Biarylmethoxy isonipecotanilides as potent and selective inhibitors of blood coagulation factor Xa

Article abstract of DOI:10.1016/j.ejps.2010.11.010

New chloro-substituted biarylmethoxyphenyl piperidine-4-carboxamides were synthesized and assayed in vitro as inhibitors of the blood coagulation enzymes factor Xa (fXa) and thrombin. An investigation of effects of the amidine and isopropyl groups attache

Full text of DOI:10.1016/j.ejps.2010.11.010

European Journal of Pharmaceutical Sciences 42 (2011) 180–191
Contents lists available at ScienceDirect
European Journal of Pharmaceutical Sciences
journal homepage: www.elsevier.com/locate/ejps
Biarylmethoxy isonipecotanilides as potent and selective inhibitors of blood
coagulation factor Xa
Gianfranco Lopopoloa, Filomena Fiorellaa, Modesto de Candiaa, Orazio Nicolottia, Sophie Martelb,
Pierre-Alain Carruptb, Cosimo Altomarea,∗
a Dipartimento Farmaco-Chimico, University of Bari, Via E. Orabona 4, I-70125 Bari, Italy
b School of Pharmaceutical Sciences, University of Geneva, University of Lausanne, Quai Ernest-Ansermet 30, CH-1211 Geneva 4, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
New chloro-substituted biarylmethoxyphenyl piperidine-4-carboxamides were synthesized and assayed
in vitro as inhibitors of the blood coagulation enzymes factor Xa (fXa) and thrombin. An inves-
tigation of effects of the amidine and isopropyl groups attached at the piperidine nitrogen and
5-(halogenoaryl)isoxazol-3-yl groups as biaryl substituents led us to identify new compounds which
proved to be selective fXa inhibitors, with inhibition constants in the low nanomolar range. The
most potent compound 21e, that incorporates 2-Cl-thiophen-5-yl group as the P1 motif and 1-
isopropylpiperidine P4 group, inhibited fXa with Ki value of 0.3 nM and very high selectivity over thrombin
and some other tested serine proteases, achieving moderate levels of anticoagulant activity in the low
micromolar range, as assessed by the prothrombin time clotting assay (PT2 = 3.30 ␮M). Based on reli-
able docking simulations, molecular modeling provided a rationale for interpreting structure–activity
relationships. The predicted binding modes highlighted the structural requirements for addressing the
subsites S1 and S4 of the fXa enzyme.
Received 14 September 2010
Received in revised form 24 October 2010
Accepted 18 November 2010
Available online 26 November 2010
Keywords:
Piperidine-4-carboxamide
Factor Xa inhibitor
Anticoagulant
Thromboembolism
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
may have greater antithrombotic potential (Ansell, 2007). Indeed,
fXa inhibitors do not affect the existing levels of thrombin, allow-
Among the novel antithrombotics in development (Arepally
and Ortel, 2006; Krynetskiy and McDonnell, 2007), small-molecule
orally available inhibitors of proteases in the blood coagulation cas-
cade, mainly factor Xa (fXa) and thrombin, have shown improved
efficacy and lesser side effects, such as bleeding, associated with
the currently used therapies (warfarin, heparins) (Mackman, 2008;
Weitz and Linkins, 2007), which require continuous monitoring
for accurate dosing. Due to its position at the confluence of the
intrinsic and extrinsic pathways of the coagulation cascade, fXa
has emerged as an attractive target for the development of potent
and safer anticoagulant drugs. Together with fVa and calcium
ions on a phospholipid surface, fXa forms the prothrombinase
complex, which is responsible for the conversion of prothrom-
bin to thrombin, the final effector of coagulation. Despite initial
efforts in the discovery of oral anticoagulant drugs focused on
the development of small-molecule direct inhibitors of thrombin
(i.e., the oral DTIs), recently accumulated evidence has suggested
that early inhibition in the coagulation cascade at the level of fXa
ing the small amounts of remaining thrombin after fXa inhibition
to activate thrombin receptors and preserve primary haemostatic
functions (Leadley, 2001). Preclinical studies suggest that selective
inhibitors of fXa may possess a wider therapeutic index than DTIs
(Wong et al., 2009). Recent reviews described the various chemo-
types that were employed in the evolution of fXa inhibitors (de
Candia et al., 2009a; Pinto et al., 2010).
The availability of numerous fXa crystallographic structures,
paved the road to the use of structure-based drug design and mod-
eling techniques that have been successfully employed to develop
new oral fXa inhibitors (Maignan and Mikol, 2001). These tech-
niques have been critical in achieving both enzyme selectivity and
oral bioavailability. FXa contains a serine protease domain in a
trypsin-like closed ␤-barrel fold encompassing the catalytic triad
Ser195–His57–Asp102 and two essential subsites, S1 and S4. The
most potent inhibitors adopt L-shaped binding conformations, ori-
enting two almost orthogonal P1 and P4 groups to fill the S1 and
S4 pockets in the enzyme binding site.
Early fXa inhibitors brought as primary anchoring points benza-
midine (e.g., otamixaban 1; Fig. 1), naphtylamidine or other basic
groups, which in the protonated form interact with Asp189 at
the bottom of the S1 pocket, whereas the P4 aromatic moiety is
∗
Corresponding author. Tel.: +39 080 5442781; fax: +39 080 5442230.
E-mail address: altomare@farmchim.uniba.it (C. Altomare).
0928-0987/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejps.2010.11.010

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
181
O
N
O
O
H
N
P4
O
NH2
NH
P1
1, Otamixaban (FXV-673)
N
F
SO 2CH3
F
N
CF3
N
NH
O
CF3
NH
O
N
N
N
N
P4
NH2
NH2
O
N
P1
2a, DPC423
2b, Razaxaban (DPC906)
O
N
O
N
O
N
O
O
O
NH2
N
O
N
HN
N
O
P4
S
Cl
O
P1
3, Rivaroxaban (BAY 59-7939)
4, Apixaban (BMS-562247-1)
Fig. 1. Structures of five representative factor Xa inhibitors – the so-called xabans – bearing as the P1 interacting moiety a basic benzamidine group (otamixaban), less basic
amidine isosters (DPC423, razaxaban), and neutral substituted aryl groups (rivaroxaban, apixaban).
embedded into the S4 site, that is an aromatic box lined by the
aromatic side chains of Tyr99, Phe174 and Trp215 (Al-Obeidi and
Ostrem, 1999). Due to the limited oral bioavailability often associ-
ated with the first generation amidine-based fXa inhibitors, efforts
were made to replace the amidine group with less basic, such as in
DPC423 (2a) that was the first oral clinical candidate (Pinto et al.,
2001), or nonpolar neutral groups (Lam et al., 2003). These efforts
enabled the discovery of second-generation of fXa inhibitors with
improved PK properties. They include compounds bearing either
less basic amidine isosters, such as razaxaban 2b (Quan et al., 2005),
or neutral P1 substituents, such as rivaroxaban 3 (Roehrig et al.,
2005) and apixaban 4 (Pinto et al., 2007) (Fig. 1).
The chlorothiophene moiety in rivaroxaban is buried inside the
S1 pocket, with chlorine pointing towards the center of the Tyr228
aromatic ring. The gain in binding energy due to this hydrophobic
contact, that is the so-called chloro binding mode, balances the lack
of electrostatic/H-bond interactions between amidine and Asp189.
Interestingly, it has been shown that even in the presence of a ben-
zamidine plus a chloroaryl or chloroheteroaryl group, fXa-selective
inhibitors engage the enzyme binding site by orienting the neutral
group into the S1 pocket (Lumma et al., 1998). A similar binding
pose has been reported for apixaban 4, with the 4-methoxyphenyl
substituent occupying the same space of the chlorothiophene P1
moiety.

