Potent, nonpeptide inhibitors of human mast cell tryptase. Synthesis and biological evaluation of novel spirocyclic piperidine amide derivatives

Article abstract of DOI:10.1016/j.bmcl.2008.01.093

We have explored a series of spirocyclic piperidine amide derivatives (5) as tryptase inhibitors. Thus, 4 (JNJ-27390467) was identified as a potent, selective tryptase inhibitor with oral efficacy in two animal models of airway inflammation (sheep and guinea pig asthma models). An X-ray co-crystal structure of 4 · tryptase revealed a hydrophobic pocket in the enzyme's active site, which is induced by the phenylethynyl group and is comprised of amino acid residues from two different monomers of the tetrameric protein.

Full text of DOI:10.1016/j.bmcl.2008.01.093

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2114–2121
Potent, nonpeptide inhibitors of human mast cell tryptase.
Synthesis and biological evaluation of novel spirocyclic
piperidine amide derivatives
Michael J. Costanzo,a,* Stephen C. Yabut,a Han-Cheng Zhang,a Kimberley B. White,a
Lawrence de Garavilla,a Yuanping Wang,a Lisa K. Minor,a Brett A. Tounge,a
Alexander N. Barnakov,a Frank Lewandowski,a Cynthia Milligan,a John C. Spurlino,a
William M. Abraham,b Victoria Boswell-Smith,c Clive P. Pagec and Bruce E. Maryanoff a,*
aResearch and Early Development, Johnson & Johnson Pharmaceutical Research & Development, Welsh & McKean Roads,
Spring House, PA 19477-0776, USA
bUniversity of Miami at Mount Sinai Medical Center, Miami Beach, FL 33104, USA
cThe Sackler Institute of Pulmonary Pharmacology, King’s College London, School of Biomedical and Health Sciences, London
WC2R 2LS, UK
Received 30 November 2007; revised 21 January 2008; accepted 23 January 2008
Available online 30 January 2008
Abstract—We have explored a series of spirocyclic piperidine amide derivatives (5) as tryptase inhibitors. Thus, 4 (JNJ-27390467)
was identified as a potent, selective tryptase inhibitor with oral efficacy in two animal models of airway inflammation (sheep and
guinea pig asthma models). An X-ray co-crystal structure of 4 Æ tryptase revealed a hydrophobic pocket in the enzyme’s active site,
which is induced by the phenylethynyl group and is comprised of amino acid residues from two different monomers of the tetrameric
protein.
Ó 2008 Elsevier Ltd. All rights reserved.
The serine protease tryptase (EC 3.4.21.59) is an impor-
tant mediator of mast cell-dependent allergic and
inflammatory responses.1 This trypsin-like, heparin-sta-
bilized, homotetrameric enzyme is stored in mast cell
granules and released upon degranulation.1 Inhibitors
of tryptase have therapeutic potential for treating aller-
gic or inflammatory disorders such as asthma2 and
inflammatory bowel disease.3
not a pure examination of tryptase inhibition. A similar
concern applies to the human clinical studies with APC-
366, which is an inhibitor of tryptase, trypsin, and
thrombin.6 Another issue with 1 is its lack of oral bio-
availability. Thus, we have continued to search for po-
tent tryptase inhibitors with good enzyme selectivity
and oral efficacy.
A nonpeptide series of tryptase inhibitors from Aventis
Pharmaceuticals (now Sanofi-Aventis),7 represented by
key compound 2,7d attracted our attention. In particu-
lar, we were intrigued by the fact that certain potent
derivatives showed favorable enzyme selectivity and
good oral bioavailability.7a We modeled des-fluoro ana-
logue 3 in the active site of human b-tryptase and deter-
mined that a model of a spirocyclic congener, 4, in a
computed low-energy conformation, had good overlap
(Fig. 1). Consequently, we decided to explore such spiro-
cyclic piperidine amide derivatives in more detail. Here-
in, we report on the synthesis and biological assessment
of compounds with general structure 5 (Table 1). From
From our work with a series of a-ketoheterocycle-based
tryptase inhibitors,4 RWJ-58643 (1; RWJ-56423 for the
L-Arg diastereomer) was selected for proof-of-concept
evaluation in humans. In patients with allergic rhinitis,
RWJ-58643 demonstrated efficacy when administered
intranasally.5 However, since 1 inhibits both tryptase
and trypsin with nearly equal potency, this study was
Keywords: Tryptase; Inhibitor; Spiropiperidine; Asthma.
