A β-tryptase inhibitor with a tropanylamide scaffold to improve in vitro stability and to lower hERG channel binding affinity

Article abstract of DOI:10.1016/j.bmcl.2011.12.127

Tropanylamide was investigated as a possible scaffold for b-tryptase inhibitors with a basic benzylamine P1 group and a substituted thiophene P4 group. Comparing to piperidinylamide, the tropanylamide scaffold is much more rigid, which presents less oppor

Full text of DOI:10.1016/j.bmcl.2011.12.127

Bioorganic & Medicinal Chemistry Letters 22 (2012) 1606–1610
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A b-tryptase inhibitor with a tropanylamide scaffold to improve in vitro
stability and to lower hERG channel binding affinity
Guyan Liang a, , Yong Mi Choi-Sledeski a, Patrick Shum a, Xin Chen a, Gregory B. Poli a, Vasant Kumar a,
⇑
Anne Minnich b, Qingping Wang c, Joseph Tsay b, Keith Sides b, Jiesheng Kang c, Ying Zhang a
a Molecular Innovative Therapeutics, Sanofi Pharmaceuticals, Route 202-206, Bridgewater, NJ 08807, USA
b Fibrosis and Wound Repair, Sanofi Pharmaceuticals, Route 202-206, Bridgewater, NJ 08807, USA
c Disposition, Safety, and Animal Research, Sanofi Pharmaceuticals, Route 202-206, Bridgewater, NJ 08807, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Tropanylamide was investigated as a possible scaffold for b-tryptase inhibitors with a basic benzylamine
P1 group and a substituted thiophene P4 group. Comparing to piperidinylamide, the tropanylamide scaf-
fold is much more rigid, which presents less opportunity for the inhibitor to bind with off-target proteins,
such as cytochrome P450, SSAO, and hERG potassium channel. The proposed binding mode was further
confirmed by an in-house X-ray structure through co-crystallization.
Received 30 November 2011
Revised 22 December 2011
Accepted 27 December 2011
Available online 3 January 2012
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Serine protease
Asthma
COPD
hERG
Tryptase
Inhibitor
Molecular modeling
Structure based drug design
X-ray
Benzylamine
Human b-tryptase is a mast cell-specific trypsin-like serine pro-
tease that is reported to have a strong correlation with inflamma-
tory and allergic disorders, especially asthma and chronic
between the inhibitor and the enzyme constitutes a significant
portion of the overall binding affinity, which leads to a concern
of lacking selectivity and its potential consequence of long-term
in vivo toxicity. As a result, most recent publications avoided this
class of inhibitors. The second class contains a basic P1 group that
interacts with D189 in the S1 pocket through an H-bonding net-
work. This class of inhibitors is most popular in number and is
widely available for almost all trypsin-like serine proteases closely
related to b-tryptase. The third class of inhibitors contains a non-
basic P1 group and does not have a direct H-bonding interaction
with D189. Due to decreased basicity, this class of inhibitors often
has more desirable pharmacokinetic properties, especially oral bio-
availability. However, so far all published inhibitors in this class
are for a subset of serine proteases with an alanine at the 190 posi-
tion, such as fXa and Thrombin. To our knowledge, no inhibitor
with a non-basic P1 group and without a reactive moiety has been
reported for serine proteases with a serine at the 190 position such
as b-tryptase.
obstructive pulmonary disease (COPD).1–4 Although both
a- and
b-tryptases are selectively expressed in mast cell, only b-tryptase
appears to be activated intracellularly and stored in secretary gran-
ules. Upon mast cell activation, the active form b-tryptase stabi-
lized by negatively charged heparin is secreted and triggers
allergic and inflammatory responses.
