Design, synthesis and SAR of novel ethylenediamine and phenylenediamine derivatives as factor Xa inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.01.132

We previously reported on a series of cyclohexanediamine derivatives as highly potent factor Xa inhibitors. Herein, we describe the modification of the spacer moiety to discover an alternative scaffold. Ethylenediamine derivatives possessing a substituent at the C1 position in S configuration and phenylenediamine derivatives possessing a substituent at the C5 position demonstrated moderate to strong anti-fXa activity. Further SAR studies led to the identification of compound 30h which showed both good in vitro activity (fXa IC₅₀ = 2.2 nM, PTCT2 = 3.9 μM) and in vivo antithrombotic efficacy.

Full text of DOI:10.1016/j.bmcl.2011.01.132

Bioorganic & Medicinal Chemistry Letters 21 (2011) 2133–2140
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and SAR of novel ethylenediamine and phenylenediamine
derivatives as factor Xa inhibitors
Kenji Yoshikawa a, , Toshiharu Yoshino a, Yoshihiro Yokomizo b, Kouichi Uoto a, Hiroyuki Naito a,
⇑
Katsuhiro Kawakami a, Akiyoshi Mochizuki a, Tsutomu Nagata b, Makoto Suzuki b, Hideyuki Kanno c,
Makoto Takemura a, Toshiharu Ohta a
a R&D Division, Daiichi Sankyo Co., Ltd, 1-16-13, Kita-Kasai, Edogawa-ku, Tokyo 134-8630, Japan
b R&D Division, Daiichi Sankyo Co., Ltd, 1-2-58, Hiromachi, Shinagawa-ku, Tokyo 140-8710, Japan
c Daiichi Sankyo RD Associe Co., Ltd, 1-16-13, Kita-Kasai, Edogawa-ku, Tokyo 134-8630, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We previously reported on a series of cyclohexanediamine derivatives as highly potent factor Xa inhib-
itors. Herein, we describe the modification of the spacer moiety to discover an alternative scaffold. Eth-
Received 31 October 2010
Revised 27 January 2011
Accepted 28 January 2011
Available online 1 February 2011
ylenediamine derivatives possessing
a substituent at the C1 position in S configuration and
phenylenediamine derivatives possessing a substituent at the C5 position demonstrated moderate to
strong anti-fXa activity. Further SAR studies led to the identification of compound 30h which showed
both good in vitro activity (fXa IC50 = 2.2 nM, PTCT2 = 3.9
l
M) and in vivo antithrombotic efficacy.
Keywords:
Anticoagulant
Ó 2011 Elsevier Ltd. All rights reserved.
Factor Xa inhibitor
Phenylenediamine
Ethylenediamine
Thromboembolic disorders are the leading cause of morbidity
and mortality in developed countries.1 These include acute myo-
cardial infarction (MI) or ischemic stroke in arterial circulation,
and deep vein thrombosis (DVT) or pulmonary embolism (PE) in
venous circulation. Anticoagulants currently in clinical use for
the prevention and treatment of these diseases include unfraction-
ated heparin, low molecular weight heparins (LMWHs), the
indirect fXa inhibitor fondaparinux, and the direct thrombin inhib-
itors (DTIs) argatroban, bivalirudin, and hirudin. Although the effi-
cacies of these drugs have been confirmed, all require parenteral
administration. For oral administration, vitamin K antagonists
(VKAs) such as warfarin have been the mainstay of anticoagulants
therapy for more than 60 years. However, VKAs have substantial
drawbacks (slow onset and offset of action, inconvenience of fre-
quent monitoring, numerous drug and food interactions, and so
on) that limit their clinical use.2 Thus, a novel anticoagulant that
can be administered more conveniently and safely has been de-
sired. Factor Xa (fXa) is a key serine protease located at the conver-
gence of the intrinsic and extrinsic pathways catalyzing the
conversion of prothrombin to thrombin. It is anticipated that selec-
tive inhibition of fXa will provide an antithrombotic effect by
blocking the amplified formation of thrombin without compromis-
ing normal hemostasis.3 Therefore, fXa is a particularly attractive
target and extensive clinical trials are now ongoing for a number
of clinical candidates.4–9
We have previously reported on the orally active potent fXa
inhibitor compound A,10 which was chosen as a clinical candidate
(Fig. 1). The X-ray structure analysis revealed the binding mode
of compound A in which the indole moiety fitted into the S1 site,
5-methyl-4,5,6,7-tetrahydrothiazolo[5,4-c]pyridine moiety fitted
into the S4 site, and the cyclohexane diamine acts as a spacer con-
necting these two key elements. The result of the SAR study of the
S1 moiety of compound A has been disclosed recently.11,12 In addi-
tion to the S1 modification, we concurrently conducted an explor-
atory study of the spacer moiety aiming to discover an alternative
O
N
S4 moiety
S
Cl
O
R
Cl
O
S
N
N
H
N
O
HN
N
H
N
N
HN
N
N
H
O
H
O
R
S1 moiety
Cl
Compound A
anti-fXa activity IC50 2.3 nM 13
S
N
H
PTCT2 in plasma
Solubility pH 1.2
pH 6.8
0.33 µM 14
N
HN
N
N
µg/ml
µg/ml
124
14
2.8
H
O
Log D (pH 6.8)
⇑
Corresponding author. Tel.: +81 3 5696 7436.
