SAR and X-ray structures of enantiopure 1,2-cis-(1R,2S)-cyclopentyldiamine and cyclohexyldiamine derivatives as inhibitors of coagulation Factor Xa

Article abstract of DOI:10.1016/j.bmcl.2007.06.029

In the search of Factor Xa (FXa) inhibitors structurally different from the pyrazole-based series, we identified a viable series of enantiopure cis-(1R,2S)-cycloalkyldiamine derivatives as potent and selective inhibitors of FXa. Among them, cyclohexyldiam

Full text of DOI:10.1016/j.bmcl.2007.06.029

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4419–4427
SAR and X-ray structures of enantiopure
1,2-cis-(1R,2S)-cyclopentyldiamine and cyclohexyldiamine
derivatives as inhibitors of coagulation Factor Xa
Jennifer X. Qiao,* Chong-Hwan Chang, Daniel L. Cheney, Paul E. Morin, Gren Z. Wang,
Sarah R. King, Tammy C. Wang, Alan R. Rendina, Joseph M. Luettgen,
Robert M. Knabb, Ruth R. Wexler and Patrick Y. S. Lam
Bristol-Myers Squibb Company, Research and Development, PO Box 5400, Princeton, NJ 08543-5400, USA
Received 11 May 2007; revised 3 June 2007; accepted 5 June 2007
Available online 10 June 2007
Abstract—In the search of Factor Xa (FXa) inhibitors structurally different from the pyrazole-based series, we identified a viable
series of enantiopure cis-(1R,2S)-cycloalkyldiamine derivatives as potent and selective inhibitors of FXa. Among them, cyclohexyl-
diamide 7 and cyclopentyldiamide 9 were the most potent neutral compounds, and had good anticoagulant activity comparable to
the pyrazole-based analogs. Crystal structures of 7-FXa and 9-FXa illustrate binding similarities and differences between the five-
and the six-membered core systems, and provide rationales for the observed SAR of P1 and linker moieties.
Ó 2007 Elsevier Ltd. All rights reserved.
Coagulation factor Xa (FXa), a serine protease posi-
tioned at the convergence of the intrinsic and the extrin-
sic pathways in the coagulation cascade, is an attractive
target for developing novel, orally active anticoagulants
for the treatment and prevention of thrombotic diseases.
Selective FXa inhibitors might be more efficacious, have
fewer bleeding risks, and a more favorable safety/effi-
cacy ratio compared with thrombin inhibitors.1–3 Sev-
eral small-molecule, orally active FXa inhibitors,
including pyrazole-based razaxaban4 and apixaban,5
have entered clinical development, and apixaban is
undergoing evaluation in phase 3 studies in various
indications.
F3C
F
HCl
H2NOC
O
N
N
H
N
N
N
N
N
N
N
N
O
O
NH2
O
N
OMe
Razaxaban FXa Ki = 0.19 nM
PT EC2x = 1.9 μM
Apixaban FXa Ki = 0.08 nM
PT EC2x = 3.8 μM
FXa potency in the binding assay and modest anticoag-
ulant activity in the prothrombin time (PT) assay (FXa
Ki = 1.5 nM, PT EC2x = 6.3 lM), and was inactive
against all the other serine proteases tested (trypsin,
thrombin, urokinase, plasmin, aPC, tPA, plasma kalli-
krein, and chymotrypsin).
During the search for diverse scaffolds for FXa inhibi-
tion, compound 1 (Table 1) was identified as a lead
bearing a simple ethylene diamine core with a N-phenyl-
pyridone P4 group. The phenylpyridone group, as the
unsaturated derivative of the phenylpiperidinone group
found in apixaban,5 was a potent P4 moiety in the pyr-
azole-based series.6 Compound 1 showed good in vitro
Figure 1 depicts the X-ray structure of the 1-FXa com-
7
˚
plex with R = 0.24 at 1.8 A resolution. In this complex,
1 adopts an L-shaped conformation, similar to that ob-
served in the pyrazole-based analogs previously dis-
closed,4,5 with the chlorothiophene group occupying
the S1 pocket and the phenylpyridone group bound to
the S4 pocket. The P1 amide is coplanar with the thio-
phene ring, while the P4 amide is slightly tilted from
the phenyl ring.
Keywords: Factor Xa inhibitors; (1R,2S)-cycloalkyldiamine deriva-
tives; SAR; X-ray structure.
