Novel factor Xa inhibitors based on a 2-carboxyindole scaffold: SAR of P4 substituents in combination with a neutral P1 ligand

Article abstract of DOI:10.1016/j.bmcl.2004.06.019

A series of novel, highly potent 2-carboxyindole-based factor Xa inhibitors is described. Structural requirements for P4 ligands in combination with a neutral biaryl P1 ligand were investigated with the 2-carboxyindole scaffold. A diverse set of P4 substituents was identified, which, in conjunction with a biaryl P1 ligand, gave highly potent factor Xa inhibitors, which were also selective versus other proteases and efficacious in various antithrombotic secondary assays.

Full text of DOI:10.1016/j.bmcl.2004.06.019

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4197–4201
Novel factor Xa inhibitors based on a 2-carboxyindole scaffold:
SAR of P4 substituents in combination with a neutral P1 ligand
*
Marc Nazar ꢀe , Melanie Essrich, David W. Will, Hans Matter, Kurt Ritter,
Matthias Urmann, Armin Bauer, Herman Schreuder, J o€ rg Czech, Martin Lorenz,
Volker Laux and Volkmar Wehner
Aventis Pharma Deutschland GmbH, D-65926 Frankfurt am Main, Germany
Received 27 April 2004; accepted 8 June 2004
Available online
Abstract—A series of novel, highly potent 2-carboxyindole-based factor Xa inhibitors is described. Structural requirements for P4
ligands in combination with a neutral biaryl P1 ligand were investigated with the 2-carboxyindole scaffold. A diverse set of P4
substituents was identified, which, in conjunction with a biaryl P1 ligand, gave highly potent factor Xa inhibitors, which were also
selective versus other proteases and efficacious in various antithrombotic secondary assays.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
amidine mimics addressing the same polar interaction
5
points at the cost of lower affinity.
The serine protease factor Xa (fXa) has emerged as a
key target for the control and prevention of thrombo-
genesis associated with diseases like myocardial infarc-
tion, stroke, deep vein thrombosis and pulmonary
The third is to seek new types of interaction, such as the
chloro binding mode, recently characterized by X-ray
cocrystallization, by incorporating a neutral chloroaryl
7
P1 ligand. In the accompanying communication we
6
1
embolism. Its unique position along the blood coagu-
lation cascade linking the intrinsic and extrinsic activa-
tion pathways combined with its role in thrombin
activation has led to the expectation that inhibitors of
fXa should be highly effective antithrombotics with
fewer side effects and difficulties requiring monitoring of
drug levels, slow onset of action, mode of administration
and potentially life-threatening bleeding encountered
described the optimization of the neutral P1 ligand of
fXa inhibitors based on a 2-carboxyindole scaffold. We
succeeded in generating a set of equipotent inhibitors
with structurally diverse neutral P1 substituents. These
inhibitors were also shown to have improved Caco-2
permeability compared to amidine/guanidine com-
pounds. We next turned our attention to the P4 ligand
of fXa inhibitors based on the 2-carboxyindole scaffold.
Interestingly, the S4 pocket of fXa is more conforma-
tionally flexible than the S1 pocket and thus P4 ligands
with the correct orientation may bind by induced fit. We
intended therefore to exploit this potential for structural
flexibility by mapping the SAR of the P4 ligand in
combination with a neutral P1 group interacting with
the rigid S1 pocket of fXa.
2
with current available treatment. Among the plethora
of highly active selective fXa inhibitors discovered many
share an amidine or guanidine moiety designed to bind
in the S1 pocket of the enzyme. In general the presence
of these highly basic groups results in poor oral
absorption and have hindered the development of an
3
orally bioavailable drug. Three main strategies have
been exploited to meet this challenge: the first is to mask
4
the amidine as a prodrug; the second is to reduce the
basicity of the amidine by replacing it with various
8
9
In addition, we and others have observed that the
potency of fXa inhibitors in more clinically-relevant
assays of antithrombotic activity, such as activated
partial thromboplastin time (APTT), prothrombin time
(PT) and dilute PT (dPT), is not solely determined by
Keyword: Factor Xa inhibitors.