*
Corresponding authors. Tel.: +1 215 6285648 (M.J.C.); tel.: +1 215
6285530; fax: +1 215 6284985 (B.E.M.); e-mail addresses:
mcostanz@prdus.jnj.com; bmaryano@prdus.jnj.com
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.01.093

M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
2115
end-game. Key intermediate 6 was acylated with 5-
(phenylethynyl)furan-2-carboxylic acid by using bis
(2-oxo-3-oxazolidinyl)phosphonic chloride (BOP-Cl) as
an activating agent and the phthalimide was removed
with methylhydrazine to give 4, which was isolated
and purified as a benzoic acid salt.
The syntheses of 5a–d are depicted in Scheme 2. Re-
moval of the Boc group from 12 with trifluoroacetic
acid followed by coupling with 5-(phenyleth-
ynyl)furan-2-carboxylic acid using BOP-Cl generated
nitrile 13. Pinner reaction of 13 with hydrochloric acid
in ethanol at 5 °C, followed by neutralization with 3 N
NaOH afforded ethyl imidate 14 as the free base,
which was reacted with either anhydrous ammonia,
methylamine or hydrazine. Purification by reverse-
phase HPLC then yielded 5a, 5b, and 5c, respectively.
Carboxamide 5d was isolated by reverse-phase HPLC
as a side-product from the reaction of nitrile 13 with
hydroxylamine at 90 °C in a mixture of ethanol/1,4-
dioxane/water (3:1:1).
Figure 1. Stick models of three-dimensional structures of 3 (green;
enzyme-bound conformation from computer-assisted modeling) and 4
(yellow; energy-minimized conformation).
this study, we determined that 4 (JNJ-27390467) is a po-
tent, selective tryptase inhibitor with oral efficacy in ani-
mal models of airway inflammation.
Most of the compounds that we investigated as tryptase
inhibitors were spiro-dihydrobenzofuran analogues of 4,
namely 5a–k and 5q–s (Table 1). For comparison, we
also examined some dihydroindole analogues, 5l–o,
and indane analogue 5p. The synthesis of 4, 5e–k, and
5q–s necessitated key intermediate 6, for which we de-
vised an efficient synthetic route (Scheme 1). Thus,
Boc-piperidinone 7 was reacted with phenylsulfinyl-
methyl lithium to give hydroxyl sulfoxide 8, which was
subjected to a [2,3]-rearrangement8 according to the pro-
cedure described by MacCoss et al.9 Allyl alcohol 9 was
converted to the unstable chloride 10, which was cou-
pled with 2-bromo-4-cyanophenol to give ether 11.
The crucial radical-based cyclization10 of 11 to 12 was
effected in high yield in 0.02 M solution by using tribu-
tyltin hydride and 2,20-azobis(isobutyronitrile). The ni-
trile group in 12 was hydrogenated with Raney nickel,
the resultant primary amine was protected as a phthali-
mide (Pht), and the Boc group was removed to yield 6,
which was isolated as a hydrochloride salt. The prepara-
tion of 4 serves as a representative example of the
The synthesis of spiro-indoline derivatives 5l–o required
key intermediate 15 (Scheme 3). Thus, spiropiperidine
1611 was N-formylated to give 17, which was converted
to 18 by a Tscherniac–Einhorn reaction12 involving N-
(hydroxymethyl)phthalimide and concentrated sulfuric
acid. The N-methyl group was eliminated by using a-
chloroethyl chloroformate (ACE-Cl)13 in the presence
of Proton Sponge [1,8-bis-(dimethylamino)naphtha-
lene], and the resulting ACE-carbamate was readily
cleaved to yield 15, as a dihydrochloride salt. For 5l–o,
key intermediate 15 was acylated with the N-succinim-
idyl ester14 of 5-(phenylethynyl)furan-2-carboxylic acid
to furnish 19 [Ar = 5-(PhC„C)-2-furyl]. Compound 5l
was obtained by removing the phthalimide with meth-
ylhydrazine. In the case of 5m–o, 19 was functionalized
on the aniline nitrogen, as shown, and the final product
was obtained by removing the phthalimide with meth-
ylhydrazine. The four target compounds were purified
by reverse-phase HPLC and isolated as trifluoroacetate
salts.
