Design, synthesis, and in vitro biological activity of indole-based factor Xa inhibitors

Article abstract of DOI:10.1016/S0960-894X(00)00138-4

A series of indole and carbazole based inhibitors of factor Xa (FXa) has been investigated. The most potent compound inhibits FXa with a K(i) of 0.2 nM and has 900- and 750-fold selectivity over thrombin and trypsin, respectively. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00138-4

Bioorganic & Medicinal Chemistry Letters 10 (2000) 957±961
Design, Synthesis, and In Vitro Biological Activity of Indole-based
Factor Xa Inhibitors
Damian O. Arnaiz,* Zuchun (Spring) Zhao, Amy Liang, Lan Trinh, Marc Whitlow,
Sunil K. Koovakkat and Kenneth J. Shaw
Discovery Research, Berlex Biosciences 15049 San Pablo Ave. PO Box 4099, Richmond, CA 94804, USA
Received 13 December 1999; accepted 22 February 2000
AbstractÐA series of indole and carbazole based inhibitors of factor Xa (FXa) has been investigated. The most potent compound
inhibits FXa with a Ki of 0.2 nM and has 900- and 750-fold selectivity over thrombin and trypsin, respectively. # 2000 Elsevier
Science Ltd. All rights reserved.
Introduction
photochemical instability. We sought to replace the
cycloheptanone core of 1 with stable sca€olds that
could maintain the U-shaped conformation of this
molecule. This work ultimately led to the discovery of
ZK 807834 (CI 1031) (2), a highly potent, selective, e-
cacious and orally active inhibitor of FXa.4 Molecular
modeling studies suggested that a bis-aryl substituted
indole such as 3 could bind to FXa in a similar manner
as 1, and in this paper we describe the development of a
series of potent FXa inhibitors based on this template.
The interruption of the coagulation cascade is of pri-
mary importance in the prevention and treatment of
thrombotic disorders. The serine protease factor Xa
(FXa) links the intrinsic and extrinsic pathways of the
coagulation cascade. The prothrombinase complex,
formed by the combination of FXa with factor Va,
Ca2+ and phospholipids, catalyzes the conversion of
prothrombin to thrombin. Thrombin causes blood clot-
ting by induction of platelet aggregation and conversion
of ®brinogen to insoluble ®brin. Thrombin inhibitors
have been extensively studied as anticoagulants, but
have shown a tendency to prolong bleeding at plasma
concentrations close to the e€ective dose.1 Since FXa
inhibitors do not a€ect platelet function, they may have
a wider therapeutic index than thrombin inhibitors
have. This is supported by studies with the naturally
occurring proteinaceous FXa inhibitor tick anti-
coagulant peptide (TAP), which has been shown to be
ecacious in several animal models of thrombosis with
less bleeding when compared to other antithrombotic
agents.2
Chemistry
The indole analogues in Table 1 (7a±f) were prepared
from commercially available 6-nitroindole (4), and a
representative synthesis is illustrated in Figure 2, route
A. Arylation of the indole nitrogen with sodium hydride
and 4-¯uorobenzonitrile at 60 ꢀC followed by reduction
of the nitro group a€orded 5.5 Alkylation with 7-(bro-
momethyl)-2-naphthalenecarbonitrile6 then a€orded 6.
Finally, the nitrile groups were converted to amidines
by the Pinner7 method to a€ord indole 7f.
Early studies in our laboratory led to the identi®cation
of Z,Z-2,7-bis-(4-amidinobenzylidene) cycloheptanone
(Z,Z-BABCH, 1) as the active isomer in a series of con-
formationally rigid bis-benzamidine inhibitors (Fig. 1).3
Z,Z-BABCH inhibits human FXa with a Ki of 0.66 nM,
but has limited potential for development due to its
The carbazole analogues in Tables 2 and 3 (11a±d, 14a±b)
were synthesized from 2-hydroxycarbazole (8) in a
similar manner as the indole analogues. A representa-
tive synthesis of 11c is shown in Figure 2, route B.
Selective arylation of the carbazole oxygen with 3-
¯uorobenzonitrile followed by alkylation of the indole
nitrogen with 7-(bromomethyl)-2-naphthalenecarboni-
trile a€orded 10. Pinner reaction then a€orded 11c.
Alternatively, 8 was treated with hydroxypiperidine
under Mitsunobu8 conditions (route C) to a€ord 12,
*Corresponding author. Tel.: +1-510-669-4457; fax: +1-510-669-
4310; e-mail: damian_arnaiz@berlex.com
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00138-4

