Ketene aminal-based lactam derivatives as a novel class of orally active FXa inhibitors

Article abstract of DOI:10.1016/j.bmcl.2005.08.107

N,N′-Disubstituted ketene aminals are good bioisosteres of thiourea functional groups. We report the design and synthesis of a novel class of ketene aminal-based lactam derivatives as potent and orally active FXa inhibitors.

Full text of DOI:10.1016/j.bmcl.2005.08.107

Bioorganic & Medicinal Chemistry Letters 15 (2005) 5453–5458
Ketene aminal-based lactam derivatives as a novel class of
orally active FXa inhibitors
Yan Shi,* Jing Zhang, Philip D. Stein, Mengxiao Shi,à Stephen P. OÕConnor,
Sharon N. Bisaha, Chi Li, Karnail S. Atwal, Gregory S. Bisacchi,§ Doree Sitkoff,
Andrew T. Pudzianowski, Eddie C. Liu, Karen S. Hartl, Steven M. Seiler, Sonia Youssef,
Thomas E. Steinbacher, William A. Schumacher, Alan R. Rendina, Jeffrey M. Bozarth,
Tara L. Peterson, Ge Zhang and Robert Zahler
Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ 08543-5400, USA
Received 27 June 2005; revised 26 August 2005; accepted 30 August 2005
Available online 5 October 2005
In memory of Dr. Steven M. Seiler.
Abstract—N,N0-Disubstituted ketene aminals are good bioisosteres of thiourea functional groups. We report the design and synthe-
sis of a novel class of ketene aminal-based lactam derivatives as potent and orally active FXa inhibitors.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Inhibition of the trypsin-like serine protease factor Xa
(FXa) has emerged as a key point of intervention in
the blood coagulation cascade for the development of
antithrombotic agents.1 Within the cascade, factor Xa
functions at the point where the intrinsic and extrinsic
coagulation pathways converge,2 and FXa is the key en-
zyme responsible for thrombin activation.
guanidine or amidine as S1 or S4 binding substrates;
their pharmacokinetic profiles are thus limited due to
low oral bioavailability and short plasma half life.
We have been interested in the development of novel,
orally bioavailable FXa inhibitors.3 As a result of
our early program effort,4 it was found that compound
1 is an inhibitor of human FXa (IC50 = 110 nM).
Although it contains a thiourea, its novel structure is
intriguing due to the absence of an active site directing
guanidine or amidine moiety in S1, and its nonpeptidic
nature. Also, compound 1 was found to be relatively
selective for FXa when tested against a panel of tryp-
sin-like serine proteases.4 Based on this lead, we em-
barked on synthetic efforts to explore alternate
bioisosteres to the thiourea motif that could adopt a
L-shaped conformation and would be more chemically
stable and amenable to a final drug candidate. Herein,
we describe a series of ketene aminal-based lactam
derivatives as a novel class of orally bioavailable FXa
inhibitors.5
Factor Xa contains a deep S1 and a box-like S4 recog-
nition site at the enzymeÕs active site. Potent FXa
inhibitors generally require both an S1 and an S4 bind-
ing element which are connected through L-shaped or
other ÔbentÕ scaffolds.1 Many factor Xa inhibitors in
the literature contain basic pharmacophores such as
Keywords: Serine protease; Factor Xa inhibitors; Antithrombotic
agents; 2-Methylbenzofuran; Thiourea; Bioisosteres; N,N0-Disubsti-
tuted ketene aminal; 2,2-Dicyanoketene aminal; 2-Nitroketene aminal;
Orally active.
*
Corresponding author. Tel.: +1 609 818 4124; fax: +1 609 818
3450; e-mail: yan.shi@bms.com
O
Present address: Hoffman-La Roche Inc., 340 Kingsland Street,
Nutley, NJ 07110, USA.
H
H
N
N
N
N
à
Present address: Wyeth Research, 401N Middletown Road, Pearl
River, NY 10965, USA.
O
S
§
Present address: AstraZeneca, 35 Gatehouse Dr., Waltham, MA
02451, USA.