Due to its pivotal role in allergic and inflammatory actions, b-
tryptase has attracted widespread interest and has been pursued
as a therapeutic target for more than two decades. A rich collection
of small organic inhibitors has been reported in both scientific
journals and patent publications.5–14 Historically, three classes of
inhibitors have been reported for broadly-defined trypsin-like ser-
ine proteases including fXa, FVIIa, Thrombin, and tryptases. The
first class contains a reactive functional group that can form a
covalent bond with the catalytic serine (S195). While this process
can be either reversible or irreversible, the covalent bonding
Vast majority of b-tryptase inhibitors published so far has a basic
P1 group. The pKa value of those basic P1 groups plays an important
role in the b-tryptase binding because the basic P1 group directly
interacts with the de-protonated carboxlate side chain of D189.
⇑
Corresponding author. Tel.: +1 908 231 4573.
E-mail address: guyan.liang@sanofi-aventis.com (G. Liang).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.12.127

G. Liang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1606–1610
1607
The functional group with a higher pKa value carries a higher posi-
tive partial charge and tends to have a stronger interaction with
D189. However, for the same reason, those groups with higher
pKa and stronger partial charge often associate with undesirable
properties, such as poorer absorption and lower oral bioavailability.
Therefore, we focused our effort on P1 groups like benzylamines
which have a lower pKa than benzamidine or benzamidine-like
moieties. To overcome the challenge of the potency loss due to
the lower pKa value, various scaffolds were designed, synthesized,
and tested for their suitability of binding to b-tryptase and their ef-
fect on ADMET properties. The present Letter describes our study of
an inhibitor with a tropanylamide scaffold linker, a benzylamine P1
group, and a substituted thiophene P4 group, as shown in Figure 1.
One objective of the present design was to preserve the binding
of the compound in the b-tryptase catalytic site. As exhibited by X-
ray structures, b-tryptase is a donut-shaped tetramer with four
substrate binding sites facing toward to the central pore of the tet-
ramer in a pair-wise distribution.15–17 Within each pair, the S4
pockets of the two substrate binding sites face each other and
are located at the interface of two adjacent b-tryptase monomers.
Although each monomer has its own catalytic center, the shape of
the S4 pocket would change significantly without the tetrameric
folding, which would lead to a significantly decreased binding
affinity and selectivity to both the endogenous substrate and the
artificial peptide substrate used for screening.
Another objective of the present design was to decrease the
blockage of the cardiac potassium channel encoded by the human
Ether-à-go-go Related Gene (hERG). The hERG potassium channel
plays a critical role for QT interval prolongation which may lead
to ventricular fibrillation and cause sudden cardiac death. Several
well-known drugs including one of our own, terfenadine, have
been withdrawn from the market due to hERG channel blockade
that causes prolongation of the QT interval. As a result, screening
against hERG blocking property has become an important prophy-
lactic measure during pharmaceutical development, to which our
b-tryptase inhibitors are no exception. Representatives of our lead
compounds with a benzylamine P1 group and a piperidinylamide
scaffold were tested for hERG blockade and some of them exhib-
ited a moderate inhibition. It would be more desirable if the hERG
channel binding could be further decreased without compromising
other properties.
essential for their binding with b-tryptase. The hydrophobic/aro-
matic center from the hERG binding pharmacohore maps onto
the thiophene ring (P4) of compound 1, which is also important
for this series of inhibitors. The strategy for the present design
was to modify the scaffold linker between the P1 and P4 groups
so that the binding to the hERG channel can be interrupted without
removing either benzylamine (the basic nitrogen in the hERG phar-
macophore) or the thiophene ring (the hydrophobic/aromatic cen-
ter in the hERG pharmacophore).
To serve this purpose, a tropanyl ring was introduced in the
place of piperidine so that the scaffold can accommodate less con-
formational flexibility beyond what is required for b-tryptase bind-
ing. Therefore, the new scaffold may have a lower chance to bind
with hERG channel and other off-target proteins without losing
b-tryptase activity. A molecular dynamics (MD) simulation was
carried out to compare the difference in terms of conformational
flexibility between the two scaffolds using CHARMM force field
implemented in Discovery StudioÒ from AccelrysÒ. A trajectory of
10 ns was recorded after a heating period and a period of 1 ns
equilibration.
