N-Arylaminonitriles as bioavailable peptidomimetic inhibitors of cathepsin B

Article abstract of DOI:10.1016/j.bmcl.2003.08.006

To improve the pharmacokinetics of a previously reported series of dipeptidyl nitrile cathepsin B inhibitors, the P₂-P₃ amide group was replaced with an arylamine. Further optimization of this template resulted in highly potent and s

Full text of DOI:10.1016/j.bmcl.2003.08.006

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4121–4124
N-Arylaminonitriles as Bioavailable Peptidomimetic
Inhibitors of Cathepsin B
Paul D. Greenspan,* Kirk L. Clark, Scott D. Cowen, Leslie W. McQuire,
Ruben A. Tommasi, David L. Farley, Elizabeth Quadros, David E. Coppa,
Zengming Du, Zheng Fang, Huanghai Zhou, John Doughty,
Karen T. Toscano, Andrew M. Wigg and Siyuan Zhou
Novartis Institute of Biomedical Research, One Health Plaza, East Hanover, NJ 07936, USA
Received 30 April 2003; revised 31 July 2003; accepted 6 August 2003
Abstract—To improve the pharmacokinetics of a previously reported series of dipeptidyl nitrile cathepsin B inhibitors, the P2–P3
amide group was replaced with an arylamine. Further optimization of this template resulted in highly potent and selective inhibitors
with excellent oral availability.
# 2003 Elsevier Ltd. All rights reserved.
Cathepsin B (cat B), a lysosomal cysteine protease, has
been implicated in the pathology of a number of
important human diseases. Cat B is upregulated in the
fold selectivity versus cathepsins S and L. Unfortu-
nately, compound 2 and its analogues were found to
have very poor pharmacokinetic properties and could
not be used for in vivo testing. It was hypothesized that
the highly peptidic nature of compounds such as 2 was
contributing to this problem, perhaps through amide
hydrolysis or low intestinal permeability. This report
describes our efforts to use a structure-guided approach
to partially depeptidize these compounds, resulting in
highly potent and selective cathepsin B inhibitors with
much more favorable pharmacokinetic properties.
1
synoviumof patients with rheumatoid arthritis
and
selective cat B inhibitors have been found to block pro-
2
teoglycan release fromarticular cartilage. Cat B has
also been implicated in the progression and invasion of
tumors, by aiding in extracellular matrix breakdown.3
More recently, cat Bꢀ/ꢀ mice have been found to be
resistant to TNF-a mediated hepatocyte apoptosis,
revealing a potential contributory role in various types
of liver injury.4 It also appears that cat B facilitates
CD8+ T-cell mediated cytotoxicity, by protecting these
5
cells fromtheir own lethality. This may indicate a cen-
tral role for cat B in host defense. The wide-ranging
functions of this enzyme therefore make it an attractive
drug discovery target.
Our group has previously reported on a series of dipep-
tidyl nitrile inhibitors of cat B.6 These compounds were
designed with a carboxylate-containing substituent at
the P1 position, whi0ch was capable of interacting with
His110 within the S2 site of cat B. Thus, the moderately
potent (50 nM) but nonselective compound 1 was opti-
mized to compound 2, a 6 nM cat B inhibitor with 100-
The X-ray structure of a compound 1–cat B complex6
revealed hydrogen-bonding interactions between the
backbone of the enzyme and both NH’s of the inhibitor,
as well as the P2 (phenylalanyl) carbonyl group. How-
ever, the P3 (diphenylacetyl) carbonyl does not appear
to play any role in cat B binding (Fig. 1). With this in
*Corresponding author at present address: MilleniumPharmaceuticals
Inc., 75 Sidney Street, Cambridge, MA 02139, USA. Tel.: +1-617-444-
1691; fax: +1-617-551-8908; e-mail: paul.greenspan@mpi.com
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.08.006

