Design and synthesis of selective keto-1,2,4-oxadiazole-based tryptase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.04.013

Using a scaleable, directed library approach based on orthogonally protected advanced intermediates, we have prepared a series of potent keto-1,2,4-oxadiazoles designed to explore the P₂ binding pocket of human mast cell tryptase, while buildin

Full text of DOI:10.1016/j.bmcl.2006.04.013

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3434–3439
Design and synthesis of selective keto-1,2,4-oxadiazole-based
tryptase inhibitors
James T. Palmer,* Robert M. Rydzewski, Rohan V. Mendonca, David Sperandio,
Jeffrey R. Spencer, Bernard L. Hirschbein, Julia Lohman, Jeri Beltman,
Margaret Nguyen and Liang Liu
Celera Genomics, 180 Kimball Way, South San Francisco, CA 94080, USA
Received 2 March 2006; revised 31 March 2006; accepted 3 April 2006
Available online 27 April 2006
Abstract—Using a scaleable, directed library approach based on orthogonally protected advanced intermediates, we have prepared
a series of potent keto-1,2,4-oxadiazoles designed to explore the P2 binding pocket of human mast cell tryptase, while building in a
high degree of selectivity over human trypsin and other serine proteases.
Ó 2006 Elsevier Ltd. All rights reserved.
Mast cells are involved in allergic disease by virtue of
pro-inflammatory mediators stored inside the secretory
granules. Upon degranulation, mucosal and respiratory
mast cell subpopulations release numerous effectors
including the tetrameric serine protease tryptase. With
trypsin-like specificity, tryptase processes substrates
with immediate and long-term effects with all of the
physiological responses associated with asthma. For
example, tryptase activates kininogens, leading to the
potent bronchoconstrictor bradykinin, and it is capable
of amplifying its own signal through stimulating mast
cells to further degranulate. Tryptase also processes
vasoactive intestinal peptide, or down-regulating this
bronchodilatory agent, and it behaves as a potent mito-
gen toward smooth muscle cells, endothelial cells, and
fibroblasts. This effect on smooth muscle cells is of
primary concern for long-term sufferers of acute
asthma.1–14 Crystallization of tryptase in 1998 by Pere-
ira and colleagues revealed a homotetrameric structure
that was amenable to structure-guided design.15 Previ-
ously, we have described the preparation of a-ketohet-
erocycles elaborated with prime-side (P0) binding
moieties to afford potent and selective tryptase inhibi-
tors.16 The best-characterized example in this series (1,
Fig. 1) showed potency of 5.4 nM for tryptase. Our goal
in this program required that we improve selectivity over
trypsin and other enzymes, and to improve drug-like
properties within this series. We also sought to develop
a robust synthetic strategy that would allow rapid ana-
logue preparation and large-scale synthesis of these
inhibitors to support in vivo studies (See Table 1).
Our improved synthesis of [1,2,4]oxadiazole-containing
tryptase inhibitors took advantage of orthogonal
protecting groups as illustrated in Scheme 1. Commer-
cially available N-a-benzyloxycarbonyl-N-e-tert-butyl-
oxycarbonyl-L-lysine 2 was converted to its N-a-
allyloxycarbonyl derivative 4 via the Weinreb-amide 3.
Reduction to the aldehyde 5 followed by treatment with
acetone cyanohydrin afforded the diastereomeric cyano-
hydrins 6a,b. Subsequent protection as the diastereo-
meric TBS ethers 7a,b and then conversion to the N-
hydroxy-amidines 8a,b in the presence of 50% aqueous
hydroxylamine in ethanol at 50 °C provided the needed
material for analogue preparation, without the need for
intermediate chromatography.
Taking advantage of our distal pocket binding knowl-
edge on the prime side of the core, we then prepared
further advanced intermediates suitable for analogue
exploration. The use of the 3,4-dichlorophenethyloxy-
phenyl group (Scheme 2) gave us not only a modest
increase in potency over the lead structure 1, but also
provided us with a more facile, higher-yielding synthesis
that again avoided chromatography. Mitsunobu-based
coupling between 3,4-dichlorophenylethanol 9 and
Keywords: Tryptase; Serine protease; Inhibitor; Heterocycle; Keto-
heterocycle; Selective; Scaleable process; Pharmacokinetics.
