Inhibitors of the tissue factor/factor VIIa-induced coagulation: Synthesis and in vitro evaluation of novel specific 2-aryl substituted 4H-3,1-benzoxazin-4-ones

Article abstract of DOI:10.1016/S0968-0896(00)00129-2

The synthesis of a series of novel 2-aryl substituted 4H-3,1-benzoxazin-4-ones and their evaluation as specific inhibitors of the Tissue Factor (TF)/Factor VIIa (FVIIa)-induced pathway of coagulation is reported. Inhibitory activities (IC₅₀ val

Full text of DOI:10.1016/S0968-0896(00)00129-2

Bioorganic & Medicinal Chemistry 8 (2000) 2095±2103
Inhibitors of the Tissue Factor/Factor VIIa-Induced Coagulation:
Synthesis and In Vitro Evaluation of Novel Speci®c 2-Aryl
Substituted 4H-3,1-Benzoxazin-4-ones
Palle Jakobsen,a,* Bo Ritsmar Pedersena and Egon Perssonb
aMedicinal Chemistry Research, Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Maaloev, Denmark
bProtein Chemistry 3 Novo Nordisk A/S, Novo Nordisk Park, DK-2760 Maaloev, Denmark
Received 23 March 2000; accepted 18 May 2000
AbstractÐThe synthesis of a series of novel 2-aryl substituted 4H-3,1-benzoxazin-4-ones and their evaluation as speci®c inhibitors
of the Tissue Factor (TF)/Factor VIIa (FVIIa)-induced pathway of coagulation is reported. Inhibitory activities (IC50 values) in the
range 0.17 to >40 mM on the activation of Factor X (FX) by the TF/FVIIa complex were found for compounds having one or two
electronegative substituents such as F, Cl and NO2 in the 2-aryl substituent. Di€erent substitutions both electron-attracting and donating
groups were allowed in the 5, 6, 7 and 8 positions. Several of the compounds showed a selectivity ratio towards FX and thrombin of >50,
thus being the ®rst small molecules described as potential drugs for oral antithrombotic treatment without side e€ects such as bleeding
which is observed especially with thrombin inhibitors. The best substituent pattern being the 2-aryl group substituted with: 2-F; 2,6-
F2; or 2-FX; 6-Cl; together with electronegative substitution in the 5, 6, 7, or 8 positions. 2-Heteroaryl substituents like thienyl and
furanyl were of low activity while some 2-(2-chloro-3-pyridyl) derivatives had inhibitory activity <10 mM and a good selectivity.
# 2000 Elsevier Science Ltd. All rights reserved.
Introduction
FVIIa and TF appear to be the principal initiators of
blood coagulation in vivo.1
Blood coagulation is a complex process involving various
blood components or factors that eventually give rise to
a ®brin clot. Activated Factor X (FXa) is required to
convert prothrombin to thrombin which then converts
®brinogen to ®brin as a ®nal stage in forming a ®brin clot.
It is often desirable to inhibit the coagulation cascade in
a patient. Anticoagulants such as heparin, coumarin and
its derivatives, indandione derivatives, low-molecular-
weight thrombin or FXa inhibitors, or other agents may
be used. Treatment with heparin and other anticoagulants
may, however, have undesirable side e€ects, for example,
bleedings. Inhibition at the initial stage of blood coagula-
tion, i.e. of the FVIIa/TF activity, results in signi®cantly
less bleeding.2 In addition, clinically available antic-
oagulants act throughout the body, rather than acting
speci®cally at the site of injury where the coagulation
cascade is active. Other known anticoagulants comprise
thrombin and Factor Xa inhibitors derived from blood-
sucking organisms. Antithrombins, hirudin, hirulog and
hirugen are recombinant proteins or peptides derived
from the leach Hirudo medicinalis, whereas the Factor
Xa inhibitor antistatin and the recombinant derivative
rTAP are tick-derived proteins. Inhibitors of platelet
aggregation such as monoclonal antibodies or synthetic
peptides, which interfere with the platelet receptor gpIIb/
IIIa, are also e€ective as anticoagulants. These agents are,
however, not suitable for oral administration. Speci®c
protein inhibitors of the FVIIa/TF activity are the
There are two pathways that promote the activation of
Factor X (FX). The `intrinsic pathway' refers to those
reactions leading to thrombin formation through utili-
sation of factors present only in blood. A series of protease-
mediated activations generate Factor IXa which, in
conjunction with Factor VIIIa, cleaves FX into FXa.
In the `extrinsic pathway' Factor VIIa (FVIIa) and its
cofactor, Tissue Factor (TF), exert the same proteolysis.
TF is a membrane-bound protein and does not normally
circulate in plasma. Upon vessel disruption, however, it
is exposed and forms a complex with FVIIa to catalyse
FX or Factor IX activation in the presence of Ca2+ and
phospholipid (1). While the relative importance of the
two coagulation pathways in hemostasis is unclear,
*Corresponding author.
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00129-2

2096
P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
physiological inhibitor TFPI (Tissue Factor Pathway
Inhibitor) and the inactivated FVII (FVIIai) which binds
to TF and compete with FVIIa
DMF using triethylamine or pyridine as base. However,
the yields varied considerably (0% to 100%), depending
on the substitution pattern in both the anthranilic acid
and the aryl carboxylic acid or acid chloride. Syntheses
using an aromatic ortho aminocarboxylic acid ester and
an aryl carboxylic acid chloride as starting materials per-
forming the ring closure with sulphuric acid3 gave very
low yields, and the same was found for syntheses using an
anthranilic acid derivative and an arylcarboxylic acid
performing the reaction with HOBT and EDAC, prob-
ably due to the presence of two carboxylic acids which
could be activated by HOBT.4 9 The reaction is depicted
in Scheme 1.
There is still a need for improved compositions having
anticoagulant activity and which can be administered
orally or otherwise non-intravenously at relatively low
doses without producing undesirable side e€ects. We found
that some 2-arylsubstituted 4H-3,1-benzoxazin-4-ones
showed speci®c inhibitory activity on FVIIa-TF in in
vitro assays, and therefore seem to be potential drug
candidates for blood coagulation regulation, suitable for
oral administration and presumably without the above
mentioned side e€ects. We report here the synthesis and
SAR of some 2-arylsubstituted 4H-3,1-benzoxazin-4-ones
in in vitro systems of TF/FVIIa-catalyzed activation of
FX, as well as their e€ects on the amidolytic activity of
FXa and thrombin and the prolongation of clotting
time caused by selected compounds. 4H-3,1-Benzoxazin-
4-one-derivatives are known to act as serine protease inhi-
bitors, but very few of the reported compounds contain an
aryl group connected directly to the ring system together
with substituents in the 5, 6, 7, or 8 positions.3 9 None of
the previously described compounds have been reported
to speci®cally inhibit the TF/FVIIa-induced coagulation
pathway, but rather to act as thrombin and FXa inhibi-
tors. Similar structures, often containing a 2-amino sub-
stituent instead of the 2-aryl substituent, are reported to
act as inhibitors of HSV-1 protease,10 C1r serine pro-
tease,11 and human cytomegalovirus protease.12 A solid-
phase synthetic approach towards 2-amino substituted
4H-3,1-benzoxazin-4-ones has been reported.13
In order to conveniently prepare a series of analogues for
the SAR investigation, a solution phase parallel synthesis
procedure was developed (Scheme 2).