E-mail address: yoshikawa.kenji.t6@daiichisankyo.co.jp (K. Yoshikawa).
Figure 1. Structure of compound A and design for alternative spacer scaffolds.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.01.132

2134
K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
spacer scaffold. In this report, the SAR of the spacer moiety will be
discussed. We designed ethylenediamine and phenylenediamine
derivatives as alternative scaffolds which have less complexity
(Fig. 1).
In the synthesis of unsubstituted ethylenediamine and phenyl-
enediamine derivatives, mono-protected diamines (1a–c, and 10)
were acylated with carboxylate 2 or carboxylic acid 4 giving
mono-acylated intermediates (3a, b, 5a, b, and 11). Then, the Boc
group was removed and the resulting amine was acylated with 4
or 2 to afford diamides 6–9, and 12 (Scheme 1).
Mono-substituted ethylenediamine derivatives were synthe-
sized as depicted in Scheme 2. Both 1-substituted derivatives
(19, 20, and 21) and 2-substituted derivatives (22 and 23) were
synthesized via key intermediate 16, which was prepared by the
reduction of carboxylic acid 14, mesylation of the resulting alcohol
15, and substitution with azide.
Phenylenediamine derivatives having a substituent at the 5-po-
sition were synthesized as outlined in Scheme 3. Acylation of
5-substituted-1-nitro-2-aminobenzenes (26a–j) by 1-benzenesul-
fonyl-5-chloroindole-2-carbonyl chloride (25), followed by
O
S
2
S
N
N
N
Cl
(a)
O
CO2Li
H
(c), (b)
(c), (a)
N
R1
N
N
N
Boc
S
R2
N
R1
N
R1
3a R1 = R2 = H
3b R1 = H, R2 = Me
Cl
Boc
R2
N
N
N
N
R2
H
O
1a R1 = R2 = H
1b R1 = H, R2 = Me
1c R1 = R2 = Me
6 R1 = R2 = H
(b)
HO2C
Boc
7 R1 = H, R2 = Me
8 R1 = Me, R2 = H
9 R1 = R2 = Me
N
R1
Cl
N
N
R2
N
H
H
O
4
5a R1 = Me, R2 = H
5b R1 = R2 = Me
Cl
Cl
(b)
(d), (a)
O
Boc
Boc
S
N
N
N
H
H
H
NH2
HN
N
N
HN
N
N
10
H
H
11
12
O
O
Scheme 1. Reagents and conditions: (a) lithium 5-methyl-4,5,6,7-tetrahydrothiazolo[5,4-c]pyridine-2-carboxylate (2), EDCÁHCl, HOBt, DMF, rt, 1–3 days; (b) 5-chloroindole-
2-carboxylic acid (4), EDCÁHCl, HOBt, DMF, rt, 1–3 days; (c) satd HCl, MeOH, or satd HCl, EtOH, rt, 1 h; (d) TFA, DCM, rt, 1 h.
O
O
O
O
O
OBn
O
(a)
OBn
O
(b)
Cl
OBn
OMe
OMe
(c), (d), (e)
(c), (f)
Boc
H2N
Boc
Boc
N
Boc
N
N
N
OH
H
H
OH
H
OH
H
N3
16a (S)
16b (R)
HN
17a (S)
17b (R)
N
13a (S)
14a (S)
14b (R)
H
15a (S)
15b (R)
O
(c), (g)
MeO
O
O
O
Cl
S
O
R
1
N
R
O
Cl
S
H
O
S
N
N
HN
2
N
(h), (g)
(h), (f)
17a (S)
17b (R)
Boc
18a (S)
18b (R)
H
18a (S)
18b (R)
N
N
HN
N
N
H
N
N
HN
H
O
N
19a (S)
19b (R)
H
O
R = OMe
22a (S)
22b (R)
(i)
(j)
R = OMe
20a (S)
20b (R)
R = OH
(i)
23a (S)
23b (R)
R = OH
21a (S)
R = NMe2
Scheme 2. Reagents and conditions: (a) Boc2O, TEA, dioxane, H2O, 0 °C to rt, overnight; (b) (i) ClCO2iPr, TEA, THF, 0 °C, 30–60 min; (ii) NaBH4, H2O, rt, 1 h; (c) 10% Pd/C, H2,
MeOH or THF, rt, 1–3 h; (d) TMSCHN2, MeOH, THF, 0 °C, 1 h; (e) (i) MsCl, TEA, DCM, À78 to 0 °C, 1 h; (ii) NaN3, DMF, 65 °C, 1 h; (f) 4, EDCÁHCl, HOBt, DIPEA, DMF, rt, overnight;
(g) 2, EDCÁHCl, HOBt, DIPEA, DMF, rt, overnight; (h) HCl, 1,4-dioxane, MeOH, rt, 1 h; (i) LiOH, THF, H2O, rt, 1 h; (j) Me2NHÁHCl, EDCÁHCl, HOBt, TEA, DMF, rt, overnight.
