The discovery of fluoropyridine-based inhibitors of the factor VIIa/TF complex - Part 2

Article abstract of DOI:10.1016/j.bmcl.2005.10.076

The activated factor VII/tissue factor complex (FVIIa/TF) is known to play a key role in the formation of blood clots. Inhibition of this complex may lead to new antithrombotic drugs. A fluoropyridine-based series of FVIIa/TF inhibitors was discovered whi

Full text of DOI:10.1016/j.bmcl.2005.10.076

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1060–1064
The discovery of fluoropyridine-based inhibitors of the factor
VIIa/TF complex—Part 2
Jeffrey T. Kohrt,* Kevin J. Filipski, Wayne L. Cody, Cuiman Cai, Danette A. Dudley,
Chad A. Van Huis, J. Adam Willardsen, Lakshmi S. Narasimhan, Erli Zhang,
Stephen T. Rapundalo, Kamlai Saiya-Cork, Robert J. Leadley and Jeremy J. Edmunds
Pfizer Global Research and Development, Michigan Laboratories, 2800 Plymouth Road, Ann Arbor, MI 48105, USA
Received 6 September 2005; revised 21 October 2005; accepted 21 October 2005
Available online 11 November 2005
Abstract—The activated factor VII/tissue factor complex (FVIIa/TF) is known to play a key role in the formation of blood clots.
Inhibition of this complex may lead to new antithrombotic drugs. A fluoropyridine-based series of FVIIa/TF inhibitors was discov-
ered which utilized a diisopropylamino group for binding in the S2 and S3 binding pockets of the active site of the enzyme complex.
In this series, an enhancement in binding affinity was observed by substitution at the 5-position of the hydroxybenzoic acid side-
chain. An X-ray crystal structure indicates that amides at this position may increase inhibitor binding affinity through interactions
with the S10/S20 pocket.
Ó 2005 Elsevier Ltd. All rights reserved.
The activated factor VII/tissue factor complex (FVIIa/
TF) has been shown to play a key role in the formation
of blood clots.1 Theoretically, inhibition of the FVIIa/
TF complex formation or inhibition of its active site
which is responsible for cleavage of factor X (FX) to
activated factor X (FXa) will disable the key extrinsic
coagulation pathway, but keep intact the intrinsic path-
way for the maintenance of normal hemostasis. This
may lead to enhanced antithrombotic drugs with de-
creased bleeding side effects.2 Several strategies exist
for disruption of the FVIIa/TF complex formation
and function,3 but work still remains incomplete for
the development of small molecule, FVIIa/TF active-site
inhibitors as anticoagulant therapies.4
Figure 1. Proposed binding mode for the fluoropyridine-based inhib-
itors in the active site of the FVIIa/TF complex.
In a previous publication, we reported on the discovery
of a fluoropyridine-based series of FVIIa/TF inhibitors
containing 5-substituted hydroxybenzoic acid side-
chains.5 Molecular modeling suggested that a substitu-
ent at the 5-position of this sidechain increased
inhibitor binding affinity by occupying the S10/S20
pocket of the enzyme complex (Fig. 1).5 This series of
inhibitors also utilized a racemic 1-methyl-3-phenyl-pro-
pylamino pharmacophore for binding in the S2 and S3
binding pockets, respectively, in the active site of the en-
zyme complex.
The presence of this racemic amine complicated the iso-
lation of analogs that also contained chiral 5-amido-
hydroxybenzoic acid sidechains due to the formation
of a mixture of diastereomeric final products that could
not be easily separated. Also, synthetic complications
Keywords: Factor VIIa inhibitor; Tissue factor; Serine protease
inhibitor.