Appropriately substituted 2-aminobenzoic acids were
dissolved in toluene, triethyl amine was added and the
mixture shaken for 10 min at room temperature. Subse-
quently, substituted benzoic acid chloride was added
and the mixture shaken overnight at room temperature.
Ethyl acetate and HCl (0.1N) were added followed by
shaking. The organic layer was separated, saturated
NaHCO3 added, and the mixture shaken for 2 min. The
organic layer was separated and ®ltered through a silica-gel
column (Sep-Pack plus), the column washed with di-
chloromethane and the collected organic phases evapo-
rated. The identity and purity were checked by LC-MS,
>90% of the experiments gave >75% yield as esti-
mated from the ELS measurement.
Chemistry
Results
2-Aryl substituted 4H-3,1-benzoxazin-4-ones were con-
veniently prepared from optionally substituted anthranilic
acids and an aryl carboxylic acid chloride in toluene or
The compounds were evaluated for inhibitory activity on
the TF/FVIIa-induced activation of FX using a two-step
amidolytic assay. Compounds showing good inhibitory
activity in that assay were further tested for activity on
FXa and thrombin in order to sort out structures with
the best selectivity. Selected compounds were further
tested for their ability to prolong clotting time in a
standard in vitro clotting assay. Selected data for 2-aryl
substituted 4H-3,1-benzoxazin-4-ones are depicted in
Table 1.
Scheme 1. Synthetic pathway used for the preparation of compounds
2, 5, 17±20, 22, 24, 27, 28, 30, 32, 33, 36, 38, 44, 59 and 60. i: Ar(R2)-
COCl, Et3N, Toluene; ii: Ar(R2)COOH, HOBT, EDAC, DMF.
The data revealed a broad variety of selectivity, ranging
from compounds having a similar inhibitory e€ect on
Scheme 2. Solution phase parallel synthesis approach for the preparation of compounds 3, 4, 6±9, 12±16, 21, 25, 26, 29, 31, 39, 40, 42, 43, 46 50 and 61:

P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
2097
Table 1. IC50 values (mM) for the inhibition of TF/FVIIa-catalyzed
activation of FX and the inhibition of FXa and thrombin amidolytic
activity
and 7-tri¯uoromethyl substitution in the benzoxazine
ring, respectively, and 2-¯uoro or 2,6-di¯uoro substitution
in the 2-phenyl substituent. These compounds also pro-
longed the TF-induced clotting time, the ratios being
1.32, 1.31, and 1.51 times the control clotting time. The
best inhibitors of the TF/FVIIa-induced activation of
FX were substituted with electronegative substituents
like Cl, NO2, CF3, F in the 5, 6, or 7 positions in the ben-
zoxazine ring together with a 2,6-di¯uoro substitution in
the 2-phenyl group. However, these compounds also
possessed good inhibitory e€ects on the FXa amidolytic
activity resulting in a speci®city factor below 50, while
their activities at the active site of thrombin were low.
Also a 6-methyl substituent (compound 20) inhibited
the FX activation at sub-mM concentrations, however,
the inhibitory e€ect on thrombin was also good (8 mM).
Compounds carrying-NH2,-OMe or-COOMe in the
benzoxazine ring (compounds 4, 26, 28, 32 and 33) gave
a 3-fold to 10-fold higher IC50 in the FX activation
assay than found for the electronegative substituents,
while the IC50 for FXa and thrombin was often found
to be >200 mM (i.e. no e€ect).
Compound
no.
R1
R2
FX FXa Thrombin
activation IC50 IC50
IC50 mM
mM
mM
1
2
3
4
5
6
7
8
H
2-F
2-F
2-F
2-F
2-F
2-F
2-F
2-F
2-F,3-F
2-F
2-F
2-F,3-F
2-F,3-F
2-F
4.7
3.6
3.1
3.0
1.2
5.6
7.2
17
18
25
30
32
138
>200
37
>200
180
138
56
145
34
40
>200
56
>200
17
105
>200
199
>200
NT
NT
NT
NT
>200
>200
>200
8
98
>200
NT
>200
134
30
>200
149
156
>200
>200
>200
>200
NT
6-Me
6-CF3
6-OMe
5-Cl,8-Cl
5-COOMe
6-F,7-F
5-NO2
6-F,7-F
6-Br,8-Br
7-Cl
>200
26
9
>200
>200
>200
>200
NT
NT
NT
NT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
6-Me
6-NO2
5-CF3
8-CF3
5-COOEt
5-Cl
5-NO2
5-Cl,8-Cl
6-Me
40
>40
>40
>40
0.17
0.35
0.36
0.60
0.63
0.82
0.90
1.0
1.4
1.5
1.6
3.2
4.1
4.3
4.5
22
2-F
2-F
Selected data for the inhibitory e€ect of 2-heteroaryl
substituted 4H-3,1-benzoxazin-4-ones is presented in
Table 2.
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-F
2-F,6-Cl
2-F,6-Cl
4.9
6.3
6.3
The compounds having a heteroaromatic substituent in
position 2 by and large showed a 10-fold higher IC50
value than found with a substituted phenyl group in
position 2. This probably re¯ects the di€erence in elec-
tronegativity for the two substituent groups.
22
6-CF3
7-NO2
5-F
16
112
6
6-NO2
7-CF3
6-OMe
6-NHAc
6-NH2
5-COOMe
5-Me
8-CF3
7-NH2
5-NH2
6-Me
19
174
154
35
154
112
89
Compound 47, however, had a good pro®le in agreement
with the 6,7-halogen disubstituted 2-phenyl compounds 7,
9 and 10.
80
>200
>200
10
6.0
110
15
The compounds giving the best prolongation in the clot-
ting time (17, 18, 20, 28 and 47) were substituted in the 2-
position with a 2,6-di¯uorophenyl group (or a 2-chloro-3-
pyridyl for compound 47), while the substitution on the
benzoxazin ring was Cl, NO2, Me, NH2 and F (Table 3).