4122
P. D. Greenspan et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4121–4124
Figure 1. Schematic representation of H-bonding pattern of nitrile
inhibitors to cat B, and proposed modification of P3 amide.
Scheme 2. Reagents and conditions: (a) NaNO2, NaOAc, HOAc;
(b)ꢀTsOH, EtOH, reflux; (c) Tf2O, pyridine; (d) ArNH2, THF; (e)
LiOH, THF, water; (f) ArBr or ArI, (P(o-tol)3)2PdCl2, CuI, TEBA,
K2CO3, TEA, DMF, water.
mind, we designed compounds that would maintain the
relatively acidic P2 NH, but would lack the P3 carbonyl
group. To this end, replacement of the amide with an N-
arylamine appeared to be attractive for several reasons:
(1) the aniline NH was expected to have similar H-bond
donating properties to the amide NH, (2) the aryl group
still provided a hydrophobic substituent for binding in
the S3 pocket, and (3) the modification was not expected
to affect the P1 or P2 substituent orientations, so that
reoptimization of these portions of the molecule would
probably not be necessary.
In the early stages of this effort, the N-aryl amino acids 3
were prepared through displacement of the corresponding
triflated R-hydroxyester 8, which were themselves pre-
pared from R-amino acids 7 using standard methodology
(Scheme 2). Compounds 14–16 were prepared in this
manner (Table 1). Later, however, the more convenient
Pd(0)/Cu(I)-catalyzed amino acid arylation conditions
reported by Ma were employed using the appropriate
aryl bromide or iodide as the coupling partner.8 Com-
pounds 17–20 and 23 were prepared in this manner.
The desired compounds 6 were assembled in one of two
ways (Scheme 1). When Ar=Ph, N-arylation was car-
ried out on the pre-assembled amino nitrile 5 using tri-
phenylbismuth diacetate (path b), as described by
Barton.7 For substituted aromatic groups, the N-aryl
amino acid 3 was prepared and then coupled with an
amino nitrile 4 (path a). Preparations of amino nitriles 4
and 5 were carried out as previously described.6 To gen-
erate compounds with a carboxylic acid on the P1 sub-
stituent, esters 4 or 5 were prepared (R0=CO2allyl) and
deprotected in the final step.6 The corresponding P1-sub-
stituted tetrazoles were introduced with a cyanoethyl pro-
tecting group and were revealed using DBU in CH2Cl2.
In order to provide adequate throughput for assessment
of oral exposure, compounds were subjected to a single
timepoint plasma drug level measurement as an initial
screen. To accomplish this, compounds were dosed
orally to rats.9 Plasma concentrations were then mea-
sured 4 h post-dosing, using HPLC/MS.
Data for the compounds are presented Table 1. We
were gratified to find that replacement of the P3
amide with a simple phenyl group (compound 10)
resulted in a nanomolar cat B inhibitor. Although this
compound was somewhat less potent than 2, its 4-h
Scheme 1. Reagents and conditions: (a) EDCI, HOBt, NMM,
CH2Cl2; (b) Ph3Bi(OAc)2, Cu(OAc)2, CH2Cl2.
Figure 2. Molecular model of compound 16 bound to cat B, showing
key H-bonds between enzyme and inhibitor.

P. D. Greenspan et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4121–4124
4123
plasma concentration was much higher than that of 2
(1.1 mM for 10 vs 0.09 mM for 2).
To improve the in vitro potency of this series, we
examined a model of compound 10 bound to cat B,
which was derived fromthe compound 1–cat B X-ray
structure.6 It appeared fromthe model that an H-bond
acceptor attached to the meta-position of the arylamine
would be correctly positioned to interact with the back-
bone NH of Gly198 (Fig. 2). A meta-sulfone substituent
at this position resulted in a 4-fold increase in potency
As expected, the P2 and P1 substituent SAR for the N-
aryl amines was identical that of the P3-acyl series:6
removal (11) or repositioning (12) of the P1 carboxylate
resulted in reduced potency, as did removal of the 3-
methyl group on the P2 phenyl ring (13).
Table 1. Structures and data for compounds discussed in paper
Compd
Ar
R
Z
Cat B IC50 (nM)
Cat L IC50 (nM)
Cat S IC50 (nM)
Plasma concn.a (mM)
10
11
12
13
14
15
Ph
Ph
Ph
Me
Me
Me
H
Me
Me
3-COOH
H
4-COOH
3-COOH
3-COOH
3-COOH
49
194
282
832
13.3
8.6
1800
220
2800
6100
—
2000
1200
3900
2600
—
1.1
—
—
—
0.037
—
Ph
3-(MeSO2)–Ph
3-(COOMe)–Ph
210
710
16
17
18
19
20
Me
Me
Me
Me
Me
3-COOH
3-COOH
3-COOH
3-COOH
3-COOH
6.5
5.3
5.3
9.3
4.9
—
—
—
—
0.44
0.07
560
780
0.11
>1000
>1000
>1000
>1000
0.038
0.015
21
22
Ph
Ph
Me
Me
3-OH
3-Tetrazole
>1000
—
—
—
—
—
0.23
16.5
23
24
Me
Me
3-Tetrazole
5.0
4.1
2400
1100
1000
1000
0.037
0.056
3-COOH, 4-F
25
26
Ph
Ph
Me
Me
3-COOH, 4-F
3-COOH, 4-Cl
12.2
8.7
1400
1400
1600
1700
5.27
1.88
— No data available for this compound.
aPlasma concentration is measured 4 h after a 30 mg/kg po dose.