1 IC50 = 110 nM
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.08.107

5454
Y. Shi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5453–5458
The N,N0-disubstituted 2-nitroketene aminal 2 is known
to be an excellent bioisostere of N,N0-disubstituted
thiourea 3 in histamine H2-receptor antagonist studies.6
For example, the nitroketene aminal in Ranitidine7 is a
bioisostere of the thiourea functional group in Metia-
mide8 (Scheme 1). Similar to its thiourea analog
1,4 (S)-3-(1-(m-toluidino)-2-nitrovinylamino)-1-(2-oxo-
2-(pyrrolidin-1-yl)ethyl))azepan-2-one (compound 6),
is predicted to prefer the anti–syn1 conformation by
gas-phase semi-empirical calculations.9 Scheme 2 shows
the four lowest energy conformers for both 6 and N-
phenyl-N0-methyl 2-nitroketene aminal 5. The other
possible conformers are disfavored due to the internal
steric interactions between nitrogen substituents and
the nitro group. It is clear that both compounds prefer
anti–syn1 conformation with the aromatic group (Ar)
anti to the nitroketene group and the alkyl group (R)
syn to the ketene. This conformation affords the best
combination of relief from steric crowding between
the nitro and the Ar or R groups, and the stabilization
by intramolecular hydrogen bonding between the nitro
group and an aminal NH. The anti–anti conformations
are disfavored because of steric interactions between
the R and Ar groups. The syn–anti1 conformation is
less favored than anti–syn1 probably due to the steric
interaction of the R group and the hydrogen atom from
Ar-N–H. This effect becomes more pronounced for 6
with a bulkier R group relative to 5. In addition, the
hydrogen bonding between the nitro group and the aryl
N–H enhances the planarity of the aryl N–H by mak-
ing it part of a 6-membered hydrogen bond ring,
whereas when the nitro is hydrogen bonded to the alkyl
N–H, the Ar-N–H nitrogen becomes more pyramidal.
Thus, the Ar-N–H groups in anti–syn1 conformers have
better conjugation with the ketene group than their
syn–anti1 conformers.
Electronically, both thiourea10 and ketene aminal11
have large dipole moments that have been attributed
to considerable resonance contributions from dipole
canonical structures, which allows the ketene aminal
to exist in a rapid rotameric equilibrium around the
ketene double bond in polar solvents.12 Thus, the ke-
tene double bond can adopt two conformations in
binding to enzymes. Electron-withdrawing substituents
on the ketene group (X and Y in 4) also behave sim-
ilarly to the thione carbonyl (S@C) by reducing the
electron density of the conjugated amino groups. As
a result, the ketene aminal serves as a neutral bio-iso-
stere to thiourea.6b,13
O2N
X
Y
S
R2
R2
R2
HN
R1
N
HN
R1
N
HN
R1
N
H
H
H
2
3
4
NO2
S
N
HN
S
N
S
N
H
N
H
N
H
N
H
O
Ranitidine
Metiamide
Scheme 1. The thioureas and ketene aminals.
syn-anti1
anti-anti1
anti-syn1
anti-anti2
O2N
O2N
NO2
H
NO2
H
H
H
R
Ar
H
H
N
N
R
N
N
H
N
H
N
R
N
N
R
Ar
Ar
Ar
5
6
2.5 kcal/mol
4.5 kcal/mol
0.0 kcal/mol
0.0 kcal/mol
1.5 kcal/mol
4.9 kcal/mol
2.9 kcal/mol
5.8 kcal/mol
O
N
5: Ar = Ph, R = Me
6: Ar = 3-Me-Ph,
R =
N
O
Scheme 2. Relative energies of conformers of compounds 5 and 6.