30
0.9
1.2
1.2
2.2
2.8
0.32
3.9
6.8
7.3
10
5-F
NT
36
NT
>40
97
6-Me 2-Cl,3-Cl
6-I
2-Cl
2-Cl,6-Cl
2-NO2
2-OMe
2-OMe,5-Cl
6-Me
5-NO2
5-NO2
H
5-NO2 2-OCOMe
6-NO2 2-OCOMe
2-OCF3
2-Br
2-Me 2-OCOMe
63
<1.6
5.8
30
26
49
>200
6
>200
19
NT
Discussion
>200
>200
NT
NT
>200
The expected mechanism of action for the serine pro-
tease inhibition by benzoxazinone-type of compounds is
deactivation by means of reaction of the benzoxazinone
lactone with the active site serine, probably under the
formation of a drug-enzyme ester followed by its hydro-
lysis.7,14 This mechanism was supported by data obtained
for activity in a TF/FVII assay and in a FVII amidolytic
assay in which we found that the benzoxazinones gave
inhibition of the same order of magnitude as found in
the FX activation assay (not shown).
6-Me
6-Cl
12
1.5
33
all three enzymes to compounds active only in the FX
activation assay. Compounds with a detectable inhibitory
e€ect on FXa will also appear to e€ect FX activation and
FVIIa/TF-initiated clot formation, without necessarily
inhibiting FVIIa directly. Thus true IC50-values for the
FVIIa inhibition as measured by the FX activation
assay are obtained only for compounds which do not
inhibit Fxa.
The corresponding open anthranilic acid amide deriva-
tives (the `open form' of the benzoxazinone ring) were
found without inhibitory e€ects. These compounds could
be prepared from the corresponding benzoxazinones by
hydrolysis with NaOH in methanol followed by careful
acidi®cation with an equimolar amount of HCl at room
The best compounds regarding inhibitory activity and
selectivity were 5, 22 and 25 with 5,8-dichloro, 7-nitro,

2098
P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
Table 2. IC50 values (mM) for the inhibition of TF/FVIIa-catalyzed activation of FX, FXa and thrombin
Compound no.
R1
R3
FX-activation
IC50 mM
FX amidolytic
IC50 mM
Thrombin
IC50 mM
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
6-F,7-F
5-Me
2-Cl-3-pyridyl
2-Cl-3-pyridyl
2-Cl-3-pyridyl
2-Cl-3-pyridyl
4-pyridyl
5.0
7.1
14
21
27
33
33
16
25
34
>40
27
35
40
>40
>200
>200
98
>200
30
5-NO2
6-NO2
6-Br6-OMe
7-OMe
6-Br
7-Cl6-OMe
7-Ome
6-Br
6-Me
H
6-Me
6-Me
6-F,7-F
9.7
140
>200
NT
NT
NT
NT
>200
NT
NT
>200
>200
>200
>200
>200
>200
NT
>200
>200
NT
3-Cl,5-CF3-2-pyridyl
3-Cl,5-CF3-2-pyridyl
5-NO2-2-furanyl
2-furanyl
2-furanyl
5-NO2-2-furanyl
2-thienyl
2-thienyl
3-Br-2-thienyl
2-thienyl
NT
>200
NT
NT
Nt
temperature. Prolonged standing or higher temperatures
gave a mixture of the `open form' and the benzox-
azinone. The `open form' was also accessible from the
reaction of anthranilic acid and arylcarboxylic acid
chloride in pyridine at 30 ꢀC. However, also this
method gave mixtures of the `open' and ring-closed
forms (data not shown). Taking this expected mechanism
into account, it seems reasonable to expect that electro-
negative substituents on the 2-aryl group as well as on
the 5, 6, 7, or 8 positions of the benzoxazinone ring will
make the ring carbonyl more reactive towards the serine
OH-group of the enzyme.
Table 3. E€ects on TF/FVIIa-induced clotting. The ratio of the
clotting times with (330 mM) and without test compound is given
Compound no.
Clotting ratio
1
2
3
5
1.45
1.47
1.59
1.32
1.60
>3
2.58
1.68
>3
7
17
18
19
20
22
25
28
33
38
40
42
43
47
49
50
55
59
1.31
1.51
>3
Compounds without substitution in the 2 or 6-positions
in the 2-aryl substituent did not inhibit the TF/FVIIa-
catalyzed FX activation at concentrations below 40 mM
(data not shown).
1.06
1.88
1.25
1.14
1.22
2.32
1.52
1.16
1.35
1.76
Introduction of an ester functionality at the 5-position
(compounds 6, 16 and 29) or at the 2-position in the 2-
phenyl substituent (compounds 42, 43 and 46) was per-
formed in order to clarify if this group might, in some
way, compete with the ester formation between the serine-
OH and the benzoxazinone ring, either by steric hindrance
of the drug-enzyme reaction or by participating in the
ester formation with the enzyme. These compounds, how-
ever, gave results of the same magnitude and selectivity as
found for both electronegative and electron-donating
substituents at the same position. Consequently, no proof
in favour of that hypothesis was obtained.
while the IC50 for inhibition of FXa was 10-fold higher.
This is in contrast to the ®ndings for compounds 18, 39
and 40 for which the inhibitory e€ects were best in the
TF/FVIIa-catalyzed FX activation assay.
2-Pyridyl and 4-pyridyl substituents were found to be of
low activity in the activation assay, in the former case
probably because the ring nitrogen occupied an ortho
position and no electron-attracting groups were positioned
in an `ortho' position, and in the latter because no ortho
electronegative substituent was present. For the thienyl
and furanyl substituted compounds, only compound
60 had an electronegative substituent in the ortho
position, however, in this case a bromo substituent was
no good compared to compound 45 (IC50 values in the
FX activation assay being 40 and 1.5 mM, respectively).
Compounds substituted with 2-thienyl, 2-furanyl or 2-,
3- or 4-pyridyl in the 2-position of the benzoxazine ring
(compounds 47±61) showed the same pattern as found
for the 2-phenyl substituted compounds. IC50 values of
<10 mM were observed only for 47 and 48 carrying an
electronegative ortho-substituent in the 2-substituent
together with 6,7-di¯uoro or 5-Me substitution. Com-
pound 49 had a surprisingly good e€ect on thrombin
(9.7 mM), i.e. equipotent for thrombin and FVIIa/TF,

P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
Table 4. Analytical data for compounds prepared by solution phase parallel synthesis
2099
No.