182
G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
Thr98
Tyr99
Glu97
F
F
O
S4
H
Ser195
F
O
O
HN
Phe174
Trp215
Gly216
N
O
H
H
O
Asp189
Tyr228
5
Ki(fXa) = 57 nM
Ki (thr) > 250000 nM
S1
Fig. 2. Schematic representation of the binding mode of the highly potent and selective fXa inhibitor 5, built up on the benzyloxyphenyl piperidine-4-carboxamide scaffold
(de Candia et al., 2009b), into the binding site of fXa; S1 and S4 pockets and several residues important for ligand binding are depicted. The inhibition constant values (Ki)
against factor Xa and thrombin of compound 5, which proved to be also a moderate inhibitor (IC50 = 65.7 ␮M) of ADP-induced platelet aggregation, are shown.
Rivaroxaban 3 (Roehrig et al., 2005), apixaban 4 (Pinto et al.,
2007) and other orally administered selective fXa inhibitors
(Haginoya et al., 2004) have progressed to advanced phases of clin-
ical studies (de Candia et al., 2009a).
In our previous work, we described new benzyloxy
isonipecotanilide derivatives with dual function against
thrombosis. In particular, the 4-CF3 derivative of N-[3-(1,1ꢀ-
2010), may exploit the arylhalogen binding mode in the S1 pocket
of fXa (Maignan et al., 2003). The effects on the activity of the basic
amidine and isopropyl groups appended at the piperidine nitrogen
(N1), as well as the ortho-substituted isomers of the biarylmethoxy
derivatives, were also investigated.
Herein, we report synthesis, blood coagulation factors’ enzyme
inhibition, in vitro anticoagulant activity, structure–activity rela-
tionships (SARs) and molecular modeling of biarylmethoxy
isonipecotanilide derivatives, highlighting the improvement in
their fXa inhibition potency and selectivity, as well as in antico-
agulant activity as assessed by clotting assays.
biphenyl-4-ylmethoxy)phenyl]piperidine-4-carboxamide
(5,
Fig. 2) proved to be a potent and selective fXa inhibitor (Ki = 57 nM),
showing additional antiplatelet effects with inhibition potency in
the micromolar range (de Candia et al., 2009b).
Docking calculations suggested that compound 5 and its analogs
bound human fXa through two main interactions: (i) salt bridge,
strengthen by H-bonds, between the piperidinium group and the
carboxylate of Asp189 at the bottom of the S1 subsite and (ii) H-
bond between the amide carbonyl of the ligand/s and the catalytic
residue Ser195. The consistent values of docking scores highlighted
the relevant role played by hydrophobic interactions occurring
between the distal phenyl group and the S4 aromatic pocket, rein-
forced by two middle-range contacts between C(sp3)–F and CO
moieties of Thr98 and Glu97.
Albeit interesting as in vitro fXa inhibitor and antiplatelet agent,
compound 5 was too lipophilic. Its log P (ca. 6) far exceeded
the value of 4, which is suggested as the maximum calculated
threshold for any fXa direct inhibitor that possesses good phar-
macokinetics and in vivo anticoagulant activity (Remko, 2009).
As a general trend, compounds with comparable binding affin-
ity show better anticoagulant activity if their lipophilicity is
lower.
With the aim of improving either anticoagulant activity or
physicochemical profile of the biphenylmethoxy isonipecotanilide
derivatives, we designed and synthesized new analogs, exploring
less lipophilic variants of the 1,1ꢀ-biphenyl moiety in compound
5. Hence, one phenyl group of the biphenyl moiety was replaced
by the isoxazolyl group, and the 4-CF3-phenyl (distal) group with
4-Cl-phenyl and 2-Cl-thiophen-5-yl groups, which, as shown for
a number of novel inhibitors (de Candia et al., 2009a; Pinto et al.,
2. Materials and methods
Melting points were determined by using the capillary method
on a Stuart Scientific SMP3 electrothermal apparatus and are not
corrected. IR spectra were recorded using KBr disks on a Perkin-
Elmer Spectrum One FT-IR spectrophotometer (Perkin-Elmer Ltd.,
Buckinghamshire, UK), and the most significant absorption bands
expressed in cm−1 are listed. 1H NMR spectra were recorded at
300 MHz on a Varian Mercury 300 instrument. Chemical shift val-
ues are expressed in ı and the coupling constants J in Hertz.
Abbreviations are as follows: s, singlet; d, doublet; t, triplet; q,
quartet, dd, doublet of doublets; m, multiplet; br, broad. Signals
due to NH and OH protons were located by deuterium exchange
with D2O. Mass spectra were recorded on Agilent GC–MS 689-973.
Elemental analyses (C, H, N) were performed on an Euro EA3000
analyzer (Eurovector, Milan, Italy) using the Analytical Laboratory
Service of the Dipartimento Farmaco-Chimico of the University of
Bari, and the results agreed to within 0.40% of the theoretical val-
ues. Chromatographic separations were performed on silica gel 60
for column chromatography (Merck 70-230 mesh, or alternatively
15-40 mesh for flash chromatography).
Several compounds were synthesized according to known pro-
cedures with slight modifications; their melting points and spectral
data were in full agreement with those reported in literature, and
no effort was made at this stage to optimize the yields. Unless

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
183
Br
(A)
HO
OH
B
a
b
+
Br
X
O
X
a, X = 4'-F
b, X = 4'-Cl
c, X = 3'-Cl
X
6a-c
7a-c
(B)
O
HO
Br
O
N
N
N
O
O
O
e
c, d
f
a, Ar = 4-F-C6H4
Ar
b, Ar = 4-Cl-C6H4
c, Ar = 3-Cl-C6H4
d, Ar = 4-(OCH3)-C6H4
e, Ar = 2-Cl-C4H2S
Ar
Ar
Ar
9a-e
10a-e
8a-e
Scheme 1. Reagents and conditions: (a) K2CO3, TBAB (tetrabutylammonium bromide), Pd(OAc)2, PEG400, 110 ◦C, 16 h; (b) NBS (N-bromosuccinimide), AIBN (azobisisobu-
tyronitrile), CCl4, reflux, 20 h; (c) nBuLi, dry THF, 0 ◦C, 10 min, then diethyloxalate, dry THF, rt, 3 h; (d) NH2OH·HCl, abs EtOH, reflux, 6 h; (e) NaBH4, abs. EtOH, dry THF, 0 ◦C,
4 h; (f) CBr4, PPh3, dry CH2Cl2, 4 h.
otherwise stated, all chemicals and solvents were purchased from
Sigma–Aldrich.
2.1. General procedure for the synthesis of X-substituted
4-(bromomethyl)-1,1ꢀ-biphenyls (7a–c)
General synthesis procedures, along with physicochemical data,
elemental analyses (C, H, N), IR and 1H NMR spectra of some repre-
sentative newly synthesized intermediates (8a and 10a, Scheme 1;
18b and 20b, Scheme 3) and target compounds (12b, 13a and 14b,
Scheme 2; 20b, Scheme 3), are reported below. Characterization
data for all the other newly synthesized compounds (8b–e, 12c,
13b, 13c, 15a, 15b, 15e, 16a, 16b, 16e, 17a–e, 18e, 19e, 20e, 21b,
21e, 22b, 22e) are reported in Supporting Information available on
the journal website.
Compounds 7a–c were prepared in two steps. Using X-
substituted (X = 4-F, 4-Cl and 3-Cl) phenylboronic acids and
4-bromotoluene as the starting materials, a Suzuki cross-coupling
reaction was carried out through a reported green chemistry pro-
cedure (Liu et al., 2006), that allowed the intermediate biphenyl
derivatives 6a–c to be obtained in high yields (>90%). The spec-
tral data of 6a (de Candia et al., 2009a,b), 6b and 6c (Miguez et al.,
2007) were in full agreement with those already reported. In the
O
N
O
H
N
O
c, d
e, b
X
.HCl
13a-c
14b
N
H
O
HN
.HCl
X
O
12a-c
7a-c
N
O
H
H2N
N
a, b
.HCl
NH
O
Cl
N
H
OH
N
O
Boc
N
10a-e
11
N
O
O
H
N
a, b
.HCl
O
c, d
e, b
16a, b, e
Ar
N
N
O
O
H
HN
.HCl
O
N
Ar
N
O
O
H
15a-e
H2N
N
.HCl
NH
Ar
17a-e
Scheme 2. Reagents and conditions: (a) NaH, dry DMF, rt, 4 h; (b) HCl gas, CHCl3, 0 ◦C, 30 min; (c) acetone, Na(CN)BH3, MeOH, rt, 24 h; (d) HCl (1.25 M) in MeOH, 0 ◦C, 30 min;
(e) 1,3-bis(tert-butoxycarbonyl)-2-methyl-2-thiopseudourea, HgCl2, dry DMF, 0 ◦C to rt, 24 h.

184
G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
O
N
H
N
O
.HCl
N
O
O
O
e, f
OH
O2N
H2N
N
N
Ar
H
O
21b, e
Boc
O
HN
O
a
b
+ 9b, e
.HCl
O2N
N
O
c, d
N
O
N
O
OH
g, d
Ar
Ar
O
Ar
N
H
H2N
N
O
20b, e
18b, e
19b, e
NH .HCl
N
O
Ar
22b, e
Scheme 3. Reagents and conditions: (a) DIAD (diisopropylazodicarboxylate), PPh3, dry THF, rt, 48 h; (b) SnCl2·2H2O, DMF, rt, overnight; (c) DCC (dicyclohexylcarbodiimide),
HOBt (1-hydroxybenzotriazole), dry THF, rt, 48 h; (d) HCl gas, CHCl3, 0 ◦C, 30 min; (e) acetone, Na(CN)BH3, MeOH, rt, 24 h; (f) HCl (1.25 M) in MeOH, 0 ◦C; (g) 1,3-bis(tert-
butoxycarbonyl)-2-methyl-2-thiopseudourea, HgCl2, dry DMF, 0 ◦C to rt, 24 h.
second step, to a solution of 6a (or 6b or 6c) in CCl4, a stoichiometric
amount of N-bromosuccinimmide (NBS) and a catalytic amount of
azobisisobutyronitrile (AIBN) were added, and the reaction mixture
was refluxed for 20 h. After cooling to room temperature, the insol-
uble succinimide was filtered off and the solvent was evaporated
under reduced pressure to provide the target bromomethyl deriva-
tives 7a–c as crude yellow solids (yields in the 60–95% range, as
assessed by GC–MS), which were used in the subsequent reactions
without further purification. The spectral data of 7a (X = 4ꢀ-F; de
Candia et al., 2009b), 7b (X = 4ꢀ-Cl; Van Duzer and Roland, 1994) and
7c (X = 3ꢀ-Cl; Van Duzer and Roland, 1994) were in full agreement
with the reported ones.
compound 8a (1.85 g, 49% yield) as white solid, which was used in
the subsequent reaction.
IR (cm−1) 1733, 1247, 1139, 842, 771, 812. 1H NMR (CDCl3) ı
7.80 (dd, 2H, J = 8.8, 5.2 Hz), 7.19 (t, 2H, J = 8.5 Hz), 6.88 (s, 1H), 4,47
(q, 2H, J = 7.2 Hz), 1.44 (t, 3H, J = 7.2 Hz).
2.3. General procedure for the synthesis of
3-(bromomethyl)-5-arylisoxazoles (10a–e)
According to Scheme 1(B), the ethyl carboxylates 8a–e were
converted by reaction with NaBH4 into the corresponding methyl
alcohols 9a–e (yields in the range 85–95%), which were used in
the subsequent bromination reaction without further purification.
All the crude [5-arylisoxazol-3-yl]methanol products (9a–e) were
converted to the corresponding 3(bromomethyl)-5-arylisoxazole
derivatives (10a–e) under mild conditions through the Appel reac-
tion (Appel, 1975), using CBr4 and triphenylphosphine (PPh3), with
good yields (60–95%, as assessed by GC–MS). The preparation of 3-
(bromomethyl)-5-(4-fluorophenyl)isoxazole 10a is reported as an
example.
2.2. General procedure for the synthesis of ethyl
5-(aryl)isoxazole-3-carboxylates (8a–e)
Compounds 8a–e were synthesized through known procedures
(Nazaré et al., 2004, 2005), with slight modifications. The prepara-
tion of 8a is described as representative example.
2.2.1. Ethyl 5-(4-fluorophenyl)isoxazole-3-carboxylate (8a)
To a stirred solution of 1-(4-fluorophenyl)ethanone (1.96 ml,
18.1 mmol) in 60 ml of dry THF, cooled at a 0 ◦C and under
argon atmosphere, a 2.5 M solution of n-butyl lithium in n-hexane
(8.69 ml, 21.7 mmol) was added dropwise. After 10 min of stirring
at 0 ◦C, diethyl oxalate (2.95 ml, 21.7 mmol) was added dropwise,
and stirring was continued at room temperature for 3 h. The reac-
tion was then quenched by addition of aqueous 2 M HCl (13.6 ml,
27.1 mmol) and the mixture was partitioned between water and
EtOAc. The aqueous phase was extracted twice with EtOAc and the
combined organic layers were washed with brine, dried (Na2SO4)
and concentrated under reduced pressure to afford an oil residue.
Hydroxylamine hydrochloride (1.38 g, 19.9 mmol) was added to the
obtained residue dissolved in 50 ml of abs EtOH, and the mixture
was refluxed for 6 h. After cooling to room temperature, the sol-
vent was removed by evaporation under reduced pressure, and the
obtained residue was dissolved in EtOAc and washed with water.
The organic layer was dried (Na2SO4) and concentrated under
reduce pressure. Purification by silica gel chromatography of the
residue, eluting with petroleum ether/EtOAc (90:10, v/v), provided
2.3.1. 3-(Bromomethyl)-5-(4-fluorophenyl)isoxazole (10a)
To a cooled solution of ethyl 5-(4-fluorophenyl)isoxazole-3-
carboxylate 8a (0.50 g, 2.13 mmol) in a mixture of 6 ml of dry THF
and 6 ml of absolute ethanol, NaBH4 (0.12 g, 3.19 mmol) was added
portionwise. The reaction mixture was stirred at 0 ◦C for 4 h, then
quenched by addition of aqueous 1 N HCl, and partitioned between
EtOAc and water. The aqueous phase was twice extracted with
EtOAc, and the combined organic layers were washed with brine,
dried (Na2SO4) and concentrated under reduced pressure to pro-
vide 9a as white solid (0.35 g, 85% yield). To a solution of the product
9a (0.35 g, 1.81 mmol) in 15 ml of CH2Cl2 cooled at 0 ◦C, CBr4 (0.66 g,
1.99 mmol) and PPh3 (0.52 g, 1.99 mmol) were added, and the mix-
ture was stirred at room temperature for 24 h. After the removal
of the solvent under reduced pressure, the residue was purified by
silica gel chromatography, using the petroleum ether/EtOAc mix-
ture (90:10, v/v), and compound 10a obtained as pale brown solid
(0.29 g, 62% yield).

