well, yielding a final inhibitor concentration of 10 µM. Enzyme
activity was determined by measuring fluorescence of the
liberated aminomethylcoumarin at 460 nM using 355 nM
excitation, at 20 min. Percent enzyme inhibition was deter-
mined by comparison of this activity to that of a solution
containing no inhibitor. Compounds that showed >50% inhibi-
tion at 10 µM were subsequently subjected to a dose response
curve analysis, and IC50’s were determined.
1
169-170 °C (0.32 g, 85% yield). H NMR (300 MHz, CDCl3):
δ 7.22-7.38 (8 H, m); 6.98-7.18 (6H, m); 6.85 (1H, br s); 6.80
(1H, d, J ) 7.1 Hz); 6.19 (1H, d, J ) 6.5 Hz); 4.87 (1H, s); 4.83
(2H, d, J ) 6.1 Hz); 3.79 (1H, dd, J ) 17.3, 6.0 Hz); 3.63 (1H,
dd, J ) 17.3, 6.0 Hz); 3.02 (1H, dd, J ) 14.03, 6.5 Hz); 2.89
(1H, dd, J ) 14.0, 8.0 Hz); 2.27 (3H, s). MS (m/z): (M + H)
412. Anal. Calcd for C26H25N3O2: C, H, N.
Ch em istr y. Melting points (mp) were determined on a
Thomas-Hoover melting point apparatus and are uncorrected.
1H NMR spectra were recorded on a Bruker DPX-300 spec-
trometer. Mass spectroscopy (MS) analyses were perfomed on
a Micromass Platform electrospray mass spectrometer, using
positive ionization. Microanalyses were performed at Robert-
son Laboratory, Inc., Madison, NJ . All organic solvents used
were of anhydrous grade. All reactions were run under a
positive pressure of nitrogen unless otherwise stated. Chro-
matographic separations were performed with silica gel 60.
Compounds were dried by suspension of anhydrous MgSO4,
followed by vacuum filtration.
N-(ter t-Bu t oxyca r b on yl)-3-m et h ylp h en yla la n in e (2).
To 1.1 g of sodium hydride (1.82, 60% dispersion in mineral
oil, washed with hexane, 46 mmol) suspended in anhydrous
dimethylformamide (DMF, 125 mL), at room temperature
under a N2 atmosphere, was added diethyl tert-butoxycarbo-
nylaminomalonate (1, 11.1 g, 40 mmol, 10.2 mL) dropwise.
After 2 h at ambient temperature, m-methylbenzylbromide (7.5
g, 40 mmol, 5.5 mL) was added dropwise, and the reaction was
stirred at ambient temperature for 2 h. The reaction mixture
was adjusted to pH 5 with acetic acid (0.4 mL), and 2.0 mL of
water was added. It was stirred in a 150 °C bath under gentle
reflux for 16 h. After the solution cooled to room temperature,
NaHCO3 (4.55 g, 54 mmol) dissolved in water (250 mL) was
added followed by 2.5 mL of Alcalase 2.4 L (Novo Nordisk
BioChem, ID no. 119878). After 2.5 h at 25-27 °C, 150 mL of
Compounds 19-32 and 34-37 were prepared in a manner
similar or identical to that for compound 3.
N -Cya n om e t h yl-N r-m e t h yl-N r-(d ip h e n yla ce t yl)-3-
m eth ylp h en yla la n in a m id e (5). To a solution of NaH (60%
in mineral oil, 39 mg, 0.98 mmol) in DMF (2 mL) was added
a solution of 4 (0.30 g, 0.98 mmol) in DMF (3 mL) dropwise,
over 5 min. The mixture was then stirred at room temperature
for 10 min, after which time iodomethane (0.14 g, 61 uL, 0.98
mmol) was added in one portion. After it was stirred for 16 h,
the solution was quenched with water (5 mL) and poured into
saturated aqueous LiCl (50 mL). The mixture was extracted
with ethyl acetate (3 × 20 mL), and the combined organic
phases were washed with saturated aqueous LiCl (2 × 50 mL),
dried over MgSO4, and evaporated. The residue was chromato-
graphed (silica, 10% ethyl acetate/hexane) to yield N-(tert-
butoxycarbonyl)-N-methyl-3-methylphenylanine ethyl ester as
1
a white solid (0.19 g, 60% yield). H NMR (300 MHz, CDCl3,
mixture of rotamers): δ 7.12-7.21 (1H, m); 6.91-7.06 (3H,
m); 4.93 (0.5H, dd, J ) 11.4, 4.5 Hz); 4.56 (0.5H, dd, J ) 12.2,
4.5 Hz); 3.72 (1.5H, s); 3.71 (1.5H, s); 3.19-3.31 (1H, m); 2.90-
3.04 (1H, m); 2.71 (1.5H, s); 2.70 (1.5H, s); 2.32 (3H, s); 1.39
(4.5H, s); 1.32 (4.5H, s). This material was dissolved in ethyl
acetate (15 mL) and cooled to 0 °C. HCl was then bubbled into
the solution for 1 min. The solution was warmed to room
temperature, stirred for 0.5 h, and then evaporated to dryness.

Dipeptidyl Nitriles as Potent and Selective Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26 4531
The residue was taken up in CH2Cl2 (15 mL), and N-
methylmorpholine (0.19 mL, 0.178 g, 0.18 mmol) was added,
followed by diphenylacetyl chloride (0.14 g, 0.59 mmol), and
the solution was stirred at room temperature for 2 h, after
which time it was washed with 1 N HCl (50 mL) and brine
(50 mL), dried (MgSO4), and evaporated to yield 0.21 g of
material. This residue was next dissolved in a 2:1 tetrahydro-
furan (THF)/methanol mixture (3 mL), followed by addition
of 1 N aqueous NaOH (0.56 mL, 0.56 mmol). After the solution
was stirred for 1 h at room temperature, the THF and
methanol were evaporated, and the residue was acidified with
1 N HCl. The aqueous phase was then extracted with ethyl
acetate (3 × 30 mL). The organic phases were washed with
water (50 mL) and brine (50 mL), dried (MgSO4), and evapo-
rated to yield 0.20 g of crude product, which was taken up in
CH2Cl2 (10 mL), followed by the addition of N-methylmorpho-
line (0.29 mL, 0.26 g, 2.62 mmol) and HOBt (88 mg, 0.58 mmol)
and EDCI (0.15 g, 0.79 mmol). After it was stirred for 16 h at
room temperature, the solution was washed with 1 N HCl (50
mL), saturated aqueous NaHCO3 (50 mL), and brine (50 mL),
dried (MgSO4), evaporated, chromatographed (5% diethyl
ether/CH2Cl2), and recrystallized (ethyl acetate/hexane) to
yield 5 as a white solid, mp 153-154 °C (84 mg, 32% yield for
was stirred for 15 min, ammonia gas was bubbled into the
solution at a moderately vigorous rate for 15 min at -10 °C.