*
Corresponding author. Tel.: +49-69-305-15150; fax: +49-69-331-399;
e-mail: marc.nazare@aventis.com
the K value of the inhibitor on isolated fXa enzyme, but
i
0
960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.019

4198
M. Nazar ꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4197–4201
is driven by a complex interplay of several parameters,
such as protein binding and polarity. In order to be able
to tune these parameters we sought to generate a set of
interchangeable P4 substituents to be combined with the
O
O
N
H
Route 1
a)
Route 2 (a)
Route 3
b), (c)
7
P1 ligands we reported previously.
(
(
O
O
N
2
. Chemistry
O
O
R
4
O
N
R
N
5 H
N
H
R
Scheme 1 illustrates the representative synthesis of
prototypic building blocks 1, 2, 3 serving as putative P4
substituents. The N-alkyl derivatives 1 were obtained by
subjecting commercially available N-Boc protected 4-
aminopiperidine to standard reductive amination or
alkylation procedures followed by acid catalyzed Boc
deprotection. The corresponding 4-pyridyl compound 2
was prepared by ipso substitution of 4-chloropyridine
and subsequent cleavage of the Boc-protecting group. A
similar sequence extended by a catalytic dehalogenation
led to pyrimidine 3.
(b), (c)
4
O
N
R
N
H
NBoc
(
b), (c)
(a)
6
(
d), (e)
O
N
R
N R'
H
Scheme 2 shows the synthetic routes to the inhibitors
described. Pre-assembly of the P4 ligands (1–3, and their
analogues) followed by coupling to the 2-carboxyindole
scaffold, which could be pre-derivatized (route 1) or
post-derivatized (route 2) with the biaryl P1 ligand, led
to the derivatives 7–33. Alternatively, Boc-protected 4-
aminopiperidine could be attached to the scaffold to give
compound 6. Following deprotection of the Boc group
the terminal substituent could be introduced by, for
example, reductive amination (route 3). Other P4
ligands were constructed and introduced analogously by
one of these routes.
7-33
Scheme 2. Reagents and conditions: (a) R–Br, NaH, DMF, 80 °C, 1 h;
b) LiOH, THF/H
O, 60 °C, 2 h; (c) amine, TOTU, N-ethylmorpho-
line, CH Cl , or BOP–Cl, NEt , CH Cl ; (d) HCl, methanol, rt, 2 h; (e)
aldehyde or ketone Na(CN)BH , acetonitrile, rt, 16 h or alkyl halide,
NEt , DMF, rt, 4–16 h.
(
2
2
2
3
2
2
3
3
scaffold were kept constant throughout. The biaryl P1
ligand, which binds in the S1 pocket via the chloro
binding mode driven by water displacement, was pre-
viously found to be one of the most potent in a series of
7
related ligands. Initially the 4-aminopiperidine ‘spacer’
was held constant, and simple alkyl terminal substitu-
ents were introduced. Subsequently, more polar termi-
nal groups were introduced onto the 4-aminopiperidine
moiety, as well as heterocycles, which had previously
been found to be active in combination with other
3
. Results and discussion
Figure 1 shows the strategy employed to probe the SAR
of the P4 ligand. The P1 group and the carboxyindole
6
scaffolds. Ring-contracted analogues of the 4-amino-
piperidine moiety were introduced, both as simple alkyl
derivatives and derivatized with heterocycles. Finally, a
variety of other related P4 ligands with, for example,
reversed, extended or contracted spacers were investi-
gated.
NHBoc
(a), (b)
H N
N
alkyl
1
2
A:
B:
N
H
Table 1 summarizes the results for fXa inhibitors 7–15
with P4 substituents composed of a 4-aminopiperidine
‘spacer’ and simple alkyl/cycloalkyl terminal substitu-
ents. Compounds 7–9 show that of the simple un-
branched alkyl termini, ethyl has the optimal size.