Y. Shi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5453–5458
5455
Early efforts in our FXa program focused on analogs of
1 where the thiourea moiety is replaced by a 2-nitroke-
tene, with an initial focus on the aromatic moiety (com-
pounds 7–18). These compounds were prepared by
sequential displacement of the two methylthio groups
of 1,1-bismethylthio-2-nitroethene with relevant
amines.14
However, this compound is about 3-fold more potent
than 8, a compound without a meta-substitution on
the phenyl ring. SAR studies indicate that other small
meta-substituted phenyl derivatives are of comparable
potency to 7 (for example, compounds 9 and 10), while
some para-substituents, with the exception of a meth-
oxy, result in the loss of activity (compare compounds
11, 12, and 13). The 3-methyl-4-methoxy substituted
phenyl compound 14 has an IC50 similar to that of 11.
As shown in Table 1, replacement of the thiourea group
in 1 with a 2-nitroketene aminal gives compound 7 with
an IC50 of 6400 nM, which is 58-fold less active than 1.
Replacement of the phenyl ring in 8 with a 2-naphthyl
group gives 16, which is 16-fold more potent, indicating
that a bicyclic moiety in this region may be a better
pharmacophore. Compound 15, which is a bicyclic ana-
log of 14, provides modest increase in potency. The 5-
benzofuranyl group (17) results in a further increase in
the potency. Finally, the 5-(2-methylbenzofuranyl)
derivative provides the most potent compound in this
Table 1. SAR of the aromatic binding element
Ar
HN
O
N
H
N
N
O
15
series (18), with an IC50 of 51 nM and an EC2·PT of
O2N
50 lM.
7-18
Similar SAR (data not shown) is observed when 2,2-
dicyanoketene aminal is employed as the thiourea surro-
gate. Compound 19 (see Table 2) is about 2-fold more
potent than its direct analog of nitroketene aminal 18,
with an IC50 of 30 nM and EC2·PT of 19 lM.
Compound
Ar
IC50 (nM)a
7
8
3-Me-Ph
Ph
6400
16500
6700
9
3-MeO-Ph
3-Me2N-Ph
4-MeO-Ph
4-Me-Ph
4-Me2N-Ph
10
11
12
13
14
15
16
17
18
2100
3100
The proposed binding mode of the ketene aminal com-
pounds with FXa is illustrated with compound 19 in
Figure 1.16 The ketene aminal group provides the L-
shaped scaffold seen in many FXa inhibitors, allowing
the 2-methylbenzofuran group to fit deep in the S1 pock-
et and the acylpyrrolidine group to reside in S4. One of
the nitrile groups is in close contact with the disulfide
bond; the high degree of polarizability of both these
groups could provide a strong favorable van der Waals
component to the free energy of binding. The other
>34000
>34000
3400
3-Me-4-MeO-Ph
2,3-Dihydro-benzofuran-5-yl
Naphthalene-2-yl
Benzofuran-5-yl
1500
1000
539
51
2-Methyl-benzofuran-5-yl
a IC50 are measured against human factor Xa utilizing the cleavage of a
synthetic substrate S-2222. Data are the average of two independent
determinations.
Table 2. SAR of the ketene substituents Scheme 3
O
H
N
H
N
N
N
O
O
Y
X
Compound
X
Y
IC50 (nM)a
EC2·PT (lM)b
19
27
28
29
30
31
32
33
34
35
36
37
38
39
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
CN
H
CN
30
17
19
13
23
31
60
10
21
24
nd
62
36
7
CONH2
CO2Me
CO2Et
20
24
CO2t-Bu
CO2(CH2)4OH
57
10
CO-4-methylpiperidine
SO2Me
SO2-i-Pr
57
46
647
13
CO-4-CI-Ph
CO-4-MeO-Ph
CONHCONH2
PO(OEt)2
13
13
658
10
>60
31
CO-2-thiazole
a IC50 are measured against human factor Xa utilizing the cleavage of a synthetic substrate S-2222.
b Concentration of inhibitor required to double the prothrombin based clotting time in human plasma; data are the average of two independent
determinations.