The solution was then warmed to room temperature and
stirred for 30 min. The reaction mixure was cooled to 0 °C,
and 1 N HCl (500 mL) was added. The organic phase was
washed with 1 N HCl (2 × 300 mL), washed with saturated
NaHCO3 (500 mL), dried (MgSO4), and then evaporated in
vacuo to yield 8 as a clear oil (14.53 g, 21% for 2 steps). 1H
NMR (300 MHz, CDCl3): δ 7.95-8.03 (2H, m); 7.39-7.55 (2H,
m); 6.42 (1H, br s); 5.95-6.12 (1H, m); 5.62 (1H, br s); 5.40
(1H, dq, J ) 17.1, 1.5 Hz); 5.29 (1H, dq, J ) 10.4, 1.5 Hz);
4.82 (1H, dt, J ) 5.8, 1.5 Hz); 4.64 (1H, d, J ) 11.9 Hz); 4.56
(1H, d, J ) 11.9 Hz); 3.93 (1H, dd, J ) 9.2, 3.7 Hz); 3.60 (1H,
dd, J ) 9.2 Hz, 6.4 Hz), 1.44 (9H, s).
(s)-2-Am in o-3-(3-a llyloxyca r b on ylp h e n yl)m e t h oxy-
p r op ion itr ile (9). To dry DMF (50 mL) at 0 °C was added
oxalyl chloride (9.55 g, 6.56 mL, 75.24 mmol) slowly, via
syringe. The mixture was then stirred at 0 °C for 5 min, after
which time pyridine (12.2 mL, 150.48 mmol) was added in one
portion, followed by 8 (14.22 g, 37.62 mmol) in DMF (50 mL).
The mixture was stirred at 0 °C for 45 min, then diluted with
ethyl acetate (600 mL), washed with saturated aqueous LiCl
(3 × 600 mL), and dried over MgSO4. Evaporation of solvent,
followed by chromatography (silica, 35% ethyl acetate/hexane),
yielded crude (s)-2-(tert-butoxycarbonylamino)-3-(3-allyloxy-
carbonylphenyl)methoxypropionitrile as a clear oil (11.78 g,
87% yield). 1H NMR (300 MHz, CDCl3): δ 8.00-8.05 (2H, m);
7.60 (1H, d, J ) 8.9 Hz); 7.47 (1H, t, J ) 7.5 Hz); 5.97-6.12
(1H, m); 5.42 (dq, 1H, J ) 17.4, 1.5 Hz); 5.30 (1H, dq, J )
10.5, 1.5); 4.84 (2H, dt, J ) 5.7, 1.5 Hz); 4.77 (1H, br s); 4.67
(2H, s); 3.74 (1H, dd, J ) 9.5, 3.7 Hz); 3.67 (dd, 1H, J ) 9.5,
4.2 Hz); 1.45 (9H, s). This material was dissolved in formic
acid (125 mL), and the solution was stirred at room temper-
ature for 6 h, after which time the formic acid was evaporated
at 25 °C. The residue was then dissolved in water (50 mL),
basified with saturated aqueous NaHCO3, and extracted with
ethyl acetate (3 × 150 mL). The combined organic layers were
then washed with water (50 mL) and brine (50 mL), dried
(MgSO4), and evaporated to yield 9 as a light yellow oil, which
was utilized without further purification (7.73 g, 91% yield).
1H NMR (300 MHz, CDCl3): δ 8.00-8.05 (2H, m); 7.57-7.62
(1H, m); 7.46 (1H, t, J ) 6.9 Hz); 5.98-6.12 (1H, m); 5.42 (1H,
dq, J ) 17.4, 1.5 Hz); 5.30 (J ) 10.5, 1.5 Hz); 4.83 (2H, d, J )
5.7); 4.67 (2H, s); 3.89 (1H, br s); 3.68 (2H, dd, J ) 3.0, 1.5
Hz); 1.45 (9H, s).
1
4 steps). H NMR (250 MHz, CDCl3): δ 6.80-7.35 (14H, m);
5.38 (1H, dd, J ) 7.0, 5.7 Hz); 5.10 (1H, s); 3.94 (1H, dd, J )
17.8, 6.3 Hz); 3.63 (1H, dd, J ) 17.8, 5.7 Hz); 3.16 (1H, dd, J
) 12.0, 6.3 Hz); 2.97 (1H, dd, J ) 12.6, 6.3 Hz); 2.90 (3H, s);
2.28 (3H, s). MS (m/z): (M + H) 426. Anal. Calcd for
C
27H27N3O2: C, H, N.