Compound 10, with an isopropyl terminus, the smallest
possible branched alkyl substituent, is the most potent
compound in this series. Increasing the size of the
hydrophobic terminus, in compounds 11 and 12, as well
as in the cycloalkyl derivatives 14 and 15 was also del-
eterious to potency. Surprisingly, cyclopropyl-substi-
tuted compound 13 is nine times less potent than the
isopropyl derivative 10, although both compounds are
essentially isosteric. This can be explained by the
reduction in basicity of the cyclopropyl-substituted
piperidine nitrogen compared to its isopropyl counter-
NHBoc
(c), (b)
H N
N
N
2
3
2
N
H
NHBoc
NH2
NHBoc
(d)
(e), (b)
C:
N
N
N
N
N
N
H
N
Cl
Scheme 1. Reagents and conditions: (a) aldehyde or ketone
Na(CN)BH , acetic acid, methanol, rt, 16 h or alkyl halide, NEt , DMF,
rt, 4–16 h; (b) HCl, methanol, rt, 16 h; (c) 4-chloropyridine, n-butanol/
3
3
water/NEt
water/NEt
3
1:1:1, 100 °C, 48 h; (d) 2,4-dichloropyrimidine, n-butanol/
1:1:1, 100 °C, 48 h; (e) H , Pd/C, ethanol, acetic acid, rt, 5 h.
3
2

M. Nazar ꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4197–4201
4199
H
H
n
N
N
Alkyl
N
N Alkyl
n
Terminal alkyl variation
Ring size variation
H
N
N Polar
P4
Terminus
Terminal polar substituents
Spacer
N
O
H
n
n
H
N
N Heterocycle
N
N Heterocycle
Cl
S
N
O
Terminal heterocyclic substituents Ring size variation
P1
Other: eg. Spacer reversal/extension/contraction
Figure 1. Schematic diagram summarising the P4 ligand modification strategy.
part. The measured pK
compared to 8.6 for compound 10, due to the vinylic
a
value for compound 13 was 6.5,
logue of compound 10, and is 1000-fold less potent. A
considerable, but less dramatic drop in activity is ob-
served on extension of the linker by one methylene unit
(17). Compounds 18 and 19, which retain the tertiary
amine of the piperidine linker, but have a hydrophobic
terminus have similar activity to those compounds in
Table 1 with larger purely hydrophobic termini. Acetyl,
methylsulfonyl and urea derivatives 20–22, which were
designed to be highly polar, but approximately iso-
morphous with the isopropyl group, are also consider-
ably less active than 10, again confirming the preference
for a basic nitrogen in the P4 ligand.
11
character of the cyclopropyl group. This clearly dem-
onstrates that it is not only hydrophobic interactions,
which are important for potency in the S4 pocket, but
also the piperidine nitrogen is involved in polar inter-
actions and plays a critical role to fill the S4 cation hole.
Compound 16 in Table 2 is the ‘reversed linker’ ana-
Table 1. 2-Carboxyindole-based factor Xa inhibitors: SAR of varia-
tion of the alkyl substituent on 4-aminopiperidine P4 ligands
R1
Ring-contraction gave aminopyrrolidine derivatives (23
and 24) of reasonable potency, whereby one enantiomer
was twice as active as the other. Further ring-contrac-
tion to give aminoazetidine derivative 25 resulted in a
somewhat more significant drop in activity. Compounds
N
O
Cl
S
N
O
26 and 27 were synthesized in an attempt to compensate
Compds
R1
i
K (fXa) nM
for the shortening of the spacer compared to the original
aminopiperidine, however they brought no improve-
ment in activity.
7
N
H
N
N
40
18
8
9
N
H
The results in Tables 1 and 2 demonstrate that a rela-
tively broad range of terminal substituents is tolerated in
the S4 pocket, with very few modifications resulting in a
dramatic loss in potency. This is indicative of the con-
formational flexibility of the S4 pocket of fXa.
N
H
N
N
N
64
3
10
11
12
13
14
15
N
H
In our previous work on fXa inhibitors based on a
8
benzoic acid scaffold we found that a 4-aminopyridine
N
H
220
142
27
terminus to the P4 ligand was optimal for that scaffold.
