Novel factor Xa inhibitors based on a benzoic acid scaffold and incorporating a neutral P1 ligand

Article abstract of DOI:10.1016/j.bmcl.2004.03.059

A series of novel, highly potent, achiral factor Xa inhibitors based on a benzoic acid scaffold and containing a chlorophenethyl moiety directed towards the protease S1 pocket is described. A number of structural features, such as the requirements of the P1, P4 and ester-binding pocket ligands were explored with respect to inhibition of factor Xa. Compound 46 was found to be the most potent compound in a series of antithrombotic secondary assays.

Full text of DOI:10.1016/j.bmcl.2004.03.059

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2801–2805
Novel factor Xa inhibitors based on a benzoic acid scaffold and
incorporating a neutral P1 ligand
a
Marc Nazare, Hans Matter, Otmar Klingler, Fahad Al-Obeidi, Herman Schreuder,
a
b
a
ꢀ a
Gerhard Zoller, Jorg Czech, Martin Lorenz, Angela Dudda, Anusch Peyman,
a
a
a
a
a
€
Hans Peter Nestler,a Matthias Urmann,a Armin Bauer,a Volker Laux,a Volkmar Wehnera
and David W. Willa,*
aAventis Pharma Deutschland GmbH, D-65926 Frankfurt, Germany
bAventis Combinatorial Technologies Center, 1580 East Hanley Blvd., Tucson, AZ 85737-9525, USA
Received 3 February 2004; revised 15 March 2004; accepted 19 March 2004
Abstract—A series of novel, highly potent, achiral factor Xa inhibitors based on a benzoic acid scaffold and containing a chloro-
phenethyl moiety directed towards the protease S1 pocket is described. A number of structural features, such as the requirements of
the P1, P4 and ester-binding pocket ligands were explored with respect to inhibition of factor Xa. Compound 46 was found to be the
most potent compound in a series of antithrombotic secondary assays.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
neither a benzamidine nor a guanidine moiety, and
incorporate a neutral chlorophenyl P1 ligand.
Currently available drug treatment for thrombosis-
related diseases is less than satisfactory and fraught with
difficulties such as severe side effects requiring moni-
toring of drug levels, slow onset of action, mode of
administration and potentially life-threatening bleed-
ing.1 The trypsin-like serine protease factor Xa is an
attractive target for antithrombotic therapy as it occu-
pies a unique position at the convergence of the intrinsic
and extrinsic pathways,2 and thus affects coagulation
without directly affecting platelet function. A huge
amount of effort has been invested in the search for an
orally available factor Xa inhibitor, which should
overcome many of the disadvantages of current thera-
pies. Although many highly potent and selective inhib-
itors of factor Xa have been discovered, their major
shortcoming has often been the presence of highly basic
benzamidine or guanidine moieties used as arginine
mimetics, and designed to bind in the P1 position. These
moieties generally result in extremely poor oral bio-
availability of the inhibitors in which they are incorpo-
rated.3 Here we report the synthesis of novel, potent and
highly selective factor Xa inhibitors, which contain
Synthesis of a directed library targeting factor Xa led to
the discovery of compound 1, with a Ki of 153 nM4 (Fig.
1). This compound was initially expected to bind with the
guanidino group in the S1 pocket, and the dichloro-
phenethyl group in the S4 pocket. However the X-ray
crystal structure of factor Xa crystals with this ligand
showed clearly that 1 bound in the active site of factor Xa
with the guanidino group in the S4 pocket, the methoxy
group in the ‘ester-binding pocket’ (EBP)5, which is a
binding site subpocket adjacent to S1, and the dichlo-
rophenethyl group in the S1 pocket, whereby the para
chloro substituent displaced a water molecule normally
bound in the base of the S1 pocket. (Results will be
published elsewhere.) This ‘chloro-binding mode’ has
also been discovered independently by other researchers.6
Synthesis of analogues of 1, aimed at replacing the
arginine amide by simpler, smaller and more rigid moi-
eties gave compound 2, with a Ki of 189 nM. This com-
pound was the starting point in the current investigation.7
2. Non-guanidino/amidino P4 ligands
Keywords: Factor Xa inhibitor.
* Corresponding author. Tel.: +49-69-305-5036; fax: +49-69-331-399;
e-mail: david.will@aventis.com
Scheme 1 outlines the synthesis of a small library
of analogues of 2. Compound 4 was synthesized in
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.059