HO
O
H
N
H2N
S
O
N
N
O
N
Ac
O
X
NH
HN
NH2
2 X = F
3 X = H
1
H2N
Z
H2N
O
O
O
N
N
Y
O
O
X
R
Ph
5
4

2116
M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
Table 1. Compounds and their inhibition of tryptase and trypsina
Z
R
O
H2N
O
X
N
Y
Compound
X
Y
Z
R
Tryptaseb IC50 (nM)
Tryptaseb Ki (nM)
Trypsinc IC50 (nM)
4
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
C„C
CH2CH2
S
O
H2
H
3.6 0.3d
4.5 0.5
780 220
3.7 0.5
18,000
1400
5a
5b
5c
5d
5e
5f
O
NH
NMe
NNH2
O
H
O
H
>50,000
>50,000
39,000
12,000
38,000
12,000
5500
O
H
59
29%
2
O
H
O
H2
H2
2-F
2-Cl
2-Me
2-OMe
4-Me
4-t-Bu
4-OH
H
4.3 1.5
5.2 0.3
1.2 0.3
3.0
3.4
4.0
0
0
0
O
5g
5h
5i
O
H2
H2
O
13
1
O
H2
H2
8.3 1.0
700 160
3200 300
23 3e
13,000
13,000
23,000
>50,000
>50,000
>50,000
6700
5j
O
5k
5l
O
H2
H2
NH
NAc
NSO2Me
NSO2Ph
CH2
O
5m
5n
5o
5p
5q
5r
5s
1g
H2
H2
H
15
2
H
1.8 0.6
5.7 2.0
3.1 0.2
82 27
7.1
16
0
0
H2
H2
H
H
7700
H2
H2
H
47,000
9100
8600
10 4i
O
H
90 40
f
—
O
H2
H
28
13
7
1
22 5h
a Target compounds were purified by reversed-phase HPLC on a Kromasil C-18 column eluting with acetonitrile/water mixtures containing 0.2%
trifluoroacetic acid. All compounds were characterized by ESI-MS and 300-MHz 1H NMR.
b Inhibition of human b-tryptase according to the assay described in Ref. 18; standard error and N = 2 unless noted otherwise. For compounds with
very low potency, percent inhibition at 50 lM may be given.
c Inhibition of human pancreatic trypsin according to the assay described in Ref. 19; N = 1–4.
d N = 6.
e N = 4.
f Linker X is a direct bond.
g Reference standard. Nearly a 1:1 mixture of L-Arg and D-Arg diastereomers, the former of which (RWJ-56423) is the active species (Ki = 10
nM) (Ref. 4).
3
h N = 13 (Ref. 4).
i Ki = 6.0 1.0, N = 9 (Ref. 4).
Spiro-indane derivative 5p was prepared as outlined in
Scheme 4. Indene 2015 was doubly alkylated to form
spirocyclic piperidine 21,16 which was hydroformylated
with carbon monoxide under Pd(II) catalysis, as
shown, to give 22.17 Treatment of 22 with hydrazine
provided the corresponding hydrazone, which was
hydrogenated to the amine, in conjunction with alkene
reduction, to yield 23. The amine was protected as a
phthalimide and the Boc group was eliminated to give
24. Acylation of 24 with 3-ethyl-1-(dimethylaminopro-
pyl)carbodiimide (EDC) and removal of the phthali-
mide with hydrazine afforded 5p, which was purified
by reverse-phase HPLC and isolated as a trifluoroace-
tate salt.
very potent tryptase inhibitor, with an IC50 value of
3.7 nM; it also had excellent selectivity over trypsin
of ꢀ5000-fold. The primary benzylamine of 4 could
be replaced by a primary amidine and be virtually
equipotent (viz. 5a), although selectivity vs trypsin
was somewhat eroded (ꢀ300-fold). Whereas addition
of an N-methyl substituent to the amidine was unfa-
vorable (viz. 5b; IC50 = 780 nM), an amino substituent
had less impact on potency (viz. 5c; IC50 = 59 nM). By
contrast, a primary carboxamide group was not effec-
tive (viz. 5d; IC50 > 50,000 nM). These results are con-
sistent with the ability of this group on the benzene
ring to occupy the S1 pocket and interact with the res-
ident Asp-189.
