Exploration of the P2-P3 SAR of aldehyde cathepsin K inhibitors

Article abstract of DOI:10.1016/j.bmcl.2004.04.084

The synthesis and biological activity of a series of aldehyde inhibitors of cathepsin K are reported. Exploration of the properties of the S² and S³ subsites with a series of carbamate derivatized norleucine aldehydes substituted at the P² and P³ positions afforded analogs with cathepsin K IC₅₀s between 600nM and 130pM.

Full text of DOI:10.1016/j.bmcl.2004.04.084

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3425–3429
Exploration of the P2–P3 SAR of aldehyde cathepsin K inhibitors
Eric E. Boros,a David N. Deaton,a,* Anne M. Hassell,b Robert B. McFadyen,a
Aaron B. Miller,b Larry R. Miller,c Margot G. Paulick,a,ꢀ Lisa M. Shewchuk,b
James B. Thompson,a Derril H. Willard, Jr.d,z and Lois L. Wrighte
aDepartment of Medicinal Chemistry, GlaxoSmithKline, Five Moore Drive, Research Triangle Park, NC 27709-3398, USA
bDiscovery Research Computational, Analytical, and Structural Sciences, GlaxoSmithKline, Five Moore Drive,
Research Triangle Park, NC 27709-3398, USA
cDepartment of Molecular Pharmacology, GlaxoSmithKline, Five Moore Drive, Research Triangle Park, NC 27709-3398, USA
dDepartment of Gene Expression and Protein Biochemistry, GlaxoSmithKline, Five Moore Drive, Research Triangle Park,
NC 27709-3398, USA
eDiscovery Research Biology, GlaxoSmithKline, Five Moore Drive, Research Triangle Park, NC 27709-3398, USA
Received 26 March 2004; accepted 24 April 2004
Available online
Abstract—The synthesis and biological activity of a series of aldehyde inhibitors of cathepsin K are reported. Exploration of the
properties of the S2 and S3 subsites with a series of carbamate derivatized norleucine aldehydes substituted at the P2 and P3 positions
afforded analogs with cathepsin K IC50s between 600 nM and 130 pM.
Ó 2004 Elsevier Ltd. All rights reserved.
Cathepsin K, a lysosomal cysteine protease of the C1A
family highly expressed in osteoclasts, is crucial for the
degradation of bone matrix.1 Moreover, small molecule
cathepsin K inhibitors are able to attenuate bone
resorption in osteoporotic animal models.2 As part of a
broader goal of developing new therapeutics for osteo-
porosis, these researchers recently reported the discov-
ery and P1 SAR exploration of the aldehyde-based
cathepsin K inhibitor 1 (Boc–Nle–H) derived from
truncation of calpeptin (Cbz–Leu–Nle–H), which was
identified in a focussed screen of known serine and
cysteine protease aldehyde inhibitors.3 This report will
detail the further investigation of the SAR of this lead,
concentrating on the P2 and P3 regions.
R
R1 R2
O
Cl
H2N
nBu
R
OH
1 R2
2a-2c, 2l-2m
b
a
OH
+
R
O
3a-3c, 3l-3m
4
H
N
H
R
R1 R2
O
R
O
N
c
OH
O
R1 R2
O
nBu
O
nBu
5a-5c, 5l-5m
6a-6c, 6k-6m
Scheme 1. (a) COCl2, PhMe, 0 °C to rt; (b) iPr2NEt, dioxane; (c)
pyridine ꢀ SO3, NEt3, DMSO, CH2Cl2, )10 °C.
chloroformates were then coupled to the known amine
46 to give the alcohols 5a–c and 5l–m.7 Subsequent
Moffatt oxidation of these alcohols 5a–c and 5l–m
provided the desired aldehydes 6a–c and 6k–m after an
extractive work-up.8 The yields of the alcohols 5a–u and
aldehydes 6a–u as well as the cathepsin K activity of the
aldehyde analogs 6a–u are shown in Table 1, while the
yields of the alcohols 5v–z and aldehydes 6v–z as well as
the cathepsin K activity of the aldehyde analogs 6v–z are
shown in Table 2.