*
Corresponding author. Fax: +1 609 818 6810; e-mail: jennifer.
qiao@bms.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.029

4420
J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
Table 1. In vitro Xa potency of methyl-substituted ethylenediamine derivatives bearing a phenylpyridone P4 group
R3 R4
O
R2
2
R1
N
O
HN
P1 =
1
NH
H
S
O
N
Cl
Cl
P1
R1
P4
A
B
Compound
P1
R2
R3
R4
FXa Ki (nM)
PT EC2x (lM)
1
A
A
A
A
A
H
H
H
H
1.5
1.7
3.2
2.1
2.5
6.3
7.5
nd
12
1a ( )
1b
Me
Me
H
H
H
H
Me
H
H
H
1c ( )
1d
Me
Me
H
Me
H
H
nd
2
B
B
B
B
B
H
H
H
H
13
10
nd
nd
17
11
nd
2a ( )
2b
Me
Me
H
H
H
H
Me
H
H
H
1.3
3.9
4.8
2c ( )
2d
Me
Me
H
Me
H
H
Purified human enzymes were used. Values were averages from multiple determinations (n P 2). Ki values were measured as described in Ref. 10. PT
EC2x (the concentration of the inhibitor that doubles the prothrombin time from the control) values were measured according to Ref. 4. Same for all
tables.
Figure 2. Compound 1 is shown with the initial 2Fo À Fc electron
density map contoured at 0.7r in white mesh. Created using PyMol.8
Figure 1. Interactions between compound 1 and the FXa active site.
There is a water-mediated hydrogen-bond system involving the P1
carbonyl and Ser214 O. The P1 amide NH forms a hydrogen bond
a 5-chlorothiophene P1 group (A) resulted in com-
with Gly218 O, and the P4 amide NH forms a tight hydrogen bond
pounds having similar FXa inhibitory activity. In con-
with Gly216 O. The P1 Cl substituent forms a close contact with
trast, adding methyl group(s) in 2 bearing a larger
Tyr228 side chain.
bicyclic 3-chloroindole P1 group (B) resulted in im-
proved anti-FXa activity (2a–2d). Compound 2b with
S1 and S4 pockets are well-defined and distinctive sub-
sites of FXa. Based on experience, we believe that inter-
actions of a ligand in the FXa S1 and S4 subsites are
essential to the FXa binding affinity of the ligand. The
proper orientation of the P1 and P4 groups toward the
S1 and S4 pockets could be achieved by the selection
of an appropriate core or linker. The well-defined elec-
tron density of the flexible ethylene linker in the crystal
structure of the 1-FXa complex (Fig. 2) led us to the
strategy of improving FXa binding affinity by adding
one or more methyl groups on the ethylene linker to fur-
ther stabilize the observed gauche conformation. The
potency of the resulting methylated ethylenediamides
is shown in Table 1. Adding methyl group(s) at either
the C1 or the C2 position of the ethylene chain in 1 with
a gem-dimethyl substitution at C1 of the ethylene linker
was 10-fold more active than the unsubstituted 2. Mod-
eling studies9 suggested that FXa potency could be fur-
ther improved by modifying the methyl substituents in
2b to interact with the disulfide bridge region (S1b).
We did not investigate this series further due to the neg-
ligible solubility and relatively low anticoagulant activ-
ity of 2b, the latter of which is suggestive of high
protein binding.
Another way of constraining the floppy ethylene linker
in 1 and 2 is to tie back the C1 and the C2 CH2 groups
to form cyclicdiamine derivatives (Table 2). We found
that the cis racemic cyclohexyl analogs 3 and 5 were
slightly (threefold) more active than the trans analogs

J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
4421
Table 2. Stereochemistry of cycloalkyldiamine derivatives bearing a
phenylpyridone P4 group
17. Compound 11 with a 3-chloroindole group was the
most potent among compounds bearing chlorinated
bicyclic (hetero) aromatic P1 groups (11–16).
O
M
HN
P1 =
O
N
Adding a nitrogen in the thiophene ring of 9 resulted in
more than three orders of magnitude loss of potency
(18). Chlorine substitution and the position of the chlo-
rine are essential to the compound’s FXa inhibitory
activity.12 Replacing the chlorine atom with a methyl
group (9 vs 19), or a methoxy group (17 vs 20), and
removing the chlorine atom (11 vs 21) resulted in five-
fold to 120-fold decrease of potency. The 4-Cl-phenyl
analog 17 was 30-fold and 70-fold more potent than
the 3- and 2-Cl-phenyls (22, 23), respectively. Ortho-
or meta-substitution on the 4-chlorophenyl ring in 17
with F, Cl, OMe, Me, and SO2Me resulted in decreased
FXa potency (22–32).