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
185
2.4. General procedure for the synthesis of compounds 12a–e
and 15a–e
1H), 7.70 (dd, 2H, J = 9.3, 5.2 Hz), 7.66 (d, 2H, J = 8.3 Hz), 7.50 (d,
2H, J = 8.2 Hz), 7.43 (s, 1H), 7.28 (t, 2H, J = 8.8 Hz), 7.22–7.12 (m,
2H), 6.71 (d, 1H, J = 7.4 Hz), 5.10 (s, 2H), 3.43 (d, 2H, J = 10.5 Hz),
3.25–3.21 (m, 1H), 2.92–2.83 (m, 2H), 2.64–2.55 (m, 1H), 2.08–1.89
(m, 4H) and 1.26 (d, 6H, J = 6.6 Hz). MS (ESI, positive ion) m/z: 447.1
(M+H). Anal. Calcd. for C28H31ClN2O2 × HCl × 2.5H2O: C, 61.71; H,
6.80; N, 5.14%; found: C, 61.54; H, 6.45; N, 5.28%.
The piperidine-4-carboxamides derivatives 12a–e and 15a–e
were synthesized according to previously reported procedures (de
Candia et al., 2009a,b). The preparation of compound 12b as HCl
salt is reported as an example.
2.4.1. N-{3-[(4ꢀ-Chloro-1,1ꢀ-biphenyl-4-
2.6. General procedure for guanylation of compounds 12b (14b),
15a–e (17a–e) and 20b, e (22b, e)
yl)methoxy]phenyl}piperidine-4-carboxamide hydrochloride
(12b)
Asolutionof t-butyl4-[N-(3-hydroxyphenyl)carbamoyl]piperidine-
1-carboxylate 11 (1.56 g, 4.86 mmol) in 5 ml of dry DMF was added
to a suspension of NaH (0.14 g, 5.83 mmol) in 6 ml of dry DMF
at room temperature. After 15 min of stirring, a solution of 4-
(bromomethyl)-4ꢀ-chloro-1,1ꢀ-biphenyl (1.37 g, 4.86 mmol) in
6 ml of dry DMF was added dropwise. The reaction mixture was
stirred for 4 h at room temperature, and then quenched by adding
5 ml of 5% (w/v) KHSO4. The mixture was then diluted with water
and extracted with EtOAc. The combined organic extracts were
washed with 2 N NaOH, brine, dried (Na2SO4) and concentrated
under reduced pressure. The residue was purified by silica gel
chromatography, eluting with petroleum ether/EtOAc (90:10, v/v),
and the chromatographically pure N-BOC protected derivative
was treated in chloroform solution with HCl gas to remove the
BOC protecting group. Subsequent purification by crystallization
afforded 12b as white solid (1.45 g, 65% yield), mp 246-248 ◦C from
EtOH/EtOAc. IR (cm−1) 3142, 2955, 1661, 1226, 1160, 1137, 1091,
1036, 809, 779. 1H NMR (DMSO-d6) ı 10.11 (s, 1H), 8.72 (br, 2H),
7.70–7.66 (m, 4H), 7.58–7.49 (m, 4H), 7.42 (s, 1H), 7.21–7.11 (m,
2H), 6.70 (d, 1H, J = 7.7 Hz), 5.10 (s, 2H), 3.29 (d, 2H, J = 13.2 Hz),
2.89 (d, 1H, J = 12.4 Hz), 2.85 (d, 1H, J = 12.4 Hz), 2.65–2.58 (m, 1H),
1.97–1.94 (m, 2H) and 1.90–1.72 (m, 2H). MS (ESI, positive ion)
m/z: 421.1 (M+H). Anal. Calcd. for C25H25ClN2O2 × HCl × H2O: C,
63.16; H, 5.94; N, 5.89%; found: C, 63.55; H, 5.69; N, 6.03%.
The
preparation
of
N-{3-[(4ꢀ-chloro-1,1ꢀ-biphenyl-4-
yl)methoxy]phenyl}-1-carbamimidoylpiperidine-4-carboxamide
hydrochloride 14b is reported as a representative example.
2.6.1. N-{3-[(4ꢀ-Chloro-1,1ꢀ-biphenyl-4-yl)methoxy]phenyl}-1-
carbamimidoylpiperidine-4-carboxamide hydrochloride
(14b)
1,3-bis(tertbutoxycarbonyl)-2-methyl-2-thiopseudourea
(0.38 g, 1.31 mmol), mercury(II) chloride (0.36 g, 1.31 mmol)
and triethylamine (0.46 ml, 3.27 mmol) were added to
a
cooled solution (0 ◦C) of N-{3-[(4ꢀ-chloro-1,1ꢀ-biphenyl-4-
yl)methoxy]phenyl}piperidine-4-carboxamide
hydrochloride
12b (0.50 g, 1.09 mmol) in 8 ml of dry DMF. The reaction mixture
was stirred for 1 h at 0 ◦C and 24 h at room temperature; 10 ml of
EtOAc were then added, and stirring was continued for 15 min. The
precipitate was filtered off and washed 3 times with EtOAc. The
combined organic phases were washed with brine, dried (Na2SO4),
and concentrated under reduced pressure to afford a residue,
which was purified by silica gel chromatography, eluting with
petroleum ether/EtOAc (90:10, v/v). The pure N,Nꢀ-bis-BOC pro-
tected derivative obtained was treated with HCl gas in chloroform
solution to afford 0.21 g (39% yield) of the product HCl salt 14b as
white microcrystalline solid; mp 260–261 ◦C, from EtOH/EtOAc. IR
(cm−1) 3211, 1651, 1222, 1156, 1093, 1038, 954, 872, 810, 774. 1
H
NMR (DMSO-d6) ı 10.06 (s, 1H), 7.77–7.66 (m, 4H), 7.58–7.48 (m,
4H), 7.44 (br, 4H), 7.21–7.12 (m, 2H), 6.70 (d, 1H, J = 8.0 Hz), 5.10
(s, 2H), 3.88 (d, 2H, J = 13.6 Hz), 3.06 (d, 1H, J = 12.1 Hz), 3.02 (d, 1H,
J = 11.3 Hz), 2.68–2.61 (m, 1H), 1.88–1.83 (m, 2H) and 1.64–1.52
(m, 2H). MS (ESI, positive ion) m/z: 463.1 (M+H). Anal. Calcd. for
2.5. General procedure for the synthesis of
1-isopropylpiperidine-4-carboxamide derivatives (13a–c, 16a, b,
e and 21b, e)
The introduction of the isopropyl group on piperidine nitrogen
was achieved by one pot reductive alkylation of the piperidine-
4-carboxamide derivatives (yields: 38–95%). The preparation of
N-{3-[(4ꢀ-chloro-1,1ꢀ-biphenyl-4-yl)methoxy]phenyl}piperidine-
4-carboxamide hydrochloride 13a is reported as an example.
C
27H26ClN3O2 × HCl × H2O: C, 60.35; H, 5.84; N, 10.83%; found C,
60.74; H, 5.65; N, 11.07%.
2.7. General procedure for the synthesis of
2-[(5-arylisoxazol-3-yl)methoxy]aniline derivatives (19b, e)
2.5.1. N-{3-[(4ꢀ-Fluoro-1,1ꢀ-biphenyl-4-yl)methoxy]phenyl}-1-
(1-methylethyl)piperidine-4-carboxamide hydrochloride
(13a)
The aniline derivatives 19b and 19e, necessary for synthe-
sizing the ortho isomers of the biaryl-containing piperidine-4-
carboxamides 20b and 20e (Scheme 3), were prepared in two steps.
At first, a Mitsunobu coupling was carried out with 2-nitrophenol
and suitable 5-arylisoxazol-3-yl-methanol derivatives 9b and 9e,
to furnish the nitrophenyl compounds 18b and 18e, which were
then reduced to 19b and 19e using SnCl2 × 2H2O. The preparation
of compound 19b is reported as an example.
To a cooled solution of the already reported N-{3-[(4ꢀ-fluoro-
1,1ꢀ-biphenyl-4-yl)methoxy]phenyl}piperidine-4-carboxamide
hydrochloride (de Candia et al., 2009a,b) 12a (0.33 g, 0.76 mmol) in
15 ml of MeOH, acetone (2 ml) and Na(CN)BH3 (52 mg, 0.83 mmol)
were added. The reaction mixture was stirred at room tempera-
ture for 24 h, and then concentrated under reduced pressure. The
residue was partitioned between EtOAc and 2 N NaOH, and the
aqueous layer was separated and extracted twice with EtOAc. The
combined organic layers were dried (Na2SO4) and concentrated
under reduced pressure to afford the free base, which was con-
verted into the hydrochloride salt by treatment with 5 ml of 1 N
HCl in MeOH at 0 ◦C for 30 min. Removal of the solvent in vacuo
and purification by crystallization afforded the 0.16 g (44% yield)
of the product HCl salt 13a as white microcrystalline solid; mp
223–225 ◦C, from EtOH/EtOAc. IR (cm−1) 3137, 1669, 1224, 1158,
1035, 811, 775, 688. 1H NMR (DMSO-d6) ı 10.14 (s, 1H), 9.73 (br,
2.7.1. 5-(4-Chlorophenyl)-3-[(2-nitrophenoxy)methyl]isoxazole
(18b)
To
a
cooled (0 ◦C) stirred solution of 2-nitrophenol
(1.15 g, 8.28 mmol), PPh3 (2.17 g, 8.28 mmol), and [5-(4-
chlorophenyl)isoxazol-3-yl]methanol 9b (1.74 g, 8.28 mmol)
in 20 ml of dry THF, diisopropylazodicarboxylate DIAD (1.63 ml,
8.28 mmol), dissolved in 5 ml of dry THF, was added dropwise, and
the mixture was stirred at room temperature for about 48 h. After
removal of the solvent under reduced pressure, the oil residue
was purified by silica gel chromatography, eluting with petroleum