Two routes were utilized to synthesize P2–P3 aldehyde
analogs. The first method converted commercially
available P2–P3 alcohols 2a–c and 2l or known P2–P3
alcohol 2m4 into their chloroformates 3a–c and 3l–m
mediated by phosgene as depicted in Scheme 1.5 These
* Corresponding author. Tel.: +1-919-483-6270; fax: +1-919-315-0430;
e-mail: david.n.deaton@gsk.com
ꢀ
Present address: Department of Chemistry, University of Califor-
nia––Berkeley, Berkeley, CA 94720-1460, USA.
Deceased.
Alternatively, the commercially available P2–P3 alcohols
2d, 2f, 2i–j, and 2o–u, the previously described P2–P3
z
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.04.084

3426
E. E. Boros et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3425–3429
Table 1. P2–P3 cathepsin K inhibition and synthesis yields
H
N
R
R1 R2
O
O
O
nBu
#
R
R1
R2
5 yd. (%)
6 yd. (%)
Cat. K IC50 (nM)a
1
Me
Ph
Me
H
Me
H
65
65
86
97
93
47
74
63
45
82
86
77
67
68
88
84
88
90
90
81
64
75
99
99
78
88
67
90
76
80
95
25
88
94
88
90
80
98
99
79
85
96
98
64
51
540b
270b
600b
12
6a
6b
6c
6d
6e
6f
PhCH2
Ph(CH2)2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
C6H11CH2
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
H
H
H
H
H
Me
Me
(CH2)3
Et
2.4
2.1
0.35c
Et
6g
6h
6i
(CH2)4
(CH2)5
Me
H
2.0
1.8
100d
H
6j
Me
H
6k
6l
Me
Et
2.7
H
0.13c
1.5
6m
6n
6o
6p
6q
6r
6s
6t
nPr
iPr
H
H
0.50
6.5
iBu
PhCH2
Et
H
H
71
Et
nPr
iPr
iBu
iPr
4.0
H
nPr
iPr
0.87c
0.56
1.1
H
H
Me
iBu
iPr
6u
4.7b
a Inhibition of recombinant human cathepsin K activity in a fluorescence assay using 10 lM Cbz–Phe–Arg–AMC as substrate in 100 mM NaOAc,
10 mM DTT, 120 mM NaCl, pH ¼ 5.5. The IC50 values are the mean of two or three inhibition assays, individual data points in each experiment
were within a 2-fold range of each other.
b These entries represent an n of one.
c Individual data points in each experiment were within a 3-fold range of each other.
d This analog was synthesized from a commercially available chiral alcohol 2j. It cannot be ruled out that the activity of 6j might arise from con-
tamination by a small amount of 6i derived from starting alcohol 2j being less than 100% enantiomerically pure.
Table 2. P3 cathepsin K inhibition and synthesis yields
O
O
C
N
O
H
N
H
N
O
R
R1 R2
O
R
OH
R1 R2
2d-2j, 2n-2z
R
O
O
a
O
+
O
nBu
O
nBu
nBu
7
8d-8j, 8n-8z
#
R
5 yd. (%)
6 yd. (%) Cat. K IC50
(nM)a
OH
O
H
H
6v
6w
6x
6y
6z
3-MeO–Ph
4-MeO–Ph
2-Cl–Ph
99
69
77
81
85
98
78
77
93
99
5.4b
2.9
R
O
R1 R2
N
R
R1 R2
O
N
b, c
d
O
nBu
O
nBu
6d-6j, 6n-6z
3.1
4-Cl–Ph
3-Thiophenyl
4.6b
7.6b
5d-5j, 5n-5z
Scheme 2. (a) PhMe, 90 °C, 22–99%; (b) LiOH, THF, H2O; (c)
iPrOCOCl, NEt3, THF, )10 °C; NaBH4, H2O, 0 °C; (d) pyridine ꢀ SO3,
NEt3, DMSO, CH2Cl2, )10 °C.
