Preparation of 1-(4-methoxyphenyl)-1H-pyrazolo[4,3-d]pyrimidin-7(6H)-ones as potent, selective and bioavailable inhibitors of coagulation factor Xa

Article abstract of DOI:10.1016/j.bmcl.2006.04.044

Previously, potent factor Xa inhibitors were described based on a pyrazole core. Modifications of the pyrazole core have provided additional novel, highly potent factor Xa inhibitors. This manuscript will describe the synthesis and biological activity of

Full text of DOI:10.1016/j.bmcl.2006.04.044

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3755–3760
Preparation of 1-(4-methoxyphenyl)-1H-pyrazolo[4,3-d]pyrimidin-
7(6H)-ones as potent, selective and bioavailable
inhibitors of coagulation factor Xa
John M. Fevig,* Joseph Cacciola,ꢀ Joseph Buriak, Jr.,ꢁ Karen A. Rossi, Robert M. Knabb,
Joseph M. Luettgen, Pancras C. Wong, Stephen A. Bai,
Ruth R. Wexler and Patrick Y. S. Lam
Bristol-Myers Squibb Pharmaceutical Research Institute, PO Box 5400, Princeton, NJ 08543-5400, USA
Received 28 March 2006; revised 18 April 2006; accepted 18 April 2006
Available online 8 May 2006
Abstract—Previously, potent factor Xa inhibitors were described based on a pyrazole core. Modifications of the pyrazole core have
provided additional novel, highly potent factor Xa inhibitors. This manuscript will describe the synthesis and biological activity of
factor Xa inhibitors containing the 1H-pyrazolo[4,3-d]pyrimidin-7(6H)-one and related bicyclic cores. Many of these compounds are
potent, selective, and orally bioavailable inhibitors of coagulation factor Xa.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Thromboembolic diseases remain the leading cause of
death and disability in developed countries. This reality,
combined with the limitations of current therapies, has
led to extensive efforts to develop novel antithrombotic
agents.1 Factor Xa has become a major focus of phar-
maceutical intervention in the past decade because of
its central role in the blood coagulation cascade.2 Exten-
sive preclinical and clinical evidence has demonstrated
that inhibition of factor Xa is efficacious in both venous
and arterial thrombosis.3,4
F3C
N
F3C
N
F
F
H
H
NMe2 . HCl
N
N
N
N
N
SO2Me
O
O
N
NH2 . HCl
NH2
2, razaxaban
fXa Ki = 0.19 nM
O
N
1, DPC423
fXa Ki = 0.15 nM
F3C
N
F
H
N
N
SO2Me
O
3
MeO
Previously, it was demonstrated that a series of N-aryl-
pyrazole carboxamides, represented by the P1 benzyl-
amine analog DPC423 (1),5 were highly potent,
selective, and orally bioavailable small molecule inhibi-
tors of factor Xa (Fig. 1). Subsequently, razaxaban
(2),6a with an aminobenzisoxazole P1, was discovered
and has been shown to be efficacious in phase II deep
vein thrombosis (DVT) clinical trials.6b Furthermore,
it was demonstrated that the 4-methoxyphenyl residue
could be an effective P1 group when combined with an
fXa Ki = 3.6 nM
R
R
R
in vivo?
a
+
H2N
b
N
Ar
H
N
N
N
N
N
N
N
fused
Ar
Ar
CO2H
P1
P1
P1
bicyclic
O
O
6
5
7
4
Figure 1.
optimized pyrazole-P4 subunit, as in compound 3.7
Thus, the N-arylpyrazole carboxamides have been
shown to be a versatile and highly efficacious class of
fXa inhibitors.
Keywords: Factor Xa inhibitors; Anticoagulants; Antithrombotic
agents; Pyrazole; Bicyclic core; Pyrazolo[4,3-d]pyrimidin-7(6H)-ones.
*
Corresponding author. Tel.: +1 609 818 5285; fax: +1 609 818
3450; e-mail: john.fevig@bms.com
ꢀ
ꢁ
However, it was recognized that the common pyrazole
carboxamide 4, present in all of the compounds 1–3,
Present address: AstraZeneca, Wilmington, DE, USA.
Present address: DuPont Ag Products, Newark, DE, USA.