S
O
H
NH
N
Cl
5-Cl-thienyl
Cl
P1
P4
3-Cl-indolyl
Compound
M
Stereochemistry P1
FXa Ki
(nM)
1
2
—
—
5-Cl-thienyl
1.5
3-Cl-indolyl 13
3
cis-( )
3-Cl-indolyl
1.0
5-Chlorothiophene and 3-chloroindole in 9 and 7 were
the two best P1 groups in both the five- and the six-
membered core series (Table 4). 6-Chloronaphthyl, 6-
chlorobenzo[b]thiophenyl, and 4-chlorophenyl P1
groups provided compounds that were one to two or-
ders of magnitude less active than 9 and 7. Tables 3
and 4 show that subtle structural deviations from 5-
chlorothiophene and 3-chloroindole P1 groups reduced
FXa potency drastically, revealing the importance of
the S1 pocket of serine proteases for binding affinity.
4
5
trans-( )
cis-( )
3-Cl-indolyl
5-Cl-thienyl
3.7
7.3
6
7
trans-( )
(1R, 2S)
5-Cl-thienyl 23
1
2
3-Cl-indolyl
0.67
P4 SAR of (1R,2S)-cyclopentyldiamine and (1R,2S)-
cyclohexyldiamine derivatives. FXa activities of
(1R,2S)-cyclopentyl and cyclohexyldiamine core systems
containing a variety of P4 groups were studied using 5-
chlorothiophene and 3-chloroindole P1 groups
(Table 5). Pyridones 7, 9, and 11 and a-CH2-N-pyrrolid-
inyl phenylcyclopropyl analogs 54–56 were the most po-
tent FXa inhibitors (FXa Ki 6 1 nM) in both core
systems. Compounds containing a variety of six-mem-
bered lactam-based P4 moieties, such as piperidinones
40–42, morpholinone 43, cyclic carbamate 44, and cyclic
urea 45, were potent with Ki values less than 5 nM. The
basic 2-dimethylaminomethyl imidazole 50 and piperid-
inone 42 were also similar in potency. Compounds 46
and 47 bearing five- and seven-membered lactam P4
groups were less potent than piperidinones 40–42. The
cyclic sulfonamide in 48 was the least potent P4 group
in this study.
8
9
(1S,2R)
3-Cl-indolyl
3.8
1 2
(1R, 2S)
5-Cl-thienyl
0.43
10
(1S,2R)
5-Cl-thienyl 87
4 and 6, respectively. The cyclohexyldiamine derivative 7
(FXa Ki = 0.67 nM) bearing a 3-chloroindole P1 group
with a cis-(1R,2S) configuration was five times more
potent than its enantiomer 8, and the cis cyclopentyldi-
amine derivative 9 (FXa Ki = 0.43 nM) bearing a
5-chlorothiophene P1 group with the same (1R,2S) con-
figuration was 200-fold more potent than its enantiomer
10. Compounds 9 and 7 were threefold and 20-fold more
potent than the corresponding ethylenediamine leads 1
and 2, respectively. The absolute stereochemistry of 9,
cis-(1R,2S), was confirmed by single crystal X-ray
analysis11 using the anomalous dispersion of sulfur
and chlorine atoms. Given the potency of the cis
analogs, the SAR discussion reported herein will there-
fore focus on the cis-(1R,2S)-cyclopentyldiamine and
cis-(1R,2S)-cyclohexyldiamine core systems.
Table 5 shows that cyclohexyldiamine derivatives bear-
ing the larger 3-chloroindole P1 group were slightly
more potent than the cyclopentyldiamine counterparts.
For example, compounds 7, 50, and 54 were two- to
fivefold more potent than compounds 11, 49, and 53,
respectively. In contrast, when using the smaller 5-chlo-
rothiophene P1 group, the resulting cyclopentyldiamine
derivatives were equal or slightly more potent than the
corresponding cyclohexyldiamines (9 vs 37, fourfold;
55 vs 56, twofold; and 51 vs 52, twofold). Most com-
pounds with good binding activity (FXa Ki < 2 nM) also
demonstrated good anticoagulant activity in the clotting
time functional assay. Compared with pyrazole-based
raxazaban4 and apixaban,5 these cyclicdiamides were
less potent in FXa binding affinity (FXa Ki). However,
P1 SAR of (1R,2S)-cyclopentyldiamine and (1R,2S)-
cyclohexyldiamine derivatives bearing phenylpyridone P4
group. In the cyclopentyldiamine series (Table 3), 9 with
a 5-chlorothiophene group was the most potent among
compounds bearing monocyclic P1 groups, and it was
80-fold more potent than the 4-chlorophenyl analog

4422
J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
Table 3. P1 SAR of (1R,2S)-cyclopentyldiamine derivatives bearing a
phenylpyridone P4 group
be the result of the reduced lipophilicity, which in turn
lowers protein binding.
In general, the series of compounds in Table 5 were
highly selective against other related serine proteases
(see Table 6 for four representative compounds). The
observed high selectivity resulted from the structural
variations in the S1 and S4 pockets of the different serine
proteases (e.g., the size difference of the S4 pockets, the
Ala190Ser difference in the S1 pockets), evident in the
overlay of the publicly available crystal structures.