958
D. O. Arnaiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 957±961
Figure 1.
which was subsequently alkylated to a€ord 13. Amidine
formation resulted in concomitant cleavage of the Boc-
group a€ording 14a. Finally, treatment of 14a with
ethyl acetimidate a€orded 14b.
analogues (18c,d) were formed by hydrolysis of 16 fol-
lowed by a standard coupling reaction with the appro-
priate amines and subsequent amidine formation. The
corresponding carboxylate analogue 18e was prepared
by hydrolysis of 18a.
The indole analogues in Table 4 (18a±e) were synthe-
sized from 2-carbomethoxy-6-hydroxyindole9 (15) in a
manner similar to that described for 14b, and a repre-
sentative synthesis of 18a is illustrated in Figure 2, route
D. Indole 18b was prepared by a transesteri®cation
reaction by forming the imidate in ethanol. The amide
Results and Discussion
A series of bis-amidine substituted indoles was prepared
and evaluated (Table 1). The initial compounds in this
Table 1. 1,6-Disubstituted indolesa
R1
R2
Ki hFXa (nM)
Ki hFIIa (nM)
Ki bTrp (nM)
7a
7b
7c
7d
7e
CH2Ph-4-Amidine
CH2Ph-4-Amidine
CH2Ph-3-Amidine
CH2Ph-3-Amidine
C(O)Ph-3-Amidine
CH2Ph-4-Amidine
Ph-4-Amidine
Ph-4-Amidine
Ph-3-Amidine
Ph-4-Amidine
1050
840
340
190
160
3530
4530
3780
4000
2320
670
690
1620
410
410
7f
Ph-4-Amidine
30
570
300
aKi values for these competitive inhibitors are averaged from multiple determinations (nꢁ2) and the standard deviations are <30% of the mean. See
ref 4 for assay conditions.
Table 2. Carbazolesa
R1
R2
Ki hFXa (nM)
Ki hFIIa (nM)
Ki bTrp (nM)
11a
Ph-4-Amidine
60
1990
510
11b
11c
Ph-4-Amidine
Ph-3-Amidine
6.0
210
170
250
80
0.90
aKi values for these competitive inhibitors are averaged from multiple determinations (nꢁ2) and the standard deviations are <30% of the mean. See
ref 4 for assay conditions.

D. O. Arnaiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 957±961
959
Table 3. Amidine replacementa
thylamidine is a common substituent in many known
FXa inhibitors,6,10 and studies in our laboratories have
shown that 7-methyl-2-naphthylamidine is 10-fold more
potent than benzamidine against FXa.11 Consistent
with these results, introduction of a naphthylamidine
gave a 10-fold increase in FXa inhibitory activity (7f
versus 7c).
R1
Ki hFXa
(nM)
Ki hFIIa
(nM)
Ki bTrp
(nM)
In compound 7f, the R2 substituent is essentially locked,
and may not be optimally positioned for binding to
FXa. We considered that transposition of the R1 and R2
substituents might improve activity by allowing more
¯exibility of the R2 substituent while reducing the num-
ber of rotatable bonds of the R1 substituent. We initially
attempted this transposition on the aminoindole, how-
ever, since we were unable to introduce an aryl group
onto the 6-amino substituent, we considered using
commercially available 2-hydroxycarbazole as a new
sca€old. Comparison of carbazole 11a (Table 2) with
indole 7f demonstrates that the fused ring of the carba-
zole has minimal e€ect on FXa inhibitory activity.
Subsequent transposition of the two substituents resul-
ted in a 10-fold increase in activity (11b). Moving the R1
amidine from the para- to the meta-position further
increased the inhibitory activity, a€ording the ®rst sub-
nanomolar inhibitor in this series (11c). Moreover, 11c
has approximately 200-fold selectivity over thrombin
and 90-fold selectivity over trypsin.
11d
14a
CH2C(O)N(CH3)2
5.0
4.0
50
90
30
170
14b
1.6
40
40
aKi values for these competitive inhibitors are averaged from multiple
determinations (nꢁ2) and the standard deviations are <30% of the
mean. See ref 4 for assay conditions.
series (7a,b) inhibited FXa with a Ki of approximately 1
mM with a 3- to 5-fold selectivity over thrombin, and no
selectivity over trypsin. Similar to the results observed
with our diphenoxypyridine FXa inhibitors,4 changing
the R1-amidine from the para- to the meta-position gave
an increase in potency (7c versus 7b), and a further
increase in potency was observed by a similar transfor-
mation of the R2-amidine (7d). Comparison of 7c and
7e demonstrates that a 2-fold increase in potency was
achieved by introduction of an amide linker. Naph-
Having achieved subnanomolar FXa potency with 11c,
we next sought to replace one of the aryl amidines in an
Figure 2. (a) (i) NaH, 4-FPhCN, DMF, 60 ꢀC, (ii) SnCl2, EtOAc; (b) K2CO3, DMF, 7-(bromomethyl)-2-naphthalenecarbonitrile; (c) (i) HCl, (ii)
NH3; (d) (i) NaH, 3-FPhCN, DMF, 60 ꢀC; (e) NaH, DMF, 7-(bromomethyl)-2-naphthalenecarbonitrile, 25 ꢀC; (f) (i) HCl, (ii) NH3; (g) (i) N-Boc-4-
hydroxypiperidine, PPh3, DEAD, THF; (h) 7-(bromomethyl)-2-naphthalenecarbonitrile, K2CO3, DMF; (i) (i) HCl, (ii) NH3; (j) ethyl acetimidate,
Et3N, MeOH.