Cl
Cl
(a)
HO
Cl
N
N
SO2Ph
SO2Ph
O
O
24
25
R
R
R
5
2
R
Cl
Cl
O
Cl
(b)
(c)
1
(d)
S
N
O2N
O2N
H
H2N
N
N
HN
HN
N
NH2
N
HN
H
N
O
SO2Ph
27a-j
O
26a-j
28a-j
SO2Ph
O
29a-j
R = ester, amide
(e)
30c,d,f-j R = carboxylic acid
Scheme 3. Reagents and conditions: (a) SOCl2, DMF (cat.), CHCl3, reflux, 30–60 min; (b) 1-benzenesulfonyl-5-chloroindole-2-carbosnylchloride (25), pyridine, TEA, DMAP,
DCM, 0–60 °C, 1–6 days; (c) FeCl3, Zn, DMF, H2O, 100 °C, 10–60 min; (d) 2, EDCÁHCl, HOBt, DMF, rt–70 °C, 1–6 days; (e) LiOH or NaOH, EtOH or THF, H2O, 0 °C–rt, 4–36 h.

K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
2135
COOH
NHAc
CO2H
CONMe2
NHAc
CO2H
CONMe2
CO2Et
(a)
(c)
(b)
(d)
O2N
31
O2N
32
O2N
26b
O2N
O2N
26c
NHAc
33
NH2
NHAc
(i),(b)
NH2
34
Me CO2Et
CO2Et
CO2Et
Me CO2Et
Me CO2Et
(e)
(f)
(g),(h)
O2N
26d
NPhth
NH2
NPhth
36
NH2
NHAc
38
35
37
CO2Et
Me
Me
CO2Et
Me
Me
CO2Et
CO2Et
(j)
CO2Et
(b),(e)
(i)
(g)
O2N
40
O2N
42
O2N
O2N
26e
NHAc
39
NHAc
NPhth
NPhth
NH2
41
R1 R2
CO2Me
OH
O
CO2Me
R1 R2
26f R1 = R2 = H
O
(l)
(k)
26i R1 = R2 = Me
26j R1 = R2= - (CH2)2-
26g R1 = H, R2 = Me
26h R1 = R2 = Me
O2N
NH2
43
O2N
O2N
NH2
NH2
Scheme 4. Reagents and conditions: (a) (i) SOCl2, benzene, reflux, 3 h; (ii) Me2NH, THF, rt, overnight; (b) 1 N-HCl, EtOH, reflux, 2–5 h; (c) KNO3, H2SO4, À10 °C, 1 h; (d) satd
HCl, EtOH, reflux, 4 h; (e) phthalic anhydride, AcOH, reflux, 18 h; (f) (i) NaH, DMF, 0 °C to rt, 1 h; (ii) MeI, 0 °C to rt, 2 h; (g) H2NNH2ÁH2O, EtOH, rt, 18 h; (h) Ac2O, 130 °C, 1 h; (i)
fuming HNO3, AcOH, Ac2O, 0 °C to rt, 2–5 h; (j) (i) LiHMDS, THF, À78 °C, 30 min; (ii) MeI, THF, rt, 3 h; (k) BrCR1R2CO2Me, K2CO3, DMF, 70 °C, overnight; (l) BrCH2CR1R2CO2Me,
K2CO3, DMF, 60–70 °C, 20–24 h.
reduction of the nitro group and condensation with 2 gave com-
pounds 29a–j, which were hydrolyzed to afford carboxylic acids
(30c, d, f–j). In the case of 29a and 29e, hydrolysis in the aqueous
alkaline solution failed to yield the corresponding carboxylic acids.
The syntheses of 5-substituted-1-nitro-2-aminobenzenes
(26b–j) are shown in Scheme 4. Phenylacetic acid derivatives
(26c–e) were synthesized via the introduction of the nitro group
to the corresponding 4-(acetamido)phenylacetic acid derivatives
(33, 38, or 39). Phenoxyacetic acid derivatives (26f–h) and phen-
oxypropionic acid derivatives (26i and 26j) were synthesized by
alkylation of phenol 43.
Diamine 44 was selectively acylated with carboxylic acid 4 on
the amine meta to the ester, due to the deactivation of para-amine
by the electron withdrawing group. Then another amino group was
acylated with 2 to provide 4-substituted phenylenediamine 29k
(Scheme 5). The hydroxyl group of 43 was alkylated after the pro-
tection of the amino group. Compound 48 was reduced to give
mono-protected diamine, which was converted to diamide 29l.
The anti-fXa activity (IC50), anticoagulant activity (PTCT2), and
solubility of ethylenediamine derivatives are shown in Table 1.