O
O
NH HN
P1
O
N
Compound
P1
FXa Ki (nM)
9
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
5-Cl-2-thienyl
3-Cl-6-indolyl
5-Cl-2-indolyl
6-Cl-2-indolyl
0.43
1.0
Linker SAR. Tables 7 and 8 list FXa potency of five-
and six-membered cores with different P1 and P4 linkers.
Compounds 7 and 9 with P1 and P4 amide linkers were
the most potent. Compounds with a planar P1 linker,
such as amides 9 and 17, urea 57, and oxalamide 58,
were much more potent than those with an aplanar lin-
ker, such as methylamine 59 and sulfonamide 60.
Changing the P4 amide in 7 to a sulfonamide in 63 led
to loss of potency, while the methylsulfonyl P4 linker
analog 64 (racemic) resulted in 19-fold loss of potency.
Different SAR of P4 linker was observed between the
five- and six-membered core systems: in the five-mem-
bered core, reducing the P4 amide in 9 to a methylamine
in 61 led to 10-fold loss of potency; while in the six-
membered ring system, the corresponding methylamine
65 had similar FXa potency as parent 7. Ortho substitu-
tion on the inner phenyl ring in cyclopentyl 9 with a
fluorine atom in 62 resulted in a 10-fold drop of
FXa potency; while for the six-membered ring, FXa
potency of the fluorine analog 66 was maintained
compared with 7.
74
266
34
6-Cl-2-benzo[b]thiophenyl
5-Cl-2-benzo[b]thiophenyl
6-Cl-2-naphthyl
4-Cl-phenyl
288
27
36
2-Cl-5-thiazolyl
5-Me-2-thienyl
4-MeO-phenyl
1220
7.92
170
124
1186
2610
138
892
226
433
129
360
603
7420
8560
6-Indolyl
3-Cl-phenyl
2-Cl-phenyl
4-Cl-3-Me-phenyl
4-Cl-3-OMe-phenyl
4-Cl-3-F-phenyl
3,4-di-Cl-phenyl
4-Cl-2-F-phenyl
2,4-di-Cl-phenyl
4-Cl-2-Me-phenyl
4-Cl-2-MeO-phenyl
4-Cl-2-SO2Me-phenyl
X-ray structures of compounds 7 and 9. Cyclopentyl-dia-
mide 9 with a chlorothiophene P1 and cyclohexyl-dia-
mide 7 with a chloroindole P1 were the most potent
neutral leads in this series. X-ray crystal structures of
7-FXa13 and 9-FXa14 complexes were obtained with res-
olution at 2.2 A and 1.8 A, respectively. Figures 3 and 4
illustrate the interactions of 7 and 9, respectively, with
the FXa residues in the active site. Figure 5 shows the
overlay of the X-ray structures of FXa-bound ethylene-
diamide 1, cyclohexyldiamide 7, and cyclopentyldiamide
9. All three molecules bind in a similar extended man-
ner, with the chloroheteroaryl groups occupying the S1
pocket and the phenylpyridone group bound to the S4
pocket. The pyridone ring in each of the three molecules
forms an edge-to-face interaction with Try215, and
stacks nicely between Tyr99 and Phe174.
Table 4. P1 SAR of cyclopentyldiamine vs cyclohexyldiamine deriv-
atives bearing a phenylpyridone P4 group
X
O
O
NH HN
˚
˚
HN
S
P1
P1 =
Cl
O
Cl
3-Cl-indolyl
N
5-Cl-thienyl
Compound P1
X
FXa Ki
(nM)
11
7
3-Cl-6-indolyl
3-Cl-6-indolyl
6-Cl-2-indolyl
6-Cl-2-indolyl
–CH2–
1.0
–CH2CH2– 0.67
–CH2– 266
–CH2CH2– 14
34
6-Cl-2-benzo[b]thiophenyl –CH2CH2– 16
13
33
14
34
16
35
17
36
9