*
Corresponding author. Tel.: +1 650 624 1341; e-mail:
jpalmer@rigel.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.04.013

J. T. Palmer et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3434–3439
3435
Table 1. Enzyme inhibition and pharmacokinetics data for 1
center. For ease of preparation in subsequent steps, we
used the crystalline material (stereochemistry unas-
signed) and at a later stage repeated the sequence with
a mixture of 14a,b and obtained similar results.
O
H
O
N
N
O
N
O
O
With the a-amino group now available for derivatiza-
tion, we prepared a library of compounds to explore
the P2 residue of tryptase (Scheme 3). Through coupling
with the appropriate acylating agent (acid chloride, ami-
nocarbonyl chloride, isocyanate, etc.) in the presence of
appropriate bases, 15a,b derivatives of type 16a,b were
made. Deprotection of the silyl group with tetrabutylam-
monium fluoride gave alcohols 17a,b, which were then
oxidized using either Dess–Martin periodinane or
through a Swern procedure to give the Boc protected,
penultimate intermediate 18. Finally, HCl-mediated
deprotection of the N-e group yielded inhibitors 19–44.
1
NH2
Potency (lM)
Human B tryptase
Selectivty (lM)
Trypsin
0.0054
0.190
64
Thrombin
Plasmin
0.43
43
Kallikrein
APC
>150
>150
>150
>150
>150
>150
Chymotrypsin
Elastase
Chymase
Urokinase
Granzyme K
Compounds 19–44 were assayed against tryptase and
trypsin according to conditions outlined in Ref. 17.
Table 2 shows the results as organized by P2 binding
moiety, represented by R2, as attached through linker L.
methyl 4-hydroxyphenylacetate 10, followed by saponi-
fication of the ester, extraction of neutral by-products,
acidification, and filtration, permitted the product 11
to be isolated in near quantitative yield. 11 was convert-
ed to its N-hydroxysuccinimidoyl ester 12, which was
then coupled with 8a,b under neutral conditions to give
the intermediate esters 13a,b and then cyclized via heat-
ing in toluene, utilizing a Dean–Stark type apparatus.
The diastereomeric 1,2,4-oxadiazoles, 14a,b, were then
N-a-deprotected via Pd-catalyzed reductive cleavage
using tributyl stannane to give 15a,b, which were thus
set up for the final stages of the synthesis. Although
the diastereomeric nature of the intermediates (via the
silyloxy group) complicated intermediate analysis, we
were fortunate in discovering that one of the diastereo-
mers of 14a,b could in fact be isolated by crystallization,
although at this point we did not care which one. Subse-
quently, the silyl group was to be removed and the
secondary alcohol oxidized, thus removing the chiral
While good potency was inherent in this series, thanks
to the distal pocket binding moiety on the prime side,
the goal of achieving several 100-fold selectivity over hu-
man trypsin required an extensive analysis of different
binding elements elsewhere in the active site. In this ser-
ies, we explored the effects of aliphatic and aromatic
amides, carbamates, and ureas as linkers between the ly-
sine a-amine group and the P2-targeting group. Simple
aliphatic amides (compounds 19–25) showed the lowest
selectivity, with only the sterically hindered (and possi-
bly anomalous, within this series) pivalamide 36 display-
ing the targeted selectivity. Carbamates, both simple and
extended (26, 27, 30, and 32) showed modest (80- to 130-
fold) selectivity, suggesting that the atom next to the
linker carbonyl should bear minimal substituents. The
ureas began to show an improvement in selectivity (29,
31, and 33–35) but we still felt that the intrinsic potency
against trypsin was too high. When we introduced an
O
O
O
H
CbzHN
CbzHN
O
N
O
b, c
a
OH
N
Alloc
N
4
2
3
NHBoc
NHBoc
NHBoc
d
O
H
Si
Si
O
N
O
g
Alloc
H
H
e, f
N
NH2
N
Alloc
Alloc
CN
N
OH
NHBoc
7a, b
8a, b
5
NHBoc
NHBoc
Scheme 1. Reagents and conditions: (a) N,O-Dimethylhydroxylamine, DCC, Et3N, CH2Cl2, rt; (b) H2/10% Pd, EtOH; (c) Alloc-Cl, Et3N, THF, rt;
(d) LiAlH4, THF, 0 °C; (e) acetone cyanohydrin, NEt3, CH2Cl2, rt; (f) TBSCl, imidazole, DMAP, CH2Cl2, rt; (g) NH2OH 50 wt% in H2O, EtOH,
50 °C.