Unsubstituted ethylenediamine (6) showed moderate anti-fXa
activity (IC50 = 221 nM). N-Methyl ethylenediamine derivatives
had significantly decreased activity (7, 8, and 9). Then a substituent
was introduced on the ethylene chain. Of the four methoxycarbon-
ylmethyl derivatives (19a, 19b, 22a, and 22b), only C1-substituted
derivative in S configuration (19a) exhibited better anti-fXa activ-
ity than compound 6. As seen in the X-ray crystal structure of com-
pound A, C2 carbon is located close to the amino acid (Gln192) of
fXa (Fig. 2). Therefore, we considered that the introduction of a rel-
atively large methoxycarbonylmethyl group at the C2 position
caused steric repulsion, and thus resulted in decreased activity
(22a and 22b). In the case of substitution at the C1 position, there
is not enough space between the C1 carbon and enzyme if the sub-
stituent is in the R configuration (19b). On the other hand, if the
substituent is in the S configuration, it is supposed to be located
towards the outside of the enzyme without causing any steric
4
CO2Et
Cl
CO2Et
CO2Et
Cl
O
(a)
1
(b)
S
N
H2N
H2N
44
OH
2
H
N
N
HN
HN
NH2
N
H
N
H
45
O
29k
(e), (f)
O
OTBS
CO2Me
OTBS
O
(d)
(c)
(g), (h)
O2N
43
O2N
O2N
O2N
48
NH2
NH2
NHBoc
NHBoc
46
47
Me
CO2
CO2Me
O
O
Cl
Cl
O
(i), (b)
S
BocHN
N
H
N
N
HN
HN
N
N
49
H
SO2Ph
29l
O
O
Scheme 5. Reagents and conditions: (a) 4, EDCÁHCl, HOBt, DMF, rt, 20 h; (b) 2, EDCÁHCl, HOBt, DMF, 50–60 °C, two days; (c) TBSCl, imidazole, DMF, 0 °C to rt, 1 h; (d) (i)
NaHMDS, THF, rt, 30 min; (ii) (Boc)2O, THF, rt, overnight; (e) TBAF, THF, 0 °C, 1 h; (f) BrCH2CO2Me, K2CO3, DMF, 70 °C, overnight; (g) 10% Pd/C, H2, MeOH, rt, 3 h; (h) 25,
pyridine, DCM, 0 °C to rt, overnight; (i) TFA, anisole, DCM, 0 °C, 2 h.

2136
K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
Table 1
Assay results of ethylenediamine derivatives
Cl
O
S
Spacer
N
N
N
H
O
Compd
Spacer
R
In vitro activity
Solubility (
pH 1.2
l
g/ml)
pH 6.8
Ex vivo anti-fXa activityc (%)
a
fXa IC50 (nM)
PTCT2b (M)
O
N
5
6
4
A
2.3
0.33
124
14
62
3
N
H
1
HN
2
N
H
6
7
8
9
221
1600
4100
3945
18
—
—
—
—
—
—
—
—
—
—
—
—
—
HN
N
N
H
N
—
HN
N
—
N
O
1
19a
20a
21a
OMe
OH
NMe2
70
200
47
6.5
—
3.2
>1000
900
>1000
135
432
140
1
1
15
R
(S )
2
N
H
HN
19b
20b
O
OMe
OH
1200
2100
—
—
813
938
87
544
—
—
R
(R )
N
H
HN
R
22a
23a
O
OMe
OH
770
460
—
—
536
>1000
93
39
—
—
N
H
(S )
HN
R
22b
23b
O
OMe
OH
970
610
—
—
237
786
93
30
—
—
N
H
(R )
HN
a
b
c
Anti-fXa activity (IC50) was measured as described in Ref. 13.
Anticoagulant activity (PTCT2) was evaluated as described in Ref. 14.
The maximum anti-fXa activity in plasma after oral administration is shown. The assay method is described in Ref. 15.
case, demonstrating the utility of this spacer in terms of improving
physicochemical property.
The assay results of phenylenediamine derivatives are shown in
Table 2. Even unsubstituted derivative 12 exhibited very strong
anti-fXa activity (IC50 = 7.4 nM). A comparison of 5-substituted
derivatives (29a, f) with corresponding 4-substituted derivatives
(29k, l) demonstrates that substitution at the 5-position is prefer-
able. There seemed to be two major problems regarding these
phenylenediamine compounds. One is their weak anticoagulant
activity and the other is low solubility, especially in the neutral
pH region. To overcome these problems, we continued the modifi-
cation of a substituent at the 5-position.
As shown in Table 3, esters 29c–j showed moderate to strong
anti-fXa activity. However, their anticoagulant activity was still
not strong. Those compounds showed notably high c log P values
(3.50–4.84). As it is known that higher lipophilicity of compounds
results in higher protein binding and weaker anticoagulant activ-
ity,10 we envisioned that anticoagulant activity would be improved
by lowering the lipophilicity. Carboxylic acid derivatives (30c, d,
f–j) and carbamoyl derivative (29b) were designed for this aim.
The c log P values were significantly lower (0.71–2.63) than the
parent compounds. The carboxylic acids (30c, d, f–j) showed a
drastic increase in anti-fXa activity and also demonstrated distinct
anticoagulant activity. This is in contrast to unsubstituted
derivative 12 or ester 29f, which showed only weak anticoagulant
Figure 2. X-ray crystal structure of compound A in fXa (PDB code:2EI8).10
repulsion (19a). Similar results were also observed for the anti-fXa
activity of carboxylic acid derivatives (20a, 20b, 23a, and 23b).