2802
M. Nazarꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2801–2805
Phe174
O
CONEt2
S4
H
N
EBP
NH
O
N
H
Glu97
Gln192
NH2
Tyr99
Trp215
N
H
MeO
Cl
Gly216
N
NH
NH2
O
Ser195
MeO
Cl
O
2
1
Cl
Tyr228
S1
Cl
Asp189
Figure 1. Schematic diagram showing relative positions of S1, S4 and ester-binding pockets and the binding mode of this class of factor Xa
inhibitors.
O
O
O
O
OMe
Cl
OH
Cl
R1
Cl
OMe
(a)
(b)
Cl
(c)
Cl
MeO
MeO
MeO
MeO
O
O
O
OH
5a-h
3
4
Cl
Scheme 1. Reagents and conditions: (a) 2-(2,4-dichlorophenyl)-ethanol, DEAD, PPh3–polystyrene, THF/rt, 16 h; (b) (i) NaOH (aq), dioxan/60 °C,
1 h, (ii) HCl/water, precipitation; (c) amine, TOTU, N-ethylmorpholine, DMF.
multi-gram amounts starting from commercially avail-
able 3-hydroxy-4-methoxy-benzoic acid methyl ester.
Analogous procedures were used to synthesize all other
compounds described herein, typically on a 50 mg scale.
The Mitsunobu alkylation gave yields typically of
between 40% and 95%. Saponification proceeded in near
quantitative yield. The final amide coupling gave yields
of between 20% and 90%, depending on the combina-
tion of building blocks. Using an in-house synthesis
robot a library of 330 amides of structure 5 was syn-
thesized, using amines which were pre-selected for fit
into the S4 pocket. The eight most active compounds,
with Ki values less than 1 lM are shown in Table 1.
the 2-substituent of the S1 binder. The results are shown
in Table 2. The compounds were synthesized analo-
gously to compound 5a using different alcohols for the
Mitsunobu alkylation of the phenol of the central scaffold.
Compounds 5a–13 were also tested for their inhibition
of thrombin4 and in all cases had a Ki of >2.5 lM.
Clearly the 2-chloro substituent is important for potency
in the case where the EBP ligand is methoxy. Removal
of the chlorine substituent results in a 2-fold drop in
potency (6) and introduction of a pyridyl nitrogen at the
2-position (7) results in a 3-fold drop in potency. This
trend is essentially reversed in the case where the EBP
ligand is chlorine (compounds 8–10). Here the removal
of the 2-chloro substituent of the P1 ligand results in a
2–3-fold increase in potency.
Compound 5a was by far the most active factor Xa
inhibitor in this series, and an order of magnitude more
active than lead compound 2. With the exception of
compound 5e, the other active compounds found (5b–h)
were all minimally modified analogues of 5a, and all
considerably less active. Compound 5e was found to be
strongly metabolized by N-oxide formation on incuba-
tion with S9 liver fraction, and was not considered for
further investigation. Compound 5a was selected as the
starting point for the optimization of the P1 and EBP
binding ligands.
A slight improvement in potency is also observed when
the EBP ligand is fluorine. These findings confirm that
there is indeed interaction of the EBP ligand and the
substituent at the 2-position of the chlorophenethyl P1
ligand. When the EBP ligand is a halogen then the
2-chloro P1 substituent, which is critical for potency in
compound 5a, is unfavourable.
We next turned our attention to replacement of the
4-chloro substituent of the P1 ligand. This moiety is
critical for driving the binding of these inhibitors to
factor Xa as it displaces a bound water molecule at the
base of the S1 pocket. The results in Table 2 showed that
for monosubstituted P1 ligands more potent factor Xa
inhibitors are obtained when a halogen is selected as the
EBP ligand. We thus elected to synthesize and test
inhibitors with a halogen in the EBP combined with a
selection of monosubstituted P1 ligands. The results are
shown in Table 3. Clearly a para chloro substituent
3. Structural requirements of the P1 ligand
As a first step in our investigation of P1 ligands, we
chose to examine the role of the chlorine in the 2-posi-
tion of the dichlorophenethyl group. This substituent is
not involved in the displacement of water.
We simultaneously modified the EBP ligand since the
proximity of the S1 pocket and the EBP opens up the
possibility of interaction between the EBP ligand and

M. Nazarꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2801–2805
2803
Table 1. Benzoic acid-based factor Xa inhibitors with Ki values <1 lM
from directed library of non-guanidino/amidino P4 ligands4
Table 3. SAR of water-displacement moiety in P1 ligand
O
O
N
H
R1
N
X
MeO
N
O
R
O
Cl
Compound
X ¼ Cl X ¼ F
R
Ki (FXa)/nM
X ¼ Cl
X ¼ F
Cl
9
12
24
25
13
40
Compound
R1
Ki (FXa)/nM
Cl
HN
5a
18
N
N
14
15
36
71
Br
F
Me
N
HN
HN
5b
5c
57
N
N
102
203
N
N
N
110
16
17
26
27
124
314
206
307
Cl
N
HN
N
5d
5e
264
600
N
O
N
18
28
1552
4605
N
N
HN
HN
N
N
5f
650
950
970
Cl
19
20
29
30
1931
2494
712
OH
N
O
5g
5h
3348
N
HN
N
21
31
420
1554
H2N
Table 2. Simultaneous modification of the 2-position of the P1 ligand
and the EBP ligand
O
22
23
32
33
4531
>10,000
>10,000
O
N
O
N
H
N
R2
N
O
O
R3
A
>10,000
N
H
Cl
Compound
R2
R3
A
Ki (FXa)/nM
(9 and 12) is far superior to all other substituents. For
example, in compounds 14–16 bromo and fluoro para
substituents, as well as a para methyl substituent also
retain nanomolar activity, but are approximately 3-, 8-
and 9-times less potent, respectively, than compound 9.
Compounds 19 and 29, the meta chloro analogues of
compounds 9 and 12 are very much less potent than the
para chloro parent compounds. Interestingly, methoxy
substituents (17, 27, 20, 30) are considerably less potent
5a
6
OMe
OMe
OMe
Cl
Cl
H
C
C
N
C
C
N
C
C
N
18
41
54
41
13
20
50
40
37
7
––
Cl
H
8
9
Cl
Cl
10
11
12
13
––
Cl
H
F
F
F
––

2804
M. Nazarꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2801–2805
than halogen or methyl substituents. The use of methoxy
as an amidine replacement has previously been reported.8
but trifluoromethyl (38) is already too large. The series
40–45 show that a simple amino substituent (40) is more
potent than chloro. Systematically increasing the size of
the substituent by alkylation or acylation of the amino
function results in a reduction in potency (41–44), as
does reversing the amido function (45). Interestingly,
shifting the CONH2 substituent to position R2 in
compound 46 results in a 10-fold increase in potency
compared to compound 45. The series 5a, 47–49 shows
the deleterious effect of introducing larger analogues of
methoxy. Shifting the methoxy substituent to position
R2 (50) also results in a loss of potency, which can be
partially recovered by introducing a second methoxy
group at R1 (51). Compound 52 with two ethoxy sub-
stituents is considerably reduced in potency. Bridging
positions R1 and R2 through a methylene dioxy ring,
gave compound 53, which is also less potent than 5a.
4. Structural requirements of the EBP ligand
Table 4 shows results for inhibitors with variation of the
EBP ligands combined with the dichlorophenethyl P1
ligand. The series 34, 11, 8, 35 and 36, with increasing
size of the R1 substituent, indicates that of the halogen
substituents chlorine has the optimal size in this posi-
tion. Bromo and iodo substituents are too large, and
fluoro and H substituents too small. Other small sub-
stituents like methyl (37) and nitrile (39) are also potent,
Table 4. Structure–activity relationships in the ester-binding pocket
Compounds 54–57 show the effect of combining alkoxy
substituents with halogens in the EBP. Compound 57, as
well as compound 58 with a methyl group in R1 and a
methoxy group in R2, are the most potent factor Xa
inhibitors in this series.
O
R2
NH
R1
O
N
N
Cl
Cl
Compound
R1
R2
Ki (FXa)/nM
5. Secondary assays
34
11
8
H
H
106
50
F
Cl
H
The anticoagulant effect of factor Xa inhibitors was
determined in citrated human plasma spiked with
increasing concentrations of inhibitor. The standard
coagulation assays activated partial thromboplastin
time (APTT) and prothrombin time (PT) were mea-
sured, as well as the more sensitive dilute PT (dPT) with
diluted thromboplastin reagent.4 Surprisingly, com-
pound 46, which in the enzyme assay is approximately 2
and 5 times less potent than 5a and 57, respectively, is
around 2 and 5 times more potent than these com-
pounds in the more stringent and predictive dPT anti-
thrombotic assay, as well as being more potent in the
APTT and PT assays4 (Table 5). The reason for this
trend is not clear, especially since all three compounds
show similar plasma protein binding. In addition, 46 has
a predicted 100% intestinal absorption in the Caco-2
assay, as well as good stability in human S9 liver frac-
tion.
H
41
76
35
36
37
38
39
40
41
42
43
44
45
46
5a
47
48
49
50
51
52
53
54
55
56
57
58
Br
I
H
H
156
37
CH3
CF3
CN
NH2
NHCH3
H
H
140
44
H
H
29
48
H
NHCOH
H
75
NHCOCH3
NHCONH2
CONH2
H
H
316
1763
588
51
H
H
CONH2
H
CH3O
CH3S
CH3CH2O
iPrO
18
28
H
H
61
H
233
125
50
H
CH3O
CH3O
CH3CH2O
CH3O
CH3CH2O
OCH2O
Br
263
39
In conclusion, we have discovered and optimized a new
class of potent non-guanidino/amidino factor Xa
inhibitors, based on a simple achiral benzoic acid scaf-
fold. Many of the compounds are predicted to be orally
bioavailable and metabolically stable. The compounds
are capable of reducing the dilute prothrombin time, a
CH3O
CH3CH2O
Br
20
86
Br
CH3O
Cl
CH3
1053
10
CH3O
CH3O
13
Table 5. Antithrombotic, predicted absorption and metabolic stability of selected compounds
Compound
Ki (FXa)/nM
dPT/nM
APTT/lM
PT/lM
Caco-2a
P
app/nm sꢀ1
S9b/%
5a
57
46
18
10
51
237
521
111
7.30
14.35
2.86
3.66
6.95
2.29
47.8
0.51
12.4
96
86
90
a Apparent permeability coefficient.
b Percent compound remaining after 2 h incubation with human S9 liver fraction.