3436
J. T. Palmer et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3434–3439
Cl
Cl
Cl
9
O
HO
b, c
O
O
Cl
O
O
OH
a
O
10
Cl
Cl
d
HO
Cl
O
Cl
N
O
O
O
O
O
12
11
OTBS
H
NH2
e
N
O
Alloc
f
O
13a, b
N
8a, b
Cl
O
Cl
NHBoc
OTBS
N
OTBS
N
H
N
H2N
O
O
g
N
Alloc
N
Cl
O
O
Cl
Cl
Cl
15a, b
14a, b
NHBoc
NHBoc
Scheme 2. Reagents and conditions: (a) PPh3, DIAD, CHCl3; (b) NaOH, THF/H2O; (c) Et2O extraction, followed by acidification to pH 3;
(d) N-hydroxysuccinimide, DCC, THF, 70 °C; (e) THF, rt; (f) toluene, reflux, Dean–Stark trap; (g) Pd(PPh3)4, BuSn3H, THF, rt.
R2-L-X
a
OTBS
N
OTBS
N
H
R2
L
N
O
H2N
O
N
N
O
Cl
Cl
O
Cl
Cl
16
15a, b
L = CO, NR'CO etc.
NHBoc
NHBoc
b
OH
H
R2
L
N
O
c
N
N
O
Cl
Cl
O
NHBoc
O
O
H
N
H
R2
L
O
R2
L
N
O
d
N
N
N
N
Cl
Cl
O
Cl
Cl
18
19-44
NH2.HCl
NHBoc
Scheme 3. Reagents and conditions: (a) Et3N, THF etc.; (b) Bu4NF, THF; (c) Dess–Martin periodinane, CH2Cl2, rt, or DMSO, (COCl)2, Et3N,
CH2Cl2, ꢀ78 °C; (d) 4.0 M HCl/dioxane, rt.
aromatic amide at R2, we achieved the desired increase
in selectivity but not at the cost of potency, as exempli-
fied by compounds 38–44. In one case the heteroaromat-
ic example 39 showed a modest increase in trypsin
potency, so fine tuning of this substituent was clearly
called for. Ultimately, the p-fluorobenzamide 44
displayed both the highest potency (2.5 nM) and
selectivity (440-fold) over trypsin that we had yet seen
in this program. The 3,4-difluorobenzamide 43 was
equally selective, and we profiled this compound further
against an extended panel of serine proteases. This
derivative showed greater than 100-fold selectivity
against over a dozen serine proteases in both trypsin
and chymotrypsin-like sub-families. Table 3 indicates
the results obtained.
An additional goal of this program was to improve
pharmacokinetic parameters, for instance increasing
terminal half-life and lowering clearance of these com-
pounds when dosed in vivo, so that they could be
used in animal models of disease. We made the initial
choice of the 3,4-dichlorophenethyl group in the distal
pocket to block oxidative pathways wherein unsubsti-
tuted phenyl groups might be susceptible to oxidation
and rapid clearance. The same logic was applied in
the use of fluorobenzamides at the P2 position; to this

J. T. Palmer et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3434–3439
Table 2. Enzyme inhibition data for compounds 19–50
3437
Linker (L)
H
O
R2
N
O
N
O
N
O
Cl
Cl
NH2.HCl
Compound
R2
L type
Tryptase
0.0046
Trypsin
0.044
Selectivity
10
19
Aliphatic amide
Aliphatic amide
Aliphatic amide
20
21
0.0068
0.0098
0.12
0.18
18
18
n-Pentyl
22
23
24
25
26
27
Aliphatic amide
Aliphatic amide
Aliphatic amide
Aliphatic amide
Carbamate
0.0065
0.022
0.3
46
50
51
76
81
83
O
1.1
O
0.0037
0.0029
0.0047
0.005
0.19
0.22
0.38
0.415
CF3
O
Carbamate
O
O
28
(Hetero)aromatic amide
0.046
3.9
85
29
30
31
32
Urea
0.0096
0.005
0.032
0.017
0.85
0.5
3.8
2.2
89
100
119
129
N
O
Carbamate
Urea
O
N
O
Carbamate
33
34
Urea
Urea
0.018
2.4
1.4
133
152
N
N
0.0092
35
36
37
Urea
0.0057
0.00375
0.011
0.87
0.625
1.9
153
167
173
O
N
t-Bu
Aliphatic amide
Carbamate
t-Bu
F
O
38
Aromatic amide
0.0062
1.1
177
O
39
40
(Hetero)aromatic amide
Aromatic amide
0.0023
0.0058
0.48
1.7
209
293
F
F
41
42
Aromatic amide
Aromatic amide
0.00325
0.00655
1.15
2.55
354
389
F
Cl
(continued on next page)

3438
J. T. Palmer et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3434–3439
Table 2 (continued)
Compound
R2
L type
Tryptase
0.0076
Trypsin
3.3
Selectivity
434
F
43
44
Aromatic amide
F
Aromatic amide
0.0025
1.1
440
F
Inhibition constants (Ki(app.)) are given in lM.