a Inhibition of recombinant human cathepsin K activity in a fluores-
cence assay using 10 lM Cbz–Phe–Arg–AMC as substrate in 100 mM
NaOAc, 10 mM DTT, 120 mM NaCl, pH ¼ 5.5. The IC50 values are
the mean of two or three inhibition assays, individual data points in
each experiment were within a 2-fold range of each other.
b These entries represent an n of one.
OH
OH
H
N
H
N
a
O
O
O
nBu
O
nBu
alcohols 2e,9 2g–h,10 2n,11 and 2w–y,12–14 or the novel
alcohols 2v15 and 2z16 were coupled to the known iso-
cyanate 717 to give the carbamates 8d–j and 8n–z as
shown in Scheme 2.18 Then, hydrolysis of the resulting
esters 8d–j and 8n–z, followed by sodium borohydride-
mediated reduction of the in situ generated mixed
anhydrides provided the alcohols 5d–j and 5n–z.19 The
cyclohexyl alcohol 5k was synthesized by hydrogenation
5i
5k
Scheme 3. (a) H2, RhCl3ÆH2O, Aliquatâ 336, ClCH2CH2Cl, H2O
(77%).
of the aryl ring of alcohol 5i catalyzed by rhodium under
phase transfer conditions as displayed in Scheme 3.20

E. E. Boros et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3425–3429
3427
Once again, the alcohols 5d–j and 5n–z were oxidized to
the desired aldehydes 6d–j and 6n–z.
drawing groups (6x–y) were all well tolerated. The ab-
sence of strong effects on the inhibitory activity by these
P3 changes suggests that the interactions of these sub-
stituents with the protein are quite weak.
The lead aldehyde 1 contains no P3 residue. With an X-
ray crystal structure of cathepsin K unavailable at the
start of this work, a traditional structure activity rela-
tionship study was employed. It was thought that the
incorporation of an appropriately placed aryl sub-
stituent could result in favorable p–p stacking interac-
tions with potential aryl groups in the S3 subsite or the
active site trough that accommodates the substrate
backbone. To test this hypothesis, a potential S3-binding
phenyl ring was attached to the amino acid backbone
with tethers of varying lengths.21 The benzyl, phenethyl,
and phenpropyl derivatives 6a–c (cathepsin K IC50s of
540, 270, and 600 nM, respectively, as shown in Table 1)
did not differ significantly from each other in their
ability to inhibit cathepsin K, nor did they offer
increased potency over the aldehyde lead 1.
Since the introduction of larger P2 moieties resulted in
increased inhibitory potency in the gem-disubstituted
analogs and only the pro-(S) substituent was important
for activity, further exploration of bigger P2 alkyl
groups in the more active diastereomer series was war-
ranted. To that end, analogs of 6i bearing larger P2 alkyl
groups were synthesized. The ethyl- and iso-propyl-
containing analogs 6l (IC50 ¼ 0.13 nM) and 6n
(IC50 ¼ 0.50 nM) were more potent inhibitors of
cathepsin K than the methyl analog 6i, whereas the
n-propyl- and iso-butyl-containing derivatives 6m
(IC50 ¼ 1.5 nM) and 6o (IC50 ¼ 6.5 nM) were approxi-
mately equipotent to the methyl analog 6i. However,
introduction of the larger benzyl group in compound 6p
(IC50 ¼ 71 nM) resulted in a significant decrease in
inhibitory activity. It is probably too large to be rea-
sonably accommodated in the S2 pocket.