5456
Y. Shi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5453–5458
To further explore the SAR of ketene substituents, addi-
tional compounds based on 5-(2-methylbenzofuranyl) as
the S1 pharmacophore were synthesized according to
Scheme 3. Compounds 19, 27–32, 37, and 38 were syn-
thesized by the stepwise displacement of both methyl-
thio groups of ketene dithioacetals 21, which was
prepared by the reaction of the anion of 20 with carbon
disulfide and then methylation of both thio-groups.17
Compounds 33–36, 39 were prepared according to a
one-pot procedure we developed for N,N0-disubstituted
ketene aminals.12 For example, treatment of 2-(meth-
ylsulfonyl)acetonitrile (X = CN, Y = SO2Me in 25) with
sodium hydride and reaction of the resulting anion with
2-methylbenzofuran-5-isothiocyanate 23 affords an
intermediate thioamide anion (26). Subsequent reaction
of 26 with caprolactam amine 24 and EDCI (1-ethyl-3-
[3-(dimethylamino)propyl]carbodiimide hydrochloride)
provides 32.
Figure 1. Model of 19 (cyan) in FXa (magenta). For clarity, only key
residues are shown.
As shown in Table 2, compounds with differently substi-
tuted ketene aminals generally have similar FXa inhibi-
tory activities to 19, except for compounds 34 and 38.
The SAR is consistent with our binding model shown
in Figure 1, where the ketene aminal adopts an anti-
syn conformation with one ketene substituent pointing
into the solvent. Modification of the ketene substituents
leads to only slight improvement in anticoagulation
activities (EC2·PT).
nitrile lies against the caprolactam ring and points to-
ward solvent. Another key feature of this proposed
binding mode is a hydrogen bond formed between the
caprolactam carbonyl and the backbone NH of glycine
˚
216 (3.3 A in the model). From this model, it is clear
that the methyl group could provide a key van der
Waals interaction with Tyr228 and enhance the binding
activity, which is reflected in the 10-fold increase in IC50
for compound 18 compared to 17; it also could be pro-
posed that variations of the ketene substitution groups
may provide compounds with improved physiochemical
properties without strongly impacting their anti-FXa
activity.
Similar to the lead compound 1,4 the ketene aminals de-
scribed above are selective FXa inhibitors relative to
related trypsin-like serine proteases. Table 3 shows the
selectivity profiles of compound 31. Furthermore, com-
pound 31 showed activity in rats after systemic adminis-
tration and measurement of ex vivo clotting time.18 As
O
N
H
N
H
N
MeS
Y
SMe
CN
a,b,c
d,e
N
Y
CN
O
O
Y
CN
19, 27-32, 37, 38
20
21
O
H2N
SCN
N
f
H2N
N
O
O
O
22
23
24
O
N
H
N
H
N
H
N
SNa
i
g,h
N
O
Y
X
O
O
Y
X
Y
X
26
25
33-36, 39
Scheme 3. The synthesis of ketene aminals 19, 27–39. Reagents and conditions: (a) KOH, H2O, rt; (b) CS2; (c) Me2SO4; (d) 22, EtOH, 70 ꢁC; (e) 24,
EtOH, 70 ꢁC; (f) 1,10-thiocarbonyldi-2(1H)-pyridone, CH2Cl2; (g) NaH, DMF; (h) 23, 60 ꢁC; (i) 24, 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide
hydrochloride.

Y. Shi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5453–5458
5457
Table 3. Selectivity profile of 31
References and notes
Human enzymes
Ki (nM)a
1. For reviews, see: (a) Kunitada, S.; Nagahara, T.; Hara, T.
Handb. Exp. Pharmacol. 1998, 397; (b) Vacca, J. P. Annu.
Rep. Med. Chem. 1998, 33, 81; (c) Zhu, B.-Y.; Scarbor-
ough, R. M . Curr. Opin. Cardiovasc. Pulm. Ren. Investig.
Drugs 1999, 1, 63; (d) Ewing, W. R.; Pauls, H. W.; Spada,
A. P. Drugs Future 1999, 24, 771; (e) Fevig, J. M.; Wexler,
R. R. Annu. Rep. Med. Chem. 1999, 34, 81; (f) Betz, A.
Expert Opin. Ther. Patents 2001, 11, 1007.