Allyl 3-(Iod om eth yl)ben zoa te (7). A solution of 3-(chlo-
romethyl)benzoic acid 6 (50.0 g, 0.293 mol), potassium carbon-
ate (48.61 g, 0.352 mol), and allyl bromide (50.7 mL, 0.586
mol) in acetone (500 mL) was refluxed for 2 h, after which
time the solution was cooled to room temperature and filtered
through Celite. The filtrate was evaporated, and the residue
was chromatographed (silica, 5% ethyl acetate/hexane) to yield
54.74 g of a mixture of allyl 3-(chloromethyl)benzoate and allyl
3-(bromomethyl)benzoate as a clear oil. This was taken up in
acetone (500 mL), and sodium iodide (46.56 g, 0.311 mol) was
added in one portion. The mixture was stirred for 6.5 h, after
which time the mixture was filtered. The filtrate was evapo-
rated, and the residue was dissolved in diethyl ether (500 mL),
washed with water (1 × 200 mL), 5% sodium sulfite solution
(1 × 200 mL), and brine (1 × 200 mL), dried over magnesium
sulfate, and evaporated to yield 7 as a white solid, which was
used directly (66.66 g, 76% yield for 2 steps). 1H NMR (300
MHz, CDCl3): δ 8.06 (1H, t, J ) 1.8 Hz); 7.94 (1H, dt, J ) 7.9,
1.5 Hz); 7.58 (1H, dt, J ) 7.9, 1.5 Hz); 7.38 (1H, d, J ) 7.6
Hz); 5.95-6.12 (1H, m); 5.42 (1H, dq, J ) 17.1, 1.5 Hz); 5.31
(1H, dq, J ) 10.4, 1.5 Hz); 4.83 (2H, dt, J ) 6.8, 1.2 Hz); 4.48
(3H, s).
N-[2-[(3-Ca r boxyp h en yl)m eth oxy]-1(S)-cya n oeth yl]-3-
m et h yl-Nr-(2,4-d iflu or ob en zoyl)-L-p h en yla la n in a m id e
(10). To a solution of 2 (3.22 g, 11.54 mmol) and 9 (3.0 g, 11.54
mmol) in CH2Cl2 (100 mL) was added N-methylmorpholine (3.8
mL, 34.6 mmol), followed by HOAt (1.88 g, 13.8 mmol) and
EDCI (3.32 g,17.3 mmol), and the solution was stirred over-
night at room temperature The solution was then washed with
1 N HCl (100 mL), saturated NaHCO3 (100 mL), and brine
(100 mL), dried (MgSO4), evaporated, and chromatographed
(silica, 25% ethyl acetate/hexane) to yield N-[2-[(3-(allyloxy-
carbonyl)phenyl)methoxy]-1(S)-cyanoethyl]-3-methyl-NR-(tert-
butoxycarbonyl)-L-phenylalaninamide as a white solid (4.36 g,
72% yield). 1H NMR (300 MHz, CDCl3): δ 8.02 (1H, d, J ) 7.4
Hz); 7.96 (1H, s); 7.43-7.55 (2H, m); 7.26-7.22 (1H, m); 7.02
(3H, m); 6.56 (1H, d, J ) 8.6 Hz); 5.98-6.12 (1H, m); 5.42 (1H,
dd, J ) 17.3, 1.2 Hz); 5.30 (1H, dd, J ) 17.3, 1.0 Hz); 5.20
(1H, br s); 4.99-5.04 (1H, m); 4.84 (1H, dd, J ) 5.7 Hz); 4.59
(2H, s); 4.31 (1H, q, J ) 4.31); 3.68 (1H, dd, J ) 9.8, 3.7 Hz);
3.56 (1H, dd, J ) 9.8, 4.2 Hz); 3.01 (2H, d, J ) 6.01 Hz); 2.30
(3H, s); 1.39 (9H, s). This compound was taken up in formic
acid (50 mL) and stirred at room temperature for 6 h, after
which time the formic acid was evaporated at 25 °C. The
residue was then dissolved in water (50 mL), basified with
saturated aqueous NaHCO3, and extracted with ethyl acetate
(3 × 50 mL). The combined organic layers were then washed
with water (50 mL) and brine (50 mL), dried (MgSO4), and
evaporated to yield N-[2-[(3-(allyloxycarbonyl)phenyl)methoxy]-
1(S)-cyanoethyl]-3-methyl-L-phenylalaninamide as a clear oil,
which was utilized without further purification (3.5 g, 100%
O-[[3-(Allyloxycar bon yl)ph en yl]m eth yl]-N-(ter t-bu toxy-
ca r bon yl)-L-ser in a m id e (8). Sodium hydride (7.5 g, 60% in
mineral oil, 187 mmol) was washed with dry pentane (2 × 20
mL) to remove the mineral oil and then suspended in anhy-
drous DMF (150 mL). To this suspension at 0 °C was added
N-butoxycarbonyl-L-serine (19.2 g, 93.7 mmol) dropwise with
vigorous stirring. The mixture was stirred for an additional 5
min at 0 °C and then at room temperature for 30 min. The
solution was cooled back down to 0 °C, and a solution of 7 (28.3
g, 93.7 mmol) in DMF (150 mL) was added dropwise over 15
min. The mixture was then warmed to room temperature for
30 min. DMF was evaporated under high vacuum (bath
temperature < 40 °C), and the residue was diluted with water
(500 mL) and washed with ether (2 × 400 mL). The aqueous
layer was then acidified with 1 N HCl to pH 3 and extracted
with ethyl acetate (3 × 100 mL). The organic layers were
washed with brine (200 mL) and then dried (MgSO4) and
evaporated to yield crude O-[[3-(allyloxycarbonyl)phenyl]-
methyl]-N-(tert-butoxycarbonyl)-L-serine as a yellowish oil
(11.43 g). A solution of this material (22.6 g, 59.6 mmol) and
N-methylmorpholine (19.7 mL, 179 mmol) in CH2Cl2 (400 mL)
was cooled to -10 °C, and isobutyl chloroformate (8.52 mL,
65.6 mmol) was added dropwise over 10 min. After the solution

4532 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26
Greenspan et al.
1
yield). H NMR (300 MHz, CDCl3): 8.11-8.17 (1H, m); 8.00-
3.03 (2H, m); 1.46 (9H, s). Mosher ester formation of free amine
revealed >97% ee.