The compounds of interest, 4 and 5a–s, were first
evaluated for inhibition of human b-tryptase by using
a chromogenic assay (Table 1).18 As a counterscreen
for selectivity, we also tested these compounds for
inhibition of human pancreatic trypsin by using a
chromogenic assay.19 Prototype 4 turned out to be a
With the primary amine held fixed, we looked at some
substituents on the phenyl ring of the alkyne subunit
(Table 1). Substitution of the 2-position with fluoro
(5e), chloro (5f), or methyl (5g) had a little effect on
the single-digit nanomolar potency, although the po-
tency of the 2-methoxy compound (5h) was a bit atten-

M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
2117
Ph
O
OH
S
OH
PhS(O)Me
LiN( -Pr)2, THF
KOt-Bu
cryst. from
MeCN
prep.
O
i
HPLC
(55%)
t
-BuOH, Δ
N
N
Boc
N
Boc
Boc
−75 to 23 ºC
(95%)
8
7
9
Br
O
NC
Cl
NC
Br
OH
SOCl2
cryst. from
PhMe, 60 ºC
K2CO3, Me2CO
Bu4N+I-, Δ
i
-PrOH
N
Boc
N
Boc
(67%)
10
11
(use quickly)
Boc
N
1) Bu3SnH, AIBN
PhMe, Δ
1) phthalic anhydride
Ra Ni, 50 psig H2
3 N NaOH, EtOH
(100%)
PhMe,
i-PrNEt2, Δ
HCl
NC
2) prep. HPLC
(85%)
2) prep. HPLC
(88%)
i
-PrOH
(100%)
O
12
H • HCl
N
HO2C
O
Ph
1) MeNHNH2, EtOH, 40 ºC
4
Pht-N
BOP-Cl, (
i
-Pr)2NEt
2) PhCO2H, cryst.
from EtOH
O
CH2Cl2
6
(75%)
(50%)
Scheme 1.
NC
O
1) CF3CO2H/CH2Cl2 (1:4)
1) HCl, EtOH/dioxane/H2O, 5 ºC
12
O
N
Ph
2) BOP-Cl, (i-Pr)2NEt, CH2Cl2
2) 3 N NaOH to pH 7-8
(99%)
HO2C
O
O
Ph
13
(95%)
NH
EtO
O
1) anhyd NH3, MeNH2
or N2H4, EtOH
5a (66%)
5b (72%)
5c (33%)
O
N
Ph
2) prep. HPLC
O
14
1) HONH2•HCl, NaHCO3
EtOH/dioxane/H2O
(3:1:1), 90 ºC
13
5d (28%)
2) prep. HPLC
Scheme 2.
uated. Spirocyclic piperidine amides 5e (Ki = 3.0 nM)
and 4 (Ki = 3.7 nM) were equipotent, whereas the values
reported7d for congeners 3 and 2 (the corresponding
piperidine amides) differed by 3.6-fold (Ki’s of 8.7 and
31 nM, respectively). 4-Methyl spiropiperidine amide
5i was in a comparable potency range (IC50 = 8.3 nM),

2118
M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
Me
Me
Me
Pht-N
N
N
N
ACE-Cl
HCO2H
Pht-NCH2OH
H2SO4
Proton Sponge
PhMe, Δ
(CH2Cl)2, Δ
N
H
N
N
CHO
CHO
18
(99%)
(93%)
16
17
O
Ar
N
CO2-Suc
Pht-N
O
NH • 2 HCl
Pht-N
Ph
HCl
MeOH, Δ
THF
N
H
N
H
(68%)
(77%)
15
19
AcCl, or
MsCl, or
PhSO2Cl
1) MeNHNH2
EtOH, 40 ºC
2) prep. HPLC
Et3N
THF, 0 ºC
1) MeNHNH2
5m (54%)
5n (65%)
5o (72%)
(63%)
EtOH, 40 ºC
5l
2) prep. HPLC
Scheme 3.
Boc
N
PdBr2(PPh3)2
(3-ClPh)3P, HCO2Na
Br
Br
BocN(CH2CH2Cl)2
LiN(SiMe3)2, THF, 0 ºC
CO, DMF, 110 ºC
20
(51%)
(33%)
21
Boc
Boc
N
1) phthalic anhydride
-Pr2NEt, PhMe, Δ
NH2
N
OHC
1) N2H4, EtOH
i
2) PtO2, H2, EtOH
(29%)
2) CF3CO2H, CH2Cl2
(63%)
22
23
CO2H
O
Pht-N
NH
Ph
1) N2H4, EtOH, Δ
2) prep. HPLC
(34%)
5p
EDC, DMF
(53%)
24
Scheme 4.
but the 4-tert-butyl (5j; IC50 = 700 nM) and 4-hydroxy
(5k; IC50 = 3200 nM) analogues were much less potent.