6-Cl-2-benzo[b]thiophenyl –CH2–
The chlorothiophene groups of 1 and 9 and the chloro-
indole group of 7 are buried deeply inside the S1 pocket
(Fig. 5). The chlorine atoms occupy essentially the same
position at the bottom of the S1 pocket, displacing a
structurally conserved water and forming close contacts
with Tyr228. The average ClÁ Á ÁTyr228CZ distance is
6-Cl-2-naphthyl
6-Cl-2-naphthyl
4-Cl-phenyl
–CH2–
–CH2CH2– 20
–CH2– 36
–CH2CH2– 37
27
4-Cl-phenyl
5-Cl-2-thienyl
5-Cl-2-thienyl
–CH2–
–CH2CH2– 1.8
0.43
˚
3.6 A. Such Cl–Tyr228 interaction (presumably
37
Ar–ClÁ Á Áp interaction) is also observed in other inhibi-
tor-FXa complexes.12 The narrow P1 SAR could be at
least partially attributed to the high degree of directional
preference15 of the Ar–ClÁ Á Áp interactions. The overall
conformation of the cyclopentyldiamide 9 is very similar
the most potent compounds such as 7, 9, 38, 40, 54–56
showed comparably good anticoagulant activity (PT
EC2x 2–3 lM) in human plasma, which we assume to

J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
Table 5. P4 SAR of cyclopentyl- and cyclohexyl-diamine derivatives
4423
X
O
O
HN
S
NH HN
P1 =
P1
Cl
Cl
5-Cl-thienyl
3-Cl-indolyl
P4
R4
Compound
P4
P1
X
FXa Ki (nM)
PT EC2x (lM)
O
N
9
5-Cl-thienyl
5-Cl-thienyl
3-Cl-indolyl
3-Cl-indolyl
5-Cl-thienyl
3-Cl-indolyl
5-Cl-thienyl
3-Cl-indolyl
3-Cl-indolyl
3-Cl-indolyl
–CH2–
0.43
1.8
1.7 (aPTT = 3.6)
O
N
37
11
7
–CH2CH2–
–CH2–
6.5
O
N
1.0
5.5
O
N
–CH2CH2–
–CH2–
0.67
1.1
3.2 (aPTT = 4.4)
O
N
38
39
40
41
42
43
2.3
6.0
3.5
nd
nd
nd
N
N
O
N
–CH2CH2–
–CH2–
0.94
1.5
O
N
O
N
–CH2–
2.6
O
N
–CH2CH2–
–CH2–
2.6
O
N
4.2
O
O
N
O
44
45
46
47
48
3-Cl-indolyl
3-Cl-indolyl
3-Cl-indolyl
3-Cl-indolyl
3-Cl-indolyl
–CH2CH2–
–CH2CH2–
–CH2CH2–
–CH2CH2–
–CH2CH2–
5.0
4.0
8.1
nd
O
N
NH
nd
O
nd
N
O
N
17
nd
O2
S
107
nd
N
(continued on next page)

4424
J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
Table 5 (continued)
Compound
P4
P1
X
FXa Ki (nM)
PT EC2x (lM)
CH2NMe2
49
50
51
52
3-Cl-indolyl
–CH2–
7.1
nd
N
N
CH2NMe2
N
3-Cl-indolyl
5-Cl-thienyl
5-Cl-thienyl
–CH2CH2–
–CH2–
2.3
2.8
4.4
nd
nd
nd
N
CH2NMe2
N
N
CH2NMe2
N
–CH2CH2–
N
N
N
N
53
54
55
3-Cl-indolyl
3-Cl-indolyl
5-Cl-thienyl
–CH2–
2.4
10
–CH2CH2–
–CH2–
0.44
0.51
2.8
2.1
N
56
5-Cl-thienyl
–CH2CH2–
0.95
2.9
Raxazaban
Apixaban
0.19
0.08
1.9
3.8
Table 6. Human enzyme selectivity profile
Enzyme Ki (nM)
7
9
40
55
FXa
Thrombin
Trypsin
0.67
>12,000
>5000
0.43
7660
1.5
8871
0.51
3601
>5000
>5000
>5000
Factor VIIa
Factor XIa
aPC
>16,000
>15,000
>21,000
>22,000
5502
>16,000
>15,000
>21,000
>22,000
>20,000
>14,000
>16,000
>20,000
>10,000
>10,000
>21,000
>22,000
>20,000
>14,000
>10,000
>20,000
>10,000
>10,000
>21,000
>22,000
>20,000
>14,000
>10,000
>20,000
Plasmin
tPA
Urokinase
Plasma kallikrein
Chymotrypsin
>14,000
8458
>20,000
All Ki’s were obtained from purified human enzymes and were averaged from multiple determinations (n P 2). See Ref. 4 for more details.