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.04.044

3756
J. M. Fevig et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3755–3760
NC
N
can potentially be hydrolyzed in vivo to liberate a
pyrazole carboxylic acid 5 and a biarylaniline 6
(Fig. 1). This was of concern due to the known mutage-
nicity of many anilines/biarylanilines.8 Furthermore, it
was discovered that there was no clear SAR in terms
of mutagenicity within this biarylaniline P4 series, neces-
sitating that any aniline of interest would require rigor-
ous purification and mutagenicity testing. Thus, in
efforts to discover back-up compounds to razaxaban,
several strategies were pursued that were designed to
avoid the potential formation of an aniline fragment.
One such strategy involved incorporating the amide
within a bicyclic core structure 7. This was expected to
provide protection from in vivo aniline formation by
rendering the amide less likely to be hydrolyzed and also
by requiring an additional step (cleavage of the N–b
bond) to liberate the free aniline. This manuscript
describes the initial success in realizing this strategy,9
where it was discovered that several bicyclic cores
maintained fXa potency and served as effective mimics
of the pyrazole carboxamide moiety when 4-methoxy-
phenyl was utilized as the P1 recognition group.7,10
NC
N
MeO2C
N
N3
O
NH2
N
O
Y
Y
a-c
H
N
d,e
Br
N
N
N
N
CO2Me
Br
MeO
MeO
MeO
10
14 (Y=H, F)
15 (Y=H, F)
R3
N
Y
N
SO2NH-t-Bu
N
N
(OH)2B
SO2NH-t-Bu
i
16
O
20-24
f
b,i
i,j
17 (R3=CO2Me)
18 (R3=CONH2)
19 (R3=CN)
MeO
25
g
h
26
Scheme 2. Reagents and conditions: (a) NaNO2, TFA, 0 ꢁC; then
NaN3 (75%); (b) LiOH, THF/H2O (90–95%); (c) (COCl)2, CH2Cl2;
then 4-bromoaniline or 4-bromo-2-fluoroaniline (70–80%); (d) SnCl2Æ-
H2O, MeOH, 65 ꢁC (55–65%); (e) 95% HCO2H, reflux (80–85%); (f)
16, Pd(PPh3)4, Na2CO3, benzene, H2O, 80 ꢁC (75–80%); (g) NH4OH,
dioxane, 80 ꢁC (50–60%); (h) POCl3, benzene, reflux (50%); (i) TFA,
CH2Cl2, 25 ꢁC (40–60%); (j) H2 (1 atm), 10% Pd/C, EtOH, TFA (50%).
The initial bicyclic core example, the 1H-pyrazolo-
[4,3-d]pyrimidin-5,7(4H,6H)-dione 13, was prepared as
described in Scheme 1. Diazotization of p-anisidine 8
was followed by treatment with malonitrile to afford
the dicyanohydrazone 9. Treatment of this hydrazone
with methyl bromoacetate and K2CO3 at 90 ꢁC gave
the 4-aminopyrazole 10 by N-alkylation and in situ ring
closure. Weinreb coupling with the highly optimized
biphenylaniline 117 afforded the amide 12 in a low an
irreproducible yield, which was subsequently cyclized
to compound 13, also in low yield, by treatment with
carbonyl diimidazole.
coupling via the acid chloride afforded amides 14.
Reduction of the azide was accomplished with tin(II)
chloride hydrate in refluxing methanol, conditions that
also effected the conversion of the nitrile to a methyl
ester. The resulting 4-aminopyrazole-5-carboxamide
was cleanly cyclized to the 1H-pyrazolo[4,3-d]pyrimi-
din-7(6H)-one bicyclic core 15 by refluxing in formic
acid. Suzuki coupling with boronic acid 166a smoothly
furnished the protected biphenylsulfonamide 17. The
pyrazole ester 17 was converted to the primary amide
18 by treatment with ammonium hydroxide in dioxane,
and the amide was subsequently dehydrated with POCl3
in refluxing benzene to afford the nitrile 19. Compounds
17–19 were deprotected with TFA to give compounds
20–24. The ester 17 was converted to the carboxylic acid
25 by hydrolysis and subsequent TFA deprotection.
TFA-catalyzed deprotection of 19 followed by nitrile
reduction furnished the primary amine 26.
The initial 1H-pyrazolo[4,3-d]pyrimidin-7(6H)-one
examples 20–26, with a biphenylsulfonamide P4 group,
were prepared as shown in Scheme 2. A more stepwise
approach was taken to avoid the low-yielding Weinreb
amide coupling. Aminopyrazole 10 was diazotized and
treated with sodium azide to form a 4-azidopyrazole,
where the azide functionality essentially served as an
amine protecting group. Ester hydrolysis and amide
The
analog 32 was prepared as described in Scheme 3.