The torsion angles between the P1 exiting vector (C1–C2) and
several P4 exiting vectors (N3–C4, C4–C5, and C5–C6) were ana-
lyzed by sampling 5000 frames out of the 10-ns MD trajectory
and plotting them against the simulation time, as shown in Figure
2. For the piperidinylamide scaffold, some major conformational
transitions took place during the course of the 10-ns MD simula-
tion. As shown in Figure 2(a), the torsion angle of C1–C2–N3–C4
dramatically changed its value six times within 10 ns, which corre-
sponds to the conformational changes of the ring from a chair con-
formation to a boat or a twisted chair/boat conformation. This
conformational flexibility is much less significant, for the tropa-
nylamide scaffold. As shown in Figure 2(d), the same torsion angle,
C1–C2–N3–C4, for the tropanylamide scaffold remained around
the same value through the entire course of the 10-ns MD simula-
tion without any major change in value. Additionally, the tropa-
nylamide scaffold is more rigid than the piperidinylamide
scaffold even without considering the major chair-to-boat confor-
mational transition. For instance, by only including the frames of
the chair conformation, the root mean square (RMS) deviation of
the torsion angle C1–C2–N3–C4 is 16.0° for piperidinylamide scaf-
fold and 11.3° for the tropanylamide scaffold.
Numerous hERG models have been published either based on
homology models derived from bacterial potassium channel or
based on structures of known hERG inhibitors.18–20 While the reli-
ability of the quantitative prediction ranged from poor to moder-
Similar comparisons were made for the torsion angles of C1–C2–
C4–C5 and C1–C2–C5–C6, which lead to the same conclusion that
the tropanylamide scaffold is conformationally more rigid than
the piperidinylamide scaffold. Both torsion angles for the tropa-
nylamide scaffold stayed within a small range, while those for the
piperidinaylamide scaffold each underwent a major change in value
(120° or more) which correspond to a major conformational transi-
tion. Similarly, the rigid nature of the tropanylamide scaffold is also
reflected through the smaller RMS deviations of those torsion an-
gles without including the major conformational changes. The
torsion angle C1–C2–C4–C5 has RMS deviations of 11.5° and 9.7°
for piperidinylamide and tropanylamide, respectively, while the
ate, they more or less agreed on
a general hERG binding
pharmacophore which requires a basic nitrogen (positive charge
group or H-bonding group) and one or two hydrophobic/aromatic
groups. The distance between the basic nitrogen and the hydro-
phobic center, which is important for many hERG inhibitors, pre-
fers a separation of 6–8 Å. In the present case, the basic nitrogen
in our b-tryptase inhibitors, the benzylamine P1 group, is dictated
by the H-bonding interaction with D189, which is absolutely
O
O
H2N
H2N
S
S
N
N
O
O
Br
Br
1
2
Figure 1. b-Tryptase inhibitors with benzylamine P1 group and an substituted thiophene P4 group. Whereas compound 1 has a piperidinylamide scaffold, compound 2
changes the scaffold to tropanylamide while preserving the rest of the molecule.

1608
G. Liang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1606–1610
O
C4
180
120
60
180
120
60
180
NH2
S
N3
C5
C6
O
boat
C2
120
60
C1
Br
chair
0
0
0
-60
-120
-180
-60
-120
-180
-60
-120
-180
0
2
4
6
8
10
0
2
4
6
8
10
0
2
4
6
8
10
Time (ns)
(a)
Time (ns)
Time (ns)
(c)
(b)
O
C4
N3
180
120
60
180
120
60
180
120
60
NH2
S
C5
C6
O
C2
C1
Br
0
0
0
-60
-120
-180
-60
-120
-180
-60
-120
-180
0
2
4
6
8
10
0
2
4
6
8
10
0
2
4
6
8
10
Time (ns)
Time (ns)
Time (ns)
(d)
(e)
(f)
Figure 2. Torsion angle trajectories of relevant exiting vectors during a course of 10 ns molecular dynamics simulation during which 5000 frames were sampled. Curves a, b,
and c are for the simulation of the piperidinylamide scaffold (compound 1); curves d, e, and f are for the one of the tropanylamide scaffold (compound 2).