2. (a) Davie, E. W.; Fujikawa, K.; Kisiel, W. Biochemistry
1991, 30, 10363; (b) Butenas, S.; VanÕt Veer, C.; Cawthern,
K.; Brummel, K. E.; Mann, K. G. Blood Coagul. Fibri-
nolysis 2000, 11, S9.
Factor Xa
Trypsin
Thrombin
12
>5000
6300
Plasma kallikrein
Activated protein C
Factor IXa
Factor VIIa
Chymotrypsin
Urokinase
Plasmin
>19400
>21500
>10000
>55000
9420
>14000
>22000
16000
tPA
a KiÕs are calculated from the IC50 values assuming competitive inhi-
bition versus the low molecular weight synthetic substrates using the
relationship, Ki = IC50/(1 + [S]/Km), where [S] is the substrate con-
centration in the assay and Km is the Michaelis constant for that
substrate. Data are the average of two independent determinations.
3. For orally bioavailable Factor Xa inhibitors, see: (a) Lam,
P. Y. S.; Clark, C. G.; Li, R.; Pinto, D. J. P.; Orwat, M. J.;
Galemmo, R. A., Jr.; Fevig, J. M.; Teleha, C. A.;
Alexander, R. S.; Smallwood, A. M.; Rossi, K. A.;
Wright, M. R.; Bai, S.; He, K.; Luettgen, J. M.; Wong,
P. C.; Knabb, R. M.; Wexler, R. R. J. Med. Chem. 2003,
46, 4405; (b) Quan, M. L.; Lam, P. Y. S.; Han, Q.; Pinto,
D. J. P.; 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. J. Med. Chem. 2005, 48, 1729; (c) Ueno, H.; Yokota,
K.; Hoshi, J.; Yasue, K.; Hayashi, M.; Hase, Y.; Uchida,
I.; Aisaka, K.; Katoh, S.; Cho, H. J. Med. Chem. 2005, 48,
3586; (d) Pruitt, J. R.; Pinto, D. J. P.; Galemmo, R. A., Jr.;
Alexander, R. S.; Rossi, K. A.; Wells, B. L.; Drummond,
S.; Bostrom, L. L.; Burdick, D.; Bruckner, R.; Chen, H.;
Smallwood, A.; Wong, P. C.; Wright, M. R.; Bai, S.;
Luettgen, J. M.; Knabb, R. M.; Lam, P. Y. S.; Wexler, R.
R. J. Med. Chem. 2003, 46, 5298, and references cited
therein.
5
31
Compound , I.D.
4
3
2
1
31
Compound , I.V.
30 mg/kg I.D. n=2
10 mg/kg I.V. n=2
4. Bisacchi, G. S., et al., unpublished results.
5. This work was presented in part by: Shi, Y.; Sitkoff, D.;
Zhang, J.; Grazier, N.; Stein, P. D.; Atwal, K. S.; Bisacchi,
G. S.; Zhang, J.; Liu, E. C. -K.; Seiler, S. M.; Hartl, K. S.;
Schumacher, W. A.; Youssef, S. A.; Klei, H.; Pudzianow-
ski, A. T.; Kish, K.; Yanchunas, J. In Abstract of Papers,
226th National Meeting of the American Chemical
Society, New York, NY; American Chemical Society:
Washington, DC, 2003; Abstract MEDI-081.
0
20
40
60
80
100
120
min
6. (a) Silverman, R. B. The Organic Chemistry of Drug
Design and Drug Action; Academic Press, New York,
1992, pp 88; (b) Durant, G. J.; Emmett, J. C.; Ganellin, C.
R.; Miles, P. D.; Parsons, M. E.; Prain, H. D.; White, G.
R. J. Med. Chem. 1977, 20, 901; (c) Black, J. W.; Durant,
G. J.; Ganellin, C. R. Nature 1974, 248, 65; (d) Bradshaw,
J. C.; Ganellin, C. R.; Parsons, M. E. J. Int. Med. Res.
1975, 3, 86; (e) Domschke, W.; Lux, G.; Domschke, D.
Lancet 1979, 8111, 320.