186
G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
ether/EtOAc (90:10, v/v), to provide compound 18b (1.65 g, 60%
yield) as pale-yellow solid. IR (cm−1) 1615, 1529, 1347, 1277,
1257, 1094, 1003, 864, 835, 799, 740. 1H NMR (CDCl3) ı 7.88 (d,
1H, J = 8.3 Hz), 7.74 (d, 2H, J = 6.6 Hz), 7.55 (t, 1H, J = 8.8 Hz), 7.45 (d,
2H, J = 6.6 Hz), 7.23 (d, 1H, J = 8.0 Hz), 7.10 (t, 1H, J = 8.3 Hz), 6.74 (s,
1H) and 5.35 (s, 2H).
(Milan, Italy). Enzymes and substrates were obtained as fol-
lows: for fXa kinetic studies, 2 nM human factor Xa and 0.04 mM
S-2765 (Z-d-Arg–Gly–Arg–p-nitroanilide) from Chromogenix AB-
Instrumentation Laboratories (Milan, Italy); for fIIa kinetic studies,
0.41 U/ml bovine thrombin from Sigma–Aldrich (Milan, Italy)
and 0.050 mM S-2238 (d-Phe–Pip–Arg–p-nitroanilide) from Chro-
mogenix AB-Instrumentation Laboratories (Milan, Italy); for ␣-CT
kinetic studies, 0.4 ␮g/ml bovine ␣-chymotrypsin and 0.185 mM N-
succinyl–Ala–Ala–Pro–Phe–p-nitroanilide from Sigma–Aldrich Co.
(Milan, Italy). In kinetic studies with fXa or thrombin, 500 ␮l of the
enzyme solution in a 10 mM Tris buffer (pH 8), containing 0.15 M
NaCl and 0.1% of PEG 6000, was mixed with 10 ␮l of DMSO solution
of the test compound, its final concentration ranging from 100 ␮M
to 0.1 nM; DMSO, derived from stock solutions, did not exceed a
final concentration of 1% (v/v) (10 ␮l of DMSO in control assays).
The enzymes were incubated with the test compound or its sol-
vent (DMSO) for 30 min at 25 ◦C, and the reactions were initiated
by the addition of 500 ␮l of the appropriate substrate solution in
buffer. Increase in absorbance due to proteolytic cleavage of the
substrate was monitored at 25 ◦C and 405 nm for 5 min.
In kinetic studies with ␣-CT, 845 ␮l of 50 mM Tris buffer (pH
7.5), containing 50 mM of CaCl2, were added to 50 ␮l of enzyme
solution in buffer and 5 ␮l of solution of the test compound (final
concentration ranging from 10 nM to 0.25 mM) or its solvent. After
10 min of incubation at 25 ◦C, 100 ␮l of the substrate solution in
buffer was added and the absorbance increase at 405 nm was
monitored at 25 ◦C for 5 min. Initial velocities were determined
by analysis of the initial linear portion of the kinetic curve, and
values were plotted against log inhibitor concentration. IC50 val-
ues were obtained by nonlinear (sigmoid) regression calculation
on dose–response plots in at least three different determinations,
and used in the Cheng–Prusoff equation for calculating inhibition
constants Ki (Cheng and Prusoff, 1973).
2.7.2. 2-{[5-(4-Chlorophenyl)isoxazol-3-yl]methoxy}aniline
(19b)
SnCl2 × 2H2O (6.19 g, 27.4 mmol) was added to a solution of
18b (1.65 g, 5.49 mmol) in 15 ml of DMF. The reaction mixture was
stirred overnight at room temperature, and then poured into ice
and alcalinized to pH 12 by adding aqueous 2 N NaOH. The resulting
suspension was filtered, and the collected solid was washed with
Et2O. The aqueous phase was extracted twice with Et2O, and the
combined organic phases were washed with brine, dried (Na2SO4),
and concentrated under reduced pressure. The oil residue was
purified by silica gel chromatography, eluting with petroleum
ether/EtOAc (60:40, v/v), to yield 0.99 g of the product 19b (60%) as
pale-red solid.
IR (cm−1) 3442, 3358, 1613, 1508, 1460, 1435, 1280, 1217, 1095,
1055, 835, 818, 734. 1H NMR (CDCl3) ı 7.72 (d, 2H, J = 9.1 Hz), 7.44
(d, 2H, J = 9.1 Hz), 6.92–6.82 (m, 2H), 6.77–6.70 (m, 2H), 6.63 (s, 1H),
5.22 (s, 2H), 3.83 (br, 2H).
2.8. General procedure for the synthesis of substituted N-(2-{[5-
arylisoxazol-3-yl]methoxy}phenyl)piperidine-4-carboxamides
hydrochloride (20b, e)
The preparation of compound 20b is reported as an example.
2.8.1. N-(2-{[5-(4-Chlorophenyl)isoxazol-3-
yl]methoxy}phenyl)piperidine-4-carboxamide hydrochloride
(20b)
N-Hydroxybenzotriazole hydrate (0.44 g, 3.28 mmol), N,Nꢀ-
dicyclohexylcarbodiimide (0.68 g, 3.28 mmol) were added to a
cooled solution of 1-(t-butoxycarbonyl)piperidine-4-carboxylic
acid (0.82 g, 3.61 mmol) in 15 ml of THF. After stirring for 30 min,
2-{[5-(4-chlorophenyl)isoxazol-3-yl]methoxy}aniline 19b (0.99 g,
3.28 mmol) was added, and the reaction mixture was stirred at
room temperature for 48 h. Then, the formed dicyclohexylurea was
filtered off and the filtrate was concentrated under reduced pres-
sure. The obtained residue was dissolved in EtOAc and sequentially
washed with 5% (w/v) NaHCO3, 1 N HCl and brine. The organic
phase was dried (Na2SO4) and concentrated under reduced pres-
sure, to provide the N-BOC protected derivative, which was treated
with HCl gas in chloroform solution for removing the BOC pro-
tecting group. Purification by crystallization of the crude product
afforded 0.80 g of compound 20b (50% yield) as white microcrys-
talline solid; mp 211–215 ◦C from EtOH/EtOAc. IR (cm−1): 3129,
1664, 1205, 1092, 1052, 800, 766, 754. 1H NMR (DMSO-d6) ı 9.22
(s, 1H), 8.84 (br, 1H), 8.67 (br, 1H), 7. 88 (d, 2H, J = 9.1 Hz), 7.78
(d, 1H, J = 7.3 Hz), 7.61 (d, 2H, J = 8.5 Hz), 7.23 (s, 1H), 7.18 (d, 1H,
J = 7.3 Hz), 7.09 (t, 1H, J = 7.0 Hz), 6.94 (t, 1H, J = 7.0 Hz), 5. 29 (s,
2H), 3.25 (d, 2H, J = 12.4 Hz), 2.90–2.75 (m, 3H) and 1.94–1.73
(m, 4H). MS (ESI, positive ion) m/z: 412.0 (M+H). Anal. Calcd. for
2.10. In vitro coagulation assays
Pooled lyophilized human plasma (100 ␮l) and PT reagent
(200 ␮l) (Futura System s.r.l., Formello, Rome, Italy) were used.
Inhibitor or solvent (DMSO, max 1%, v/v) was added to plasma
and incubated for 3 min at 37 ◦C. Clotting times (s) were deter-
mined in a coagulometer (Behnk electronic GmbH, Norderstedt,
Germany) and compared with those from the appropriate human
control plasma. Each measurement was performed in triplicate,
and the concentrations of the test compound which caused a 2-
fold prolongation of the prothrombin time (PT2) were calculated
from concentration–response curves.
2.11. Determination of lipophilicity parameters by hydrophilic
interaction liquid chromatography (HILIC)
Retention measurements were performed at 23 ◦C by using a
ZIC-pHILIC column (sulfoalkylbetaine phase on a polymeric sup-
port, 10 cm × 4.6 mm, 5 ␮m) from SeQuant (Umeå, Sweden). All
the measurements were made at temperature of 23 ◦C, flow-rate
1.0 ml/min on a Merck Hitachi EliteLaChrom liquid chromatograph
(Merck, Darmstadt, Germany, and Hitachi Instrument, Inc., San
Jose, CA, USA) equipped with a L-2200 auto sampler, a L-2130
pump, and a L-7614 degasser. Detection was performed using a
L-2400 UV detector operating at 230 nm for all the analyzed com-
pounds. The chromatographic system was controlled by a EzChrom
Elite System Manager software version 3.1.7.
C
22H22ClN3O3 × HCl × H2O: C, 56.66; H, 5.40; N, 9.01%; found: C,
56.27; H, 5.30; N, 8.71%.
2.9. Inhibition assays of factor Xa and related serine proteases
The target compounds were assessed for their inhibitory activ-
ity toward factor Xa (fXa), thrombin, and ␣-chymotrypsin (␣-CT)
by kinetic analysis using chromogenic substrates monitored at
405 nm. All buffer salts were obtained from Sigma–Aldrich Co.
Phoebus software 1.0 (Sedere, Centre Analyze, Orleans, France)
was used to prepare buffer (trifluoroacetic acid/ammonium) at pH
2 and ionic strength (I) equal to 100 mM. The pH of the mobile
phase were measured both before (wwpH) and after the adjunction