Since the primary determinant of specificity of papain
family endopeptidases is the S2 pocket,22 the gem-
dimethyl P2 group of lead 1 was combined with the
phenyl P3 moiety of 6b to provide inhibitor 6d
(IC50 ¼ 12 nM). This analog was ꢁ4-fold more potent
than analog 1 and ꢁ20-fold more potent than analog 6b.
The gem-diethyl P2 analog 6f was then prepared and
found to be even more active (IC50 ¼ 2.1 nM). To further
investigate the dimensions of the S2 pocket, the cyclo-
butyl, cyclopentyl, and cyclohexyl P2 moieties were
incorporated with the P3 phenyl group to afford inhib-
itors 6e and 6g–h. The cyclobutyl and cyclohexyl-con-
taining compounds 6e (IC50 ¼ 2.4 nM) and 6h
(IC50 ¼ 2.0 nM) were similar in inhibitory potency to the
gem-diethyl analog 6f, whereas the cyclopentyl-con-
taining compound 6g provided a 6-fold increase in
potency (IC50 ¼ 0.35 nM).
In an attempt to avoid the added synthetic complexity
associated with the incorporation of a second stereo-
center, a group of symmetrical alcohols was used to
prepare a series of inhibitors containing achiral P2–P3
moieties (compounds 6q–u). The diethyl-, di-n-propyl-,
di-iso-propyl-, and di-iso-butyl-containing analogs 6q–t
were more potent than the dibenzyl analog 6p discussed
above, and exhibited activity similar to the P2-methyl
P3-benzyl analog 6i. Addition of a methyl group to
afford the methyl, di-iso-propyl-containing derivative 6u
(IC50 ¼ 4.7 nM) did not enhance activity over 6s
(IC50 ¼ 0.56 nM). Presumably, in these symmetrical
inhibitors, one of the side chains occupies the S2
pocket, whereas the second side chain lies in the
active site trough that accommodates the substrate
backbone.
Although the incorporation of geminal disubstitution in
the P2 region afforded potent cathepsin K inhibitors, it
was not clear if both P2 substituents were required for
favorable interactions with the enzyme. To address the
contribution of each substituent to protease inhibition,
the diastereomers 6i and 6j were synthesized. The (S)-
methyl derivative 6i (IC50 ¼ 1.8 nM) exhibited substan-
tially more inhibitory activity versus cathepsin K than
its (R)-epimer (IC50 ¼ 100 nM). Furthermore, analog 6i
was more potent than the gem-dimethyl derivative 6d,
demonstrating that the interaction of the inhibitor
with cathepsin K in the S2 pocket is stereospecific.
Furthermore, a second geminal substituent is not only
unnecessary, but can be detrimental to the inhibitory
activity.
Selectivity is a critical concern when designing potential
therapeutics to avoid undesired side effects. Given the
high degree of homology between the lysosomal cysteine
cathepsins, achieving reasonable selectivity can be a
daunting task. To ascertain the selectivity of this class of
aldehyde inhibitors, selected analogs were screened
against the commercially available cathepsins B and L
with data shown in Table 3. Gratifyingly, all of these
aldehyde inhibitors exhibit more inhibitory activity
versus cathepsin K than against cathepsins B and L.
Although several of the inhibitors were greater than 100-
fold selective for cathepsin K versus the other two
enzymes, in general, the cycloalkyl-containing inhibitors
6e, 6g, and 6h were the more selective. The cyclohexyl-
containing derivative 6h is >6000-fold selective for
cathepsin B and >4000-fold selective for cathepsin L.
Interestingly, the diethyl-bearing analog 6f (cathepsin L
IC50 ¼ 110 nM) is a substantially more potent inhibitor
of cathepsin L than are the cycloalkyl-containing ana-
logs 6e, 6g, and 6h, whereas the mono-ethyl compound
6l (cathepsin L IC50 ¼ 15 nM) is even more potent.
Analog 6l is also a reasonably good inhibitor of
cathepsin B (cathepsin B IC50 ¼ 72 nM).