Figure 2. The iv and id dosing of 31 in rat.
shown in Figure 2, intravenous bolus injection of 10 mg/
kg of 31 led to a 1.38-fold increase in prothrombin time
(PT) after 2 h, and intraduodenal dosing of 30 mg/kg of
31 resulted in a 1.35-fold increase in PT after 2 h. Com-
pound 31 also partially protected against lethality in
mice induced by intravenous injection of RusselÕs viper
venom (which causes FXa activation and lethality due
to pulmonary thrombosis). Vehicle-treated mice died
within 1.0 0.1 min (n = 16) after RusselÕs viper venom
injection, while survival times were increased to
16.0 8.4 min (n = 5) after oral administration of
2 mg/kg of 31.19
7. In Ranitidine crystals, the 2-nitroketene aminal functional
group was found to adopt either a syn–anti or an anti–syn
conformation depending on its crystal forms. The ketene
˚
˚
C@C bond length is between 1.357 and 1.433 A; the two
C–N bond lengths are between 1.315 and 1.431 A, with an
´
N–C–N angle between 119.5ꢁ and 121.4ꢁ. See: (a) Kojic-
ˇ
ˇ
´
ˇ ´
Prodic, B.; Ruzic-Toros, Z.; Toso, R. Acta Crystallogr.
1982, B38, 1837; (b) Ishida, T.; In, Y.; Inoue, M. Acta
Crystallogr. 1990, C46, 1893; (c) Hempel, A.; Camerman,
N.; Mastropaolo, D.; Camerman, A. Acta Crystallogr.
2000, C56, 1048.
In summary, we have discovered a series of novel, orally
active FXa inhibitors. These compounds exhibit potent
and highly selective anti-FXa activity. The SAR studies
led to the 2-methylbenzofuran as the critical S1 binding
element, and that the ketene aminals are good bioisos-
teres for the thiourea functional group.
8. In a Metiamide crystal, the thiourea functional group was
found to have an anti-syn conformation with the methyl
group cis to the thiocarbonyl sulfur. The C@S bond
˚
distance is 1.711 A; the two C–N bond lengths are 1.338
˚
and 1.325 A, with an N–C–N angle of 118ꢁ. See: Prout,

5458
Y. Shi et al. / Bioorg. Med. Chem. Lett. 15 (2005) 5453–5458
K.; Critchley, S. R.; Ganellin, C. R.; Mitchell, R. C.
J. Chem. Soc., Perkin Trans. 2 1977, 68.
forcefield. All residues in FXa were fixed with the
exception of Tyr99, Phe174, Cys191, Cys220, Trp215,
Ser195, Gln192, Arg143, Glu146, and Gly218.
9. (a) Dewar, M. J. S.; Jie, C.; Yu, J. Tetrahedron 1993, 49,
5003; (b) AMPAC 6.0, Semichem, Shawnee KS 66216,
1997. The AMPAC 6.0 program system was used to carry
out geometry optimizations on the structures shown in
Scheme 2.
17. Henriksen, L. Acta Chem. Scand. 1996, 50, 432.
18. Compound testing protocol: male Sprague–Dawley rats
(320–390 g) were fasted overnight and then anesthetized
with sodium pentobarbital (50 mg/kg, ip). The trachea
was cannulated with PE-205 tubing to assure airway
patency. Catheters (PE-50) were placed in the right
carotid artery for blood withdrawal and in the left
jugular vein for saline infusion (25 lL/min throughout
the experiment) and for iv dosing of test compound. For
intestinal delivery (id) of test compound, the small
intestine was exposed via a midline laparotomy and a
dosing catheter (PE-50) was inserted into the duodenum
at the level of the bile duct. Animals received compound
by either iv (10 mg/kg) or id (30 mg/kg) route in a 1 mg/
mL volume of saline vehicle followed by a 0.3 mL saline
flush. Arterial blood samples (0.5 mL) were withdrawn
into 3.8% Na-citrate (1/10; v/v) for ex vivo prothrombin
time (PT) determination before (0 min control), and at
30, 60, 90, and 120 min after test compound dosing. The
10. Kumler, W. D.; Fohlen, G. M. J. Am. Chem. Soc. 1942,
64, 1944.