8.06 (2H, m); 7.42-7.59 (2H, m); 7.20 (1H, t, J ) 8.0 Hz); 7.97-
7.09 (3H, m); 5.97-6.11 (1H, m); 5.42 (1H, dd, J ) 17.3, 1.2
Hz); 5.30 (1H, dd, J ) 17.3, 1.0 Hz); 5.05-5.12 (1H, m); 4.83
(2H, dt, J ) 5.7, 1.5 Hz); 4.68 (1H, d, J ) 12.1); 4.62 (1H, d, J
) 12.1); 3.75 (1H, dd, J ) 9.4, 3.7 Hz); 3.66 (1H, dd, J ) 9.4,
3.8 Hz); 3.58 (1H, dd, J ) 9.8, 4.5 Hz); 3.20 (1H, dd, J ) 12.7,
4.2 Hz); 2.62 (1H, dd, J ) 12.7, 8.5 Hz); 2.31 (3H, s). To a
solution of this product (0.25 g, 0.59 mmol) in CH2Cl2 (20 mL)
was added N-methylmorpholine (0.18 g, 0.20 mL, 1.78 mmol),
followed by 2,4-difluorobenzoyl chloride (0.10 g, 0.073 mL, 0.59
mmol). After the solution was stirred for 2 h at room temper-
ature, the solution was washed with 1 N HCl (25 mL),
saturated NaHCO3 (25 mL), and brine (20 mL), dried over
MgSO4, evaporated, and chromatographed (silica, 1% methanol/
CH2Cl2) to yield N-[2-[(3-(allyloxycarbonyl)phenyl)methoxy]-
1(S)-cyanoethyl]-NR-(2,4-difluorobenzoyl)-L-3-methylphenyl-
alaninamide as a white solid (0.263 g, 79% yield). 1H NMR
(300 MHz, CDCl3): 7.95-8.05 (2H, m); 7.90 (1H, s); 7.15-7.50
(4H, m); 6.76-7.10 (6H, m); 5.97-6.11 (1H, m); 5.42 (1H, dd,
J ) 17.3, 1.2 Hz); 5.30 (1H, dd, J ) 17.3, 1.0 Hz); 4.97-5.05
(1H, m); 4.80-4.89 (3H, m); 4.54 (2H, m); 3.62 (1H, dd, J )
9.8, 3.3 Hz); 3.56 (1H, dd, J ) 9.8, 4.1 Hz); 3.20 (1H, dd, J )
13.6, 6.4 Hz); 3.08 (1H, dd, J ) 13.6, 7.6 Hz); 2.30 (3H, s). A
solution of this product (0.43 g, 0.76 mmol) and morpholine
(0.66 g, 0.66 mL, 7.62 mmol) in THF (25 mL) was deoxygenated
with bubbling N2 for 10 min, followed by addition of Pd(PPh3)4
(0.088 g, 0.076 mmol), and the solution was stirred at room
temperature for 30 min. THF was evaporated, and the residue
was dissolved in ethyl acetate (50 mL), washed with 1 N HCl
(50 mL) and brine (30 mL), dried (MgSO4), evaporated, and
chromatographed (silica, 2-4% MeOH/CH2Cl2) to yield 10 as
an off-white solid, mp 126-128 °C (0.245 g, 60% yield). 1H
NMR (300 MHZ, CDCl3): 8.02 (1H, s); 7.94 (1H, d, J ) 7.6
Hz); 7.67-7.75 (1H, m); 7.59-7.64 (1H, m); 7.45 (1H, t, J )
7.9 Hz); 6.99-7.15 (6H, m); 5.05 (1H, t, J ) 5.3 Hz); 4.78 (1H,
t, J ) 6.5 Hz); 4.67 (2H, s); 3.74 (2H, d, J ) 5.3 Hz); 3.14 (1H,
dd, J ) 13.5, 6.4 Hz); 3.01 (1H, dd, J ) 7.9 Hz); 2.29 (3H, s).
MS (m/z): (M + NH4+) 539. Anal. Calcd for C28H24F2N3O5: C,
H, N.
(S)-2-Am in o-5-(3-ca r b om et h oxyp h en yl)p en t a n a m id e
(13). To a solution of 12 (24.2 g, 69.6 mmol) in EtOH (540 mL)
and THF (540 mL) was added 10% Pd/C (10.8 g), and the
solution was hydrogenated under atmospheric pressure for 2
h. Filtration through Celite followed by evaporation yielded
(S)-2-amino-5-(3-carbomethoxyphenyl)pentanoic acid as a clear
1
oil (24.38 g, 99%). H NMR: δ 7.81-7.90 (2H, m); 7.30-7.38
(2H, m); 5.03 (br d, J ) 8.3 Hz); 4.32-4.40 (1H, m); 3.90 (3H,
s); 2.63-2.72 (2H, m); 1.62-1.90 (4H, m); 1.45 (9H, s). To a
solution of this material and N-methylmorpholine (21.1 g, 22.9
mL, 0.21 mol) in CH2Cl2 (300 mL) at -10 °C was added
isobutyl chloroformate (9.45 g, 9.02 mL, 6.94 mol) dropwise
over 10 min, and the solution was stirred at -10 °C for 10
min. Ammonia gas was then bubbled through the solution at
a moderately vigorous rate for 15 min, and the solution was
warmed to room temperature over 2 h. The solution was then
washed with 1 N HCl (3 × 1 L) and saturated NaHCO3 (300
mL). The organic phase was evaporated, and the residue was
crystallized from ether/hexane to yield (S)-2-(tert-butoxycar-
bonylamino)-5-(3-carbomethoxyphenyl)pentanamide as a white
1
solid (17.6 g, 73%). H NMR (300 MHz, CDCl3): δ 7.81-7.90
(2H, m); 7.30-7.38 (2H, m); 6.07 (1H, br s); 5.48 (1H, br s);
5.00 (1H, br d, J ) 8.6 Hz); 4.06-4.13 (1H, m); 3.89 (3H, s);
2.60-2.72 (2H, m); 1.62-1.90 (4H, m); 1.42 (9H, s). To a
portion of this material (3.0 g, 8.6 mmol) in ethyl acetate (100
mL) at 0 °C was bubbled HCl gas for 15 min, after which time
solvent was evaporated, yielding 13 as a white solid (2.44 g),
1
which was taken on directly to the subsequent step. H NMR
(MeOH): δ 7.82-7.95 (2H, m); 7.33-7.50 (2H, m); 3.90 (3H,
s); 3.82-2.92 (1H, m); 2.76 (2H, t, J ) 6.5 Hz); 1.73-1.98 (4H,
m).