The negative effect of a bulky hydrophobic group or a
polar group at this phenyl 4-position indicates that there
are critical limitations in the tryptase binding mode for
this molecular segment (vide infra).
The spiro-indoline (5l–o) and spiro-indane (5p) ana-
logues of 4 were also potent tryptase inhibitors, with
IC50 values in the range of 3–23 nM. Thus, tryptase
inhibitory potency is not particularly sensitive to the
chemical nature of this spirocyclic ring or to the indole
N-substituents, which may be related to the fact that

M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
2119
this portion of the inhibitor ligand is solvent-exposed
in the tryptase complex, according to computer model-
ing and an X-ray co-crystal structure of 4 Æ tryptase
(vide infra).
(Ascaris suum delivered by aerosol) and 4 h after antigen
challenge (N = 2). Airway resistance (SRL) was moni-
tored for 8 h following antigen challenge and airway hy-
per-reactivity to carbachol (PC400) was measured at
24 h. Compound 4 completely blocked the late-phase re-
sponse (4–8 h post antigen) and the increase in airway
hyper-reactivity at 24 h, but had no effect on the acute
early-phase response (0-4 h post antigen). Analysis of 4
in the sheep plasma indicated that its levels ranged from
0.4 to 1.0 lM during this efficacy study.
Isosteres 5q and 5r have an ethylidene or a sulfur atom,
respectively, in place of the ethynyl linker in 4. Although
the spacing for these two modifications is similar to that
for an ethynyl group, the inhibitory potencies for 5q and
5r suffered by ꢀ20-fold relative to 4. This finding sug-
gests that a rigid linear directionality is important for
this molecular fragment to achieve optimal binding
within the tryptase active-site cleft. Interestingly, differ-
ent comparisons were found in the reported piperidine
amide series (viz. 3),7a in that the 3-(phenylethynyl)benzoyl
analogue was 2-fold more potent than the 3-(phenyl-
ethyl)benzoyl analogue, and the [5-(phenylethynyl)-3-
pyridinyl]carbonyl analogue was 12-fold less potent than
the [5-(phenylethyl)-3-pyridinyl]carbonyl analogue.
Although 5s has the W linker eliminated, it displayed
a 3-fold potency improvement (IC50 = 28 nM) com-
pared to 5q and 5r.
Compound 4 (benzoate salt) was also tested in a model
of allergic asthma with ovalbumin-sensitized guinea
pigs.21 Oral doses of 4 (20 mg/kg) were administered
twice daily over 3 days, and once on Day 4 at 2 h prior
to ovalbumin challenge. Thus, 4 was found to reduce to-
tal lung resistance (RL) by 90% and to increase dynamic
lung compliance (Cdyn) by 65% relative to vehicle-trea-
ted controls. In addition, there was a 20% reduction in
the total white blood cell count in the broncho-alveolar
lavage (BAL) fluid, which signifies anti-inflammatory
activity.
Compound 4 was found to be a full competitive
inhibitor of tryptase (Ki = 3.7 nM) and trypsin (Ki =
2500 nM), with a favorable selectivity ratio of 675 on
the basis of Ki values. Enzyme inhibition studies con-
ducted with 4 and additional serine proteases revealed
a generally high level of selectivity. At 50 lM, 4 inhib-
ited factor Xa, chymase, and cathepsin G less than
30%; and at 10 lM, 4 inhibited thrombin, plasmin, tis-
sue-type plasminogen activator (tPA), factor VIIa, kalli-
krein, urokinase-type plasminogen activator (uPA),
streptokinase, human leukocyte elastase, and chymo-
trypsin less than 15%. Since 4 possesses a primary
benzylamine group, we examined its inhibition of mono-
amine oxides A and B. It was a weak inhibitor of MAO-
A and MAO-B with IC50 values of 23 and 72 lM,
respectively. We also evaluated 4 as a substrate for
MAO-B and found that it was quite stable: 95% of 4
remaining at 30 min (compared with 13% remaining at
30 min for benzylamine itself).