Compound
Analytical detailsa
3
4
6
7
8
9
2-(2-Fluoro-phenyl)-6-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
2-(2-Fluoro-phenyl)-6-methoxy-4-H-3,1-benzoxazin-4-one
2-(2-Fluoro-phenyl)- 4-H-3,1-benzoxazin-4-oxo-5-carboxylic acid methyl ester
6,7-Di¯uoro-2-(2-¯uoro-phenyl)- 4-H-3,1-benzoxazin-4-one
2-(2,3-Di¯uoro-phenyl)-5-nitro-4-H-3,1-benzoxazin-4-one
m/za: 310 RT: 15.48, ELS: 1%
m/z: 272 RT: 14.10, ELS: 8%
m/z: 300 RT: 12.97, ELS: 00%
m/z: 278 RT: 14.78, ELS: 97%
m/z: 305 RT: 13.70, ELS: 95%
m/z: 296 RT: 14.91, ELS: 98%
m/z: 274 RT: 15.1, ELS: 82%
m/z: 305 RT: 14.26, ELS: 80%
m/z: 310 RT: 15.47, ELS: 81%
m/z: 310 RT: 15.15, ELS: 99%
m/z: 314 RT: 13.73, ELS: 95%
m/z: 328 RT: 15.26, ELS: 82%
m/z: 328 RT: 15.51, ELS: 91%
m/z: 290 RT: 13.8/11.45, ELS: 80%
m/z: 318 RT: 12.82, ELS: 67%
m/z: 328 RT: 14.73, ELS: 94%
m/z: 314 RT: 13.11, ELS: 81%
m/z: 299 RT: 12.81, ELS: 61%
m/z: 327 RT: 12.88, ELS: 84%
m/z: 327 RT: 13.37, ELS: 54%
m/z: 296 RT: 14.11, ELS: 54%
m/z: 295 RT: 12.75, ELS: 96%
m/z: 273 RT: 12.61, ELS: 100%
m/z: 304 RT: 11.64, ELS: 82%
m/z: 304 RT: 11.94, ELS: 88%
m/z: 266 RT: 14.68, ELS: 98%
2-(2,3-Di¯uoro-phenyl)-6,7-di¯uoro-4-H-3,1-benzoxazin-4-one
2-(2,3-Di¯uoro-phenyl)-5-methyl-4-H-3,1-benzoxazin-4-one
2-(2,3-Di¯uoro-phenyl)-6-nitro-4-H-3,1-benzoxazin-4-one
12
13
14
15
16
21
25
26
29
31
39
40
42
43
46
47
48
49
50
61
2-(2-Fluoro-phenyl)-5-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
2-(2-Fluoro-phenyl)-8-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
2-(2-Fluoro-phenyl)-4-oxo-4-H-3,1-benzoxazin- 5-carboxylic acid ethyl ester
2-(2,6-Di¯uoro-phenyl)-6-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
2-(2,6-Di¯uoro-phenyl)-7-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
2-(2,6-Di¯uoro-phenyl)-6-methoxy-4-H-3,1-benzoxazin-4-one
2-(2,6-Di¯uoro-phenyl)-4-oxo-4-H-3,1-benzoxazin-5-carboxylic acid methyl ester
2-(2,6-Di¯uoro-phenyl)-8-tri¯uoromethyl-4-H-3,1-benzoxazin-4-one
5-Nitro-2-(2-nitro-phenyl)- 4-H-3,1-benzoxazin-4-one
2-(2-Methoxy-phenyl)-5-nitro-4-H-3,1-benzoxazin-4-one
Acetic acid 2-(5-nitro-4-oxo-4-H-3,1-benzoxazin-2-yl)-phenyl ester
Acetic acid 2-(6-nitro-4-oxo-4-H-3,1-benzoxazin- 2-yl)-phenyl ester
Acetic acid 2-(5-methyl-4-oxo-4-H-3,1-benzoxazin-2-yl)-phenyl ester
2-(2-Chloro-pyridin-3-yl)-6,7-di¯uoro-4-H-3,1-benzoxazin-4-one
2-(2-Chloro-pyridin-3-yl)-5-methyl-4-H-3,1-benzoxazin-4-one
2-(2-Chloro-pyridin-3-yl)-5-nitro-4-H-3,1-benzoxazin-4-one
2-(2-Chloro-pyridin-3-yl)-6-nitro-4-H-3,1-benzoxazin-4-one
6,7-Di¯uoro-2-thiophen-2-yl-4-H-3,1-benzoxazin-4-one
am/z: The molecular ion (M+1) from the LC-MS investigation; RT: Retention time (min); ELS: The purity estimated from the electrospray (positive
ion) measurement.
Conclusion
column was eluted with a linear gradient of 5±90% A, 85±
0% B and 10% C for 15min at a ¯ow rate of 1 mL/min
(solvent A=acetonitrile, solvent B=water and solvent,
C=0.1% tri¯uoroacetic acid in water). Microanalyses
were performed by Novo Nordisk Analytical Department.
The synthetic approach described easily yielded a series
of 2-aryl substituted 4H-3,1-benzoxazin-4-ones in fair to
good yields by standard synthetic methods and through
a solution-phase parallel fashion. The 2-aryl substituted
4H-3,1-benzoxazin-4-ones described comprise several
compounds with excellent activity and selectivity thus
being the ®rst small molecules described having that
pro®le. Good inhibition requires the presence of at least
one electron-attracting substituent in the ortho position
of the 2-aryl or heteroaryl substituent, the best structures
also bearing one or two electronegative substituents in
the 5, 6, 7 or 8 positions. The best compounds as
regards both inhibitory activity of the TF/FVIIa-cata-
lyzed FX activation and best selectivity (5, 22 and 25)
were all found within this pattern. Simple heterocycles
like thienyl, furanyl and 2- and 4-pyridyl as the 2-sub-
stituents gave compounds of low activity while 2-(3-
pyridyl) substituents were found to be of similar activity
as the phenyl substituents (47).
FX activation assay
The compounds were dissolved in DMSO and mixed
with FVIIa (produced in-house) in 50 mM Hepes, pH
7.4, containing 0.1 M NaCl, 5 mM CaCl2, and 1 mg/mL
bovine serum albumin (1+5). 30mL of this mixture was
then mixed with 45mL TF (American Diagnostica, relipi-
dated in PC/PS vesicles) and 25 mL of FX (Enzyme
Research Laboratories), all in a Ca2+-containing bu€er.
This gave ®nal concentrations of 100 pM FVIIa, 5 pM
TF, 175 nM FX, and various concentrations of the
compounds. After a 5 min incubation, the FVIIa/TF-
catalyzed activation of FX was terminated by an addition
of a 50 mL bu€er containing enough EDTA to give an
excess over the Ca2+ ions present. 50mL of a 2-mM solu-
tion of S-2765 (Chromogenix, 2 mM in water) was then
added and the FXa formed was allowed to hydrolyze the
substrate for 10 min during which the absorbance at
Experimental
405 nM was continuously monitored in
a SPEC-
General procedures. Melting points (uncorrected) were
measured on a Buchi 535. NMR spectra were recorded
on a Bruker AVANCE DPX 200 Mhz or 300 MHz
instrument operating at room temperature. Mass spectra
were obtained on a Finnigan Mat TSQ 70 apparatus using
a direct inlet system. The HPLC-MS analyses were per-
formed on a PE Sciex API 100 LC/MS System using a
WatersTM 3 mmÂ150 mm, 3.5 m, C-18 Symmetry column
and positive ion spray with a ¯ow rate at 20mL/min. The
TRAmaxTM 340 plate reader. The slope of the absorbance
curve was compared to that of a control where only
DMSO was added to FVIIa/TF/FX.