1H-pyrazolo-[3,4-d]-pyridazin-7(6H)-one
core
F
O
O
O
O
F3C
N
F3C
N
F3C Br
N
NC CN
H2N
NC
N
SO2Me
CO2Et
NH2
NH2
CO2Et
N
NH
NH
a
b
28
N
N
Me
CO2Et
N
CO2Me
+
11
a
~(1:1)
OMe
c
30
desired
29
undesired
OMe
OMe
MeO
MeO
MeO
8
10
9
27
b
NC
N
NC
N
H
N
SO2NH-t-Bu
NH2
O
(OH)2B
F
F
Me
Me
F3C
N
F3C
N
H
d
N
N
O
N
N
N
N
N
N
SO2Me
SO2Me
16
O
N
N
SO2NH2
c,d
O
O
Br
MeO
MeO
13
12
31
MeO
32
MeO
Scheme 1. Reagents and conditions: (a) NaNO2, HCl/H2O, 0 ꢁC; then
malonitrile, NaOAc, MeOH/H2O, 0 ꢁC (70–80%); (b) BrCH2CO2Me,
K2CO3, DMF, 90 ꢁC (35%); (c) 11, AlMe3, CH2Cl2, 40 ꢁC (10–20%);
(d) CDI, THF (20%).
Scheme 3. Reagents and conditions: (a) 28, NaOEt, EtOH (75%); (b)
4-bromophenylhydrazine, EtOH, reflux (80%); (c) 16, Pd(PPh3)4,
Na2CO3, benzene, H2O, 80 ꢁC (80%); (d) TFA, CH2Cl2, 25 ꢁC (60%).

J. M. Fevig et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3755–3760
3757
EtO2C
EtO2C
N
The readily available 1-ethylidene-2-phenylhydrazine
NH2
N3
O
NH2
2711 was treated with ethyl 2,4-dioxovalerate 28 in the
presence of ethanolic sodium ethoxide to afford a good
yield of a separable 1:1 mixture of pyrazole regioisomers
29 and 30. The desired 30 was smoothly condensed with
4-bromophenylhydrazine in refluxing ethanol to give the
1H-pyrazolo-[3,4-d]-pyridazin-7(6H)-one core 31, which
crystallized out of solution upon cooling. Suzuki cou-
pling with the boronic acid 16 and TFA deprotection
afforded 32.
Y
c-e
H
N
f,g
a,b
N
N
N
CO2Me
Br
OMe
MeO
R5
MeO
EtO2C
34 Y=H,F
8
33
EtO2C
R5
N
CHO
N
Y
Y
(OH)2B
N
N
N
N
N
N
CHO
36
O
O
Br
h
37 R5=H
35a R5=H
MeO
HO
MeO
35b R5=Me
Compounds with
a basic N,N-dialkylaminomethyl
residue appended onto the terminal aryl ring of the P4
were prepared as described in Scheme 4. The pyrazole
diester 33 was prepared analogously to compound 10,
except that ethyl cyanoacetate was used instead of
malonitrile. The azidoamide 34 was also prepared as
described previously, with the key step being the
selective hydrolysis of the methyl ester with lithium
hydroxide. Tin(II) chloride reduction of 34 proceeded
cleanly to afford the 4-aminopyrazole, which was
cyclized as before with formic acid to give 35a
(R5 = H). Alternatively, cyclization of the 4-aminopy-
razole in refluxing glacial acetic acid produced 35b
(R5 = Me). Suzuki coupling of 35a with 2-form-
ylphenylboronic acid 36 afforded the biphenyl aldehyde
37. Reductive amination to introduce the N,N-dialkyl-
aminomethyl group was generally unsatisfactory. A
three-step protocol was more efficient, involving sodium
borohydride reduction of the aldehyde, phosphorus
tribromide-mediated bromination of the alcohol, and
displacement of the bromide with an appropriate cyclic
or acyclic secondary amine, to afford compounds 38. As
previously, the pyrazole ester was readily converted to
the primary amides 39 with ammonium hydroxide and
then subsequently dehydrated in one case (39a) to form
the nitrile 40a.