We obtained a co-crystal of 4 and human b2-tryptase
that was suitable for X-ray crystallographic study.22 In
23
˚
the solved structure (2.0 A), there was one ligand per
monomer unit of the tetrameric protein assembly
(Fig. 2). The benzylamine group (lower left) is located
in the S1 pocket, as expected, and the amide carbonyl
is hydrogen bonded to the Gly-219 Na. The piperidine
ring is in a chair conformation and the 5-membered-ring
ether oxygen is directed toward the solvent. Interest-
ingly, the phenyl group of the rigid phenylethynyl moi-
ety disrupted a hydrogen bond found in the original
crystal structure reported by Pereira et al.24 (Fig. 3), be-
tween the carboxylate of Glu-217 and the hydroxyl of
The pharmacokinetics of 4 in rats and dogs were favor-
able for oral administration. In rats, 4 had an F value of
100%, an oral Cmax of 1.1 lM, an oral t1/2 of 7 h, and a
large volume of distribution (Vd) of 17 L/kg. In dogs, 4
had an F value of 65%, an oral Cmax of 0.9 lM, an oral
t1/2 of 27 h, and a volume of distribution of 16 L/kg. The
half-life of 4 in a human liver microsome preparation
was >100 min, with 100% of 4 remaining after 60 min.
Despite having a benzylamine group, 4 was quite stable
in the whole blood of humans, rats, dogs, sheep, and
guinea pigs. Its inhibition of cytochrome P450 in iso-
lated human liver microsomes was weak, with IC50 val-
ues of >20 lM for isozymes 1A2, 2C9, 2C19, 3A4, and
2D6.
We evaluated 4 (benzoate salt) for in vivo efficacy in
conscious, antigen-sensitized allergic sheep, which
served as a useful asthma model.2b,20 This inhibitor
was administered by oral gavage at 30 mg/kg twice daily
on Day 1, and on Day 2 at 2 h prior to antigen challenge
Figure 2. View of one monomer of the 4 Æ tryptase complex, with 4
(stick model; green with standard coloring scheme for heteroatoms)
occupying the tryptase active site (Connelly surface; white with
standard coloring scheme for heteroatoms).

2120
M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
Figure 3. Superimposed active-site views of one monomer of 4 Æ tryptase and one monomer from the original tryptase X-ray structure reported by
Pereira et al.24 to illustrate the induced-fit hydrophobic pocket. The phenyl substituent of 4 (stick model; yellow with standard coloring scheme for
heteroatoms) is situated in tryptase (ribbon representation) such that the side chains of Tyr-95 (green) and Glu-217 (green) are displaced from the
position that they adopt in the original X-ray structure,24 wherein the hydroxyl of Tyr-95 (blue) is hydrogen bonded (red dashed line) to the
carboxylate of Glu-217 (blue).
Tyr-95 in an adjacent subunit of the tetramer (i.e., in
‘‘monomer B’’). The displacement of the side chains of
Glu-217 and Tyr-95 engendered a shallow hydrophobic
binding pocket that is defined by the following residues:
Thr-96 and Trp-215 of ‘‘monomer A’’; Pro-60A and
Tyr-95 of ‘‘monomer B’’. This induced-fit binding mode
is not evident in the co-crystal structures reported for
the related piperidine amide series with different acyl
groups.7a The much weaker potency of the 4-tert-butyl
(5j) and 4-hydroxy (5k) analogues of 4 can be explained
by stereoelectronic problems connected with occupancy
in this shallow hydrophobic pocket. The linear, rigid
phenylethynyl group in 4 may be advantageous relative
to the more flexible bioisosteric groups in 5q and 5r be-
cause of the way these different acyl groups are able to
capitalize on this induced pocket.
Acknowledgment
We thank Dr. Lei Qiao for the preparation of 5a.
References and notes
1. (a) Sommerhoff, C. P.; Schaschke, N. Curr. Pharm. Des.
2007, 13, 313; (b) Cairns, J. A. Pulmon. Pharmacol. Ther.
2005, 18, 55.
2. (a) Clark, J. M.; Van Dyke, R. E.; Kurth, M. C. Prog.
Respir. Res. 2001, 31, 170; (b) Clark, J. et al. Am. J.
Respir. Crit. Care Med. 1995, 152, 2076.