Thrombin activity assay
170 mL thrombin (Boehringer Mannheim, 3.5 nM) in
the bu€er described above, 10 mL test compound in
DMSO (or DMSO alone in the control) and 20 mL

2100
P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
S-2238 (Chromogenix, 10 mM in water) were mixed.
The hydrolysis of S-2238 was monitored as described
for the FX activation assay.
puri®ed by column chromatography on silica-gel using
heptane:ꢀethyl acetate (4:1) as eluent. Yield 0.31 g (42%);
mp 140 C; MS 311/309 (M+) (Cl isotope pattern).
Calculated for C14H6Cl2FNO2: C, 54.22%; H, 1.95%;
N, 4.52%. Found: C, 53.53%; H, 1.91%; N, 4.38%. 1H
NMR (CDCl3, ppm) 8.20 (1H, double t); 7.78 (1H, d);
7.65±7.52 (1H, m); 7.48 (1H, d); 7.35±7.15 (2H, m).
FXa activity assay
170 mL FXa (Enzyme Research Laboratories, 1.2 nM in
bu€er) in the bu€er desribed above, 10mL test compound
and 20mL S-2765 (Chromogenix, 10mM in water) were
mixed. The hydrolysis of S-2765 was monitored as
described for the FX activation assay.
5-Chloro-2-(2,6-di¯uoro-phenyl)-4-H-3,1-benzoxazin-4-
one (17). 6-Chloroanthranilic acid (0.566 g), 2,6-di¯uor-
obenzoyl chloride (0.93 mL) and triethyl amine:toluene
(1/1) (18 mL) were reacted by heating to 50 ꢀC for 2 days.
Extraction between ethyl acetate (100 mL) and HCl (2N,
100 mL), followed by separation of the organic layer,
drying over MgSO4, and ®nally ®ltering and evaporation
gave a crude product that was dissolved in warm THF
(20 mL) and precipitated with hexane. The resulting
mixture was further puri®ed on a silica-gel column using
dichloromethane as eluent. The isolated fraction was
dissolved in warm THF (20 mL) and precipitated with
hexane, yielding 17 (0.65 g, 70%); mp. 176 ꢀC; MS m/e 293/
295 (M+). Calculated for C14H6Cl2FNO2: C, 57.62%; H,
2.06%; N, 4.77%. Found: C, 57.35%; H, 2.06%; N,
FVIIa/TF-initiated clotting assay
The test compounds, 20 mM in DMSO, were diluted in
citrated normal human plasma just before the analysis
(1+19) and placed in the sample carousel of an ACL
300 Research coagulometer. 55 mL sample (compound
in plasma) was mixed with 55 mL of thromboplastin
(Innovin, Dade; diluted 1:500 after dissolution) and incu-
bated for 5 min. The clotting reaction was started by add-
ing 55 mL of a 25 mM CaCl2 solution, yielding a ®nal
compound concentration of 0.33mM. The ratio between
the clotting time in the presence and absence of test com-
pounds was used to quantify the anticoagulant e€ect.
1
4.60%. H NMR (DMSO, ppm) 7.93 (1H, t); 7.82±7.65
(3H, m); 7.38 (1H, d); 7.32 (1H, d).
The following compounds were commercially available
from Maybridge (compounds 1, 10, 11, 23, 34, 35, 37,
51, 52, 53, 54, 57 and 58), CSC (compounds 41, 55 and
56), and SPECS (compound 45).
2-(2,6-Di¯uoro-phenyl)-5-nitro-4-H-3,1-benzoxazin-4-one
(18). 6-Nitroanthranilic acid (0.6 g), 2,6-di¯uorobenzoyl
chloride (0.93 mL) and triethyl amine:toluene (1:1)
(18 mL) were reacted by heating to 50 ꢀC for 2 days.
Extraction between ethyl acetate (100 mL) and HCl (2N,
100 mL), followed by separation of the organic layer,
drying over MgSO4, and ®nally ®ltering and evaporation
gave a crude product that was dissolved in warm THF
(20 mL) and precipitated with hexane. The resulting
mixture was further puri®ed on a silica-gel column using
dichloromethane as eluent. The isolated fraction was
dissolved in warm THF (20 mL) and precipitated with
hexane, yielding 18 (0.475g, 40%); mp 172 ꢀC; MS m/e
304 (M+); LC-MS m/e 305 (M+1); ELS-purity 100%.
C14H6F2N2O4: C, 56.25%; H, 2.02%; N, 7.03%. Found:
2-(2-Fluoro-phenyl)-6-methyl-4-H-3,1-benzoxazin-4-one
(2). 2-Fluorobenzoic acid (3.3 mmol, 0.52 g) was mixed
with HOBT (3.3 mmol, 0.44 g) and EDAC (6.6 mmol,
1.26 g) in CH2Cl2 (5 mL). The mixture was stirred for 1 h
and subsequently 2-amino-5-methylbenzoic acid (3.3
mmol, 0.5 g) in DMF (5 mL) was added dropwise (10 min).
After stirring for 3 h, HCl (4N, 10 mL) was added and the
mixture extracted with CH2Cl2 (3Â10 mL). The collected
organic phases were extracted with HCl (4N, 10 mL) and
then twice with water (10 mL). The organic phases were
dried (MgSO4), ®ltered and evaporated to dryness. The
crude product was puri®ed on silica-gel using CH2Cl2 as
eluent. Yield of 2 (30%); LC-MS 256 (M+1); RT
14.61 min; ELS-purity 100%; mp 144 ꢀC. Calculated for
C15H10FNO2: C, 70.58%; H, 3.95%; N, 5.49%. Found: C,
1
C, 55.27%; H, 1.99%; N, 8.96%. H NMR (DMSO,
ppm) 8.22±8.05 (2H, m); 7.97 (1H, dd); 7.85±7.68 (1H, m)
7.36 (2H, t).