corresponding RMS deviations for C1–C2–C5–C6 are 12.2° and
10.5°.
triflate 4 was prepared using N-phenyl bistrifluoromethanesulf-
onamide, and was subsequently converted to the corresponding
trimethyltin analog 5 via a palladium-catalyzed reaction with
1,1,1,2,2,2-hexamethyl-distannane. Stille coupling of trimethytin
5 and bromide 6 afforded compound 7 in good yield. After hydro-
genation and acid deprotection of compound 7, amine 9 was cou-
pled with 4-bromo-3-methyl-5-propoxy-thiophene-2-carboxylic
acid in the presence of EDCI and TEA to give compound 10 as a
Docking models of compounds 1 and 2 in the catalytic site of b-
tryptase indicate that inhibitors with a tropanylamide scaffold can
preserve all essential interactions with residues in the substrate
binding pocket of b-tryptase. As shown in Figure 3, the amino
nitrogen of the benzylamine P1 group interacts with the side chain
of D189 and the backbone carbonyl oxygen of G219 through an H-
bonding network, which is a signature interaction for trypsin-like
serine proteases. The tropanyl ring is located close to the backbone
of G219 and E217 without interacting with catalytic triad formed
by H57, D102, and S195. This binding mode is identical to the
one of the inhibitor with the corresponding piperidinylamide scaf-
fold. The carbonyl oxygen of the tropanylamide scaffold has an H-
bonding interaction with the backbone NH of G219. The thiophene
mixture of
a and b isomers, which were separated using silica
gel chromatography. Each isomer was deprotected separately with
potassium carbonate to give the final product 2.
Based on in silico models, the
a isomer was preferred energeti-
cally in the substrate binding pocket of b-tryptase. The predicted
binding modes of compounds 1 and 2 were well superimposed
on top of each other with a negligible RMS deviation. Those binding
modes were later confirmed by in-house X-ray structures that
showed highly similar electron density map for the conserved part
of the molecules. An NMR study (NOE/ROE in DMSO at 600 MHz)
was also consistent with this extended conformation, on the
average.
ring of the P4 group
p-stacks with the amide side chain of Q98
while the propanol of the P4 group has a hydrophobic contact with
the side chain of Y95. This binding mode is identical to the one
with the piperidinylamide scaffold, which can help to retain the
binding affinity to b-tryptase based on in silico models.
Compound 2 was prepared from commercially available start-
ing material 3 via seven-step synthesis, as shown in Scheme 1. Enol
The inhibition of human b-tryptase was measured in a chromo-
genic assay with 50 ng/mL b-tryptase and 500
lM substrate S-
2366 (Diapharma), -pyroglutamyl- -prolyl- -arginine-p-nitroani-
L
L
L
line hydrochloride. The IC50 of the inhibitor was determined by fit-
ting the inhibition–concentration data to a sigmoid curve and the
corresponding Ki value of the inhibitor was derived using the com-
petitive inhibition equation, Ki = IC50/(1 + [S]/Km). As shown in Ta-
ble 1, the inhibitor with the tropanylamide scaffold has a Ki of
62 nM, which represents a threefold decrease comparing to the
one with the piperidinylamide scaffold. Even if the binding mode
of the tropanylamide scaffold is almost identical to the one with
the piperidinylamide scaffold in both in silico models and X-ray
structures, it is important to recognize that the models and X-ray
structures may not fully reflect the nature of the ligand–protein
interaction under the assay condition.
The metabolic stability of compounds was measured in vitro
using human and rat liver microsomal fractions prepared by centri-
fugation. Liver microsome fractions, which contain most of the
phase I biotransformation enzymes, such as cytochrome P450s,
were incubated with the compound of interest for 20 min before
the reaction was stopped. The remaining of the compound was
Figure 3. Stereo view of b-tryptase catalytic pocket with a modeled binding mode
of compound 2. Binding poses were generated based on in-house b-tryptase X-ray
structures using GlideÒ from SchrödingerÒ. The carbon atoms of one b-tryptase
monomer are colored in green while those of the adjacent monomer are colored in
brown. The carbon atoms of the inhibitor are colored in white. The inhibitor bound
in the adjacent binding pocket is only partially visible on the top right corner.