3. Tremaine, W. J. et al. Aliment. Pharmacol. Ther. 2002, 16,
407; Isozaki, Y. et al. Scand. J. Gastroenterol. 2006, 41,
944; Yoshida, N.; Isozaki, Y.; Takagi, T.; Takenaka, S.;
Uchikawa, R.; Arizono, N.; Yoshikawa, T.; Okanoue, T.
Aliment. Pharmacol. Ther. Symp. Ser. 2006, 2, 249.
4. Costanzo, M. J. et al. J. Med. Chem. 2003, 46, 3865.
5. Erin, E. M. et al. Clin. Exp. Allergy 2006, 36, 458.
6. (a) Clark, J. M.; Moore, W. R.; Tanaka, R. D. Drugs
Future 1996, 21, 811 (Update: Anon. Drugs Future 1998,
23,, 903); (b) Rice, K. D.; Tanaka, R. D.; Katz, B. A.;
Numerof, R. P.; Moore, W. R. Curr. Pharm. Design 1998,
4, 381; (c) Burgess, L. E. Drug News Perspect. 2000, 13,
147.
7. (a) Levell, J. et al. Bioorg. Med. Chem. 2005, 13, 2859; (b)
Hopkins, C. R. et al. Bioorg. Med. Chem. Lett. 2005, 15,
2734; (c) Astles, P. C. et al. PCT Int. Appl., WO
2001090101, 2001.; (d) Pauls, H. W.; Aldous, S. C.;
Merriman, G. H.; Farr, R. A.; Sledeski, A. W. PCT Int.
Appl., WO 2004060884, 2004.
In conclusion, we have explored a series of spirocyclic
piperidine amides, several of which are very potent tryp-
tase inhibitors. In particular, 4 (JNJ-27390467) was
found to be a potent, selective tryptase inhibitor with
oral efficacy in animal models of airway inflammation.
In a sheep asthma model, 4 effectively blocked the
late-phase and airway hyperreactivity responses. In a
guinea pig asthma model, 4 prevented the increase in
airway resistance. An X-ray co-crystal structure of
4 Æ tryptase revealed an interesting induced-fit binding
mode, involving the phenylethynyl group of 4 and a
hydrophobic pocket that is formed by amino acid side
chains of two adjacent monomers in the tetrameric pro-
tein assembly. Compound 4 also possesses many desir-
able attributes of a potential clinical candidate.
8. Evans, D. A.; Andrews, G. C. Acc. Chem. Res. 1974, 7,
147.
9. MacCoss, M. et al. PCT Int. Appl., WO 9429309, 1994.

M. J. Costanzo et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2114–2121
2121
10. Chen, M.-H.; Abraham, J. A. Tetrahedron Lett. 1996, 30,
5233.
11. Ong, H. H.; Profitt, J. A.; Fortunato, J.; Glamkowski, E.
J.; Ellis, D. B.; Geyer, H. M.; III; Wilker, J. C.; Burghard,
H. J. Med. Chem. 1983, 26, 981.
12. Zaugg, H. E. Synthesis 1984, 85; Zaugg, H. E. Synthesis
1984, 181; Zaugg, H. E.; Martin, W. B. Org. React. 1965,
14, 52.
13. Olofson, R. A.; Martz, J. T.; Senet, J. P.; Piteau, M.;
Malfroot, T. J. Org. Chem. 1984, 49, 2081.
14. Xie, J.-S.; Huang, C. Q.; Fang, Y.-Y.; Zhu, Y.-F.
Tetrahedron 2004, 60, 4875.
15. Young, J. R. et al. Bioorg. Med. Chem. Lett. 2002, 12, 827.
16. Efange, S. M. N.; Khare, A. B.; Foulon, C.; Akella, S. K.;
Parsons, S. M. J. Med. Chem. 1994, 37, 2574.
Devices). The IC50 experiments were conducted by fixing
the enzyme and substrate concentrations (1.8 nM [E]/
60 lM [S]) and varying the inhibitor concentration; the Ki
experiments were conducted by fixing the enzyme concen-
tration (1.8 nM) and varying the inhibitor and substrate
concentration. Changes in absorbance at 405 nm were
monitored with the software program Softmax Pro, on
addition of enzyme without inhibitor at 22 °C for 15 min.