Based on this result, compound 21a, which possesses the
dimethylcarbamoylmethyl group at the C1 position in the S config-
uration, was synthesized. It showed modest anti-fXa activity
(IC50 = 47 nM) and anticoagulant activity (PTCT2 = 3.2
lM). The
solubility of ethylenediamine derivative was improved in every

K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
2137
Table 2
Comparison of 4- and 5-substituted phenylenediamine derivatives
Cl
O
S
Spacer
N
N
N
H
O
Compd
Spacer
In vitro activity
Solubility (lg/ml)
c log Pc
PTCT2b
(lM)
pH 1.2
pH 6.8
a
fXa IC50 (nM)
5
6
4
3
12
7.4
>5 (Â1.12)d
—
—
3.71
N
1
HN
2
H
CO2Et
29a
29k
29f
690
>20 (Â1.07)
>20 (Â1.06)
>20 (Â1.68)
2.7
1.1
51
<1
<1
<1
4.67
4.88
3.50
N
H
HN
CO2Et
CO2Me
>10,000
5.6
N
H
HN
O
N
H
HN
MeO2C
O
29l
110
>20 (Â1.14)
8
<1
3.57
N
H
HN
a,b
Refers to Table 1.
c
c log P values were calculated using c log P daylight version 4.83, which was supplied by Daylight Chemical Information Systems, Inc.
Figure in parenthesis represents the prolongation ratio of prothrombin time (PT) at the highest concentration of the compound.
d
Table 3
Assay results of 5-substituted phenylenediamine derivatives
Cl
O
S
Spacer
N
N
N
H
O
c
Compd
Spacer
R
In vitro activity
Solubility (
pH 1.2
l
g/ml)
c log P
Ex vivo anti-fXa
activity (%)
d
a
fXa IC50 (nM)
PTCT2b
(
lM)
pH 6.8
O
N
N
A
2.3
0.33
124
NDf
14
3.00
62
1
H
HN
O
N
29b
7.7
4.0
NDf
2.63
N
H
HN
29c
30c
OEt
OH
170
1.9
>20 (Â1.10)e
250
22
<1
16
4.13
0.71
0
1
O
3.7
R
R
R
N
H
NH
O
29d
30d
OEt
OH
65
1.1
>20 (Â1.71)
—
—
—
—
4.44
1.24
23
5
4.8
N
H
HN
O
29e
OEt
100
>20 (Â1.43)
—
—
4.84
8
N
H
HN
(continued on next page)

2138
K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
Table 3 (continued)
c
Compd
Spacer
R
In vitro activity
Solubility (
pH 1.2
l
g/ml)
c log P
Ex vivo anti-fXa
activity (%)
d
a
fXa IC50 (nM)
PTCT2b
(
lM)
pH 6.8
29f
30f
OMe
OH
5.6
2.5
>20 (Â1.68)
51
12
<1
<1
3.50
0.75
32
3
R
O
3.1
O
N
H
HN
O
29g
30g
OMe
OH
29
3.5
6.2
4.8
61
256
<1
825
3.81
1.21
9
0
R
O
N
H
HN
O
29h
30h
OMe
OH
72
2.2
10
3.9
37
71
<1
362
4.12
1.37
6
0
R
O
O
N
H
HN
O
29i
30i
OMe
OH
120
1.9
>20 (Â1.38)
11
4
<1
9
4.62
1.72
9
5
7.2
R
R
N
H
HN
O
29j
30j
OMe
OH
100
2.7
12
5.6
22
4
<1
1
4.25
1.36
13
4
O
N
H
HN
a–e
Refers to Tables 1 and 2.
The measurement of solubility failed due to insufficient solubility of the compound.
f
activity despite their strong anti-fXa activity. In addition, com-
pounds 30g and 30h demonstrated significantly improved solubil-
ity in the neutral pH region. Carbamoyl derivative 29b also showed
increased anti-fXa and anticoagulant activity compared to ester
29a. However, the physicochemical property of 29b was problem-
atic (the measurement of solubility failed due to its insufficient
solubility).
X-ray crystal structure of compound 12 in fXa was obtained
with 2.0 Å resolution. Figure 3 shows the superposition of the crys-
tal structures for compound 12 in fXa (green) and compound A in
fXa (white). The S1 moiety and S4 moiety of both compounds
overlap very well, while the positions of the spacer moieties differ
from each other. Another distinct feature lies in the conformation
of the amide connecting the S1 moiety and the spacer moiety. In
compound A, indole nitrogen and amide nitrogen are located in
the antiperiplanar position, while in compound 12 indole nitrogen
and amide oxygen are located in the antiperiplanar position. As the
result of those changes, two essential pharmacophores of com-
pound 12 were placed in appropriate positions.
The key interactions of compound 12 with fXa are similar to
those of compound A. As shown in Figure 4, there are hydrophobic
interaction in the S4 site, hydrogen bonding between indole NH
and carbonyl oxygen of Gly218 (2.88 Å), and chloro-aromatic inter-
action with Tyr228 in the S1 site (The distance between Tyr ring
centroid and Cl is 3.51 Å). There are additional interactions that
are characteristic to compound 12. One is the interaction of the
benzene moiety with the subsite known as S1b.17 The benzene
C3 atom and sulfur atom of Cys220 are in close contact (the dis-
tance between C3 and S is 3.71 Å). The other interactions are
hydrogen bondings of the amide NH of compound 12 to the back-
bone carbonyls of Gly218 (3.23 Å) and Gly216 (3.33 Å).18 Those
hydrogen bondings could be formed because of the inversion of
amide configuration as described above (Fig. 3). These interactions
presumably contribute to the strong anti-fXa activity of phenyl-
enediamine derivatives.