5,8-Dichloro-2-(2,6-di¯uoro-phenyl)- 4-H-3,1-benzoxazin-
4-one (19). 2-Amino-3,6-dichlorobenzoic acid (0.125 g),
2,6-di¯uorobenzoyl chloride (0.168 mL) and triethyl
amine:toluene (1:1) (1,5 mL) were reacted by heating
to 50 ꢀC for 2 h. HCL (0.2N, 1 mL) was added and the
organic layer separated and rinsed by pressing through
a silica-gel column (Sep Pack). Evaporation resulted in
the isolation of 19 (0.20 g, 10%); mp 190 ꢀC; MS m/e
328 (M+); LC-MS 329 (M+1); RT 14.85; ELS-purity
99,1%. C14H5)Cl2F2NO2: C, 51.25%; H, 1.54%; N,
1
70.41%; H, 3.98%; N, 5.39%. H NMR (CDCl3, ppm)
8.08 (1H, d t), 7.98 (1H, d), 7.82±7.57 (3H, m), 7.57±7.32
(2H, m), 2.45 (3H, s).
5,8-Dichloro-2-(2-¯uoro-phenyl)-4H-3,1-benzoxazin-4-
one (5). 2-Amino-3,6-dichlorobenzoic acid (0.5 g) and
triethyl:toluene (1:1) (20mL) were mixed. 2-Fluor-
obenzoyl chloride (0.77 g) was added dropwise under
cooling and stirring. The mixture was subsequently heated
to room temperature and stirred until the disappearance
of the starting material was observed on TLC (silica-gel)
using heptane:ethyl acetate (4:1) as eluent. After cooling
on ice, the precipitate was ®ltered o€ and the ®ltrate
evaporated to dryness. The residue was partitioned
between methylene chloride and potassium carbonate
(2N), the organic layers separated, dried over magnesium
sulfate and evaporated to dryness. The raw product was
1
4.27%. Found: C, 51.69%; H, 1.95%; N, 4.02%. H
NMR (DMSO, ppm) 8.08 (1H, d); 7.88±7.68 (2H, m);
7.35 (2H, t).
2-(2,6-Di¯uoro-phenyl)-6-methyl-4H-3,1-benzoxazin-4-one
(20). 2-Amino-5-methylbenzoic acid (0.5 g) and triethyl
amine (10 mL) were mixed in dry toluene (20 mL). 2,6-
Di¯uorobenzoyl chloride (1.28 g) was added dropwise

P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
2101
under cooling and stirring. The mixture was subse-
quently heated to room temperature for 2 days. After
cooling on ice, the precipitate was ®ltered o€ and the
®ltrate evaporated to dryness. The residue was par-
titioned between methylene chloride and potassium
carbonate (2N), the organic layers separated, dried over
magnesium sulfate and evaporated to dryness. The raw
product was puri®ed by column chromatography on
silica-gel using heptane/ethyl acetate as eluent. Yield of
20 (0.85 g, 95%); mp 156 ꢀC; LC-MS 274 (M+1); RT
14.22 min; ELS-purity 100%. C15H9F2NO2: C, 65.69%;
H, 3.30%; N, 5.15%. Found: C, 66.18%; H, 3.33%; N,
5.06%. 1H NMR (DMSO, ppm) 8.03 (1H, d); 7.90±7.60
(3H, m); 7.35 (2H, t); 2.50 (3H, s).
6-Amino-2-(2,6-di¯uoro-phenyl)-4-H-3,1-benzoxazin-4-
one (28). 6-Nitro-2-(2,6-di¯uoro-phenyl)-4-H-3,1-ben-
zoxazin-4-one 24 (50 mg) was dissolved in acetic acid
(5 mL) under N2, PtO2 (2.5 mg) was added and the mix-
ture was hydrogenated with H2. Reaction time 2 h. The
reaction mixture was ®ltered through Hy¯o1 which was
rinsed afterwards with ethyl acetate. The combined
organic phases were evaporated to dryness and subse-
quently treated three times with toluene followed by eva-
poration. The resulting mixture was dissolved in THF and
precipitated with hexane resulting in 28 (12 mg, 20%); LC-
MS 275 (M+1); RT 10.28 min; ELS-purity 96%. 1H
NMR (MeOH, ppm) 7.70±7.50 (1H, m); 7.45 (1H, d); 7.36
(1H, d); 7.26±7.08 (3H, m).
2-(2,6-Di¯uoro-phenyl)-7-nitro-4-H-3,1-benzoxazin-4-one
(22). 4-Nitroanthranilic acid (0.6 g), 2,6-di¯uorobenzoyl
chloride (0.93 mL) and triethyl amine:toluene (1:1)
(18 mL) were reacted by heating to 50ꢀC for 2 days.
Extraction between ethyl acetate (100mL) and HCl (2N,
100 mL), followed by separation of the organic layer,
drying over MgSO4, and ®nally ®ltering and evaporation
gave a crude product which was dissolved in warm THF
(20 mL) and precipitated with hexane twice to get the pure
product 22 (0.42 g, 41%); mp 187 ꢀC; MS 304 (M+). C14
H6F2N2O4: C, 55.28%; H, 1.99%; N, 9.21%. Found: C,
2-(2,6-Di¯uoro-phenyl)-5-methyl-4-H-3,1-benzoxazin-4-
one (30). 2-Amino-6-methylbenzoic acid (0.498 g), 2,6-
di¯uorobenzoyl chloride (0.93 mL) and triethyl amine:
toluene (1:1) (20 mL) were reacted by heating to 50 ꢀC
for 2 h. Extraction between ethyl acetate (100 mL) and
HCl (2N, 100 mL), followed by separation of the
organic layer, drying over MgSO4, and ®nally ®ltering
and evaporation gave a crude product which was
re-dissolved in toluene (10 mL) and precipitated with
hexane (5 mL) resulting in 30 (0.45 g, 50%); mp 156 ꢀC;
MS 273 (M+); LC-MS 274 (M+1); RT 14.34 min; ELS-
purity 100%. C15H9F2NO2: C, 65.94%; H, 3.32%; N,
1
54.85%; H, 2.07%; N, 9.39%. H NMR (DMSO, ppm)
8.38 (3H, s, broad); 7.88±7.68 (1H, m); 7.35( 2H, t).
1
5.13%. Found: C, 65.24%; H, 3.27%; N, 4.85%. H
NMR (DMSO, ppm) 7.90±7.67 (2H, m); 7.55 (2H, t);
7.35 (2H, t).