In spite of the results shown in Tables 1 and 2, which
would suggest that a pyridyl terminus to the P4 ligand
would be too large to be highly active, we synthesized
the series of compounds shown in Table 3. Remarkably,
compound 28, which incorporates the 4-aminopiper-
idine spacer with a 4-pyridyl terminal substituent, has a
N
H
N
N
N
H
K value of 14 nM. Replacement of the 4-aminopyridyl
i
N
H
N
N
71
motif by the corresponding pyrimidine (29) resulted in a
sevenfold loss in potency. Extension of the linker by one
methylene unit (30) also resulted in a drop in activity,
but the effect was not as pronounced as that caused by
N
H
106

4200
M. Nazar ꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4197–4201
Table 2. 2-Carboxyindole-based factor Xa inhibitors: SAR of non-
aromatic P4 ligands
Table 3. 2-Carboxyindole-based factor Xa inhibitors: SAR of aro-
matic P4 ligands
R1
R1
N
O
N
O
Cl
S
Cl
S
N
N
O
O
Compds
R1
K
i
(fXa) nM
Compds
R1
i
K (fXa) nM
16
17
18
19
20
21
22
23
24
25
26
N
N
H
2997
161
131
95
28
N
H
N
N
N
14
N
2
9
95
N
H
N
N
N
H
N
N
30
31
32
33
64
18
N
H
N
H
N
OH
O
N
N
N
H
N
N
N
H
N
N
O
2900
12
N
H
763
125
1075
103
48
O
N
H
N
N
O
N
H
N
S
O
NH2
N
H
N
O
N
H
N
Table 4. Selectivity profile for selected 2-carboxyindole-based factor
Xa inhibitors
a
Compds fXa
/lM
Thrombin Trypsin
/lM
Kallikrein t-Pa
/lM
N
H
N
K
i
K
i
K
i
/lM
K
i
i
K /lM
7
8
0.040
0.018
0.003
0.027
0.048
0.014
0.018
0.012
>10
>10
2.76
8.22
>10
>10
3.163
>10
>100
>100
>100
>100
>100
>100
>100
>100
>10
>10
>10
>10
>10
>10
8.54
>10
>100
>100
>100
>100
>100
>100
64.01
>100
N
H
N
225
345
10
1
2
2
3
3
3
4
8
1
3
N
H
N
H
N
a
Human tissue plasminogen activator.
27
839
N
Table 5. Anticoagulant activity of selected compounds
a
b
b
Compds
fXa
/lM
dPT
lM
APTT
lM
PT
lM
K
i
the same modification in the alkyl series (cf. 10 and 17).
Even more remarkably, compound 31, in which the 4-
aminopiperidine spacer is reversed, is approximately
equipotent to 28. This is in stark contrast to the dra-
matic 1000-fold drop in affinity caused by the same
modification in the alkyl series described above (cf. 10
and 16). This result also demonstrates that there is no
essential NH-interaction of the carboxyindole amide
with the enzyme. The critical role played by the 4-amino-
pyridyl motif in this series, can be seen clearly by com-
parison of compounds 31 and 32 where replacement of
the amino nitrogen by an oxygen results in a more than
7
8
9
0.040
0.018
0.064
0.003
0.027
0.048
0.014
0.018
0.012
1.554
0.377
1.563
0.426
1.799
0.625
0.232
0.198
0.379
40.65
13.79
38.18
5.16
30.69
9.02
15.58
5.41
1
1
2
0
3
4
27.53
16.48
19.58
13.72
6.04
28
31
33
8.260
11.63
4.380
5.78
6.20
a
Concentration required for 50% inhibition of dilute prothrombin
10
time (dPT).
b
Concentration required to double plasma clotting time of APTT and
PT, respectively.
1
60-fold drop in affinity. Ring-contraction of the spacer

M. Nazar ꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4197–4201
4201
to an azetidine (33) in combination with the 4-pyridyl
terminus resulted in an inhibitor equipotent to 28.
References and notes
1
2
3
. Leadley, R. J. Curr. Top. Med. Chem. 2001, 1, 151.
. Kaiser, B. Drugs Future 1998, 23, 423.
The surprising results shown in Table 3 tend to confirm
the conformational flexibility of the S4 pocket and sug-
gest that the P4 ligands may indeed bind by induced fit.
. Pauls, H. W.; Ewing, W. R. Curr. Top. Med. Chem. 2001,
, 83.
1
4
. (a) Song, Y.; Clizbe, L.; Bhakta, C.; Teng, W.; Wong,
P.; Huang, B.; Tran, K.; Sinha, U.; Park, G.; Reed,
A.; Scarborough, R. M.; Zhu, B.-Y. Bioorg. Med. Chem.
Lett. 2003, 13, 297; (b) Gustafsson, D.; Elg, M. Thromb.
Res. 2003, 109, S9.