*
Corresponding author. Fax: +1 734 622 3909; e-mail:
jeffrey.kohrt@pfizer.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.10.076

J. T. Kohrt et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1060–1064
1061
resulted from the need to protect this primary amine
with a para-methoxybenzyl-protecting group. This pro-
tecting group was necessary so that a more facile route
to new TF/VIIa inhibitors with elaborated hydroxy-
benzoic acid sidechains could be investigated via palladi-
um-catalyzed cross coupling reactions.5
ronic acids.8 Alternatively, when the desired boronic
acid was not available, Stille coupling conditions using
aryl or heteroarylstannanes were utilized.9 The benzyl
sidechain was formed via a Negishi coupling procedure
employing benzylzinc chloride.10 Finally, benzamide
formation relied upon the palladium-catalyzed carbox-
ylation of 6 in the presence of carbon dioxide11 to form
a 5-carboxylic acid intermediate, which was subjected to
a HATU-mediated6 coupling reaction with primary
amines. Hydrogenation of the oxadiazole intermediate
over Raney nickel under acidic conditions revealed the
amidine.6 Saponification of the methyl ester with lithium
hydroxide provided analogs 7–39. All inhibitors were
isolated by reverse-phase preparative HPLC as their
trifluoroacetate salts.
Herein, we report the discovery of a series of FVIIa/TF
inhibitors based on a fluoropyridine template that utiliz-
es diisopropylamine as a pharmacophore capable of
binding in both the S2 and S3 pockets of the FVIIa/
TF enzyme complex. This modification extended our
ability to identify potent FVIIa/TF inhibitors via the
introduction of a wider range of 5-substituted hydrox-
benzoic acid sidechains that may bind in the S10/S20
pocket of the active site of the FVIIa/TF complex.
In a search for new fluropyridine-based FVIIa/TF inhib-
itors containing non-chiral amines as a P2/P3 pharma-
cophore, we identified 7 (IC50 = 240 nM) as a potential
lead (Table 1). This new inhibitor contained diisopropyl-
amine as an alternative for the 1-methyl-3-phenyl-pro-
pylamino group that was shown in a previous series to
fill the S2 and S3 pockets of the enzyme active site.5
Although this initial lead was a weak FVIIa/TF inhibi-
tor, it was anticipated that the binding affinity of this
series could be enhanced by incorporation of a substitu-
ent at the 5-position of the hydroxybenozic acid
sidechain. From previous experience, this type of modi-
fication led to better binding of the benzoic acid side-
chain in the S10/S20 pocket of the enzyme (Fig. 1) and
provided inhibitors with enhanced inhibition of the
FVIIa/TF complex.5
The formation of the aryl ether intermediate 4 utilized a
two-step procedure where diisopropylamine was first
added to pentafluoropyridine 2 in the presence of
triethylamine followed in a separate step by the addition
of 3-(5-methyl-[1,2,4]oxadiazol-3-yl)-phenol 36 with
potassium carbonate (Scheme 1). Subsequent addition
of 5-iodo-methyl salicylate 57 and cesium carbonate pro-
duced the penta-substituted pyridine 6. This was the key
substrate for the introduction of further functionality
onto the 5-position of the hydroxybenzoic acid
sidechain.
The formation of the 5-aryl or heteroaryl analogs relied
upon a Suzuki coupling of 6 and aryl or heteroarylbo-
Replacing the 5-methoxy substituent of 7 with dimeth-
ylamino 8 (IC50 = 130 nM) resulted in an increase in
binding affinity. When a methylene group was used as
a linker to extend the dimethylamino residue deeper into
the S10/S20 pocket, a large loss in potency was observed
for 9. A much more potent inhibitor was realized by
replacing the methyldimethylamino group of 9 with
the benzyl group of 10 (73 nM). Further increases in
binding affinity were observed with phenyl 11 (36 nM).
It was observed that binding affinity was reduced from
phenyl 11 when the distal phenyl substituent was
replaced with other heteroaryl rings as exemplified in
12–15.
From molecular modeling, it was predicted that the
addition of substituents at the 3-position of the distal
phenyl ring of 11 would increase binding affinity for this
series of inhibitors by extending residues deeper into the
S10/S20 pocket. Contrary to our model, the 3-substituted
analogs 16–18 displayed less potency with respect to 11.
Binding affinity was further decreased with analogs such
as 2-toluyl 19 and 2-naphthyl 20.