11. (a) Sandstro¨m, J. Top. Stereochem. 1983, 14, 83; (b)
Ericsson, E.; Marnung, T.; Sandstro¨m, J.; Wennerbeck, I.
J. Mol. Struct. 1975, 24, 373; (c) Isaksson, G.; Sandstro¨m,
J . Acta Chem. Scand. 1973, 27, 1183; (d) Bernhardt, P. V.;
Koch, R.; Moloney, D. W. J.; Shtaiwi, M.; Wentrup, C.
J. Chem. Soc., Perkin Trans. 2 2002, 515.
12. Shi, Y.; Zhang, J.; Grazier, N.; Stein, P. D.; Atwal, K. S.;
Traeger, S. C.; Callahan, S. P.; Malley, M. F.; Galella, M.
A.; Gougoutas, J. Z. J. Org. Chem. 2004, 69, 188.
13. (a) Dumanovic, D.; Juranic, I.; Dzeletovic, D.; Vasic, V.
M.; Jovanovic, J. J. Pharm. Biomed. Anal. 1997, 15, 1667;
(b) Degim, T.; Zaimoglu, V.; Akay, C.; Degim, Z .
Farmaco 2001, 56, 659; (c) Clark, G. J.; Berry, S. G.;
Hutchinson, J. H. Anal. Chem. 1973, 45, 1751; (d)
Bodnarchuk, N. D.; Yatsishin, A. A. Zh. Org. Khim.
1977, 13, 954.
14. For the synthesis of N,N0-disubstituted 2-nitroketene
aminals, see: (a) Niemers, E.; Knorr, A.; Garthoff, B.
U.S. Patent 4567188, 1986; (b) Manley, P. W.; Quast, U.
J. Med. Chem. 1992, 35, 2327; For the synthesis of 2,2-
dicyanoketene aminals, see: (c) Gompper, R.; To¨pfl, W.
Chem. Ber. 1962, 95, 2871.
15. IC50 are measured against human factor Xa utilizing the
cleavage of a synthetic substrate S-2222. The EC2·PT is the
concentration of inhibitor required to double the pro-
thrombin based clotting time in human plasma.
PT was measured using
a Amelung KC4A micro-
coagulation analyzer (Heinrich Amelung GmbH, Lemgo,
Germany) and the standard procedure described for
Dade Thromboplastin-C reagent (Baxter Healthcare
Corp., Miami, FL).
19. Swiss Webster mice (24–38 g) were anesthetized with
sodium pentobarbital (100 mg/kg, ip). RussellÕs viper
venom (Sigma Chemical, St. Louis, MO) was prepared
in 0.9% saline and injected into the tail vein at a dose of
7 lg/mouse in a volume of 0.1 mL given 15 min after
induction of anesthesia. This dose of venom was found to
be uniformly fatal in 105 mice with death occurring in an
average of 1.5 1.9 min ( SD, range was 1 to 12 min with
88% of mice dying within 2 min). Histological evaluation
of several venom-treated mice revealed that early death
was due to occlusion of pulmonary capillaries with fibrin
aggregates, probably caused by initiation of blood coag-
ulation by venom activation of factor X to factor Xa. A
2 mg/kg dose of test compounds or saline vehicle was
administered by oral gavage 2 h before the venom
injection. Survival was monitored out to 30 min after
venom injection.
16. Initial binding mode was obtained based on an in-house
FXa crystal structure of a compound identical to 19 but
with an acylguanidine linker in place of the ketene aminal.
From this starting point, the compound was minimized in
the published FXa crystal structure 1fjs (Adler, M.;
Davey, D. D.; Phillips, G. B.; Kim, S. H.; Jancarik, J.;
Rumennik, G.; Light, D. L.; Whitlow, M. Biochemistry
˚
2000, 39, 12534. 1.92 A resolution) using Discover v. 98.0
in the InsightII package (Accelrys, Inc.) with the CFF