i,j,k
Y
n,o
EtO2C
N
N
O
Me
N
O
N
N
N
X
N
N
Br
41 R5=Me
MeO
R3
MeO
38a-c
j,p
l
Me
N
O
H2NOC
N
N
N
O
N
Y
N
N
X
N
Br
42a R3=-CH2-(1-imidazolyl)
42b R3=-CH2-(1-tetrazolyl)
42c R3=-CH2-(2-tetrazolyl)
MeO
MeO
NC
39a-c
m (39a)
q,r
R3
Me
N
O
N
O
N
N
N
N
N
N
N
N
43a-c
40a
MeO
MeO
NC
NC
N
R5
NH2
N
O
H
s,t
R5
N
N
N
N
N
O
Br
Br
Also in Scheme 4, conversion of 35b to the alcohol 41 was
accomplished by hydrolysis and subsequent reduction of
the acid via sodium borohydride reduction of an acid-
derived mixed anhydride. Treatment of 41 with phospho-
rus tribromide afforded the bromide, which was displaced
by imidazole to give 42a, and by tetrazole to give an insep-
arable mixture of the two tetrazole regioisomers 42b and
42c. Suzuki coupling of 42 with boronic acid 36 was
followed by reductive amination with pyrrolidine and
sodium triacetoxyborohydride, which in this case gave
serviceable yields of compounds 43. The isomers 43b
and 43c were separable by HPLC. Compound 46 was
prepared from 31 by procedures described above.
45a (R5=H)
44
MeO
MeO
45b (R5=Me)
NC
N
O
h,u
N
N
NMe2
40b (R5=H)
N
40c (R5=Me)
MeO
Scheme 4. Reagents and conditions: (a) NaNO2, HCl/H2O, 0 ꢁC; then
ethyl cyanoacetate, NaOAc, MeOH/H2O, 0 ꢁC (85%); (b)
BrCH2CO2Me, K2CO3, DMF, 90 ꢁC (25%); (c) NaNO2, TFA, 0 ꢁC;
then NaN3 (90%); (d) LiOH, THF/H2O (75%); (e) (COCl)2, CH2Cl2;
then 4-bromoaniline or 4-bromo-2-fluoroaniline (80–90%); (f) SnCl2–
H2O, MeOH, 65 ꢁC (50–60%); (g) 95% HCO2H, reflux (R5 = H,
80–85%), or HOAc, reflux (R5 = Me, 80–85%); (h) 36, Pd(PPh3)4,
Na2CO3, benzene, H2O, 80 ꢁC (75–85%); (i) NaBH4, MeOH (60–70%);
(j) PBr3, CH2Cl2 (80–90%); (k) HNRR0, CH3CN, (60–70%); (l)
NH4OH, dioxane, 80 ꢁC (50–60%); (m) POCl3, benzene, reflux
(50%); (n) KOH, MeOH/THF/H2O, (85%); (o) IBCF, NMM, THF,
0 ꢁC; then NaBH4, MeOH, (75%); (p) imidazole, DMF or tetrazole,
K2CO3, DMF (80–90%); (q) 36, Pd(PPh3)4, K3PO4, dioxane, 80 ꢁC,
(55–75%); (r) pyrrolidine, NaBH(OAc)3, HOAc, THF (30–50%); (s)
N,N-DMF dimethyl acetal, 100 ꢁC (R5 = H), or N,N-dimethylaceta-
mide dimethyl acetal, 100 ꢁC (R5 = Me); (t) 95% HCO2H, reflux
(R5 = H, 80% two steps), or HOAc, reflux (R5 = Me, 75% two steps);
(u) HNMe2, NaBH(OAc)3, HOAc, THF (40–50%).
Finally in Scheme 4, attempted ring closure of the
cyanopyridine 4412 under the standard conditions
(formic acid 85 ꢁC) or other conditions (N,N-DMF
dimethyl acetal or HC(OEt)3, 100 ꢁC) did not afford
the desired bicyclic core 45a. In all cases the expected
intermediate was formed (N-formyl or N0,N0-dimethyl-
N-formamidine) but ring closure of the amide nitrogen
onto this intermediate did not take place. It was pre-
sumed that the nitrile was having a negative effect on
either the amide reactivity or the basicity of the interme-
diates, or both. This problem was circumvented by

3758
J. M. Fevig et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3755–3760
employing an efficient two-step cyclization procedure.