Table 3. Enzyme inhibition data for compound 43
Table 4. Pharmacokinetic data for compound 44
F
F
O
H
F
N
O
O
H
N
N
O
O
N
N
O
Cl
O
N
Cl
Cl
O
44
Cl
NH2.HCl
43
NH2.HCl
Potency (lM)
Potency (lM)
Human B tryptase
Mouse tryptase
Monkey tryptase
0.0025
0.365
0.40
Human B tryptase
Mouse tryptase
Monkey tryptase
Selectivity (lM)
Trypsin
0.0076
0.365
0.40
PK (0.5 mg/kg, iv, rat, n = 3)
Cmax (lM)
CL (ml/min/kg)
1.01 0.25
45 7.9
3.3
Thrombin
Plasmin
>150
0.91
83
Vc (ml/kg)
Vss (ml/kg)
457 113
7800 1280
173 14
Kallikrein
APC
MRT (min)
AUC (lm min)
Terminal half-life (min)
PK (5 mg/kg, po, rat, n = 3)
Absorption (%)
>150
>150
>150
>150
>150
>150
>150
>150
>150
>150
>150
18 3.3
174 16
Chymotrypsin
Elastase
Chymase
12
9
135 59
Urokinase
Bioavailability (%)
Terminal half-life (min)
Granzyme K
Cathepsin G
Factor VIIa
Factor Xa
restricted compound member of the series that may be
further developed as a possible therapy for indications
implicated by excess mast-cell tryptase release.
Factor IXa
Factor XIa
end we performed pharmacokinetic studies18 on com-
pound 44, which showed a similarly high degree of
selectivity against all proteases tested. Table 4 shows
the results.
References and notes
1. Cairns, J. A. Pulm. Pharmacol. Ther. 2005, 18, 55.
2. Payne, V.; Kam, P. C. A. Anaesthesia 2004, 59, 695.
3. Fiorucci, L.; Ascoli, F. Cell. Mol. Life Sci. 2004, 61, 1278.
4. Caughey, G. H. Handbook of Proteolytic Enzymes, 2nd
ed., 2004; Vol. 2, p 1535.
5. Bradding, P. Curr. Opin. Allergy Clin. Immunol. 2003, 3,
45.
6. Newhouse, B. J. IDrugs 2002, 5, 682.
7. Abraham, W. M. Am. J. Physiol. 2002, 282, L193.
8. Gangloff, A. R. Curr. Opin. Investig. Drugs (PharmaPress
Ltd.) 2000, 1, 79.
9. Rice, K. D.; Sprengeler, P. A. Curr. Opin. Drug Discov.
Devel. 1999, 2, 463.
10. Rice, K.; Spencer, J. Expert Opin. Ther. Patents 1999, 9,
1537.
11. Elrod, K. C.; Numerof, R. P. Emerging Therapeutic
Targets 1999, 3, 203.
The long half-life values and mean residence time are
consistent with limited first-pass metabolism. Thus, we
achieved our intent of blocking metabolic hot-spots by
use of the halogenated aromatic rings. The relatively
low absorption (12%, cassette dosing) could be ex-
plained by a number of factors, including the relatively
high molecular weight (599.5) of the parent free base
of 44, and also the presence of a primary amine func-
tion. That the oral bioavailability of 9% represented a
high percentage (75%) of absorbed 44 suggested further
evidence that first-pass metabolism pathways were
curtailed.