960
D. O. Arnaiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 957±961
Table 4. 1,2,6-Trisubstituted indolesa
R
Ki hFXa (nM)
Ki hFIIa (nM)
Ki bTrp (nM)
18a
18b
18c
18d
18e
CO2CH3
CO2CH2CH3
C(O)N(CH3)2
0.2
0.6
3.5
20
180
360
1050
2350
520
150
20
90
30
60
C(O)NHCH(CO2H)CH2CO2H
CO2H
4.6
aKi values for these competitive inhibitors are averaged from multiple determinations (nꢁ2) and the standard deviations are <30% of the mean. See
ref 4 for assay conditions.
attempt to lower the overall basicity of these inhibitors.
We assumed that the naphthylamidine moiety was
binding in the S1 pocket and decided to replace the
benzamidine moiety. The results in Table 3 illustrate the
replacement of the benzamidine with acetamide (11d)
or piperidine (14a), a€ording only a 5-fold loss in
potency relative to 11c. Investigation of a related series
of inhibitors by the Daiichi group found that the (imi-
noethyl)piperidine was one of the optimal substituents
for binding in the S4 site.6 Based on these studies we
prepared carbazole 14b which inhibited FXa with
Ki=1.6 nM.
Trp-214. The unsubstituted phenyl ring is located near
the surface of the protein and may form a hydrophobic
interaction with His-57 of the active site.
Although carbazole 14b maintains the potency of 11c, it
is much less selective for FXa over thrombin. Further-
more, 14b exhibited poor in vitro anticoagulant activity,
causing a 2-fold prolongation of the prothrombin based
clotting time (2ÂPT) at a concentration of 19 mM in
human plasma. The poor anticoagulant activity was
attributed to the limited aqueous solubility of 14b,
therefore we decided to prepare a series of indoles that
maintained a substituent at the 2-position that could
mimic the e€ect of the fused ring. The results in Table 4
show that a carbomethoxy group is the optimal sub-
stituent in this series. Compound 18a (FXa, Ki=0.2
nM) is about 10-fold more potent than carbazole 14b
and shows nearly a 1000-fold selectivity over both
thrombin and trypsin. Increasing the size of the sub-
stituent (18b±d) causes a decrease in potency. Polar
substituents also had a negative e€ect, decreasing
potency about 20-fold (18e versus 18a). While indole
18a was signi®cantly more potent against FXa than
carbazole 14b, the anticoagulant activity of this com-
pound was only marginally improved (2ÂPT=7 mM).
An X-ray crystal structure of carbazole 14b in trypsin is
shown in Figure 3.12 The inhibitor binds in an extended
L-shaped conformation analogous to that observed for
our series of diphenoxy-pyridine FXa inhibitors4,13 in
trypsin and for DX-9065a in both trypsin and FXa.14
The naphthylamidine group binds in the S1 pocket,
forming a salt bridge with Asp-189. The (iminoethyl)-
piperidine moiety binds in the aryl binding site,15 which
is a shallow groove de®ned by Leu-99, Gln-175 and
Conclusion
We have outlined a series of indole and carbazole inhi-
bitors of FXa that lack the photochemical instability
of Z,Z-BABCH (1). While initial inhibitors had only
modest potency against FXa, optimization led to a
dramatic increase in potency to a€ord compounds such
as indole 18a with subnanomolar inhibitory potency
and good selectivity over other serine proteases. The
relatively low anticoagulant activity of these inhibitors
in plasma led to the development of other more
soluble templates with improved potency and anti-
coagulant activity. This work will be the subject of
future publications.
Acknowledgements
Figure 3. 1.8 A X-ray crystal structure of 14b in trypsin re®ned to an
R-factor of 17.6%.
We thank Drs. Jerry Dallas and Baiwei Lin for help
with NMR and mass spectra, respectively. We would