This binding mode also explains the preference of 5-substitu-
tion of phenylenediamine compounds. As seen in Figure 4,
Glu147 exists in the direction of the 4-position of the phenylenedi-
amine spacer, while the 5-position is exposed to the solvent. It is
presumed that the introduction of a large substituent at the 4-po-
sition caused a steric repulsion and resulted in the decrease of
activity.
Several potent compounds were orally administered to rats at a
dose of 10 mg/kg, and the anti-fXa activity in plasma was mea-
sured at 0.5, 1, 2, and 4 h after administration. The maximum
Figure 3. Superposition of the crystal structures of phenylenediamine derivative 12
in fXa (green) and compound A in fXa (white).16

K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
2139
Figure 4. (a) Important amino acids and key interactions with compound 12. (b) The binding mode of compound 12.
Table 4
Metabolic stability, permeability, and log D value
a
Compd
Metabolic stability Vini
(nmol/min/mg)
Permeability
AT ratiob
log D
A
30h
0.052
À0.007
>30c
2.8
1.3
0.24d
a
Metabolic stability was evaluated by the initial velocity of disappearance (Vini
)
in human liver microsomes.11
b
Permeability was measured using Caco-2 monolayers. Relative permeability is
shown as a ratio compared to atenolol.11
c
pH 6.0.
pH 7.4.
d
anti-fXa activity among them is shown as ex vivo anti-fXa activity
(%) in Tables 1 and 3.
Figure 5. Ex vivo anti-fXa activity after intravenous administration of compounds
30f and 30h.
Among the ethylenediamine derivatives tested, only compound
21a exhibited weak ex vivo anti-fXa activity (15%). With regard to
phenylenediamine derivatives, compound 29f exhibited modest
ex vivo anti-fXa activity (32%) that was about half of the activity
of compound A. Unfortunately, phenylenediamine derivatives pos-
sessing carboxylic acids (30c, d, f–j) or dimethylamide (29b)
showed almost no activity in ex vivo assay, even though these
compounds demonstrated strong anti-fXa and anticoagulant activ-
ity in vitro. Compounds 30g and 30h, which possess sufficient sol-
ubility, also did not exhibit ex vivo anti-fXa activity. As shown in
Table 4, compound 30h had high metabolic stability, but its perme-
ability was much lower than that of compound A. Reducing lipo-
philicity was required to improve anticoagulant activity, but it
impaired the permeability at the same time. Achieving a good bal-
ance between strong anticoagulant activity and oral activity in this
series of compounds remained a challenge.
Compounds 30f and 30h were administered intravenously to
rats at a dose of 0.4 mg/kg. These compounds exhibited strong
Figure 6. Antithrombotic effect of compound 30h in a rat DIC model. (a) Inhibition of the platelet consumption. (b) Inhibition of the formation of thrombin–antithrombin III
complex (TAT).

2140
K. Yoshikawa et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2133–2140
Sunnyvale, CA, USA) at room temperature, and the reaction velocity (OD/
min) was obtained. Anti-fXa activity (inhibition %) was calculated as follows:
Anti-fXa activity = {1 À [(OD/min) of sample/(OD/min) of control]} Â 100. The
IC50 value was obtained by plotting the inhibitor concentration against the
anti-fXa activity.
ex vivo anti-fXa activity, showing maximum inhibitory activity of
82% and 91%, respectively (Fig. 5).
The antithrombotic activity of phenylenediamine derivative
was tested in a rat disseminated intravascular coagulation (DIC)
model.19 Compound 30h was administered intravenously to rats
after the injection of tissue factor (TF) and demonstrated dose-
dependent antithrombotic activity (Fig. 6).
In summary, we have identified novel spacer scaffolds for our S1
and S4 pharmacophores. Among them, ethylenediamine spacer is
useful to improve solubility. The anti-fXa activity of ethylenedia-
mine compounds could be improved by the introduction of a sub-
stituent at the C1 position in the S configuration. In addition,
compound 12 possessing a phenylenediamine spacer exhibited
strong anti-fXa activity (IC50 = 7.4 nM). X-ray structure analysis re-
vealed that compound 12 utilized an additional subsite S1b, while
retaining interactions in S1 and S4 sites. Further exploratory study
resulted in the discovery of compound 30h which demonstrated
excellent anti-fXa activity (IC50 = 2.2 nM), good anticoagulant
14. Anticoagulant activity was evaluated with the human plasma clotting time
doubling concentration (PTCT2). The method is as follows: prothrombin time
(PT) was measured with an Amelung KC-10A micro coagulometer (MC Medical,
Tokyo, Japan). First, 50
DMSO/saline and incubated for 1 min at 37 °C. Coagulation was started by the
addition of 100 L of Thromboplastin C Plus (0.5 U/mL) to the mixture, and the
lL of plasma was mixed with 50 lL of inhibitor or 4%
l
clotting time was measured. The concentration of inhibitor required to double
the clotting time (CT2) was estimated from the concentration–response curve
by a regression analysis.