G. Liang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1606–1610
1609
F
F
F
F
O
O
S
O
Sn
i
O
iii
O
ii
O
N
N
O
O
O
N
N
F
O
F
O
HN
N
H
F
O
F
O
3
4
5
7
F
F
Br
F
6
O
F
F
F
O
N
vi
v
F
iv
HN
O
HN
N
F
F
HN
HN
S
F
F
O
O
HCl
Br
F
F
F
F
O
O
O
8
9
10
F
O
N
vii
H2N
HCl
S
Br
O
2
Scheme 1. The reaction scheme for synthesizing compound 2. Reagents and conditions: (i) KHMDS, N-phenyl bistrifluoromethanesulfonamide, THF, À78 °C, 78%; (ii)
Pd(Ph3)4, 1,1,1,2,2,2-hexamethyl-distannane, LiCl, THF, 80 °C, 38%; (iii) Pd(Ph3)4, 4, 110 °C, 69%; (iv) 10% Pd/C, H2, MeOH, rt, 90%; (v) 4 M HCl in dioxane, rt, quantitative; (vi)
4-bromo-3-methyl-5-propoxy-thiophene-2-carboxylic acid, EDCI, TEA, CH2Cl2, rt, 51%; (vii) (a) K2CO3, MeOH/H2O, (b) 2 M HCl in Et2O, 88%.
Table 1
Table 2
Potency, metabolism, and safety profiles of compound 1 and 2
Crystallographic statistics for b-tryptase/compound 2 complex (values in parentheses
correspond to the highest-resolution shell)
1
2
Data collection
b-tryptase, Ki (nM)
22
62
Liver microsome metabolism (human)
Liver microsome metabolism (rat)
hSSAO stability (remaining, 24 h)
26%
6%
70%
4.7
2.5
1.0
7%
0%
86%
5.3
6.9
Beam line
Wavelength (Å)
Space group
Cell parameters (Å)
Resolution (Å)
Redundancy
APS LS-CAT 21ID-F
0.97872
P41
a = 81.719, b = 81.719, c = 170.684
40–2.09 (2.16–2.09)
5.7 (5.7)
100 (100)
17.6 (3.4)
CYP/3A4, IC50
CYP/2D6, IC50
(
(
l
l
M)
M)
hERG, IC50
hERG, % inh. @ 1
hERG, % inh. @ 10
(lM)
l
l
M
M
14%
36%
Completeness (%)
I/r
I
Rmerge (%)
9.8 (59.7)
Refinement
Resolution (Å)
No. reflections
Percentage of Rfree
Rwork/Rfree (%)
R.M.S. deviations
Bond lengths
Bond angles
Ramachandran plot (%)
Preferred region
Allowed region
Outlier
40–2.09
65,965
5
measured and the metabolic lability was derived. As shown in Table
1, the inhibitor with the tropanylamide scaffold is significantly more
stable than the piperidinylamide counterpart for both rat and hu-
man. Only 7% of the tropanylamide inhibitor was metabolized with-
in 20 min comparing to 26% for the piperidinylamide inhibitor in the
human liver microsome study, whereas the new scaffold showed no
measurable metabolism in the rat liver microsome test.
It is hypothesized that this increased liver microsomal stability
is attributed to the fact that the tropanylamide scaffold is more ri-
gid and is less likely to bind with enzymes responsible for metab-
olism, like cytochrome P450s. This rationale is also consistent with
the observation that compound 2 exhibited better stability for hu-
man semicarbazide-sensitive amine oxidases (hSSAO), as shown in
Table 1.
Compounds 1 and 2 were evaluated as direct-acting inhibitors of
CYP/3A4 and CYP/2D6 in human liver microsomes. CYP activities
were assessed with marker substrates (midazolam for CYP3A4
and dexotromethorphan for CYP2D6) using LC–MS/MS system,
with IC50 values summarized in Table 1. The compound with the
tropanylamide scaffold showed similar or weaker CYP3A4 and
CYP2D6 inhibitions.