D. O. Arnaiz et al. / Bioorg. Med. Chem. Lett. 10 (2000) 957±961
961
also like to thank Dr. Monica Kochanny for helpful
discussion during the preparation of this manuscript.
6. Nagahara, T.; Yokoyama, Y.; Inamura, K.; Katakura, S.;
Komoriya, S.; Yamaguchi, H.; Hara, T. J. Med. Chem. 1994,
37, 1200.
7. Neilson. The Chemistry of Amidines and Imidates; Patai, S.,
Ed.; Wiley: New York, 1975; pp 385±489.
8. Mitsunobu, O.; Yamada, M. Bull. Chem. Soc. Jpn. 1967, 40,
2380.
References and Notes
9. Boger, D. L.; Cerbonne, L. R.; Yohannes, D. J. Org. Chem.
1988, 53, 5163.
10. (a) Kucnierz, R.; Leinert, H.; Von der Saal, W.; Neidlein
R.; Kehr, C. E. P. Patent 00842941, 1998. (b) Taniuchi, Y. et
al. Thromb. Haemost. 1998, 79, 543.
1. For a review of Synthetic inhibitors of FXa and thrombin
see Hauptman, J.; Sturzebecher, J. Thromb. Res. 1999, 93, 203
and references cited therein.
2. (a) Wong, P. C.; Crain Jr., E. J.; Nguan, O.; Watson, C. A.;
Racanelli, A. Thromb. Res. 1996, 83, 117. (b) Kaiser, B.;
Hauptmann, J. Cardiovasc. Drug Rev. 1994, 12, 225.
3. Shaw, K. J.; Guilford, W. J.; Dallas, J. L.; Koovakkat, S.
K.; Liang, A.; Light, D. R.; Morrissey, M. M. J. Med. Chem.
1998, 41, 3551.
11. Unpublished data from this Laboratory.
12. Protein Data Bank 1qb9.
13. Whitlow, M.; Arnaiz, D. O.; Buckman, B. O.; Davey, D.
D.; Griedel, B.; Guilford, W. J.; Koovakkat, S. K.; Liang, A.;
Mohan, R.; Phillips, G. B.; Seto, M.; Shaw, K. J.; Xu, W.;
Zhao, Z.; Light, D. R.; Morrissey, M. M. Acta Cryst. D. 1999,
55, 1395.
14. (a) For DX-9065a in trypsin see Stubbs, M. T.; Huber, R.;
Bode, W. FEBS Lett. 1995, 375, 103. (b) For DX-9065a in
FXa see Brandstetter, H.; Kuhne, A.; Bode, W.; Huber, R.;
von Der Saal, W.; Wirthensohn, K.; Engh, R. A. J. Biol.
Chem. 1996, 271, 29988.
4. Phillips, G. B.; Buckman, B. O.; Davey, D. D.; Eagen, K.
E.; Guilford, W. J.; Hinchman, J.; Ho, E.; Koovakkat, S. K.;
Liang, A.; Light, D. R.; Mohan, R.; Ng, H. P.; Post, J. M.;
Shaw, K. J.; Smith, D.; Subramanyam, B.; Sullivan, M. E.;
Trinh, L.; Vergona, R.; Walters, J.; White, K.; Whitlow, M.;
Wu, S.; Xu, W.; Morrissey, M. M. J. Med. Chem. 1998, 41,
3557.
5. Arylation with 3-¯uorobenzonitrile required 100 ꢀC.
15. Bode, W.; Turk, D.; Karshikov, A. Protein Sci. 1992, 1, 426.