Table 7. Linker SAR of cis-cyclophenyl core system
R1
O
P1
G
Z
N
Compound
P1
P1 linker (G)
P4 linker (Z)
R1
FXa Ki (nM)
9
17
57
58
59
60
61
62
5-Cl-thienyl
4-Cl-phenyl
4-Cl-phenyl
5-Cl-pyridinyl
5-Cl-thienyl
4-Cl-phenyl
5-Cl-thienyl
5-Cl-thienyl
–CONH–
–CONH–
–NHCONH–
–NHCO–
–NHCO–
–NHCO–
–NHCO–
–NHCO–
–NHCO–
–NHCH2–
–NHCO–
H
H
H
H
H
H
H
F
0.43
36
10
–NHCOCONH–
–CH2NH–
–SO2NH–
9.9
82
11500
4.2
4.7
–CONH–
–CONH–

J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
Table 8. Linker SAR of cis-cyclohexyl system
4425
R1
O
N
Z
P1
G
Compound P1
P1 linker P4 linker
(G) (Z)
R1 FXa
Ki nM
7
63
3-Cl-indolyl –CONH– –NHCO–
3-Cl-indolyl –CONH– –NHSO2–
3-Cl-indolyl –CONH– –CH2SO2–
3-Cl-indolyl –CONH– –NHCH2–
3-Cl-indolyl –CONH– –NHCO–
H
H
H
H
F
0.67
929
13
64 ( )
65
66
1.3
0.71
Figure 4. (a) Compound 9 is shown with the initial 2Fo À Fc electron
density map contoured at 1r in white mesh. Created using PyMol.8 (b)
Interactions between cyclopentyl analog 9 and the FXa active site. The
P1 amide NH forms a weak hydrogen bond with Gly218 O; while the
P4 amide NH forms a tight hydrogen bond to Gly216 O. C4-H of the
thiophene ring forms a close contact with Asp189 side chain,
presumably engaging in weak electrostatic interactions. The P1 Cl
substituent forms a close contact with Tyr228 side chain.
Figure 3. (a) Compound 7 is shown with the initial 2Fo À Fc electron
density map contoured at 1r in orange mesh. The electron density for
the putative water involved in hydrogen bonding to S214 is contoured
at 0.8r. Created using PyMol.8 (b) Interactions between cyclohexyl
analog 7 and the FXa active site. The P1 amide NH forms a weak
hydrogen bond with Gly216 O, and the P4 amide O forms a water
mediated hydrogen bond with Ser214 O. The NH of the indole ring
forms a hydrogen bond with Asp189 O and with Gly218 O,
respectively. The cyclohexyl ring adopts a chair form. The P1 Cl
substituent forms a close contact with Tyr228 side chain.
to that of the ethylenediamide 1, which shares the
same P1 group with 9. However, the thiophene ring in
9 is rotated 180 degrees from that in 1, keeping the
Figure 5. Overlay of X-ray structures of compounds 1 (yellow), 7
(white), and 9 (blue) in the active site of FXa.

4426
J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
n
chlorine atom at essentially the same position in the S1
pocket.
n
n
a,b
c,d
e
N3
NHBoc
BnO
HO
NHBoc
NH2
The overlay of the X-ray structures of 7-FXa and 9-FXa
(Fig. 5) clearly shows significant binding differences in
both the cycloalkyl core and the amide linker region
of 7 and 9, suggesting that different approaches may
be needed to optimize the five- and the six-membered
ring systems. The cyclohexyl ring in 7-FXa with a pre-
ferred bicyclic indole P1 group locates farther away
from the S1 pocket than the cyclopentyl ring in 9-FXa
bearing a smaller thiophene P1 group. Hydrogen bond-
ing possibilities of the two amide linkers of 9 and 7 are
different (see Figs. 3 and 4). In addition, the relative ori-
entation of the two amides with their adjacent aromatic
rings is different in these two compounds. In cyclopentyl
9, the P1 amide is almost coplanar with the thiophene
ring, and the P4 amide is nearly coplanar with the inner
phenyl ring; while in cyclohexyl 7, the two amides are
not at the same plane as that of the indole or the inner
n
n
n
O
O
f,g
O
h
HN
NH
NH2
n=1-2
NH
NHBoc
H2N
P4
P1
n=1-2
7,9, 11-56
P1
67
Scheme 2. Reagents and conditions: (a) Boc2O, Et3N, THF, quant.;
(b) Pd-C (5%), H2 (25 psi), EtOH, 6 h, 96%; (c) MsCl, Et3N, CH2Cl2,
0 °C, 2 h, 92%; (d) NaN3, DMF, 80 °C, overnight; (e) Pd–C (10%), H2,
EtOH, 6 h, 40% for two steps; (f) P1-COOH, HATU, DIEA, DMF,
54% (for 5-chlorothiophene-2-COOH), 62% (for 3-chloroindole-6-
COOH); (g) TFA, CH2Cl2; (h) P4-COOH, BOP, NMM, DMF, or
HATU, DIEA, DMF, 10–90% for two steps.
phenyl ring. This suggests that the amide linker region
in the cyclopentyl core system may be less tolerable to
structural changes than those in the cyclohexyl core
system.