15. Male Wistar rats were fasted overnight. Synthetic compounds were dissolved
in 0.5% (w/v) methylcellulose solution and administered orally to rats via a
stomach tube. For control rats, 0.5% (w/v) methylcellulose solution was
administered orally. The rats were anesthetized with ravonal at several time
points when blood samples were collected in the presence of trisodiumcitrate.
After the blood samples were centrifuged, the platelet poor plasma samples
were used for the measurement of their anti-fXa activity. Anti-fXa activity:
plasma (5
40 L), H2O (5
the addition of 0.75 M S-2222 (40
l
L) was mixed with 0.1 M Tris–0.2 M NaCl–0.2% BSA buffer (pH 7.4;
L), and 0.1 U/mL human fXa (10 L). A reaction was started by
L). The absorbance (OD) at 405 nm was
l
l
l
l
activity (PTCT2 = 3.9 lM), and improved solubility in neutral pH
monitored every 10 s with a SPECTRAmax 340 microplate spectrophotometer
(Molecular Devices, Sunnyvale, CA, USA) at room temperature, and the reaction
velocity (OD/min) was obtained. Anti-fXa activity (inhibition %) was calculated
as follows: anti-fXa activity = {1 À [(OD/min) of sample/(OD/min) of
control]} Â 100.
region. Compound 30h exhibited strong ex vivo anti-fXa activity
after intravenous administration and also demonstrated strong
antithrombotic effect in a rat DIC model. The findings described
in this Letter can offer possibilities for the future design of potent
fXa inhibitors.
16. The X-ray crystal structure coordinates of compound 12 in fXa have been
deposited in the Protein Data Bank (PDB code: 3Q3 K).
17. S1b pocket (also known as ‘ester binding pocket’) is composed of Gln192,
Arg143, Glu147, and Cys191-Cys220 disulfide bridge: (a) Qiao, J. X.; Wang, T.
C.; Wang, G. Z.; Cheney, D. L.; He, K.; Rendina, A. R.; Xin, B.; Luettgen, J. M.;
Knabb, R. M.; Wexler, R. R.; Lam, P. Y. S. Bioorg. Med. Chem. Lett. 2007, 17, 5041;
(b) Adler, M.; Kochanny, M. J.; Ye, B.; Rumennik, G.; Light, D. R.; Biancalana, S.;
Whitlow, M. Biochemistry 2002, 41, 15514; (c) Anselm, L.; Banner, D. W.; Benz,
J.; Zbinden, K. G.; Himber, J.; Hilpert, H.; Huber, W.; Kuhn, B.; Mary, J.-L.;
Otteneder, M.; Panday, N.; Ricklin, F.; Stahl, M.; Tohmi, S.; Haap, W. Bioorg.
Med. Chem. Lett. 2010, 20, 5313; (d) Huis, C. A. V.; Casimiro-Garcia, A.; Bigge, C.
F.; Cody, W. L.; Dudley, D. A.; Filipski, K. J.; Heemstra, R. J.; Kohrt, J. T.; Leadley,
R. J., Jr.; Narasimhan, L. S.; McClanahan, T.; Mochalkin, I.; Pamment, M.;
Peterson, J. T.; Sahasrabudhe, V.; Schaum, R. P.; Edmunds, J. J. Bioorg. Med.
Chem. 2009, 17, 2501; (e) Zhang, P.; Huang, W.; Wang, L.; Bao, L.; Jia, Z. J.;
Bauer, S. M.; Goldman, E. A.; Probst, G. D.; Song, Y.; Su, T.; Fan, J.; Wu, Y.; Li, W.;
Woolfrey, J.; Sinha, U.; Wong, P. W.; Edwards, S. T.; Arfsten, A. E.; Clizbe, L. A.;
Kanter, J.; Pandey, A.; Park, G.; Hutchaleelaha, A.; Lambing, J. L.; Hollenbach, S.
J.; Scarborough, R. M.; Zhu, B.-Y. Bioorg. Med. Chem. Lett. 2009, 19, 2179; (f)
Nazaré, M.; Will, D. W.; Matter, H.; Schreuder, H.; Ritter, K.; Urmann, M.;
Essrich, M.; Bauer, A.; Wagner, M.; Czech, J.; Lorenz, M.; Laux, V.; Wehner, V. J.
Med. Chem. 2005, 48, 4511.
18. Lilly’s group reported factor Xa inhibitors possessing 1,2-phenylenediamine or
anthranilamide moiety and similar hydrogen bondings was expected from the
computational modeling study: (a) Herron, D. K.; Goodson, T., Jr.; Wiley, M. R.;
Weir, L. C.; Kyle, J. A.; Yee, Y. K.; Tebbe, A. L.; Tinsley, J. M.; Mendel, D.; Masters,
J. J.; Franciskovich, J. B.; Sawyer, J. S.; Beight, D. W.; Ratz, A. M.; Milot, G.; Hall, S.
E.; Klimkowski, V. J.; Wikel, J. H.; Eastwood, B. J.; Towner, R. D.; Gifford-Moore,
D. S.; Craft, T. J.; Smith, G. F. J. Med. Chem. 2000, 43, 859; (b) Wiley, M. R.; Weir,
L. C.; Briggs, S.; Bryan, N. A.; Buben, J.; Campbell, C.; Chirgadze, N. Y.; Conrad, R.
C.; Craft, T. J.; Ficorilli, J. V.; Franciskovich, J. B.; Froelich, L. L.; Gifford-Moore, D.