In summary, we have demonstrated a series of highly
potent and selective human tryptase inhibitors, an effi-
cient method by which they can be prepared in library
fashion, and finally, an example of a metabolically
12. Rice, K. D.; Tanaka, R. D.; Katz, B. A.; Numerof, R. P.;
Moore, W. R. Curr. Pharm. Des. 1998, 4, 381.
13. Clark, J. M.; Moore, W. R.; Tanaka, R. D. Drugs Future
1996, 21, 811.

J. T. Palmer et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3434–3439
3439
14. Rice, K. D.; Moore, W. R. High Throughput Screen. 2000,
101.
15. Pereira, P. J. B.; Bergner, A.; Macedo-Ribeiro, S.; Huber,
R.; Matschiner, G.; Fritz, H.; Sommerhoff, C. P.; Bode,
W. Nature (London) 1998, 392, 306.
incubated at 2 nM with variable concentrations of inhib-
itor in 50 mM Tris (pH 7.4), 150 mM NaCl, 0.02% Tween
20, 1 mM EDTA, 50 lg/ml heparin, and 10% DMSO. The
reaction was initiated with substrate, Tosyl-Gly-Pro-Lys-
pNA (Centerchem), supplied at the Km (400 lM). The
change in absorbance as a function of time was monitored
at 405 nm.Trypsin—trypsin (Athens Research Institute)
was incubated at 10 nM with variable concentrations of
inhibitor in 50 mM Tris (pH 7.4), 150 mM NaCl, 1.5 mM
EDTA, 0.05% Tween 20, and 10% DMSO. The reaction
was initiated with substrate, Tosyl-Gly-Pro-Lys-pNA
(Centerchem), supplied at the Km (25 lM). The change
in absorbance as a function of time was monitored at
405 nm.
16. Sperandio, D. et al. Bioorg. Med. Chem. Lett. submitted
for publication.
17. General inhibition assays—the active site concentrations
of the various enzymes used for inhibition studies were
determined through active site titration with 4-methylum-
belliferyl p-guanidinobenzoate or were estimated from
protein concentration measurements. Inhibitor potency
measurements were performed at room temperature using
a Molecular Device’s SpectraMax 250 96-well kinetic plate
readers. Reaction velocities were monitored at varying
inhibitor concentrations by following the hydrolysis the
liberation of p-nitroaniline (A405nm). All substrates were
added at concentrations equal to or near their Km. All
reactions were performed in a total volume of 100 lLs.
Typically, enzyme and inhibitor were incubated 30 min
prior to initiation of reaction with the addition of
substrate. Control reactions in the absence of inhibitor
were performed in parallel. The velocity of the enzyme-
catalyzed reaction was determined from the linear portion
of the progress curve, typically the first 5 min of the
reaction after initiation with the addition of substrate. The
Ki values were determined by a non-linear least squares
regression fit of the experimentally derived data to the
Morrison equation for tight-binding inhibitors as
described (Kuzmic, P., et al., Anal. Biochem., 2000, 281,
p. 62). Example conditions for the tryptase and trypsin
assays are provided below, other enzyme assays for which
selectivity data are provided were performed in a similar
manner. Example enzyme assay conditions Tryptase—
tryptase (recombinant human Beta-1 from Promega) was
18. Plasma concentrations of the compounds were deter-
mined by LC/MS/MS. The plasma sample was processed
using acetonitrile precipitation, then the supernatant was
injected onto the LC column. The limit of quantitation
of the assay was 1–6 nM. Pharmacokinetic data were
analyzed by WinNonlin-Pro (Pharsight Corp.), using
compartmental and non-compartmental analysis for iv
and po data, respectively. Pharmacokinetic parameters,
including the area-under-the-curve (AUC), clearance
(CL), volume of distribution (Vd), and mean residence
time (MRT), were determined. Clearance and mean
residence time were calculated as follows: CL = dose/
AUC and MRT = Vd/CL. Oral absorption (Abs) and
bioavailability (F) in rats were evaluated in portal vein
(pv) and jugular vein (jv) cannulated animals. The
plasma concentrations of the compounds in the portal
and jugular vein were quantified and used to calculate
the area-under-the-curve (AUC). Oral absorption and
bioavailability were calculated from dose-normalized
AUC values as follows: Abs = AUCpo,pv/AUCiv,jv and
F = AUCpo,jv/AUCiv,jv
.