2-(2,6-Di¯uoro-phenyl)-6-nitro-4-H-3,1-benzoxazin-4-one
(24). 5-Nitroanthranilic acid (0.6 g) was dissolved in
triethyl amine:toluene (1:1) (18 mL) and stirred for
10 min. 2,6-Di¯uorobenzoyl chloride (1.3 g) was slowly
added under stirring resulting in the formation of a pre-
cipitate. The reaction was performed in an N2-atmo-
sphere. After stirring at room temperature for 24 h, the
mixture was extracted with saturated NaHCO3 and ethyl
acetate (20 mL), and the organic layer was separated and
evaporated. The crude product was rinsed by precipita-
tion from hot dioxane, the resulting mass transferred to
a silica-gel cloumn and eluted with dichloromethane,
the isolated fraction was dissolved in hot toluene (2 mL)
and precipitated with hexane, resulting in 24 (0.13 g, 13%);
mp 188 ꢀC; MS 304 (M+). C14H6F2N2O4: C, 55.28%; H,
1.99%; N, 9.21%. Found: C, 52.41%; H, 1.91%; N,
8.55%. 1H NMR (DMSO, ppm) 8.81 (1H, dd); 8.70 (1H,
dd); 7.95 (1H, d); 7.86±7.68 (1H, m); 7.35 (2H, t).
7-Amino-2-(2,6-di¯uoro-phenyl)-4-H-3,1-benzoxazin-4-
one (32). 7-Nitro-2-(2,6-di¯uoro-phenyl)-4-H-3,1-ben-
zoxazin-4-one 22 (0.10 g) was dissolved in acetic acid
(10 mL) under N2, PtO2 (5 mg) was added and the
mixture was hydrogenated with H2 gas. Reaction
time day. The reaction mixture was ®ltered through
Hy¯o1, which was rinsed afterwards with ethyl acetate.
The combined organic phases were evaporated to
dryness and subsequently treated three times with
toluene followed by evaporation. The resulting mixture
was puri®ed on a silica-gel column using dichloro-
methane as eluent. One fraction was collected and
identi®ed as 32 (15 mg, 20%); mp 196 ꢀC; MS 274 (M+).
C14H8F2N2O2: C, 61.32%; H, 2.94%; N, 10.22%.
Found: C, 61.44%; H, 3.45%; N, 8.91%. 1H NMR
(DMSO, ppm) 7.82 (1H, d); 7.70±7.58 (1H, m); 7.35
(2H, t); 6.85 (1H, dd); 6.75 (2H, s, broad, NH); 6.68
(1H, d).
6-Acetamido-(2,6-di¯uoro-phenyl)-4-H-3,1-benzoxazin-4-
one (27). 5-Acetamidoanthranilic acid (0.64 g) and 2,6-
di¯uorobenzoyl chloride (1.3 g) were reacted in triethyl
amine:toluene (1:1) (20 mL) by heating to 50 ꢀC for 1 h.
Ethyl acetate (50 mL) and HCl-solution (1 mL 4N in 50 mL
water) were added resulting in the formation of a pre-
cipitate. The mixture was ®ltered and the residue dissolved
in THF followed by evaporation to dryness, subsequent
dissolution in hot dioxane followed by precipitation with
hexane resulted in colourless crystals of 27 (0.98 g, 95%);
mp 246 ꢀC; MS m/e 316 (M+); LC-MS 317 (M+1); RT
10.32; ELS-purity 100%. C16H10F2N2O3: C, 60.76%; H,
3.19%; N, 8.86%. Found: C, 60.31%; H, 3.19%; N,
8.44%. 1H NMR (DMSO, ppm) 8.53 (1H, dd); 8.04 (1H,
dd); 7.72 (2H, dd); 7.35 (2H, t).
5-Amino-2-(2,6-di¯uoro-phenyl)-4-H-3,1-benzoxazin-4-
one (33). 5-Nitro-2-(2,6-di¯uoro-phenyl)-4-H-3,1-ben-
zoxazin-4-one 18 (0.4 g) was dissolved in acetic acid
(30 mL) under N2, PtO2 (20 mg) was added and the
mixture was hydrogenated with H2. Reaction time 1
day. The reaction mixture was ®ltered through Hy¯o1,
which was rinsed afterwards with ethyl acetate. The
combined organic phases were evaporated to dryness
and subsequently treated 3 times with toluene followed
by evaporation. The resulting mixture was dissolved
in THF and precipitated with hexane giving a crude
product that was puri®ed on a silica-gel column using
dichloromethane as eluent. One fraction was collected

2102
P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
(51 mg), identi®ed as 33; LC-MS 275 (M+1); RT
12.45 min; ELS-purity 100%; mp 197 ꢀC. C14H8F2N2O2:
C, 61.32%; H, 2.94%; N, 10.22%. Found: C, 61.55%;
H, 3.04%; N, 9.73%. H NMR (CDCl3, ppm) 7.55±
7.38 (2H, m); 7.05 (2H, t); 6.90 (1H, d); 6.70 (1H, d);
6.08 (2H, s, broad, NH).
6-Methyl-2-thiophen-2-yl-4H-3,1-benzoxazin-4-one (59).
2-Amino-5-methylbenzoic acid (1.5 g) and triethyl
amine:toluene (1:1) (80 mL) were mixed with 2-thie-
nylcarbonyl chloride (3.18 g). The mixture was stirred
at room temperature for 5 h and the precipitate was
®ltered o€ and the ®ltrate evaporated to dryness. The
residue was partitioned between methylene chloride and
potassium carbonate (2N), the organic layers separated,
dried over magnesium sulfate and evaporated to dry-
ness. The raw product was puri®ed by dissolution in
dioxan followed by precipitation with hexane resulting
in 59 (2.0 g, 85%); mp 156 ꢀC; MS 243 (M+); LC-MS
244 (M+1); RT 1.61 min; ELS-purity 100%.
C13H9NO2S: C, 64.18%; H, 3.73%; N, 5.76%. Found:
1
2-(2,3-Dichloro-phenyl)-6-methyl-4-H-3,1-benzoxazin-4-
one (36). 2-Amino-5-methylbenzoic acid (0.5 g) and dry
pyridine (20 mL) were mixed with 2,3-dichlorobenzoyl
chloride (1.52 g) in dry toluene (20 mL) at room tem-
perature. Subsequently, the mixture was heated to 80 ꢀC
for 1 h. The mixture was evaporated to dryness, and the
residue was partitioned between methylene chloride and
potassium carbonate (2N), the organic layers separated,
dried over magnesium sulfate and evaporated to dry-
ness. The raw product was recrystallised from toluene
yielding 36 (1.0 g, 99%); mp 166 ꢀC. C15H9Cl2NO2: C,
58.85%; H, 2.96%; N, 4.58%. Found: C, 58.88%; H,
1
C, 64.20%; H, 3.72%; N, 5.56%. H NMR (DMSO,
ppm) 7.95 (1H, dd); 7.86 (2H, s, broad); 7.70 (1H, dd);
7.50 (1H, dd); 7.27 (1H, t); 2.42 (3H, s). 13C NMR
(DMSO, ppm) 159.24, 153.28, 145.02, 139.08, 138.73,
134.65, 134.11, 132.21, 129.58, 128.53, 127.12, 117.12,
21.54.