The interactions of the inhibitors with the S3 domain of
cathepsin K were further explored. Replacement of the
P3 phenyl group by a cyclohexyl moiety as in 6k
(IC50 ¼ 2.7 nM) did not affect the inhibitory activity.
Furthermore, as shown in Table 2, a five-membered
heteroaryl ring as in 6z could replace the phenyl ring
with no loss in activity. Substitutions at the 2-, 3-, or 4-
position of the phenyl group in the P2 gem-dimethyl
series by electron donating (6v–w) or electron with-

3428
E. E. Boros et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3425–3429
Table 3. Cathepsin B, K, and L inhibition and selectivity
H
N
R
R1 R2
O
O
O
nBu
#
R
R1
R2
Et
Cat. B IC50 (nM)a
Cat. K IC50 (nM)a
Cat. L IC50 (nM)a
6e
6f
6g
6h
6i
PhCH2
PhCH2
PhCH2
PhCH2
PhCH2
C6H11CH2
PhCH2
(CH2)3
Et
3400
2200
810
2.4
2.1
0.35c
1200
110c
2000
8500
740
270b
15c
(CH2)4
(CH2)5
Me
12,000
2.0
1.8
H
H
H
460
6k
6l
Me
Et
400b
72c
2.7
0.13c
a Inhibition of recombinant human cathepsin B, K, and L activity in a fluorescence assay using Cbz–Phe–Arg–AMC as substrate (cathepsins B and K
(10 lM), cathepsin L (5 lM)) in 100 mM NaOAc, 10 mM DTT, 120 mM NaCl, pH ¼ 5.5. The IC50 values are the mean of two or three inhibition
assays, individual data points in each experiment were within a 2-fold range of each other.
b These entries represent an n of one.
c Individual data points in each experiment were within a 3-fold range of each other.
lipophilic nature of these residues, and this was con-
firmed by these SAR studies.
Other interactions also contribute to the binding of
these inhibitors. The S1 binding site is created by resi-
dues 23Gly, 24Ser, 64Gly, and 65Gly and may be described
as a wall rather than pocket since one half of the subsite
is solvent exposed. All of the polar backbone atoms that
form the wall are hydrogen bonded, which results in the
subsite having a very hydrophobic character, explaining
the preference for linear hydrophobic substituents in this
position. The S3 pocket consists of a ‘groove’ that is
formed by 60Asn, 61Asp, 66Gly, and 67Tyr. It too has
hydrophobic character and accommodates lipophilic
substituents. A covalent hemithioketal intermediate is
formed by the aldehyde of the inhibitor and the active
site cysteine (25Cys) of the protein. The OH of the
Figure 1. Active site of the X-ray co-crystal structure of compound 9
bound to cathepsin K. The cathepsin K carbons are colored yellow
hemithioketal is stabilized by three hydrogen bonds: (1)
to the backbone amide of 25Cys; (2) to the side chain of
19Gln; and (3) to the side chain of catalytic 159His. Two
with inhibitor 9 carbons colored green. The semi-transparent white
surface represents the molecular surface, while hydrogen bonds are
other hydrogen bonds formed between the inhibitor and
depicted as yellow dashed lines. Inhibitor 6n, with its carbons colored
the protease occur in the peptide recognition site; the
cyan, was modeled into the X-ray co-crystal structure, illustrating the
NH and carbonyl oxygen of the inhibitor carbamate
form hydrogen bonds to the backbone carbonyl of
161Asn and the backbone HN of 66Gly, respectively.
probable mode of binding of these truncated aldehydes inhibitors. The
coordinates have been deposited in the Brookhaven Protein Data
Bank, accession number 1SNK. This figure was generated using PY-
MOL version 0.93.
Analog 6n (IC50 ¼ 0.50 nM) is also depicted in Figure 1.