N-(5-(3-Ca r b oxyp h en yl)-1(S)-cya n op en t yl]-3-m et h yl-
Nr-(2,4-d iflu or oben zoyl)-L-p h en yla la n in a m id e (14). Com-
pounds 2 (2.51 g, 8.76 mmol) and 13 (2.44 g, 8.76 mmol) were
coupled as described for the preparation of 10 to yield a white
solid (4.44 g), which was, without purification, diluted in DMF
(25 mL) and added to the solution created by the addition of
oxalyl chloride (2.2 g, 1.52 mL, 17.4 mmol) to DMF (20 mL) at
0 °C, followed by the addition of pyridine (2.75 g, 2.81 mL,
34.8 mmol). The resulting brown solution was stirred at 0 °C
for 1.5 h, then diluted with ethyl acetate (200 mL), and washed
with saturated aqueous LiCl (3 × 300 mL), dried (MgSO4),
evaporated, and chromatgraphed (25-35% ethyl acetate/
hexane), followed by recrystallization (ether/hexane) to yield
N-(5-(3-carbomethoxyphenyl)-1(S)-cyanopentyl]-3-methyl-NR-
(tert-butoxycarbonyl)-L-phenylalaninamide as a white solid (2.5
g, 59% for 3 steps). 1H NMR (300 MHz, CDCl3): 7.87-7.92
(1H, m); 7.82 (1H, br s); 7.32-7.41 (2H, s); 7.15-7.21 (1H, m);
6.96-7.06 (2H, m); 6.17 (1H, br d, J ) 7.5 Hz); 4.77-4.85 (1H,
m); 3.92 (3H, s); 2.95-3.10 (2H, m); 2.65-2.70 (2H, m); 2.31
(3H, s); 1.66-1.76 (2H, m); 1.40 (9H, s). This material was
deprotected and acylated as described for compound 10 to yield
N-(5-(3-carbomethoxyphenyl)-1(R,S)-cyanopentyl]-3-methyl-
NR-(2,4-difluorobenzoyl)-L-phenylalaninamide as a clear oil.
1H NMR: 8.78 (1H, td, J ) 9.0, 6.8 Hz); 7.85 (1H, dt, J ) 7.2,
1.5 Hz); 7.77 (1H, br s); 7.15-7.35 (4H, m); 6.82-7.08 (4H,
m); 6.53 (1H, d, J ) 8.3 Hz); 4.73-4.85 (1H, m); 3.90 (3H, s);
3.19 (1H, dd, J ) 14.0, 6.8 Hz); 3.09 (1H, dd, J ) 14.0, 8.0
Hz); 2.60-2.67 (2H, m); 2.28 (3H, s); 1.62-1.76 (4H, m). To a
solution of this compound (0.46 g, 0.83 mmol) in THF (40 mL)
was added potassium trimethylsilanoate (316 mg, 2.5 mmol),
and the solution was stirred at room temperature overnight,
during which time substantial yellow precipitate formed. The
solvent was evaporated, and the residue was dissolved in ethyl
acetate (75 mL) and washed with 1 N HCl (75 mL), dried
(MgSO4), evaporated, and chromatographed (2% MeOH/CH2-
Cl2, 0-0.05% HOAc) to yield 14 as a white solid, mp 170-172
°C (0.30 g, 70% yield). 1H NMR (CD3OD): δ 7.82-7.87 (2H,
m); 7.65-7.74 (1H, m); 7.32-7.46 (2H, m); 6.98-7.18 (6H, m);
4.80 (1H, t, J ) 7.0 Hz); 4.71 (1H, t, J ) 7.0 Hz); 3.12 (1H, dd,
J ) 13.4, 7.1 Hz); 3.01 (1H, dd, J ) 13.4, 7.1 Hz); 2.66-2.74
Compounds 38-42, 45-49, and 50 were all prepared in a
manner similar or identical to that described for compound
10.
(S)-2-(ter t-Bu toxyca r bon yla m in o)-5-(3-ca r bom eth oxy-
ph en yl)-4-pen tyn oic Acid (12). N-Boc-propargylglycine meth-
yl ester1 11 (74.2 g, 0.33 mol), methyl 3-bromobenzoate (70.21
g, 0.33 mol), and CuI (2.47 g, 0.013 mmol) were dissolved in
triethylamine (1 L), which was deoxygenated with bubbling
N2 for 2-3 min. Bis-(triphenylphosphine)palladium(II) dichlo-
ride (4.59 g, 0.0065 mol) was added, and the solution was
refluxed for 2 h. The residue was then evaporated, diluted with
ethyl acetate (740 mL), washed with 1 N HCl (2 × 300 mL),
brine (200 mL), and saturated NaHCO3 (300 mL), dried
(MgSO4), evaporated, and chromatographed (10% ethyl acetate/
hexane) to yield (R,S)-2-(tert-butoxycarbonylamino)-5-(3-car-
bomethoxyphenyl)-4-pentynoic acid methyl ester as a clear oil
(66.7 g, 56% yield). 1H NMR (300 MHz, CDCl3): 8.03-8.06
(1H, m); 7.92-7.98 (1H, m); 7.52-7.58 (1H, m); 7.36 (1H, t, J
) 7.9 Hz); 5.39 (1H, br d, J ) 8.0 Hz); 4.52-4.62 (1H, m); 3.91
(3H, s): 3.80 (1H, s); 2.92-2.98 (2H, m); 1.45 (9H, s). To this
material in a mixture of acetonitrile (633 mL) and 0.2 M
aqueous NaHCO3 (1233 mL) was added Alcalase 2.4 L (Novo
Nordisk, 8.0 mL), and the solution was stirred vigorously at
room temperature for 2.5 h. The reaction mixture was then
evaporated at 30 °C to remove acetonitrile, and the aqueous
residue was washed with ether (3 × 350 mL). The aqueous
phase was filtered through Celite, the pH was adjusted to 3
with 6 N HCl (∼57 mL), and the solution was extracted with
ethyl acetate (9 × 100 mL). The combined organic layers were
then dried (MgSO4) and evaporated to yield 12 as a clear oil
(25.2 g, 39.4%). 1H NMR (300 MHz, CDCl3): δ 8.02-8.05 (1H,
m); 7.90-7.96 (1H, m); 7.51-7.57 (1H, m); 7.33 (1H, t, J ) 7.9
Hz); 5.42-5.46 (1H, m); 4.56-4.65 (1H, m); 3.90 (3H, s); 2.97-

Dipeptidyl Nitriles as Potent and Selective Inhibitors
J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26 4533
(2H, m); 2.29 (3H, s); 1.70-1.89 (4H, m). MS (m/z): (M + H)
volumes of protein solution and well buffer (50 mM Na acetate
pH 5.5, 100 mM KCl, and 15-20% monomethyl ether PEG
5000 at 4 °C). Crystals were cryo-protected at room temper-
ature by a 15 min, 10 step transfer to a cryobuffer solution
(50 mM Na acetate pH 5.5, 100 mM KCl, 50 mM NaCl, 20%
monomethyl ether PEG 5000, 14% ethylene glycol, 10%
glycerol) with 20 µM compound 3 and immediately flash-frozen
in a 100 K nitrogen gas cold stream (Low Temp System, MSC)
for data collection. Diffraction data were collected on an
R-AXIS-IIc image-plate system (Molecular Structures Corp.)