O
a,b
HOOC
HOOC
HOOC
HOOC
N
Synthesis. Schemes 1 and 2 outline the synthesis of
enantiopure cis-(1R,2S)-cyclopentyl and cyclohexyl-dia-
mine derivatives bearing a variety of P4 groups. The
requisite P4 acids were synthesized according to the
straightforward transformations illustrated in se-
quences A to H (Scheme 1). The Boc-protected cyclic-
diamine cores 67 were prepared from commercially
available enantiopure benzyl protected amino alcohols
via a sequence of deprotection, mesylation, azide dis-
placement, and then reduction of the azide. Two subse-
quent amide formation reactions of the P4 and the P1
acids with cores 67 generated the desired products
(Scheme 2).
MeOOC
MeOOC
MeOOC
MeOOC
I
A
O
c,b
N
I
B
C
D
n=1-3
O
N
d,e,f,b
N
NH2
NH2
O2
S
g,h,b
N
O
O
i,j,b
HOOC
HOOC
N
O
MeOOC
MeOOC
EtOOC
NH2
NH2
E
F
In summary, using both structure-based design and
traditional medicinal chemistry approaches, we identi-
fied enantiopure cis-(1R,2S)-cycloalkyl diamine deriv-
atives as a viable series of potent, selective FXa
inhibitors that are structurally different from previ-
ously reported pyrazole-based scaffolds. The 5-chloro-
thiophene and 3-chloroindole P1 groups were
identified as potent P1 fragments for FXa inhibition
in this series. Example compounds bearing a variety
of N-phenyl substituted lactam-based P4 groups and
alpha-CH2–N-pyrrolidinyl-phenylcyclopropyl P4 group
showed excellent binding affinity (<1 nM), good po-
tency in human plasma clotting assay (<5 lM), and
high selectivity against other serine proteases
(>5000-fold). Among them, the cyclopentyl-diamide
9 and the cyclohexyldiamide 7 were the most potent
neutral analogs with anticoagulant activity compara-
ble to raxazaban and apixaban. In each of the
X-ray structures of FXa-bound 9 and 7, the P1
group sits deep in the S1 subsite and there exist
interactions between the chlorine and Tyr228. The
crystal structures of 7-FXa and 9-FXa complexes
provided insights to guide further structural optimization
of the cyclicdiamine derivatives to modulate in vitro and
in vivo properties, which shall be reported in future
communications.
NH
k,l,b
N
O
N
m,n,o
HOOC
O
G
F
COOH
N
p,q,r,s
I
H
HOOC
Scheme 1. Reagents and conditions: (a) 2-Hydroxypyridine, CuI,
K2CO3, 1,10-phenanthroline, 120 °C, overnight, 82%; (b) 1 N NaOH,
MeOH, rt to 50 °C, 80%-quant.; (c) Five- to seven-membered lactams,
K2CO3, CuI, 1,10-phenanthroline, DMSO, 120–125 °C, overnight to
1 day, 25–80%; (d) Boc-Gly-OH, BOP, DIEA, DMF, rt, 1.5 h, 98%; (e)
TFA, CH2Cl2, rt, 1.5 h, quant.; (f) Glyoxal (40% aq), NaOH, H2O,
MeOH, À40 to 5 °C, 3 h, 55%; (g) [1,2]Oxathiolane 2,2-dioxide, neat,
110 °C, 1 h, 71%; (h) POCl3, reflux, 105 °C, 4 h; then 4 N NaOH, 94%; (i)
ClCOO(CH2)3Cl, THF, 0 °C to rt, 2 h; (j) NaH, THF, 0 °C, 3 h, 47% for
2 steps; (k) Cl(CH2)3NCO, THF, rt overnight; (l) NaH, THF, rt, 3.5 h,
50% for 2 steps; (m) morpholine (3.3 equiv), 120 °C, neat, 2 day, 96%; (n)
KMnO4, PhCH2N+Et3ClÀ, CH2Cl2, 50 °C, 3 h, 59%; (o) 1 N NaOH,
EtOH, 89%; (p) ClCOOEt, Et3N, CH2Cl2, 0 °C, 30 min; NaBH4,
˚
MeOH, 0 °C, 30 min; (q) NaOAc, PCC, 4 A MS, CH2Cl2, rt, 1.5 h; (r)
pyrrolidine, NaBH(OAc)3, HOAc, CH2Cl2, rt, 20 min, 41% for 3 steps;
(s) KOAc, Pd(OAc)2, dppf, DMF/H2O, CO, 60 °C, 2.5 h, 90%.

J. X. Qiao et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4419–4427
4427
5. Pinto, D. J. P.; Orwat, M. J.; Koch, S.; Rossi, K. A.;
Alexander, R. S.; Smallwood, A.; Wong, P. C.; Rendina,
R. A.; Luettgen, J. M.; Knabb, R. M.; He, K.; Xin, B.;
Wexler R. R.; and Lam, P. Y. S. J. Med. Chem. Accepted
for publication.