Scheme 1. Reagents and conditions: (a) diisopropylamine, Et3N,
CH2Cl2, rt; (b) 3, K2CO3, DMSO, 70 °C; (c) 5, Cs2CO3, DMSO,
80 °C; (d) R1 = amide; i—CO2, Pd(OAc)2, K2CO3, DMF/H2O (4:1),
110 °C; ii—amine, HATU6, Et3N, DMF; (e) R1 = aryl: ArB(OH)2,
K3PO4, Pd(PPh3)4, DMF, 110 °C or ArSn(Bu)3, PdCl2(PPh3)4, CuI,
THF, 70 °C or R1 = benzyl: PhCH2ZnCl, Pd(PPh3)4, THF, 60 °C; (f)
H2, Raney nickel, MeOH, acetic acid, water; (g) LiOH, MeCN, H2O,
rt.
In order to enhance the binding affinity of this series of
inhibitors, analogs with amide substitution at the 5-po-
sition of the hydroxybenzoic acid sidechain were investi-
gated (Table 2).5,6 The benzyl-amide 21 (IC50 = 47 nM)
showed a slight decrease in binding affinity in compari-
son to the bi-phenyl 11. Substituted benzyl-amides such
as 3-methoxybenzyl-amide 24 (17 nM) and 4-chloroben-

1062
J. T. Kohrt et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1060–1064
Table 1. FVIIa/TF inhibitors with 5-substituted benzoic acid sidechains
12
FVIIa/TF IC50 (lM)
13
FIIa IC50 (lM)
14
FXa IC50 (lM)a
FVIIa/TF inhibitors
R1
7
8
OMe
NMe2
0.24
0.13
2.4
0.33
0.34
ND
3.1
62%
54%
ND
9
CH2NMe2
CH2Ph
Ph
10
11
12
13
14
15
16
17
18
19
20
0.073
0.036
0.067
0.097
0.19
0.29
0.048
0.12
0.13
0.21
0.57
33%
78%
47%
83%
ND
0.50
0.17
0.81
ND
ND
3.1
3-Thienyl
2-Pyridyl
4-Pyridyl
2-Pyrimidyl
Ph (3-CH3)
Ph (3-CH2OH)
Ph (3-iPr)
Ph (2-CH3)
2-Naphthyl
ND
38%
44%
39%b
47%
54%
1.1
18
1.4
74%a
ND, not determined.
a All percent inhibitions are at 66 lM inhibitor concentration unless otherwise indicated.
b Percent inhibition at 200 lM inhibitor concentration.
Table 2. FVIIa/TF inhibitors with 5-amido-benzoic acid sidechains
FVIIa/TF inhibitors
R1
FVIIa/TF IC50 (lM)
FIIa IC50 (lM)
FXa IC50 (lM)a
21
22
23
24
25
26
27
28
29
30
31
CONHCH2Ph
0.047
0.15
0.18
0.67
ND
ND
69
62%
ND
ND
62%
66%
49%
ND
ND
ND
52%
71%
CONHCH2 (2-thienyl)
CONHCH2 (2-furyl)
CONHCH2Ph (3-OMe)
CONHCH2Ph (2,4-diMe)
CONHCH2Ph (2,6-diMe)
CONHCH2Ph (2-Cl)
CONHCH2Ph (2-OMe)
CONHCH2CH2Ph (3-OMe)
CONHCH2Ph (4-Cl)
CONHi-Butyl
0.017
0.027
0.031
0.18
2.5
1.8
ND
ND
ND
0.97
1.2
0.59
0.38
0.016
0.020
32
0.016
2.1
46%c
CONH
33
34
35
36
37
38
39
CONHPropyl
CONHPentyl
0.092
0.11
0.60
1.6
1.3
1.6
ND
3.7
2.7
4.1
1.5
31%b
58%
ND
ND
97%
56%
2.6
CONHCH2Cyclohexyl
CONHCH2 (2-THP)
CONHCyclopentyl
CONHCyclohexyl
CONHCyclopropyl
0.043
0.063
0.17
ND, not determined. THP, tetrahydropyran.
a All percent inhibitions are at 200 lM inhibitor concentration unless indicated.
b Percent inhibition at 66 lM inhibitor concentration.
c Racemic compound.