with “Yale-style” double focusing mirrors on an RU-2HR
generator (MSC) operating at 104 mA and 50 kV. A complete
data set was collected from a single crystal maintained at 100
K. The images were indexed and integrated using Denzo/
Scalepack.32
Str u ctu r e Deter m in a tion a n d Refin em en t. The struc-
ture was determined by molecular replacement using XPLOR33
(MSI, Inc.) with a structure of human cat B as the search
model. After independent rigid-body refinement of the 3 cat B
proteins in the asymmetric unit, the electron density for the
inhibitor was clearly seen in each of the active sites. Several
rounds of iterative model building with the program O,34
positional refinement, individual B refinement, and addition
of waters were conducted before adding the three inhibitors
to the model. The active site of one complex was fully solvent
accessible, while the active site of the other two complexes
were buried by crystal contacts. The electron density for the
inhibitor of the exposed active site indicated that the occupancy
may have been less than 1.0. The occupancy of each inhibitor
was refined, and the occupancies of 1.0, 1.0, and 0.78 were
used in the model. The final model contained 3 cat B/compound
3 complexes, and 535 waters with an R factor of 16.9% in the
resolution range of 6.0-1.9 Å. Recently, the model was further
refined in preparation for deposition and to exploit the power
of Rfree, monolayer refinement, and bulk solvent in CNX35 (MSI,
Inc.). The final CNX model was indistinguishable from the
original XPLOR model in the active site; summary statistics
for the CNX refined model are in the Supporting Information.
The model has been submitted to the Protein Data Bank,
accession no. 1GMY.36
520. Anal. Calcd for C29H27F2N3O4: C, H, N.
Compound 43 was prepared in a manner similar to com-
pound 14.
Meth yl 3-(2-Iod oeth oxy)ben zoa te (16). To methyl 3-hy-
droxybenzoate 15 (5.0 g, 32.9 mmol) in acetone (100 mL) was
added 1,2-dibromoethane (11.3 mL, 131.4 mmol) and K2CO3
(5.45 g, 39.4 mmol), and the solution was refluxed 16 h. TLC
showed an incomplete reaction, so acetone was evaporated and
DMF (100 mL) was added. The solution was then heated to
60 °C for 16 h. After the solution was cooled and filtered, the
solution was evaporated, and the residue was chromato-
graphed (25% ethyl acetate/hexane) to yield methyl 3-(2-
bromoethoxy)benzoate as a clear oil, which was then diluted
in acetone (50 mL). NaI (2.78 g, 18.52 mmol) was added, and
the solution was refluxed for 2 h. The cooled solution was
filtered, dissolved in ethyl acetate (50 mL), washed with 5%
aqueous Na2SO3 (50 mL), water (50 mL), and brine (50 mL),
dried (MgSO4), and evaporated to yield 16 as a clear oil (2.7
1
g, 27% for 2 steps). H NMR (250 MHz, CDCl3): δ 7.63-7.68
(1H, m); 7.53-7.57 (1H, m); 4.68 (2H, t, J ) 6.8 Hz); 4.92 (3H,
s); 3.44 (2H, t, J ) 6.8 Hz).
4-(3-Car bom eth oxyph en oxy)-2-am in obu tyr on itr ile (17).
A 1 M solution of Na(TMS)2 in THF (8.82 mL, 8.82 mmol) was
added to a solution of N-(diphenylmethylene)aminoacetonitrile
(1.9 g, 8.65 mmol) in THF (90 mL) at -78 °C via syringe and
stirred for 30 min. A solution of 16 (2.7 g, 8.82 mmol) in THF
(30 mL) was added dropwise, and the solution was warmed to
room temperature over 3 h. The solution was quenched with
saturated NH4Cl (50 mL) and extracted with ethyl acetate (3
× 50 mL). The organic layers were washed with water (50 mL)
and brine (50 mL), dried (MgSO4), evaporated, and chromato-
graphed (12.5% ethyl acetate/hexane). The purified product
was then diluted in Et2O (90 mL), and 1 N HCl (7.5 mL, 7.5
mmol) was added, followed by vigorous stirring for 16 h. The
ether layer was separated, and the aqeous layer was washed
with Et2O (3 × 50 mL), basified to pH 8 with saturated
NaHCO3, and extracted with ethyl acetate (3 × 50 mL). The
organic layer was washed with brine (50 mL), dried (MgSO4),
and evaporated to yield 17 as a clear oil (1.48 g, 73% for 2
steps). 1H NMR (250 MHz, CDCl3): 7.60-7.68 (1H, m); 7.52-
7.57 (1H, m); 7.53 (1H, t, J ) 6.8 Hz); 7.05-7.13 (1H, m); 4.02-
4.33 (3H, m); 3.92 (3H, s); 2.05-2.35 (2H, m).
Su p p or tin g In for m a tion Ava ila ble: Crystallographic
parameters for the structure of the cat B-3 complex. This
material is available free of charge via the Internet at http://
pubs.acs.org.
N-[2-[(3-Ca r boxyp h en oxy]-1(S)-cya n op r op yl]-3-m eth -
yl-Nr-(d ip h en yla cetyl)-L-p h en yla la n in a m id e (18). Com-
pound 17 was converted to N-[2-[(3-carbomethoxyphenoxy]-
1(S)-cyanopropyl]-3-methyl-NR-(diphenylacetyl)-L-phenylal-
aninamide, using a procedure similar to that described for 14.