The original lead for this aldehyde series identified by
focussed screening was calpeptin, Cbz–Leu–Nle–H.3 As
part of an effort to obtain X-ray co-crystal structures of
cathepsin K/inhibitor complexes, several heavy atom
calpeptin derivatives were synthesized to assist in phas-
ing the X-ray diffraction data. Co-crystallization of one
This analog has been modeled into the active site of the
enzyme utilizing the program MVP.24 The inhibitor 6n
was docked into the active site by growing the ligand in
the covalently attached state, starting at the b-carbon of
25Cys. It illustrates the probable binding mode of this
single amino acid aldehyde inhibitor class. As shown,
there is good overlap between the modeled iso-propyl P2
moiety of 6n and the bound iso-butyl P2 fragment of 9 in
the S2 pocket. The docked aldehyde inhibitor backbone
of 6n also maps well to the bound backbone of 9. The P1
substituents of these two inhibitors also show good
agreement in the S1 groove of the enzyme. The docked
inhibitor 6n has the P3 moiety exposed to solvent with
little if any interaction with the protease. Whereas this
model is in good agreement with the P3 SAR of these
inhibitors, in that changes in P3 had minimal effects on
of these analogs, para-bromo–Cbz–Leu–Nle–H
9
(IC50 ¼ 0.25 nM),23 with cathepsin K gave diffraction
quality crystals, allowing the determination of the
crystal structure. The X-ray co-crystal structure of
bromine derivative 9 bound to cathepsin K, of which the
active site is shown in Figure 1, gives some insights into
the SAR of these P2–P3 analogs described above. The
iso-butyl group is bound in a deep S2 pocket formed by
residues 67Tyr, 68Met, 134Ala, 163Ala, and 209Leu.
Hydrophobic groups should be preferred based on the

E. E. Boros et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3425–3429
3429
James, I. E.; Marquis, R. W.; Ru, Y.; Vasko-Moser, J. A.;
Smith, B. R.; Tomaszek, T.; Gowen, M. J. Bone Miner.
Res. 2001, 16, 1739.
the cathepsin K inhibitor potency, a comparison of
analogs 6l (IC50 ¼ 0.13 nM) and 6q (IC50 ¼ 4.0 nM)
shows that the P3 moiety does contribute to inhibitor
binding interactions. Although the data do not support
the original hypothesis that P3 aryl groups could form
p–p stacking interactions in S3, P3–S3 hydrophobic
interactions do account for additional inhibitory activ-
ity, suggesting that the P3 group of the docked inhibitor
6n needs further refinement.
3. Catalano, J. G.; Deaton, D. N.; Furfine, E. S.; Hassell, A.
M.; McFadyen, R. B.; Miller, A. B.; Miller, L. R.;
Shewchuk, L. M.; Willard, D. H.; Wright, L. L. Bioorg.
Med. Chem. Lett. 2004, 14, 275.
4. Berk, S. C.; Yeh, M. C. P.; Jeong, N.; Knochel, P.
Organometallics 1990, 9, 3053.
5. Vedejs, E.; Wang, J. Org. Lett. 2000, 2, 1031.
6. Lingibe, O.; Graffe, B.; Sacquet, M.-C.; Lhommet, G.
Heterocycles 1994, 37, 1469.
Aldehydes have a reputation for metabolic and chemical
instability.25 Not surprisingly, then, in vitro incubations
with rat S9 liver homogenates revealed that these alde-
hyde inhibitors were rapidly degraded.26 These results
were supported by in vivo experiments in male Han
Wistar rats. After dosing (iv dose ¼ 5 mg/kg or po
dose ¼ 10 mg/kg), the plasma concentration of a P1
derivative of aldehyde 6k fell below the assay detection
limit (<3 ng/mL) after 15 min, indicating rapid elimina-
tion in vivo. The poor pharmacokinetic properties of
this class of aldehyde inhibitors precluded further in
vivo pharmacodynamic studies. In spite of these disap-
pointing, but not unexpected results, the SAR gleaned
from this investigation and structural insight gained
from the cathepsin K/inhibitor co-crystal structure
proved valuable for the design of better inhibitors.