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
187
Table 1
Table 2
Inhibition constants for fXa and thrombin (thr) of 1,1ꢀ-biphenylmethoxy
Inhibition constants for fXa, thrombin (thr), ␣-chymotrypsin (␣-CT), and anti-
coagulant activity of the meta-substituted isoxazole-containing biarylmethoxy
isonipecotanilides 15–17.
isonipecotanilides 12–14.
b
No.
X
Ki (nM)a
log Pcalc
b
c
d
No. Ar
Kia (nM)
PT2
log PN
log Pcalc
fXa
thr
(␮M)
fXa
thr
␣-CT
c,d
12a
12b
12c
13a
13b
13c
14b
4ꢀ-F
3870c
33300
3215
139
165
16630
31
5.14
5.58
5.58
6.25
6.70
6.70
4.80
4ꢀ-Cl
3ꢀ-Cl
4ꢀ-F
d
d
d
d
e
e
e
e
e
15a 4-F-phenyl
15b 4-Cl-phenyl
15e 2-Cl-thiophen-5-yl 310
16a 4-F-phenyl
16b 4-Cl-phenyl
16e 2-Cl-thiophen-5-yl 95
50000
4680
ND
307
4.42
4.72
4.64
5.21
5.46
5.44
3.36
3.80
3.62
4.48
4.92
4.74
2.58
3.02
3.02
2.46
2.84
28200 327
4ꢀ-Cl
3ꢀ-Cl
4ꢀ-Cl
e
2280
218
20100
282
e
20730
2040
21400 196
16000 35400 71.3
18300 4240
6340
15380 1970
16000 5490
15000 5810
17a 4-F-phenyl
447
26
388
127
>500 4.99
a
In vitro inhibition constants against human fXa and bovine thr. Data are means
of at least three different determinations, each performed in duplicate (S.E.M. < 5%
of the mean).
17b 4-Cl-phenyl
17c 3-Cl-phenyl
17d 4-(OCH3)-phenyl
59100 109 5.37
226
124
122
5.29
4.92
5.25
b
Log of 1-octanol/water partition coefficient calculated using KOWWIN v. 1.67
17e 2-Cl-thiophen-5-yl 15
program (USEPA, 2007) within the EPI SuiteTMs.
c
a
Previously reported data (de Candia et al., 2009a,b).
Less than 30% inhibition at the maximum concentration tested (100 ␮M).
In vitro inhibition constants against human fXa and bovine thr and ␣-CT. Data
are means of at least three different determinations, each performed in duplicate
(S.E.M. < 5% of the mean).
d
b
Concentration required to double the prothrombin time in human plasma.
1-Octanol/water partition coefficient of neutral form calculated with Eq. (1)
of organic solvent (swpH), with the pH-meter calibrated in aqueous
buffer in both cases, even if swpH values have to be taken carefully
(Roses and Bosch, 2002).
c
(Guillot, 2010) using the difference parameter (ꢀlog k0–95
) between two iso-
cratic log k values of the basic compounds in their cationic forms, as assessed by
hydrophilic interaction liquid chromatography (HILIC; definitions and details in
Section 2).
According to Bard et al. (2009) and Guillot (2010), the differ-
ence parameter (ꢀlog k0–95) between two isocratic log k values of
the basic compounds in their cationic forms, namely log k0, that
is the log k obtained in 100% aqueous buffer (pH 2) mobile phase,
and log k95, that is the isocratic log k measured using a 95% (v/v)
ACN–aqueous buffer (pH 2) mobile phase, allows the partition
coefficient of their neutral form (log PN) to be determined. The
HILIC log k values are summarized in a table reported in Supporting
Information. This method offers fast measurements of lipophilic-
ity parameters for lipophilic strong basic compounds (Eq. (1); Bard
et al., 2009):
d
Log of 1-octanol/water partition coefficient calculated using KOWWIN v. 1.67
program (USEPA, 2007) within the EPI SuiteTMs.
e
Less than 30% inhibition at the maximum concentration tested (100 ␮M).
selected for our investigations. The GoldScore fitness function is
based on a linear combination of values relative to protein–ligand
hydrogen bond (HB), protein–ligand and ligand internal van der
Waals forces and ligand torsional strain energies. The inter-
ested reader is referred elsewhere for an in-depth description of
the GOLD docking principles (Jones et al., 1997; Verdonk et al.,
2003).
log PN = 1.07( 0.056) ꢀlog k0–95 + 0.61( 0.162)
n = 44; r2 = 0.957; s = 0.39; F = 9353
(1)
The X-ray crystal structure of human fXa in complex with
a 1H-indole-2-carboxamide-based inhibitor (PDB code: 2BOH)
(Nazaré et al., 2005) was retrieved and used as the target struc-
ture. According to reported computational procedures used by
us in ligand- and structure-based design of serine protease
inhibitors (Nicolotti et al., 2008, 2009, 2010), GOLD settings
were tuned and calibrated by docking the co-crystallized lig-
and (2BOH) within its empty crystal structures. Residues within
The experimental values demonstrated that these strong
basic compounds remain lipophilic in their neutral form
(4.12 < log PN < 5.46) but also suggested distribution coefficients
at pH 7.4 (1.57 < log D < 3.01) adequate for good pharmacokinetic
properties. The comparison between the lipophilicity parameters
obtained by HILIC and those calculated using KOWWIN v. 1.67
program (USEPA, 2007) within the EPI SuiteTMs (Tables 2 and 3)
confirm that compounds 15, 16, 17, 20, 21 and 22 are less lipophilic
than compounds 12, 13 and 14 (Table 1). The differences between
calculated and experimental log P may result mostly from the theo-
retical method used for compounds 15, 16, 20 and 21. For guanidine
derivatives 17 and 22 the larger differences may also reflect the par-
ticular retention mechanism on HILIC phases (plot in Supporting
Information). Work, out of the scope of this study, is in progress to
clarify this point.
˚
a radius of 11 A from the alpha carbon of Gly216 were consid-
ered in docking simulations. The inhibitor molecules were built
using the Sybyl fragment libraries (Tripos Associates, St. Louis,
MO, USA). Geometrical optimization and charge calculation were
performed by means of a quantum mechanical method with the
PM3 Hamiltonian. Conformational analysis and molecular over-
lays were performed according to the procedures and methods
described previously. According to Murcia et al. (2006), the basic
N1 amino and amidine groups of the ligands were protonated,
whereas the carboxylic groups in the protein residues were con-
sidered to be deprotonated. The target protein was prepared
by removing water and the co-crystallized inhibitors. Controls
were carefully carried out on the protonation state of polar
residues leaning into the active site (Böhm et al., 1999). Using
the ‘Protein Preparation’ module in Maestro 7.5 (Maestro 7.5.112;
Schröedinger, LLC: New York), light relaxation was performed to
optimise hydroxyl and thiol torsions followed by all-atom con-
strained minimization to relieve steric clashes until the RMSD
2.12. Molecular docking calculations
Unlike our previous simulations (de Candia et al., 2009b) per-
formed with the program QXP (McMartin and Bohacek, 1997),
the software GOLD (Jones et al., 1997) was herein preferred for
docking analysis because it was extensively trained on a large num-
ber of complexes selected from the CCDC/ASTEX validation set
including a large number of crystal structures of serine proteases
(Nissink et al., 2002). GOLD is equipped with two equally reliable
fitness functions, namely ChemScore and GoldScore (Verdonk et al.,
2003), the latter being the first and most widely used and thus
˚
reached 0.18 A. Using the refined crystal structures, GOLD was
set to generate 10 docking poses per run for each calculated
ligand.