zyl-amide 30 (16 nM) showed a 2.5-fold increase in
potency from 21. Disubstituted benzyl-amides like 2,4-
dimethylbenzyl-amide 25 and 2,6-dimethylbenzyl-amide
26 showed smaller increases in potency from 21. Replac-
ing the benzyl group with either a methylene-2-thienyl
22 or methylene-2-furyl 23 showed significant losses in
binding potency, while extending the alkyl chain to the
phenethyl-amide 29 also decreased potency significantly.
alkyl-amides 37–39. The most potent inhibitors in this
non-aromatic series of amides were identified as
branched alkyl-amides. For example, both isobutyl-
amide 31 (20 nM) and 1,2,2-trimethyl-propyl-amide 32
(16 nM) were more potent than the straight-chained
alkyl-amides 33 (92 nM) and 34 (110 nM).
In order to determine whether the 5-amido group of our
series was actually occupying the proposed S10/S20 bind-
ing pocket, an X-ray crystal structure of 31 in the active
site of the enzyme complex was obtained (Fig. 2).15
Analysis of these data showed that the phenylamidine
was hydrogen bonding to Asp189 and Gly219 in the
S1 pocket of the FVIIa/TF active site. The benzoic acid
was interacting through hydrogen bonding with Gly193
and His57 in the oxy-anion hole.16 The diisopropylami-
On examination of the SAR of inhibitors containing 5-
alkyl-amide sidechains, it was observed that replacing
the benzyl group of 21 with cycloalkylmethyl groups,
such as cyclohexylmethyl-amide 35 and tetrahydropyra-
nylmethyl-amide 36, resulted in a large loss of binding
affinity. An increase in potency was recovered with the
deletion of the methylene linker as exemplified by cyclo-

J. T. Kohrt et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1060–1064
1063
Figure 4. Potential binding mode of 12 in the active site of thrombin.
Figure 2. Binding interactions of 31 in the FVIIa/TF active site.
in the S1 pocket forming hydrogen bonds with
Asp189, while the 5-substituent of the hydroxybenzoic
acid sidechain now filled the distal S3 hydrophobic
pocket and the diisopropylamine occupied the S10 pock-
et (Fig. 4).17
no sidechain was shown to be filling the S2 and S3 bind-
ing pockets through hydrophobic interactions. Finally,
the isobutyl-amide sidechain filled the S10/S20 pocket.
A hydrogen bond between the amide carbonyl and the
nitrogen atom of the Gln143 allowed the alkyl group
to adopt a confirmation that allowed for maximum
van der Waals interactions with this solvent-exposed,
shallow pocket (Fig. 3). Others have reported inhibitors
with similar alkyl-amide sidechains which may also bind
in this S10/S20 pocket.4b,5,6
A fluoropyridine series of FVIIa/TF inhibitors has been
discovered which utilized a diisopropylamino sidechain
for binding in the S2 and S3 binding pockets of the ac-
tive site of the enzyme complex. An increase in binding
affinity was obtained by the incorporation of substitu-
tions at the 5-position of the hydroxybenzoic acid side-
chain. An X-ray crystal structure indicated that a 5-
amido substituted analog increased binding affinity
through hydrophobic interactions with the S10/S20 pock-
et of the enzyme complex. Work is ongoing to enhance
the binding and pharmacological activity of this series
and will be reported in due course.
Screening these TF/FVIIa inhibitors for selectivity
against FXa and thrombin (FIIa) showed that in all
cases, the selectively for TF/FVIIa was maintained over
FXa. Most analogs were also selective for TF/FVIIa
over thrombin with 7, 8, and 12 being notable excep-
tions. Molecular modeling experiments with 12 in the
active site of thrombin indicated that the moderate bind-
ing activity could potentially be due to an alternate
binding mode from that of TF/FVIIa. In the thrombin
active site, the phenylamidine of 12 continued to bind
References and notes
1. (a) Petersen, L. C.; Valentin, S.; Hedner, U. Thromb. Res.
1995, 79, 1; (b) Kirchhofer, D.; Nemerson, Y. Curr. Opin.