Key compounds were further characterized against re-
lated proteases and display a high degree of selectivity
(
Table 4). This favourable selectivity pattern can be
mainly attributed to the biarylic P1 substituent employing
a chloro binding mode and the other 2-carboxyindole
structural determinants common to all inhibitors.
5. (a) 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. J. Med. Chem.
Table 5 shows the evaluation of the anticoagulant
activity in the standard coagulation assays for activated
partial thromboplastin time (APTT) and prothrombin
time (PT), as well as the more sensitive dilute PT (dPT)
2
001, 44, 566; (b) Lam, P. Y. S.; Clark, C. G.; Li, R.;
Pinto, D. J. P.; 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. J. Med.
Chem. 2003, 46, 4405.
10
with diluted thromboplastin reagent. Again this con-
firmed our previous observation that there is no clear
correlation between the K
i
value on the isolated enzyme
6. (a) Adler, M.; Kochanny, M. J.; Ye, B.; Rumennik,
G.; Light, D. R.; Biancalana, S.; Whitlow, M. Biochem-
istry 2002, 41, 15514; (b) Maignan, S.; Guilloteau, J.-P.;
Choi-Sledeski, Y. M.; Becker, M. R.; Ewing, W. R.; Pauls,
H. W.; Spada, A. P.; Mikol, V. J. Med. Chem. 2003, 46,
and the clinically-relevant antithrombotic surrogate
parameters. For example, compound 10 is no more ac-
tive than compounds 8, 28, 31, 33 in the dPT assay in
spite of being four to six times more potent on fXa.
6
85.
7
. Nazar ꢀe , 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. Bioorg. Med.
Chem. Lett. 2004, 14, preceding paper in this issue.
doi:10.1016/j.bmcl.2004.06.020.
8. Nazar ꢀe , M.; Matter, H.; Klingler, O.; Al-Obeidi,
F.; Schreuder, H.; Zoller, G.; Czech, J.; Lorenz, M.;
Dudda, A.; Peyman, A.; Nestler, H. P.; Urmann, M.;
Bauer, A.; Laux, V.; Wehner, V.; Will, D. W. Bioorg. Med.
Chem. Lett. 2004, 14, 2801.
. Jia, Z. J.; Wu, Y.; Huang, W.; Zhang, P.; Clizbe, L. A.;
Goldman, E. A.; Sinha, U.; Arfsten, A. E.; Edwards, S. T.;
Alphonso, M.; Hutchaleelaha, A.; Scarborough, R. M.;
Zhu, B.-Y. Bioorg. Med. Chem. Lett. 2004, 14, 1221.
0. In vitro assays for determination of fXa and thrombin
activity were performed as described by Ostrem, J. A.; Al-
Obeidi, F.; Safar, P.; Safarova, A.; Stringer, S. K.; Patek,
M.; Cross, M. T.; Spoonamore, J.; LoCascio, J.
C.; Kasireddy, P.; Thorpe, D. S.; Sepetov, N.; Lebl,
M.; Wildgoose, P.; Strop, P. Biochemistry 1998, 37,
In summary, we have synthesized a series of 2-carb-
oxyindole inhibitors with a neutral biarylic P1 group
and cyclic amine-derived P4 ligands and demonstrated
that they are potent and selective inhibitors of fXa. In
this context structural requirements in the S4 pocket of
the enzyme were revealed and exploited to generate a set
of structural diverse inhibitors. Several compounds were
identified, which were capable of reducing the dilute
prothrombin time, a recognized parameter for in vivo
antithrombotic activity in the clinic. Hence, a new gen-
eration of potent and selective fXa inhibitors with a
broad variation in the P4 region and with a neutral P1
group emerged, which are promising drug substances for
the therapeutic prevention and control of thrombotic
diseases.
9
1
1
Acknowledgements
1
053.
1. (a) pK
a
values were measured by UV spectroscopy on a
We would like to thank Astrid Sihorsch, Marga
K €a mmerer-Dienst, Lynn Bayer, Dirk Timme, Norbert
Laub and Stefan Engel for their excellent technical
assistance.
Sirius SGA profiler; (b) Perrin, C. L.; Fabian, M. A.;
Rivero, I. A. Tetrahedron 1999, 55, 5773; (c) Gillaspy, M.;
Lefker, B. A.; Hada, W. A.; Hoover, D. J. Tetrahedron
Lett. 1995, 36, 7399.