188
G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
Table 3
3.2. Structure–activity relationships
Inhibition constants for fXa, thrombin (thr), ␣-CT, and anticoagulant activity of the
ortho-substituted biarylmethoxy isonipecotanilides 20–22.
The newly synthesized isonipecotamide-based compounds,
which are fully protonated under the condition of the biologi-
cal assay at pH equal to 8, were evaluated in vitro for inhibition
of fXa and thrombin and their potencies, expressed as inhibition
constants (Ki, nM), are summarized in Tables 1–3. In addition,
the meta- and ortho-substituted isoxazole-containing derivatives
(15–17 in Table 2 and 20–22 in Table 3, respectively) were assayed
for their selectivity over ␣-chymotrypsin (␣-CT), taken as a sur-
rogate determination for achieving selectivity over other enzymes
of the trypsin-like protease family for protein digestion in the gas-
trointestinal tract, and further characterized using the prothrombin
time (PT) clotting assay in human plasma (the in vitro anticoagu-
lant potency is reported as the concentration of inhibitor required
to produce a doubling of the uninhibited clotting time, PT2).
The enzyme inhibition data of the 4(or 3)-halogen-1,1ꢀ-
biphenyl-containing isonipecotanilide derivatives 12–14 (Table 1;
previously reported data of compound 12a are included for
comparison) allowed us to expand the SAR scope of molecules
encompassing compound 5 to N1-substitution with isopropyl and
amidine groups and halogen substituents (F and Cl) on the dis-
tal phenyl group. With the exception of the 3-Cl congener (13c),
which on the other hand did not show any activity of interest, the
secondary N1-isopropyl amine derivatives 13a (X = 4ꢀ-F) and 13b
(X = 4ꢀ-Cl) proved to be fXa inhibitors (Ki values equal to 139 and
165 nM, respectively) more potent than the corresponding primary
N1–H amine derivatives 12a and 12b. The guanidine derivative 14b
had 5-fold better affinity for fXa (Ki = 31 nM) compared with the
corresponding N1-isopropyl derivative 13b.
With regard to the N1-substitution, similar trends in the
fXa inhibitory potency (i.e., N-amidine > N-isopropyl > N–H) were
observed with the isonipecotanilide derivatives bearing 5-
arylisoxazol-3-yl (15–17) instead of the biphenyl moiety (Table 2).
The 4-Cl-phenyl derivatives 16b and 17b demonstrated similar
potencies to 13b and 14b in the biphenyl series, whereas replacing
the 4-Cl-phenyl group with the slightly more polar 2-Cl-thiophen-
5-yl group provided about 2-fold enhanced binding affinity to fXa
either for N1-isopropyl (16e) and N1-amidine (17e) derivatives.
Among the N1–H derivatives, 15e was the only compound with a
submicromolar fXa Ki value.
The most potent fXa inhibitor of the meta-substituted series
17e (Ki = 15 nM), that is the derivative with the N1-amidine P1
group and the 2-chlorothiophenyl P4 group, exhibited 1000-fold
selectivity over thrombin and just nearly 400-fold selectivity over
chymotrypsin. However, the N1-isopropyl analog 16e, although
less selective over thrombin and ␣-CT, showed higher anticoag-
ulant activity (PT2) than the corresponding N1-amidine derivative
17e, suggesting a role of lipophilicity in the anticoagulant activ-
ity. Compound 16e showed better membrane permeability than
17e (results not shown), as assessed with PAMPA (parallel artifi-
cial membrane permeability assay) performed in conditions which
allows the potential for gastrointestinal absorption of ionizable
drugs to be estimated (Wohnsland and Faller, 2001; Ottaviani et al.,
2008).
Kia (nM)
PT2
log PN
log Pcalc
b
c
d
No. Ar
(␮M)
fXa
thr
␣-CT
e
e
e
e
20b 4-Cl-phenyl
11740
390
190
13.3
3.30 5.03
52.8
10.6
4.40
4.12
5.19
3.24
3.06
4.36
4.18
2.46
2.28
20e 2-Cl-thiophen-5-yl 54
21b 4-Cl-phenyl
21e 2-Cl-thiophen-5-yl 0.3
22b 4-Cl-phenyl
22e 2-Cl-thiophen-5-yl
7
13700 53900
11010 60000
21200 34700
60
2
4.96
4.78
e
6824
a–e: See footnotes to Table 2.
3. Results and discussion
3.1. Chemistry
The isonipecotamide-based derivatives investigated in this
study (12–17 and 20–22) were synthesized as shown in
Schemes 1–3.
The 4-(bromomethyl)-1,1ꢀ-biphenyl intermediates 7a–c were
prepared in good yields according to Scheme 1(A). A Suzuki
cross-coupling reaction was performed through a cheap green-
chemistry procedure (Liu et al., 2006), using PEG400 as the solvent
and very low amounts of palladium acetate as catalyst, which
afforded the biphenyl derivatives 6a–c in high yields. Subsequent
radicalic bromination of compounds 6a–c yielded the target bro-
mides 7a–c. The 3-(bromomethyl)-5-arylisoxazole intermediates
10a–e were prepared according to the reaction pathway illus-
trated in Scheme 1(B). The ethyl 5-arylisoxazol-3-carboxylate
derivatives 8a–e were accessed, using slight modifications of
known procedures (Nazaré et al., 2005), through lithiation of 1-
arylethanones, carried out in dry THF at 0 ◦C, and subsequent
reaction of the generated carbanions with diethyloxalate. The
obtained dicarbonyl intermediates were treated with hydroxy-
lamine hydrochloride to afford compounds 8a–e, which were
then reduced to primary alcohols 9a–e with NaBH4. Their
bromination with CBr4 and PPh3 yielded the target derivatives
10a–e.
The meta-substituted isonipecotanilides 12–17 were synthe-
sized as shown in Scheme 2. The N-BOC protected phenol
intermediate 11 (de Candia et al., 2009a,b) was alkylated
with the bromomethyl-biaryl derivatives 7a–c and 10a–e.
Removal of the BOC-protecting group by treatment with
HCl gas afforded the desired products 12a–c and 15a, b, e
as hydrochloride salts. Their reductive amination with ace-
tone afforded N-isopropyl isonipecotanilides, which were then
converted into the respective HCl salts 13a–c and 16a, b,
e, whereas their reaction with 1,3-bis(tertbutoxycarbonyl)-2-
methyl-2-thiopseudourea and HgCl2 in dry DMF, followed by BOC
deprotection, yielded the HCl salts of the guanidine derivatives 14b
and 17a–e.
The ortho-substituted isonipecotanilides 20–22 were synthe-
sized as shown in Scheme 3. Mitsunobu reaction (Mitsunobu, 1981)
between 2-nitrophenol and (5-arylisoxazol-3-yl)methanol deriva-
tives, using diethylazodicarboxylate (DIAD) and PPh3, yielded
nitrophenylether derivatives 18b, e, which were then reduced
to 19b, e. Coupling them with N-BOC-isonipecotic acid, fol-
lowed by removal of the BOC-protecting group, yielded the target
compounds 20b, e as HCl salts, which were then converted
to compounds 21b, e and 22b, e, using the afore-mentioned
procedures.
We finally investigated the ortho isomers of the compounds
in Table 2, focusing the SAR exploration on the ortho-substituted
derivatives bearing 4-Cl-phenyl and 2-Cl-thiophen-5-yl groups
(Table 3).
A substantial improvement in the fXa inhibition potency of
the ortho-substituted isonipecotanilides over the corresponding
meta isomers was achieved with the N1-isopropyl derivatives.
Notably, compound 21e showed excellent inhibition potency in
the subnanomolar range (Ki = 0.3 nM), anticoagulant activity in
the low micromolar range (PT2 = 3.3 ␮M), and noteworthy selec-