Percent inhibition was calculated by comparing the initial
reaction velocity of samples without inhibitor to those
with inhibitor. Initial reaction velocities were analyzed by
using Microsoft Excel. Ki values were calculated in Sigma
Plot Enzyme Kinetic Module.
20. Abraham, W. M. Pulmonary Pharmacol. 1989, 2, 33.
21. Smith, N.; Johnson, F. J. Clin. Exp. Allergy 2005, 35, 522.
22. Recombinant human b-tryptase from Promega Corp.
(Madison, WI; #G5631) was deglycosylated at 23 °C for
72 h under the following conditions: 50 mM sodium
phosphate buffer (pH 7.8), 0.1 mM 1,2-ethylenediamine
N,N,N0,N0-tetraacetic acid, 0.2 mg/mL of GST-PNGase F
(Hampton Research, Aliso Viejo, CA), 5.7 mg/mL of
tryptase, and 0.5 mM 4. The GST-PNGase F was removed
by using glutathione–Sepharose 4B (Amersham Biosci-
ences, Piscataway, NJ). For crystallization, this tryptase in
20 mM 2-(N-morpholino)ethanesulfonic acid (pH 6.1) and
1 M NaCl, with 0.9 mM 4, was concentrated to 6.2 mg/
mL. Suitable crystals for X-ray crystallography were
grown by the standard vapor-diffusion hanging-drop
method.
17. Okano, T.; Harada, N.; Kiji, J. Bull. Chem. Soc. Jpn. 1994,
67, 2329.
18. The b-tryptase assays were performed with human lung
tryptase (Cortex Biochem, Inc., San Leandro, CA;
#CP3033) in aqueous buffer (10 mM Tris, 10 mM Hepes,
150 mM NaCl, 0.1% PEG 8000, pH 7.4) with the
chromogenic substrate H-D-HHT-Ala-Arg-pNa Æ 2AcOH
(American Diagnostica, Inc., Stamford, CT; #238;
Km = 580 lM) by using a microplate reader (Molecular
Devices Corp., Sunnyvale, CA). The IC50 experiments
were conducted by fixing enzyme and substrate concen-
trations (1 nM [E]/500 lM [S]) and varying the inhibitor
concentration; the Ki experiments were conducted by
fixing the enzyme concentration (1 nM) and varying the
inhibitor and substrate concentrations. Changes in absor-
bance at 405 nm were monitored with the software
program Softmax Pro (Molecular Devices), on addition
of enzyme without inhibitor at 22 °C for 15 min. Percent
inhibition was calculated by comparing the initial reaction
velocity of samples without inhibitor to those with
inhibitor. Initial reaction velocities were analyzed by using
Microsoft Excel. Ki values were calculated in Sigma Plot
Enzyme Kinetic Module (SPSS, Inc., Chicago, IL).
19. The trypsin assays were performed with human pancreatic
trypsin (Calbiochem #650275; Merck KGaA, Darmstadt,
DE) in aqueous buffer [200 mM NaCl, 50 mM Hepes,
0.5% octyl glucoside (Sigma–Aldrich O-9882), pH 7.4]
with the chromogenic substrate N-a-Z-D-Arg-Gly-Arg-
pNA Æ 2HCl (Chromogenix, Milan, IT; #S2765;
Km = 25 lM) by using a microplate reader (Molecular
23. X-ray crystallographic study of 4 Æ tryptase: The 4 Æ tryp-
tase complex crystallized in a tetragonal crystal form
˚
[space group P41 with a = b = 82.68 A, c = 170.28 A, and
˚
four molecules (one tetramer) in the asymmetric unit],
˚
which diffracted X-rays to 2.1 A resolution. The structure
was determined by molecular replacement by using a
published tryptase structure (PDB entry 1AOL). Inhibitor
4 could be unambiguously modeled in 2F0 ꢁ Fc and
F0 ꢁ Fc electron-density maps, and the structure was
˚
refined (2.3 A resolution; R, Rfree = 0.181, 0.233). Atomic
coordinates and structure factors for the 4 Æ tryptase
complex have been deposited with the Protein Data Bank
(www.rcsb.org; accession code 2ZA5).
24. Pereira, P. J. B.; Bergner, A.; Macedo-Ribeiro, S.; Huber,
R.; Matschiner, G.; Fritz, H.; Sommerhoff, C. P.; Bode,
W. Nature 1998, 392, 306.