4124
P. D. Greenspan et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4121–4124
Table 2. Pharmacokinetic parameters for compound 25
Tmax (h)
t1/2 (h)
CL (mL/min/kg)
Vss (L/kg)
F (%)
0.50ꢁ0.07
0.24ꢁ0.02
64.8ꢁ0.61
1.22ꢁ0.07
128ꢁ59
A full PK assessment in the rat was made for compound
25. Figure 3 shows that micromolar plasma concentra-
tions are maintained for at least 7 h after a 30 mg/kg po
dose, with a very high Cmax. Table 2 provides the PK
parameters for this compound, indicating excellent oral
availability (%F), although clearance and t1/2 values are
still not optimal.
Figure 3. Plasma concd (mM) versus time after 30 mg/kg po dose in rats.
(14). An even greater enhancement was observed with a
methyl ester (15). The best potencies were obtained
fromcyclized carbonyl compounds, in which the car-
bonyl is locked in the preferred conformation: lactones
(16, 18), lactam( 19), phthalimide (20), and cyclic ketone
(17). Unfortunately, none of these compounds displayed
4-h plasma levels as high as 10, and were not pursued
further. However, this endeavor established the feasi-
bility of preparing arylamines with potencies compar-
able to the best P3 amide inhibitors.
In conclusion, we have demonstrated that structure-
based peptidomimetic design has allowed for the
identification of a series of potent, selective and highly
orally bioavailable inhibitors of cathepsin B. These
compounds are currently being profiled in animal mod-
els to further delineate the role of this enzyme in disease
processes.
Our previous work established the importance of the P1
carboxylate, which is thought to forman electrostatic
0
interaction with His110 and His111 of the S2 pocket of
References and Notes
cat B (see Fig. 2).10 With this in mind, we investigated
the effect of carboxylate acidity on cat B potency. ortho-
Substitution with a fluoro or chloro group (25, 26) is
expected to decrease the pKa of the carboxylate by
approximately 1.0 and 1.3 log units,11 and increased the
potency by 4- and 5.6-fold, respectively. In addition,
these compounds maintain micromolar plasma con-
centrations at the 4-h timepoint. Replacement of the
carboxylate with a tetrazole also resulted in a 4-fold
increase in potency (22), but this compound displayed
lower plasma levels than the carboxylates. The additive
effect of the P3 and P1 modifications can be observed in
the enhanced potency of compounds 23 and 24.
1. Balaji, K. N.; Schaschke, N.; Machleidt, W.; Catalfamo,
M.; Henkart, P. A. J. Exp. Med. 2002, 196, 493.
2. Buttle, D. J.; Handley, C. J.; Ilic, M. Z.; Saklatvala, J.;
Murata, M.; Barrett, A. J. Arthritis Rheum. 1993, 36, 1709.
3. Sameni, M.; Moin, K.; Sloane, B. F. Neoplasia (New York)
2000, 2, 496.
4. Guicciardi, M. E.; Deussing, J.; Miyoshi, H.; Bronk, S. F.;
Svingen, P. A.; Peters, C.; Kaufmann, S. H.; Gores, G. J. J.
Clin. Invest. 2000, 106, 1127.
5. Hashimoto, Y.; Kakegawa, H.; Narita, Y.; Hachiya, Y.;
Hayakawa, T.; Kos, J.; Turk, V.; Katunuma, N. Biochem.
Biophys. Res. Commun. 2001, 283, 334.
6. Greenspan, P. D.; Clark, K. L.; Tommasi, R. A.; Cowen,
S. D.; McQuire, L. W.; Farley, D. L.; van Duzer, J. H.;
Goldberg, R. L.; Zhou, H.; Du, Z.; Fitt, J. J.; Coppa, D. E.;
Fang, Z.; Macchia, W.; Zhu, L.; Capparelli, M. P.; Goldstein,
R.; Wigg, A. M.; Doughty, J. R.; Bohacek, R. S.; Knap, A. K.
J. Med. Chem. 2001, 44, 4524.
As can be seen in Table 1, all compounds containing the
P1 carboxylate are selective for cat B over cat L and cat S.
The lower selectivity (as well as potency) of compound 11
supports the hypothesis that the S0 carboxylate recogni-
tion site is unique for cat B. Incorporation of esters or
lactones into the P3 phenyl group increases potencies
against all three enzymes, although the lactam and
imides are highly selective (19 and 20, respectively).
Halogen substitution on the P1 benzoate improves
potency only against cat B. Because of this, compounds
25 and 26 have >100-fold selectivity against the other
cathepsins.
7. Barton, D. H. R.; Finet, J. P.; Khamsi, J. Tetrahedron Lett.
1989, 30, 937.
8. Ma, D.; Yao, J. Tetrahedron: Asymmetry 1996, 7, 3075.
9. Compounds were dosed as a suspension in an aqueous vehi-
cle containing 20% PEG400, 0.5% CMC, and 0.25% Tween 80.
10. Turk, D.; Podobnik, M.; Popovic, T.; Katunuma, N.;
Bode, W.; Huber, R.; Turk, V. Biochemistry 1995, 34, 4791.
11. Dean, J. A. Lange’s Handbook of Chemistry, 15th ed.;
McGraw-Hill: New York, 1999.