S., ; Goodson, T., Jr.; Herron, D. K.; Klimkowski, V. J.; Kurz, K. D.; Kyle, J. A.;
Masters, J. J.; Ratz, A. M.; Milot, G.; Shuman, R. T.; Smith, T.; Smith, G. F.; Tebbe,
A. L.; Tinsley, J. M.; Towner, R. D.; Wilson, A.; Yee, Y. K. J. Med. Chem. 2000, 43,
883; (c) Masters, J. J.; Franciskovich, J. B.; Tinsley, J. M.; Campbell, C.; Campbell,
J. B.; Craft, T. J.; Froelich, L. L.; Gifford-Moore, D. S.; Hay, L. A.; Herron, D. K.;
Klimkowski, V. J.; Kurz, K. D.; Metz, J. T.; Ratz, A. M.; Shuman, R. T.; Smith, G. F.;
Smith, T.; Towner, R. D.; Wiley, M. R.; Wilson, A.; Yee, Y. K. J. Med. Chem. 2000,
43, 2087; (d) Franciskovich, J. B.; Masters, J. J.; Tinsley, J. M.; Craft, T. J.; Froelich,
L. L.; Gifford-Moore, D. S.; Klimkowski, V. J.; Smallwood, J. K.; Smith, G. F.;
Smith, T.; Towner, R. R.; Weir, L. C.; Wiley, M. R. Bioorg. Med. Chem. Lett. 2005,
15, 4838.
Acknowledgement
We are grateful to the anti-coagulant group in Biological Re-
search Laboratory for performing the biological assays. We thank
Dr. N. Sakabe, Dr. M. Suzuki, and Dr. N. Igarashi for their assistance
during diffraction data collection at BL-6B of Photon Factory, Japan.
References and notes
1. Mackman, N. Nature 2008, 451, 914.
2. Lam, Y.-Y.; Ma, T. K. W.; Yan, B. P. Int. J. Cardiol. 2010. doi:10.1016/
j.ijcard.2010.10.017.
3. Leadley, R. J., Jr. Curr. Top. Med. Chem. 2001, 1, 151.
4. de Candia, M.; Lopopolo, G.; Altomare, C. Expert Opin. Ther. Patents 2009, 19,
1535.
5. Pinto, D. J. P.; Smallheer, J. M.; Cheney, D. L.; Knabb, R. M.; Wexler, R. R. J. Med.
Chem. 2010, 53, 6243.
6. Mehta, R. S. Expert Rev. Hematol. 2010, 3, 227.
7. Palareti, G.; Cosmi, B. Expert Opin. Emerg. Drugs 2010, 15, 107.
8. Schirmer, S. H.; Baumhäkel, M.; Neuberger, H.-R.; Hohnloser, S. H.; van Gelder,
I. C.; Lip, G. Y. H.; Böhm, M. JACC 2010, 56, 2067.
9. Ahrens, I.; Lip, G. Y. H.; Peter, K. Thromb. Haemost. 2010, 104, 49.
10. Nagata, T.; Yoshino, T.; Haginoya, N.; Yoshikawa, K.; Nagamochi, M.;
Kobayashi, S.; Komoriya, S.; Yokomizo, A.; Muto, R.; Yamaguchi, M.; Osanai,
K.; Suzuki, M.; Kanno, H. Bioorg. Med. Chem. 2009, 17, 1193.
11. Yoshikawa, K.; Yokomizo, A.; Naito, H.; Haginoya, N.; Kobayashi, S.; Mochizuki,
A.; Yoshino, T.; Nagata, T.; Osanai, K.; Watanabe, K.; Kanno, H.; Ohta, T. Bioorg.
Med. Chem. 2009, 17, 8206.
12. Yoshikawa, K.; Kobayashi, S.; Nakamoto, Y.; Haginoya, N.; Komoriya, S.;
Yoshino, T.; Nagata, T.; Mochizuki, A.; Watanabe, K.; Suzuki, M.; Kanno, H.;
Ohta, T. Bioorg. Med. Chem. 2009, 17, 8221.
13. The in vitro anti-fXa activity was measured by using chromogenic substrate S-
2222 (Chromogenix, Inc.) and human fXa (Enzyme Research Laboratories).
Aqueous DMSO (5% V/V; 10
0.0625 U/mL human fXa (10
l
L) or inhibitors in aqueous DMSO (10
lL) and
l
L) were mixed with 0.1 M Tris–0.2 M NaCl–0.2%
BSA buffer (pH 7.4; 40
2222 (40 L). The absorbance (OD) at 405 nm was monitored every 10 s with a
SPECTRAmax 340 microplate spectrophotometer (Molecular Devices,
lL). A reaction was started by the addition of 0.75 M S-
19. Furugohri, T.; Shiozaki, Y.; Muramatsu, S.; Honda, Y.; Matsumoto, C.; Isobe, K.;
Sugiyama, N. Eur. J. Pharm. 2005, 514, 35.
l