The inhibitor with tropanylamide scaffold exhibits a major
improvement for hERG channel inhibition, which was measured
through whole-cell patch clamp electrophysiology. The hERG
cardiac potassium channel was expressed in Chinese hamster ovary
20.8/23.4
0.007
0.96
94.7
4.9
0.42
cells from which hERG channel currents were recorded. As shown in
Table 1, the compound with the piperidinylamide scaffold had a
moderate inhibition of the hERG channel current with an IC50 of
1.0
reduced the hERG channel current by only 14% at 1
10 M.
The proposed binding mode was further confirmed by an in-
lM. However, the compound with the tropanylamide scaffold
lM and 36% at
l
house X-ray structure at 2.09 Å resolution, where human b-tryptase
was co-crystallized with compound 2 with crystallographic statis-
tics shown in Table 2. The crystal structure was solved by molecular
replacement using PDB entry 2BM2 as a model and refined with
autoBuster.21 The electron density, as shown in Figure 4,
unambiguously defines the binding mode of the compound.
In conclusion, tropanylamide is a more rigid scaffold than pipe-
ridinylamide for b-tryptase inhibitors with a benzylamine P1 group

1610
G. Liang et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1606–1610
References and notes
1. Norris, A. Expert Opin. Invest. Drugs 2004, 13, 739.
2. Tremaine, W. J.; Brzezinski, A.; Katz, J. A.; Wolf, D. C.; Fleming, T. J.; Mordenti, J.;
Strenkoski-Nix, L. C.; Kurth, M. C.; Barish, C.; Brzezinski, A.; Fowler, F. C.;
Gaspari, M.; Katz, J. A.; Koval, G.; Riff, D.; Safdi, M.; Salzberg, B.; Tremaine, W. J.;
Wolf, D. C. Aliment. Pharmacol. Ther. 2002, 16, 407.
3. Rice, K. D.; Sprengeler, P. A. Curr. Opin. Drug Discov. Dev. 1999, 2, 463.
4. Clark, J. M.; Moore, W. R.; Tanaka, R. D. Drugs Future 1996, 21, 811.
5. Caughey, G. H.; Raymond, W. W.; Bacci, E.; Lombardy, R. J.; Tidwell, R. R. J.
Pharmacol. Exp. Ther. 1993, 264, 676.
6. Clark, J. M.; Abraham, W. M.; Fishman, C. E.; Forteza, R.; Ahmed, A.; Cortes, A.;
Warne, R. L.; Moore, W. R.; Tanaka, R. D. Am. J. Respir. Crit. Care Med. 1995, 152,
2076.
7. Rice, K. D.; Tanaka, R. D.; Katz, B. A.; Numerof, R. P.; Moore, W. R. Curr. Pharm.
Des. 1998, 4, 381.
8. Ni, L.-M.; Powers, J. C. Bioorg. Med. Chem. 1998, 6, 1767.
9. Combrink, K. D.; Guelgeze, H. B.; Meanwell, N. A.; Pearce, B. C.; Zulan, P.;
Bisacchi, G. S.; Roberts, D. G. M.; Stanley, P.; Seiler, S. M. J. Med. Chem. 1998, 41,
4854.
10. Burgess, L. E.; Newhouse, B. J.; Ibrahim, P.; Rizzi, J.; Kashem, M. A.; Hartman, A.;
Brandhuber, B. J.; Wright, C. D.; Thomson, D. S.; Vigers, G. P. A.; Koch, K. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 8348.
11. Dener, J. M.; Wang, V. R.; Rice, K. D.; Gangloff, A. R.; Kuo, E. Y. L.; Newcomb, W.
S.; Putnam, D.; Wong, M. Bioorg. Med. Chem. Lett. 2001, 11, 2325.