To a solution of this material (0.16 g, 0.27 mmol) in THF (3
mL) was added a solution of LiOH‚H2O (22 mg, 0.54 mmol) in
water (0.5 mL). The mixture was stirred for 6 h, after which
time additional water (1 mL) was added, and the solution was
stirred for an additional 1 h. The mixture was concentrated,
diluted in water (20 mL), acidified with 1 N HCl, and extracted
with ethyl acetate (3 × 30 mL). The combined organic layers
were washed with brine, dried (MgSO4), evaporated, and
chromatographed (5% MeOH/CH2Cl2, 0.05% AcOH) to yield
18 as a white solid (35 mg), mp 169-170 °C. NMR indicated
that the product was, as expected, a 1:1 mixture of diastereo-
isomers. 1H NMR (DMSO, 250 MHz): δ 7.50-7.56 (1H, m);
7.31-7.42 (2H, m); 6.82-7.28 (15 H); 5.03 (0.5H, s); 5.01 (0.5H,
s); 4.82-4.91 (1H, m); 4.44-4.56 (1H, m); 2.68-295 (2H, m);
2.08-2.28 (2H, m); 2.19 (1.5H, s); 2.13 (1.5H, s); 3.91-4.14
(2H, m). MS (m/z): (M + H) 574. Anal. Calcd for C33H33N3O5‚
0.5 H2O: C, H, N.
Cr ysta lliza tion a n d Da ta Collection . A complex was
formed by incubation of a 4-fold molar excess of compound 3
with 1.5 mg of cat B in binding buffer (20 mM sodium acetate
pH 5.5, 300 mM KCl and 75 mM NaCl) plus 2 mM 1,4-
dithiothreitol for 30 min at room temperature. The complex
was buffer exchanged four times into binding buffer with 25
µM compound 3 using an Amicon Centricon-10 concentrator
and finally concentrated to 10 mg/mL. Crystals were grown
by the hanging-drop vapor diffusion method by mixing equal
Refer en ces
(1) Turk, B.; Turk, D.; Turk, V. Lysosomal cysteine proteases: more
than scavengers. Biochim. Biophys. Acta 2000, 1477, 98-111.
(2) Chapman, H. A.; Riese, R. J .; Shi, G. P. Emerging roles for
cysteine proteases in human biology. Annu. Rev. Physiol. 1997,
59, 63-88.
(3) Barrett, A. J .; Kirschke, H. Cathepsin B, Cathepsin H, and
Cathepsin L. Methods Enzymol. 1981, 80C, 535-561.
(4) Berquin, I. M.; Sloane, B. F. Cathepsin B expression in human
tumors. Adv. Exp. Med. Biol. 1996, 389, 281-294.
(5) Petanceska, S.; Burke, S.; Watson, S. J .; Devi, L. Differential
distribution of messenger RNAs for cathepsins B, L, and S in
adult rat brain: an in situ hybridization study. Neuroscience
1994, 59, 729-738.
(6) Keyszer, G.; Redlich, A.; Haupl, T.; Zacher, J .; Sparmann, M.;
Entgethum, U.; Gay, S.; Burmester, G. R. Differential expression
of cathepsins B and L compared with matrix metalloproteinases
and their respective inhibitors in rheumatoid arthritis and
osteoarthritis:
a parallel investigation by semiquantitative
reverse transcriptase-polymerase chain reaction and immuno-
histochemistry. Arthritis Rheum. 1998, 41, 1378-1387.
(7) Buttle, D. J .; Handley, C. J .; Ilic, M. Z.; Saklatvala, J .; Murata,
M.; Barrett, A. J . Inhibition of cartilage proteoglycan release by
a specific inactivator of cathepsin B and an inhibitor of matrix
metalloproteinases. Evidence for two converging pathways of
chondrocyte-mediated proteoglycan degradation. Arthritis Rheum.
1993, 36, 1709-1717.
(8) Esser, R. E.; Angelo, R. A.; Murphey, M. D.; Watts, L. M.;
Thornburg, L. P.; Palmer, J . T.; Talhouk, J . W.; Smith, R. E.
Cysteine proteinase inhibitors decrease articular cartilage and
bone destruction in chronic inflammatory arthritis. Arthritis
Rheum. 1994, 37, 236-247.
(9) Veber, D. F.; Thompson, S. K. The therapeutic potential of
advances in cysteine protease inhibitor design. Curr. Opin. Drug
Discovery Dev. 2000, 3, 362-369.

4534 J ournal of Medicinal Chemistry, 2001, Vol. 44, No. 26
Greenspan et al.
(10) Otto, H. H.; Schirmeister, T. Cysteine Proteases and Their
Inhibitors. Chem. Rev. 1997, 97, 133-171.
(24) J ia, Z.; Hasnain, S.; Hirama, T.; Lee, X.; Mort, J . S.; To, R.;
Huber, C. P. Crystal Structures of Recombinant Rat Cathepsin
B and a Cathepsin B-Inhibitor Complex. J . Biol. Chem. 1995,
270, 5527-5533. At the time this work was commenced in 1994,
no human cat B-inhibitor complexes were publicly available.
We therefore utilized this structure, to which we had exclusive
access through a collaboration with the NRC of Canada.
(25) McMartin, C.; Bohacek, R. S. QXP: powerful, rapid computer
algorithms for structure-based drug design. J . Comput.-Aided
Mol. Des. 1997, 11, 333-344.
(11) Thompson, S. K.; Smith, W. W.; Zhao, B.; Halbert, S. M.;
Tomaszek, T. A.; Tew, D. G.; Levy, M. A.; J anson, C. A.;
D’Alessio, K. J .; McQueney, M. S.; Kurdyla, J .; J ones, C. S.;
DesJ arlais, R. L.; Abdel-Meguid, S. S.; Veber, D. F. Structure-
Based Design of Cathepsin K Inhibitors Containing a Benzyloxy-
Substituted Benzoyl Peptidomimetic. J . Med. Chem. 1998, 41,
3923-3927.