1
2.96%; N, 4.45%. H NMR (CDCl3, ppm) 7.55±7.38
(2H, m); 7.05 (2H, t); 6.90 (1H, d); 6.70 (1H, d); 6.08
(2H, s, broad, NH).
2-(3-Bromothiophen-2-yl)-6-methyl-4H-3,1-benzoxazin-
4-one (60). 2-Amino-5-methylbenzoic acid (0.5 g) and
triethyl amine (10 mL) were mixed in dry toluene
(10 mL). 3-Bromo-2-thienylcarbonyl chloride (0.72 g)
was added dropwise under cooling and stirring, the
mixture was further stirred at room temperature for
2 days and subsequently the mixture was kept at 4 ꢀC
for 30 days. The precipitate was ®ltered o€ and the
®ltrate evaporated to dryness. The residue was par-
titioned between methylene chloride and potassium
carbonate (2N), the organic layers separated, dried over
magnesium sulfate and evaporated to dryness. The raw
product was puri®ed by column chromatography on
silica-gel usingꢀ toluene as eluent. Yield of 60 (0.45 g,
48%); mp 159 C; MS 321, 323 (M+) (Br-isotope pat-
tern). NMR (CDCl3, ppm) 7.99 (1H, dd); 7.66±7.50
(2H, m); 7.49 (1H, d); 7.12 (1H, d); 2.47 (3H, s). 13C
NMR (CDCl3, ppm) 159.13, 152.37, 144.93, 139.31,
138.28, 134.15, 131.18, 128.88, 128.74, 127.36, 116.91,
115.14, 21.75.
2-(2,6-Dichloro-phenyl)-6-methyl-4-H-3,1-benzoxazin-4-
one (38). 2-Amino-5-methylbenzoic acid (0.5 g) and
triethyl amine (10 mL) were mixed in dry toluene
(20 mL). 2,6-Dichlorobenzoyl chloride (1.52 g) was
added dropwise under cooling and stirring. The mixture
was subsequently heated to room temperature for 24 h,
then heated to 80 ꢀC for 1 h. After cooling on ice, the
precipitate was ®ltered o€ and the ®ltrate evaporated to
dryness. The residue was partitioned between methylene
chloride and potassium carbonate (2N), the organic
layers separated, dried over magnesium sulfate and
evaporated to dryness. The raw product was puri®ed by
column chromatography on silica-gel using heptane/
ethyl acetate as eluent giving 38 (0.12 g, 11%); mp
175 ꢀC; MS 305, 307 (M+) (Cl isotope pattern); LC-MS
306, 308 (M+1); RT 15.75 min; ELS-purity 98%.
C15H9Cl2NO2: C, 58.85%; H, 2.96%; N, 4.58%. Found:
1
C, 57.96%; H, 2.90%; N, 4.09%. H NMR (CDCl3,
ppm) 8.10 (1H, s, broad); 7.72±7.58 (2H, m); 7.48±7.31
(3H, m); 2.54 (3H, s).
The following compounds were prepared using solution
phase parallel synthesis following the general description
below:
6-Methyl-2-(2-tri¯uoromethoxy-phenyl)-4H-3,1-benzoxa-
zin-4-one (44). 2-Amino-5-methylbenzoic acid (0.5 g)
and triethyl amine (10 mL) were mixed in dry toluene
(20 mL). 2-Tri¯uoromethoxybenzoyl chloride (1.64 g)
was added dropwise under cooling and stirring. The
mixture was subsequently heated to room temperature
for 2 days. After cooling on ice, the precipitate was
®ltered o€ and the ®ltrate evaporated to dryness. The
residue was partitioned between methylene chloride
and potassium carbonate (2N), the organic layers
separated, dried over magnesium sulfate and evaporated
to dryness. The raw product was puri®ed by column
chromatography on silica-gel using heptane/ethyl
acetate as eluent giving 44 (1.0 g, 98%); mp 70±80 ꢀC;
MS 321 (M+). C16H10F3NO3: C, 59.82%; H, 3.14%;
N, 4.36%. Found: C, 61.12%; H, 3.53%; N, 4.18%.
1H NMR (CDCl3, ppm) 8.10 (1H, dd); 8.04 (1H,
d, broad); 7.70±7.52 (3H, m); 7.49±7.36 (2H, m); 2.50
(3H, s).
The appropriately substituted 2-aminobenzoic acid
(30 mg) was dissolved in toluene (300 uL), triethyl amine
(300 uL) was added and the mixture shaken for 10 min
at room temperature. Appropriately substituted benzoic
acid chloride (2.2 equivalent) was added and the mix-
ture shaken overnight at room temperature. Ethyl ace-
tate (1 mL) and HCl (0.1N, 1 mL) were added followed
by shaking for 30 min. The organic layer was separated
and saturated, NaHCO3 (0.5 mL) added followed by
shaking for 2 min. The organic layer was separated and
®ltered through a silica-gel column (Sep-Pack plus), the
column was washed with dichloromethane (3 mL) and
the collected organic phases evaporated. The identity
and purity was checked by LC-MS using a PE Sciex
API 100 LC/MS system using Waters 3 mmÂ150 mm
3.5 u C-18 symmetry column and positive ion-spray
with a ¯ow rate at 20 uL/min (Table 4).

P. Jakobsen et al. / Bioorg. Med. Chem. 8 (2000) 2095±2103
6. Fenton, G. J. Med. Chem. 1989, 32, 265.
7. Krantz, A. J. Med. Chem. 1990, 33, 464.
2103
References
8. Rose, U. J. Heterocyclic Chem. 1991, 28, 2005.
9. Gilmore, J. L. Bioorg. Med. Chem. Lett. 1996, 6, 679.
1. Rapaport, S. I.; Rao, L. V. M. Thromb. Haemost. 1995, 74,
7.
10. Jarvest, R. L. Bioorg. Med. Chem. Lett. 1996, 6, 2463.
11. Hays, S. J. J. Med. Chem. 1998, 41, 1060.
2. Harker, L. A.; Hansen, S. R.; Kelly, A. B. Thromb. Hae-
most. 1995, 74, 464.
12. Abood, N. A. Bioorg. Med. Chem. Lett. 1997, 7, 2105.
13. Gordeev, M.; Patel, D. WO patent, 1998, 98, 33783.
14. Neumann, U.; Gutschow, M. Bioorg. Chem. 1995, 23, 72.
3. Papadopopoulos, E. P. Heterocycles 1982, 6, 1039.
4. Hamprecht, G.; Wuerzer, B. US patent 1982, 4, 315, 766.
5. Johnson, J. L.; Pattison, I. J. Heterocycl. Chem. 1986, 23, 249.