12. Hopkins, C.; Neuenschwander, K.; Scotese, A.; Jackson, S.; Nieduzak, T.; Pauls,
H.; Liang, G.; Sides, K.; Cramer, D.; Cairns, J.; Maignan, S.; Mathieu, M. Bioorg.
Med. Chem. Lett. 2004, 14, 4819.
13. Levell, J.; Astles, P.; Eastwood, P.; Cairns, J.; Houille, O.; Aldous, S.; Merriman,
G.; Whiteley, B.; Pribish, J.; Czekaj, M.; Liang, G.; Maignan, S.; Guilloteau, J.-P.;
Dupuy, A.; Davidson, J.; Harrison, T.; Morley, A.; Watson, S.; Fenton, G.;
McCarthy, C.; Romano, J.; Mathew, R.; Engers, D.; Gardyan, M.; Sides, K.;
Kwong, J.; Tsay, J.; Rebello, S.; Shen, L.; Wang, J.; Luo, Y.; Giardino, O.; Lim, H.-
K.; Smith, K.; Pauls, H. Bioorg. Med. Chem. 2005, 13, 2859.
Figure 4. Stereo view of electron density (FoÀFc at
3r, calculated at 2.09 Å
resolution before the ligand was introduced) around compound 2 in the refined b-
tryptase/compound 2 complex. The coordinates of the complex structure have been
deposited with Protein Data Bank under access code 3V7T. Carbon atoms of one b-
tryptase monomer are colored in green while those of the adjacent monomer are
colored in brown. The carbon atoms of the inhibitor are colored in white. The
inhibitor bound in the adjacent binding pocket is only partially visible on the top
right corner.
and an aromatic P4 group. This newly introduced conformational
rigidity constrains the inhibitor to a conformation close to the bio-
active conformation exhibited by in silico models and X-ray struc-
tures so that the potency to b-tryptase was largely retained.
Importantly, because the compound with the tropanylamide scaf-
fold is less flexible, it is less likely for this compound to adapt a
conformation suitable for binding to off-target proteins, which is
believed to be the reason for the improved metabolic stability, de-
creased CYP inhibition, and reduced hERG channel binding affinity.
14. Hopkins, C. R.; Czekaj, M.; Kaye, S. S.; Gao, Z.; Pribish, J.; Pauls, H.; Liang, G.;
Sides, K.; Cramer, D.; Cairns, J.; Luo, Y.; Lim, H.-K.; Vaz, R.; Rebello, S.; Maignan,
S.; Dupuy, A.; Mathieu, M.; Levell, J. Bioorg. Med. Chem. Lett. 2005, 15, 2734.
15. Pereira, P. J. B.; Bergner, A.; Macedo-Ribeiro, S.; Huber, R.; Matschiner, G.; Fritz,
H.; Sommerhoff, C. P.; Bode, W. Nature 1998, 392, 306.
16. Gabriel, B.; Stubbs, M. T.; Bergner, A.; Hauptmann, J.; Bode, W.; Stuerzebecher,
J.; Moroder, L. J. Med. Chem. 1998, 41, 4240.
17. Sommerhoff, C. P.; Bode, W.; Pereira, P. J.; Stubbs, M. T.; Sturzebecher, J.;
Piechottka, G. P.; Matschiner, G.; Bergner, A. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
10984.
18. Pearlstein, R. A.; Vaz, R. J.; Kang, J.; Chen, X.-L.; Preobrazhenskaya, M.;
Shchekotikhin, A. E.; Korolev, A. M.; Lysenkova, L. N.; Miroshnikova, O. V.;
Hendrix, J.; Rampe, D. Bioorg. Med. Chem. Lett. 2003, 13, 1829.
19. Pearlstein, R.; Vaz, R.; Rampe, D. J. Med. Chem. 2003, 46, 2017.
20. Aronov, A. M. Curr. Topics Med. Chem. 2008, 8, 1113.
Supplementary data
21. Blanc, E. R. P.; Vonrhein, C.; Flensburg, C.; Lea, S. M.; Bricogne, G. Acta
Crystallogr. D Biol. Crystallogr. 2004, D60, 2210.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.12.127.