Acknowledgments
The authors thank Dr. Steven Sheriff for depositing
1-FXa, 7-FXa, and 9-FXa complexes into PDB, Qi
Gao and Dedong Wu for single crystal X-ray structure
determination of 9, and Jeffrey M. Bozarth, Frank A.
Barbera, Tracy A. Bozarth, and Karen S. Hartl for
performing in vitro assays.
6. Pinto, D.; Quan, M.; Orwat, M.; Li, Y.; Han, W.; Qiao, J.;
Lam, P.; Koch, S. WO 2003026652 A1, 2003.
7. PDB Deposition number for 1-FXa complex is 2P93.
8. DeLano, W. L. The PyMOL Molecular Graphics System
(2002) on World Wide Web http://www.pymol.org.
9. Idea structures were evaluated with molecular dynamics
using Discover_3 with the CFF98 forcefield within the
InsightII v2000.L3 modeling software package.
www.accelrys.com.
10. (a) Knabb, R. M.; Kettner, C. A.; Timmermans, P. B. M.
W. M.; Reilly, T. M. Thromb. Haemost. 1992, 67, 56; (b)
Kettner, C. A.; Mersinger, L. J.; Knabb, R. M. J. Biol.
Chem. 1990, 265, 18289.
References and notes
1. Part of this research was presented at the 232nd ACS
National Meeting, San Francisco, CA, September 10–14,
2006, MEDI-393.
2. For recent review papers of FXa inhibitors, see: (a) Quan,
M.; Smallheer, J. Curr. Opin. Drug Discov. Devel. 2004, 7,
460; (b) Walenga, J. M.; Jeske, W. P.; Hoppensteadt, D.;
Fareed, J. Curr. Opin. Investig. Drugs 2003, 4, 272; (c)
Gould, W. R.; Leadley, R. J. Curr. Pharm. Design 2003, 9,
2337; (d) Quan, M. L.; Wexler, R. R. Curr. Top. Med.
Chem. 2001, 1, 137.
3. (a) Lassen, M. R.; Davidson, B. L.; Gallus, A.; Pineo, G.;
Ansell, J.; Deitchman, D. Blood 2003, 102, 11; (b) Straub,
A.; Pohlmann, J.; Lampe, T.; Pernerstorfer, J.; Schlem-
mer, K-H.; Reinemer, P.; Perzborn, E.; Roehrig, S.
J. Med. Chem. 2005, 48, 5900; (c) Eriksson, B. L.; Borris,
L.; Dahl, O. E.; Haas, S.; Huisman, M. V.; Kakkar, A. K.
J. Thromb. Haemost. 2006, 4, 121; (d) Hampton, T. JAMA
2006, 295, 743; (e) Wong, P. C.; Crain, E. J.; Watson, C.
A.; Zaspel, A. M.; Wright, M. R.; Lam, P. Y. S.; Pinto, D.
J.; Wexler, R. R.; Knabb, R. M. J. Pharmacol. Exp. Ther.
2002, 303, 993.
4. Quan, M. L.; Lam, P. Y. S.; Han, Q.; Pinto, D. J.; He, M.;
Li, R.; Ellis, C. D.; Clark, C. G.; Teleha, C. A.; Sun, J. H.;
Alexander, R. S.; Bai, S. A.; Luettgen, J. M.; Knabb, R.
M.; Wong, P. C.; Wexler, R. R. J. Med. Chem. 2005, 48,
1729.
11. There are three crystallographically independent mole-
cules having the same chirality in one asymmetric unit.
Three molecules are bound together by hydrogen bonding
˚
between –NH and amide groups (NÁ Á ÁO 2.917–3.128 A),
and solvent molecules (probably CH2Cl2 and H2O) are
disordering in the crystal lattice.
12. (a) Adler, M.; Kochanny, M. J.; Ye, B.; Rumnenik, G.;
Light, D. R.; Biancalana, S.; Whitlow, M. Biochemistry
2002, 41, 15514; (b) Maignan, S.; Guilloteau, J.-P.; Choi-
Sledeski, Y. M.; Becker, M. R.; Ewing, W. R.; Pauls, H.
W.; Spada, A. P.; Mikol, V. J. Med. Chem. 2003, 46, 685;
(c) Matter, H.; Will, D. W.; Nazare´, M.; Schreuder, H.;
Laux, V.; Wehner, V. J. Med. Chem. 2005, 48, 3290; (d)
Nazare, 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.
13. PDB Deposition number for 7-FXa complex is 2P94.
14. PDB Deposition number for 9-FXa complex is 2P95.
15. Saraogi, Ishu.; Vijay, V. G.; Das, Soma; Sekar, K.; Row,
T. N. Crystal Eng. 2003, 6, 69.