Biotechnol. 1996, 7, 386; (c) Bajaj, S. P.; Joist, J. H. Semin.
Thromb. Haemostasis 1999, 25, 407.
2. (a) Rapaport, S. I.; Rao, V. M. Thromb. Haemost. 1995,
74, 7; (b) Rao, L. V. M.; Exban, M. Blood Coag.
Fibrinolysis 2000, 11, 135; (c) Morrissey, J. H. Haemostasis
1996, 26, 66; (d) Himber, J.; Refina, C. J.; Burcklen, L.;
Roux, S.; Kirchhofer, D. Thromb. Haemost. 2001, 85, 475;
(e) Suleymanov, O. D.; Szalony, J. A.; Salyers, A. K.;
Lachance, R. M.; Parlow, J. J.; South, M. S.; Wood, R. S.;
Nicholson, N. S. J. Pharmacol. Exp. Ther. 2003, 306, 1115;
(f) Szalony, J. A.; Suleymanov, O. D.; Salyers, A. K.;
Panzer-Knodle, S. G.; Blom, J. D.; LaChance, R. M.;
Case, B. L.; Parlow, J. J.; South, M. S.; Wood, R. S.;
Nicholson, N. S. Thromb. Res. 2004, 112, 167.
3. (a) Girard, T. J.; Nicholson, N. S. Curr. Opin. Pharmacol.
2001, 1, 159; (b) Gallagher, K. P.; Mertz, T. E.; Chi, L.;
Rubin, J. R.; Uprichard, A. C. G. Handbook Exp.
Pharmacol. 1999, 132, 421; (c) Key, N. S.; Bach, R. R.
Thromb. Hemostasis 2001, 85, 375.
4. For a general review see: (a) Lazarus, R. A.; Olivero, A.
G.; Eigenbrot, C.; Kirchhofer, D. Curr. Med. Chem. 2004,
11, 2275; (b) Schweitzer, B. A.; Neumann, W. L.;
Rahman, H. K.; Kusturin, C. L.; Sample, K. R.; Poda,
Figure 3. Alignment of 31 in the binding pockets of the FVIIa/TF
active site.

1064
J. T. Kohrt et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1060–1064
G. E.; Kurumbail, R. G.; Stevens, A. M.; Stegeman, R. A.;
9. Li, J.-J.; Yue, W. S. Tetrahedron Lett. 1999, 40, 4507.
10. Zhu, L.; Wehmeyer, R. M.; Rieke, R. D. J. Org. Chem.
1991, 56, 1445.
11. Bumagin, N. A.; Nikitin, K. V.; Beletskaya, I. P. J.
Organomet. Chem. 1988, 358, 563.
Stallings, W. C.; South, M. S. Bioorg. Med. Chem. Lett.
2005, 15, 3006; (c) Buckman, B. O.; Chou, Y.-L.;
McCarrick, M.; Liang, A.; Lentz, D.; Mohan, R.; Mor-
rissey, M. M.; Shaw, K. J.; Trinh, L.; Light, D. R. Bioorg.
Med. Chem. Lett. 2005, 15, 2249; (d) Sagi, K.; Fujita, K.;
Sugiki, M.; Takahashi, M.; Takehana, S.; Tashiro, K.;
Kayahara, T.; Yamanashi, M.; Fukuda, Y.; Oono, S.;
Okajima, A.; Iwata, S.; Shoji, M.; Sakurai, K. Bioorg.
Med. Chem. Lett. 2005, 15, 1487; (e) Groebke Zbinden,
K.; Banner, D. W.; Ackermann, J.; DÕArcy, A.; Kirchho-
fer, D.; Ji, T.-H.; Tschopp, T. B.; Wallbaum, S.; Wever, L.
Bioorg. Med. Chem. Lett. 2005, 15, 817; (f) Klingler, O.;
Matter, H.; Schudok, M.; Donghi, M.; Czech, J.; Lorenz,
M.; Nestler, H. P.; Szillat, H.; Schreuder, H. Bioorg. Med.