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
189
6
5
4
3
3.3. Molecular modeling
Molecular modeling studies with the GOLD software (Jones
et al., 1997), successfully used by some of us in ligand- and
structure-based design of serine proteases’ inhibitors (Nicolotti
et al., 2008, 2009, 2010), were performed by docking the inhibitor
molecules into the active site of factor Xa. The X-ray crystal
structure of human fXa in complex with a highly potent 1H-indole-
2-carboxamide-based inhibitor (PDB code: 2BOH) (Nazaré et al.,
2004, 2005) was retrieved and used as target structure after remov-
ing the ligand.
Docking calculations were performed with eight highly potent
inhibitors (Ki < 100 nM), namely 14b (Table 1), 16e, 17b, 17e
(Table 2), 21b, 21e, 22b and 22e (Table 3), modeled in their pro-
tonated form. It is worth saying that preliminary GOLD docking
calculations carried out on our previous reference compound 5
matched the posing obtained by using QXP elsewhere (de Candia
et al., 2009b) as schematically shown in Fig. 2. In Fig. 4 the GOLD
top scored docking poses of 21e (A and B), 16e (C) and 22e (D)
are shown as representative binding models. The docking simula-
tion of the N1-isopropyl derivative 21b gave highly scored docking
poses almost similar to that of 21e and 16e, whereas the guanidine-
containing inhibitors 14b, 17b, 17e and 22b resulted in binding
modes similar to that of 22e.
As can be seen in Fig. 4A, in which the top scored docking pose
of the strongest fXa binder 21e (Ki = 0.3 nM) is shown, the neutral
2-chlorothiophene group fills the S1 pocket with the chlorine atom
pointing toward the center of the Tyr228 phenyl ring, whereas the
basic piperidine N1-isopropyl residue, in the protonated form, is
buried into the S4 aryl binding pocket, where it can be involved
in efficient cation–␲ interactions with the side chains of Phe174,
Tyr99 and Trp215. As predicted by our modeling study and in good
agreement with recent findings (Salonen et al., 2009), the tertiary
ammonium head is positioned in the center of the S4 aromatic
box. The meta-substituted isomer 16e, albeit being a weaker binder
(Ki = 95 nM) adopts a binding mode (Fig. 4C) similar to that of 21e,
with the chlorothiophene in S1 and the protonated N1–iPr group
in S4 pocket. The loss of the free energy of binding of the meta
compared to the ortho isomer (ꢀꢀG = 3.4 kcal mol−1) could most
likely result from the repositioning of the anilide scaffold which,
as a consequence, does not allow the tertiary ammonium head to
achieve efficient cation–␲ interactions, which need the cationic N+
center to be positioned on the normal crossing the centroids of the
aromatic rings (Salonen et al., 2009). As our computational data
proved, the free energy costs for achieving such a binding configu-
ration are lower for the ortho-substituted derivative 21e compared
to the meta isomer 16e.
5
6
7
8
9
10
pK
i
Fig. 3. Linear relationship between in vitro anticoagulant activity (−log PT2) and
fXa inhibitory activity (−log Ki) of compounds 15–22.
tivity over thrombin and ␣-CT (selectivity ratios of 36700 and
200000, respectively), as well as over other assayed serine pro-
teases, namely the blood coagulation factor VIIa, trypsin and human
leucocyte elastase (no significant inhibition at 250 ␮M concentra-
tion). The binding affinity for fXa of the ortho isomer 21e strongly
enhanced compared with the meta isomer 16e, being the esti-
mated change in free energy of binding (ꢀꢀG) approximately
equal to −3.4 kcal mol−1 and the related increase in PT2 higher
than 20-fold. In the ortho-substituted series, both the guanidine
derivatives 22b and 22e showed lower affinity (and selectiv-
ity) for fXa than that of the corresponding N1-iPr derivatives
21b and 21e, and thus a 3–4-fold drop in the PT2 parameter.
Taking into account all the available data, the PT2 parameter
and fXa inhibition constant, both in log units, resulted in a
pretty correlation (Fig. 3) with a determination coefficient (r2)
equal to 0.780:
−log PT2 = 0.46( 0.06) pKi + 0.78( 0.46)
n = 16; r2 = 0.780; s = 0.304; F = 49.6
(2)
The lipophilicity of the investigated compounds was assessed
by hydrophilic interaction liquid chromatography (HILIC), which
had proved to be suitable for accurate measures of 1-octanol/water
partition coefficients (log PN) of the neutral form of basic com-
pounds (Bard et al., 2009), such as those of this study. Indeed,
it had been demonstrated that on Zic-pHILIC stationary phase,
In contrast, compound 22e (Ki = 2 nM), as well as all the other
guanidine derivatives 14b, 17b, 17e and 22b, in the high-scored
docking solutions binds in the opposite way, with guanidine
group interacting with the Asp189 side chain at the bottom of
S1 pocket and the biaryl moiety embedded into the S4 aromatic
box.
In summary, suitable modifications introduced into the
structure of previously reported lipophilic biphenyl-containing
isonipecotanilides (de Candia et al., 2009b) enhanced their
fXa affinity, while lowering their lipophilicity by about 2 log P
units. In particular, some N1-isopropyl ortho-substituted
isonipecotanilides, bearing privileged oral bioavailability-
mediating 5-(chloroaryl)isoxazol-3-yl groups (Pinto et al., 2010),
have been identified as highly potent and selective fXa inhibitors.
Among them, the chlorothiophene-bearing compound 21e
(Ki = 0.3 nM) is the most potent and selective fXa inhibitor,
endowed with good anticoagulant activity as assessed by the PT
clotting assay (PT2 = 3.3 ␮M).
the difference between two isocratic log k values (i.e., ꢀlog k0–95
,
where k0 is the capacity factor measured with 100% pH 2 aque-
ous mobile phase and k95 is the capacity factor measured using
a 95% (v/v) ACN-pH 2 aqueous buffer mobile phase) is well cor-
related with log PN. The log P values calculated from HILIC data
of our compounds (Tables 2 and 3) suggest lipophilicity to play a
minor role in improving the anticoagulant activity. In fact, although
a pairwise comparison of biological (fXa affinity and PT2) and
log P data between the 4-Cl-phenyl-containing compounds (b) and
their less lipophilic 2-Cl-thiophen-5-yl analogs (e) would suggest
a trend (i.e., the activity improves due to the increased polarity),
no general relation was found between lipophilicity and antico-
agulant activity. Most likely, the activity increases because of the
change in the position of the chlorine atom with regard to Tyr228
in the S1 subsite (and not because of the increased polarity), as
a consequence of the replacement of the 4-Cl-phenyl with the
2-Cl-thiophen-5-yl.

190
G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
Fig. 4. GOLD highest scored docking poses of 5-(5-chlorothien-2-yl)isoxazole-containing isonipecotanilide inhibitors 21e (A and B, Ki = 0.3 nM), 16e (C, Ki = 95 nM) and 22e
(D, Ki = 2 nM) into the binding pocket of human fXa (PDB code: 2BOH). (A) Stick representation of 21e (carbon atoms in cyan) bound to fXa, with key amino acids in the S1
and S4 binding sites; hydrogen atoms are omitted for clarity and Van der Waals surfaces of the aromatic residues are shown. Representation of the docking modes of 21e
(B), 16e (C) and 22e (D) wherein the protein is rendered with the green-colored surface. The figure was drawn by PyMOL. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of the article.)
Some of the observed changes in fXa affinity and in vitro anti-
coagulant activity could also result from changes in lipophilicity,
which in this study has been measured by hydrophilic interaction
liquid chromatography (HILIC), but the log P-related parameters
show that lipophilicity plays a minor role. Molecular modeling
studies were instead helpful in understanding the SARs obtained.
As predicted by docking calculations, and in good agreement
with the available X-ray crystallographic structures of fXa/inhibitor
complexes (Nazaré et al., 2005), depending upon the substituent
appended at the piperidine nitrogen (N1), the isonipecotamide-
based fXa inhibitors examined herein can achieve two different
binding modes. The N1-isopropyl derivatives (e.g., 21e) preferably
direct the neutral 2-chlorothiophene group into the S1 pocket,
wherein the chlorine atom points toward the center of the Tyr228
phenyl ring, and the tertiary ammonium head (i.e., the proto-
nated N1-iPr residue) into the S4 box, where it undergoes efficient
cation–␲ interactions, and additional favorable C–H· · ·␲ interac-
tions by the isopropyl group as well, with the side chains of Phe174,
Tyr99 and Trp215. Despite a similar binding mode, the meta-
substituted derivative 16e is a fXa inhibitor weaker than the ortho
isomer 21e (ꢀꢀG = 3.4 kcal mol−1) due to a less efficient cation–␲
interactions. In contrast, the guanidine derivatives (e.g., 22e) adopt
an opposite binding mode, with the guanidinum group interacting
with the Asp189 carboxylate at the bottom of S1 pocket and the
biaryl moiety most likely engaged in hydrophobic interactions with
the aromatic residues of the S4 cavity. The same opposite binding
mode was predicted for the N1–H derivative 20e, which however
is a fXa binder weaker than 22e (ꢀꢀG = 1.9 kcal mol−1).
Compound 21e has been chosen as the lead from this class of
biarylmethoxy isonipecotanilide fXa inhibitors for further mod-
ifications, which will be designed exploiting the SARs and the
molecular modeling results obtained in this study.
Acknowledgment
C.A. acknowledges the financial support by the Italian Ministry
for Education Universities and Research (MIUR, Rome, Italy); PRIN
2007, grant no. 2007T9HTFB 003.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at doi:10.1016/j.ejps.2010.11.010.
References
Al-Obeidi, F., Ostrem, J.A., 1999. Factor Xa inhibitors. Expert Opin. Ther. Patents 9,
931–953.