M. Nazarꢀe et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2801–2805
2805
Kasireddy, P.; Thorpe, D. S.; Sepetov, N.; Lebl, M.;
Wildgoose, P.; Strop, P. Biochemistry 1998, 37, 1053.
5. Maignan, S.; Guilloteau, J.-P.; Pouzieux, S.; Choi-Sled-
eski, Y. M.; Becker, M. B.; Klein, S. I.; Ewing, W. R.;
Pauls, H. W.; Spada, A. P.; Mikol, V. J. Med. Chem. 2000,
43, 3226.
recognized surrogate parameter for in vivo antithrom-
botic activity in the clinic, and are therefore promising
drug substances for the treatment of thrombotic dis-
eases.
6. (a) Adler, M.; Kochanny, M. J.; Ye, B.; Rumennik, G.;
Light, D. R.; Biancalana, S.; Whitlow, M. Biochemistry
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, 685.
7. Peyman, A; Will, D. W.; Gerlach, U.; Nazare, M.; Zoller,
G.; Nestler, H.-P.; Matter, H.; Al-Obeidi, F. PCT Patent
WO 2002046159; Chem. Abstr. 2002, 137, 20300.
8. (a) Yee, Y. Y.; Tebbe, A. L.; Lineberger, J. H.; Beight, D.
W.; Craft, T. J.; Gifford-Moore, D.; Goodson, T.; Herron,
D. K.; Klimkowski, V. J.; Kyle, J. A.; Sawyer, J. S.; Smith,
G. F.; Tinsley, J. M.; Towner, R. D.; Weir, L.; Wiley, M.
R. J. Med. Chem. 2000, 43, 873; (b) Quan, M. L.; Ellis, C.
D.; He, M. Y.; Liauw, A. Y.; Woerner, F. J.; Alexander, R.
S.; Knabb, R. M.; Lam, P. Y. S.; Luettgen, J. M.; Wong,
P. C.; Wright, M. R.; Wexler, R. R. Bioorg. Med. Chem.
Lett. 2003, 13, 369.
References and notes
1. (a) Hirsh, J. N. Eng. J. Med. 1991, 324, 156; (b) Freedman,
M. D. Am. J. Ther. 1996, 3, 771.
2. Leadley, R. J., Jr. Curr. Top. Med. Chem. 2001, 1, 151.
3. (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.
2001, 44, 566; (b) Pauls, H. W.; Ewing, W. R. Curr. Top.
Med. Chem. 2001, 1, 83.
4. 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.;