(12) Yamashita, D. S.; Smith, W. W.; Zhao, B.; J anson, C. A.;
Tomaszek, T. A.; Bossard, M. J .; Levy, M. A.; Oh, H. J .; Carr, T.
J .; Thompson, S. K.; Ijames, C. F.; Carr, S. A.; McQueney, M.;
D’Alessio, K. J .; Amegadzie, B. Y.; Hanning, C. R.; Abdel-Meguid,
S.; DesJ arlais, R. L.; Gleason, J . G.; Veber, D. F. Structure and
Design of Potent and Selective Cathepsin K Inhibitors. J . Am.
Chem. Soc. 1997, 119, 11351-11352.
(26) Hasnain, S.; Hirama, T.; Huber, C. P.; Mason, P.; Mort, J . S.
Characterization of cathepsin
B specificity by site-directed
mutagenesis. Importance of Glu245 in the S2-P2 specificity for
arginine and its role in transition state stabilization. J . Biol.
Chem. 1993, 268, 235-240.
(27) Musil, D.; Zucic, D.; Turk, D.; Engh, R. A.; Mayr, I.; Huber, R.;
Popovic, T.; Turk, V.; Towatari, T.; Katunuma, N. The refined
2.15 A X-ray crystal structure of human liver cathepsin B: the
structural basis for its specificity. EMBO J . 1991, 10, 2321-
2330.
(28) Turk, D.; Podobnik, M.; Popovic, T.; Katunuma, N.; Bode, W.;
Huber, R.; Turk, V. Crystal structure of cathepsin B inhibited
with CA030 at 2.0-A resolution: A basis for the design of specific
epoxysuccinyl inhibitors. Biochemistry 1995, 34, 4791-4797.
(29) Aronson, N. N. J .; Barrett, A. J . The specificity of cathepsin B.
Hydrolysis of glucagon at the C-terminus by a peptidyldipepti-
dase mechanism. Biochem. J . 1978, 171, 759-765.
(30) Shaw, E.; Wikstrom, P.; Ruscica, J . An exploration of the primary
specificity site of cathepsin B. Arch. Biochem. Biophys. 1983,
222, 424.
(13) Brisson, J . R.; Carey, P. R.; Storer, A. C. Benzoylamidoacetoni-
trile is bound as a thioimidate in the active site of papain. J .
Biol. Chem. 1986, 261, 9087-9089.
(14) Dufour, E.; Storer, A. C.; Menard, R. Peptide aldehydes and
nitriles as transition state analogue inhibitors of cysteine
proteases. Biochemistry 1995, 34, 9136-9143.
(15) Drenth, J .; Kalk, K. H.; Swen, H. M. Binding of chloromethyl
ketone substrate analogues to crystalline papain. Biochemistry
1976, 15, 3731-3738.
(16) Picken, P. P.; Guthrie, D. J .; Walker, B. Inhibition of bovine
cathepsin B by amino acid-derived nitriles. Biochem. Soc. Trans.
1990, 18, 316.
(17) Several patent applications covering peptidyl nitrile inhibitors
of various cathepsins have appeared recently, but none contain
potency data. WO 0119816, WO 0049007, WO 0049008, WO
9924460.
(18) Falgueyret, J . P.; Oballa, R. M.; Okamoto, O.; Wesolowski, G.;
Aubin, Y.; Rydzewski, R. M.; Prasit, P.; Riendeau, D.; Rodan, S.
B.; Percival, M. D. Novel, Nonpeptidic Cyanamides as Potent
and Reversible Inhibitors of Human Cathepsins K and L. J . Med.
Chem. 2001, 44, 94-104. This very recent reference describes a
series of cyanamides that are quite distinct chemically from
aminonitriles.
(19) Berger, A.; Smolarsky, M.; Kurn, N.; Bosshard, H. R. A new
method for the synthesis of optically active amino acids and their
N derivatives via acylamino malonates. J . Org. Chem. 1973, 38,
457-460.
(20) Kijima, T.; Ohshima, K.; Kise, H. Facile optical resolution of
amino acid esters via hydrolysis by an industrial enzyme in
organic solvents. J . Chem. Technol. Biotechnol. 1994, 59, 61-
65.
(21) Abood, N. A.; Nosal, R. Zinc mediated addition of active halides
to a glycine cation equivalent. Tetrahedron Lett. 1994, 35, 3669-
3672.
(22) Laganis, E. D.; Chenard, B. L. Metal silanolates. Tetrahedron
Lett. 1984, 25, 5831-5834.
(23) O’Donnell, M. J .; Eckrich, T. M. The synthesis of amino acid
derivatives by catalytic phase-transfer alkylations. Tetrahedron
Lett. 1978, 25, 4625-4628.
(31) Steed, P. M.; Lasala, D.; Liebman, J .; Wigg, A.; Clark, K.; Knap,
A. K. Characterization of recombinant human cathepsin
B
expressed at high levels in baculovirus. Protein Sci. 1998, 2033-
2037.
(32) Otwinowski, Z.; Minor, W. Processing of X-ray Diffraction Data
Collected in Oscillation Mode. In Methods in Enzymology; Carter,
C. W., Sweet, R. M., Eds.; Academic Press: New York, 1997; Vol.
276, pp 307-325.
(33) Bru¨nger, A. T. XPLOR, Version 3.1; Yale University Press: New
Haven, CT, 1993.
(34) J ones, T. A.; Zou, J .-Y.; Cowan, S. W.; Kjeldgaard, M. Improved
Methods for the Building of Protein Models in Electron Density
Maps and the Location of Errors in these Models. Acta Crystal-
logr. 1991, A47, 110-119.
(35) Bru¨nger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros,
P.; Grosse-Kunstleve, R. W.; J iang, J .-S.; Kuszewski, J .; Nilges,
M.; Pannu, N. S.; Read, R. J .; Rice, L. M.; Simonson, T.; Warren,
G. L. Crystallography and NMR system (CNS). Acta Crystallogr.
1998, D54, 905-921.
(36) Berman, H. M.; Westbrook, J .; Feng, Z.; Gilliand, G.; Bhat, T.
N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data
Bank. Nucleic Acids Res. 2000, 28, 235-242.
J M010206Q