G. Lopopolo et al. / European Journal of Pharmaceutical Sciences 42 (2011) 180–191
191
Ansell, J., 2007. Factor Xa or thrombin: is factor Xa a better target? J. Thromb.
Haemostasis 5 (Suppl. 1), 60–64.
Appel, R., 1975. Tertiary phosphane/tetrachloromethane, a versatile reagent for
chlorination, dehydration, and P–N linkage. Angew. Chem., Int. Ed. 14, 801–811.
Arepally, G.M., Ortel, T.L., 2006. Heparin-induced trombocitopenia. N. Engl. J. Med.
355, 809–817.
Bard, B., Carrupt, P.-A., Martel, S., 2009. Lipophilicity of basic drugs measured by
hydrophilic interaction chromatography. J. Med. Chem. 52, 3416–3419.
Böhm, M., Stürzebecher, J., Klebe, G., 1999. Three-dimensional quantitative
structure–activity relationship analyses using comparative molecular field
analysis and comparative molecular similarity indices analysis to elucidate
selectivity differences of inhibitors binding to trypsin, thrombin, and factor Xa.
J. Med. Chem. 42, 458–477.
the subpockets of factor Xa reveals two binding modes for inhibitors based on
a 2-carboxyindole scaffold: a study combining structure–activity relationship
and X-ray crystallography. J. Med. Chem. 48, 4511–4525.
Nicolotti, O., Miscioscia, T.F., Carotti, A., Leonetti, F., Carotti, A., 2008. An integrated
approach to ligand- and structure-based drug design: development and applica-
tion to a series of serine protease inhibitors. J. Chem. Inf. Model. 48, 1211–1226.
Nicolotti, O., Giangreco, I., Miscioscia, T.F., Carotti, A., 2009. Improving quantitative
structure–activity relationships through multiobjective optimization. J. Chem.
Inf. Model. 49, 2290–2302.
Nicolotti, O., Giangreco, I., Miscioscia, T.F., Convertino, M., Leonetti, F., Pisani, L.,
Carotti, A., 2010. Screening of benzamidine-based thrombin inhibitors via a lin-
ear interaction energy in continuum electrostatics models. J. Comput. Aided Mol.
Des. 24, 117–129.
Cheng, Y., Prusoff, W.H., 1973. Relationship between the inhibition constant (Ki) and
the concentration of inhibitor which causes 50% inhibition (IC50) of an enzymatic
reaction. Biochem. Pharmacol. 22, 3099–3108.
Nissink, J.W., Murray, C., Hartshorn, M., Verdonk, M.L., Cole, J.C., Taylor, R., 2002. A
new test set for validating predictions of protein–ligand interaction. Proteins
49, 457–471.
de Candia, M., Lopopolo, G., Altomare, C., 2009a. Novel factor Xa inhibitors: a patent
review. Expert Opin. Ther. Patents 19, 1535–1580.
de Candia, M., Liantonio, F., Carotti, A., De Cristofaro, R., Altomare, C., 2009b. Fluori-
nated benzyloxyphenyl piperidine-4-carboxamides with dual function against
thrombosis: inhibitors of factor Xa and platelet aggregation. J. Med. Chem. 52,
1018–1028.
Guillot, A., 2010, Development of experimental methods to determine high
lipophilicity of new chemical compounds. Ph.D. Thesis, Applications to
Cyclosporin A and Bioconcentration, School of pharmaceutical sciences, Uni-
versity of Geneva, University of Lausanne.
Ottaviani, G., Martel, S., Escarala, C., Nicolle, E., Carrupt, P.-A., 2008. The PAMPA
technique as a HTS tool for partition coefficients determination in different
solvent/water systems. Eur. J. Pharm. Sci. 35, 68–75.
Pinto, D.J.P., Orwat, M.J., Wang, S., Fevig, J.M., Quan, M.L., Amparo, E., Cacciola, J.,
Rossi, K.A., Alexander, R.S., Smallwood, A.M., Luettgen, J.M., Liang, L., Aungst,
B.J., Wright, M.R., Knabb, R.M., Wong, P.C., Wexler, R.R., Lam, P.Y.S., 2001. Dis-
covery of 1-[3-(aminomethyl)phenyl]-N-[3-fluoro-20-(methylsulfonyl)-[1,10-
biphenyl]-4-yl]-3-(trifluoromethyl)-1H-pyrazole-5-carboxamide (DPC423),
a
highly potent, selective, and orally bioavailable inhibitor of blood coagulation
factor Xa. J. Med. Chem., 566–578.
Haginoya, N., Kobayashi, S., Komoriya, S., Hirokawa, Y., Furugori, T., Nagahara,
T., 2004. Orally active factor Xa inhibitors: 4,5,6,7-tetrahydrothiazolo[5,4-
c]pyridine derivatives. Bioorg. Med. Chem. Lett. 14, 2935–2939.
Jones, G., Willett, P., Glen, R.C., Leach, A.R.L., Taylor, R., 1997. Development and
validation of a genetic algorithm for flexible docking. J. Mol. Biol. 267, 727–748.
Krynetskiy, E., McDonnell, P., 2007. Building individualized medicine: prevention of
adverse reactions to warfarin therapy. J. Pharmacol. Exp. Ther. 322, 427–434.
Lam, P.Y., Clark, C.G., Li, R., Pinto, D.J., Orwat, M.J., Galemmo, R.A., Fevig, J.M., Teleha,
C.A., Alexander, R.S., Smallwood, A.M., Rossi, K.A., Wright, M.R., Bai, S.A., He,
K., Luettgen, J.M., Wong, P.C., Knabb, R.M., Wexler, R.R., 2003. Structure-based
design of novel guanidine/benzamidine mimics: potent and orally bioavailable
Factor Xa inhibitors as novel anticoagulants. J. Med. Chem. 46, 4405–4418.
Leadley, R.J., 2001. Coagulation factor Xa inhibition: biological background and ratio-
nale. Curr. Top. Med. Chem. 1, 151–159.
Pinto, D.J., Orwat, M.J., Koch, S., Rossi, K.A., Alexander, R.S., Smallwood, A., Wong,
P.C., Rendina, A.R., Luettgen, J.M., Knabb, R.M., He, K., Xin, B., Wexler, R.R., Lam,
P.Y.S., 2007. Discovery of 1-(4-methoxyphenyl)-7-oxo-6-(4-(2-oxopiperidin-
1-yl)phenyl)-4,5,6,7-tetrahydro-1H-pyrazolo[3,4-c]pyridine-3-carboxamide
(apixaban, BMS-562247),
a highly potent, selective, efficacious, and orally
bioavailable inhibitor of blood coagulation factor Xa. J. Med. Chem. 50,
5339–5356.
Pinto, D.J.P., Smalheer, J.M., Cheney, D.L., Knabb, R.M., Wexler, R.R., 2010. Fac-
tor Xa inhibitors: next-generation antithrombotic agents. J. Med. Chem. 53,
6243–6274.
Quan, M.L., Lam, P.Y., Han, Q., Pinto, D.J., He, M.Y., Li, R., Ellis, C.D., Clark, C.G.,
Teleha, C.A., Sun, J.H., Alexander, R.S., Bai, S., Luettgen, J.M., Knabb, R.M.,
Wong, P.C., Wexler, R.R., 2005. Discovery of 1-(3ꢀ-aminobenzisoxazol-
5ꢀ-yl)-3-trifluoromethyl-N-[2-fluoro-4-[(2ꢀ-dimethylaminomethyl)imidazol-
1-yl]phenyl]-1H-pyrazole-5-carboxamide hydrochloride (razaxaban), a highly
potent, selective, and orally bioavailable factor Xa inhibitor. J. Med. Chem. 48,
1729–1744.
Liu, W.J., Xie, Y.X., Liang, Y., Li, J.H., 2006. Reusable and efficient Pd(OAc)2/TBAB/PEG-
400 system for Suzuki–Miyaura cross-coupling reaction under ligand-free
condition. Synthesis 5, 860–864.
Lumma Jr., W.C., Witherup, K.M., Tucker, T.J., Brady, S.F., Sisko, J.T., Naylor-Olsen,
A.M., Lewis, S.D., Lucas, B.J., Vacca, J.P., 1998. Design of novel, potent, noncovalent
inhibitors of thrombin with nonbasic P1 substructures: rapid structure–activity
studies by solid-phase synthesis. J. Med. Chem. 1998, 1011–1013.
Mackman, N., 2008. Triggers, targets and treatment for thrombosis. Nature 451,
914–918.
Maignan, S., Mikol, V., 2001. The use of 3D structural data in the design of specific
factor Xa inhibitors. Curr. Top. Med. Chem. 1, 161–174.
Maignan, S., Guilloteau, J.P., Choi-Sledeski, Y.M., Becker, M.R., Ewing, W.R., Pauls,
H.W., Spada, A.P., Mikol, V., 2003. Molecular structures of human factor Xa com-
plexed with ketopiperazine inhibitors: preference for a neutral group in the S1
pocket. J. Med. Chem. 46, 685–690.
McMartin, C., Bohacek, R.S., 1997. QXO: powerful, rapid computer algorithms for
structure-based drug design. J. Comput. Aided Mol. Des. 11, 333–344.
Miguez, J.M.A., Adrio, L.A., Sousa-Pedrares, A., Vila, J.M., Hii, K.K., 2007. A practical
and general synthesis of unsymmetrical terphenyls. J. Org. Chem. 72, 7771–
7774.
Mitsunobu, O., 1981. The Use of Diethyl Azodicarboxylate and Triphenylphosphine
in Synthesis and Transformation of Natural Products. Synthesis, 1–28.
Murcia, M., Morreale, A., Ortiz, A.R., 2006. Comparative binding energy analysis con-
sidering multiple receptors: a step toward 3D-QSAR models for multiple targets.
J. Med. Chem. 49, 6241–6253.
Remko, M., 2009. Molecular structure, lipophilicity, solubility, absorption, and polar
surface area of novel anticoagulant agents. J. Mol. Struct. Theochem. 916, 76–85.
Roehrig, S., Straub, A., Pohlmann, J., Lampe, T., Pernerstorfer, J., Schlemmer,
K.H., Reinemer, P., Perzborn, E., 2005. Discovery of the novel antithrom-
botic agent 5-chloro-N-({(5S)-2-oxo-3-[4-(3-oxomorpholin-4-yl)phenyl]-1,3-
oxazolidin-5-yl}methyl)thiphene-2-carboxamide (BAY59-7939): an oral, direct
factor Xa inhibitor. J. Med. Chem. 48, 5900–5908.
Roses, M., Bosch, E., 2002. Influence of mobile phase acid–base equilibria on the
chromatographic behavior of protolytic compounds. J. Chromatogr. A 982, 1–30.
Salonen, L.M., Bucher, C., Banner, D.W., Haap, W., Mary, J.-L., Benz, J., Kuster, O., Seiler,
P., Schweizer, W.B., Diederich, F., 2009. Cation–␲ interactions at the active site
of factor Xa: dramatic enhancement upon stepwise N-alkylation of ammonium
ions. Angew. Chem., Int. Ed. 48, 811–814.
USEPA, 2007. Estimation Program Interfaces Suites (EPI Suite). Ver-
sion 3.2. US Environmental Protection Agency, Washington, DC,
http://www.epa.gov/opptintr/exposure/docs/episuite.htm.
Van Duzer, J.H., Roland, D.M., 1994. Indolyl substituted hydroxylamine derivatives.
U.S. Pat. No. 5,350,761.
Verdonk, M.L., Cole, J.C., Hartshorn, M.J., Murray, C.W., Taylor, R.D., 2003. Improved
protein–ligand docking using GOLD. Proteins 52, 609–623.
Weitz, J.I., Linkins, L.A., 2007. Beyond heparin and warfarin: the new generation of
anticoagulants. Expert Opin. Invest. Drugs 16, 271–282.
Nazaré, M., Essrich, M., Will, D.W., Matter, H., Ritter, K., Urmann, M., Bauer, A.,
Schreuder, H., Dudda, A., Czech, J., Lorenz, M., Laux, V., Wehner, V., 2004. Factor
Xa inhibitors based on a 2-carboxyindole scaffold: SAR of neutral PI substituents.
Bioorg. Med. Chem. Lett. 14, 4191–4195.
Nazaré, M., Will, D.W., Matter, H., Schreuder, H., Ritter, K., Urmann, M., Essrich, M.,
Bauer, A., Wagner, M., Czech, J., Lorenz, M., Laux, V., Wehner, V., 2005. Probing
Wohnsland, F., Faller, B., 2001. High-throughput permeability pH profile and high-
throughput alkane/water log P with artificial membranes. J. Med. Chem. 44,
923–930.
Wong, P.C., Crain, E.J., Watson, C.A., Xin, B., 2009. Favorable therapeutic index of
the direct factor Xa inhibitors, apixaban and rivaroxaban, compared with the
thrombin inhibitor dabigatran in rabbits. J. Thromb. Haemostasis 7, 1313–1320.