7. Wuts, P. G. M.; Pruitt, L. E. Synthesis 1989, 622.
8. Luly, J. R.; Hsiao, C. N.; BaMaung, N.; Plattner, J. J.
J. Org. Chem. 1988, 53, 6109.
9. Bernardon, C.; Deberly, A. J. Org. Chem. 1982, 47, 463.
10. Newkome, G. R.; Allen, J. W.; Anderson, G. M. J. Chem.
Educ. 1973, 50, 372.
11. Souppe, J.; Danon, L.; Namy, J. L.; Kagan, H. B.
J. Organomet. Chem. 1983, 250, 227.
12. Collins, S.; Hong, Y.; Hoover, G. J.; Veit, J. R. J. Org.
Chem. 1990, 55, 3565.
13. Ferrari, G. Farmaco, Edizione Scientifica 1960, 15, 337.
14. Langhals, H.; Range, G.; Wistuba, E.; Ruechardt, C.
Chem. Ber. 1981, 114, 3813.
15. The alcohol 2v was synthesized from 3-methoxyphenyl-
acetone and methyl magnesium bromide in 82% yield.
16. The alcohol 2z was synthesized from ethyl thiophene-
3-acetate and methyl magnesium bromide in 72% yield.
17. Effenberger, F.; Drauz, K. Angew. Chem. 1979, 91, 504.
18. Jeschkeit, H.; Losse, G.; Neubert, K. Chem. Ber. 1966, 99,
2803.
19. Kubota, M.; Nagase, O.; Yajima, H. Chem. Pharm. Bull.
1981, 29, 1169.
20. Blum, J.; Amer, I.; Zoran, A.; Sasson, Y. Tetrahedron
Lett. 1983, 24, 4139.
21. Waters, M. L. Curr. Opin. Chem. Biol. 2002, 6, 736.
22. Turk, D.; Guncar, G.; Podobnik, M.; Turk, B. Biol. Chem.
1998, 379, 137.
23. Calpeptin derivative 9 was synthesized by a carbodiimide
coupling of the known 4-bromo-benzyloxycarbonylleucine
with amino alcohol 4, followed by a Moffatt oxidation
of the alcohol to provide the desired aldehyde in 78%
yield.
In summary, a series of aldehyde inhibitors of cathepsin
K with varied substituents at the P2 and P3 positions
were synthesized. Starting from aldehyde lead 1, a P3
moiety was introduced into the inhibitor and the P2
substituent was optimized to produce picomolar
cathepsin K inhibitors such as 6g, 6l, and 6s. Analog 6g
was also very selective for inhibiting cathepsin K versus
cathepsins B and L. Information gained from these SAR
studies proved useful in the design of other cathepsin K
inhibitors containing more metabolically stable thiol-
reactive groups. The properties of these inhibitors will
be reported in due course.
24. Lambert, M. H. In Practical Application of Computer-
Aided Drug Design; Charifson, P. S., Ed.; Marcel Dekker:
New York, NY, 1997; p 243.
References and notes
25. Blume, H.; Oelschlaeger, H. Arzneim.-Forsch. 1978, 28,
956.
26. Iwatsubo, T.; Hirota, N.; Ooie, T.; Suzuki, H.; Shimada,
N.; Chiba, K.; Ishizaki, T.; Green, C. E.; Tyson, C. A.;
Sugiyama, Y. Pharmacol. Ther. 1997, 73, 147.
1. Yamashita, D. S.; Dodds, R. A. Curr. Pharm. Des. 2000,
6, 1.
2. Stroup, G. B.; Lark, M. W.; Veber, D. F.; Bhattacharyya,
A.; Blake, S.; Dare, L. C.; Erhard, K. F.; Hoffman, S. J.;