Chem. Lett. 2004, 14, 3715; (g) Parlow, J. J.; Kurumbail,
R. G.; Stegeman, R. A.; Stevens, A. M.; Stallings, W. C.;
South, M. S. J. Med. Chem. 2003, 46, 4696; (h) Olivero, A.
G.; Eigenbrot, C.; Goldsmith, R.; Robarge, K.; Artis, D.
R.; Flygare, J.; Rawson, T.; Sutherlin, D. P.; Kadkhoda-
yan, S.; Beresini, M.; Elliott, L. O.; DeGuzman, G. G.;
Banner, D. W.; Ultsch, M.; Marzec, U.; Hanson, S. R.;
Refino, C.; Bunting, S.; Kirchhofer, D. J. Biol. Chem.
2005, 280, 9160.
12. The ability of compounds to act as inhibitors of the
catalytic activity of the human FVIIa/TF complex was
assessed by determining the concentration of a test
substance that inhibited by 50% (IC50) the ability of a
complex of human recombinant FVIIa/TF to cleave the
fluorogenic substrate SN-17c. SN-17c (California Peptide)
is a short peptide (13 mer) linked to an ANS (6-amino-1-
naphthalene-(n-butyl) sulfonamide) fluorophore. Typical-
ly, the reaction consisted of a mixture of 5 nM FVIIa
(NovoNordisk), 25 nM human recombinant TF (Berlex
Biosciences), and 200 lM SN-17c in a buffer of 20 mM
Hepes, 20 mM CaCl2 and 0.1% BSA in a total assay
volume of 45 ll at pH 7.5. The rate of change in relative
fluorescent units (RFUs) was determined at 37 °C by
measuring the fluorescence change over 15 min. using a
SpectraMax Gemini spectrophotometer at an excitation
wavelength of 455 nm and read at 450 nm.
13. Chi, L.; Peng, Y.-W.; Gibson, G.; Hicks, G.; Mertz, T. E.;
Rapundalo, S.; Janiczek, N.; Edmunds, J. J.; Leadley, R.
J. Cardiovasc. Pharmacol. 2004, 44, 493.
14. Chi, L.; Gibson, G.; Peng, Y.-W.; Potoczak, R.;
Bousley, R.; Brammer, D.; Riley, J.; Kreuser, S.;
Rekhter, M.; Chen, J.; Leadley, R. J. Thromb. Haemost.
2004, 2, 85.
5. Kohrt, J. T.; Filipski, K. J.; Cody, W. L.; Cai, C.; Dudley, D.
A.; Van Huis, C. A.; Willardsen, J. A.; Rapundalo, S. T.;
Saiya-Cork, K.; Leadley, R. J.; Narasimhan, L. S.; Zhang,
E.; Whitlow, M.; Adler, M.; McLean, K.; Chou, Y.-L.;
McKnight, C.; Arnaiz, D. O.; Shaw, K. J.; Light, D. R.;
Edmunds, J. J. Biorg. Med. Chem. Lett. 2005, 15, 4752.
6. (a) Kohrt, J. T.; Filipski, K. J.; Rapundalo, S. T.; Cody,
W. L.; Edmunds, J. J. Tetrahedron Lett. 2000, 41, 6041; (b)
Menzler, S.; Bikker, J. A.; Horwell, D. C. Tetrahedron
Lett. 1998, 39, 7619.
15. X-ray crystallographic data was deposited in the RCSB
database (www.rcsb.org) as PDB ID code 2c4f.
16. Eigenbrot, C. Curr. Protein Peptide Sci. 2002, 287.
17. (a) Srivastava, S.; Goswami, L. N.; Dikshit, D. K. Med.
Res. Rev. 2005, 25, 66; (b) Bode, W.; Turk, D.; Karshikov,
A. Protein Sci. 1992, 1, 426; (c) Matthews, J. H.;
Krishnan, R.; Costanzo, M. J.; Maryanoff, B. E.; Tulin-
sky, A. Biophys. J. 1996, 71, 2830.
7. Kometani, T.; Watt, D. S.; Ji, T.; Fitz, T. J. Org. Chem.
1985, 50, 5384.
8. Watanabe, T.; Miyaura, N.; Suzuki, A. Synlett 1992, 208.