Thus, treatment of 44 with N,N-DMF dimethyl acetal
at 100 ꢁC afforded the expected N0,N0-dimethyl-N-form-
amidine intermediate, which does not close to the
desired bicyclic under the conditions in which it is
formed. However, when this intermediate is isolated
and subjected to refluxing formic acid, smooth ring clo-
sure to 45a occurs. Therefore, the N0,N0-dimethyl-N-
formamidine intermediate is more reactive to ring clo-
sure relative to the N-formyl intermediate, but requires
acidic conditions for ring closure to occur. Likewise,
compound 44 was treated sequentially with N,N-dime-
thylacetamide dimethyl acetal at 100 ꢁC and then with
refluxing glacial acetic acid to prepare 45b. Compounds
45a,b were treated as previously disclosed to afford com-
pounds 40b,c, respectively.
Table 1. Bicyclic cores with biphenylsulfonamide P4
R3
F3C
N
Me
N
O
Y
N
N
N
N
N
N
SO2NH2
SO2NH2
O
MeO
MeO
32
20-26
R3
The SAR for the initial set of compounds, where the P4
was a biphenylsulfonamide residue, is shown in Table 1.
The initial bicyclic example, 13, illustrated the feasibility
of this strategy. This compound was equipotent with 3, a
highly optimized analog in the 4-methoxyphenyl P1 ser-
ies,7 and more potent than the uncyclized 12, thereby
suggesting that the bicyclic core effectively orients the
biphenyl residue for binding in the P4 pocket. However,
the poor potency of 13 in the in vitro aPTT clotting as-
say indicated that the 1H-pyrazolo[4,3-d]pyrimidin-
5,7(4H,6H)-dione core might suffer from very high
protein binding. The 1H-pyrazolo[4,3-d]pyrimidin-
7(6H)-one, the major focus of this manuscript, appears
to be a more promising bicyclic core. Compounds 20–
24 all have equal or more potent binding than 13, and
also appear to have significantly lower protein binding
as indicated by their more potent activities in the aPTT
assay, which are comparable to that of razaxaban and
more favorable than those of 3 and 13. The data show
that ester, primary amide, and nitrile are well tolerated
as 3-substituents and appear to be suitable replacements
for the CF3 group. Further SAR around the 3-substitu-
ent demonstrated that the carboxylic acid 25 and the
Compounda
Y
fXa Kib
(nM)
aPTT
IC2·c
(lM)
PT
IC2·c
(lM)
2
3
—
—
—
—
—
—
—
—
H
F
0.19
3.5
8.4
3.0
2.8
1.8
1.7
1.2
1.9
38
6.1
2.1
104
12
13
20
21
22
23
24
25
26
32
197
7.1
CO2Me
CO2Me
CONH2
CONH2
CN
H
F
8.7
9.8
5.8
H
H
H
—
CO2H
CH2NH2
—
22
2.8
a All final compounds were purified by prep HPLC and gave satis-
factory spectral data.
b Ki values were obtained from purified human enzyme and were
averaged from multiple determinations (n = 2), as described in Ref.
6a.
c The aPTT (activated partial thromboplastin time) and PT (pro-
thrombin time) in vitro clotting assays were performed in human
plasma as described in Ref. 6a.
Table 2. Incorporation of a basic P4 group
R3
Me
F3C
R5
N
Y
N
N
N
N
N
N
N
X
N
O
O
MeO
38-43
MeO
46
Compounda
R3
R5
Y
X
fXa Kib (nM)
aPTT IC2·c (lM)
PT IC2·c (lM)
2
—
CO2Et
—
H
—
H
F
—
0.19
3.8
3.6
45
6.1
1.8
2.1
38a
38b
38c
39a
39b
39c
40a
40b
40c
43a
43b
43c
46
N-Pyrrolidine
N-Pyrrolidine
OH
CO2Et
CO2Et
H
H
H
H
F
CONH2
CONH2
H
N-Pyrrolidine
N-Pyrrolidine
1.1
0.74
1.5
1.7
2.1
1.8
2.1
1.4
1.9
2.8
4.0
3.9
5.0
2.2
3.4
3.6
H
CONH2
CN
H
H
H
H
H
H
H
H
—
3-(R)-OH-N-pyrrolidine
N-Pyrrolidine
NMe2
H
CN
CN
H
Me
Me
Me
Me
—
NMe2
–CH2-(1-imidazolyl)
–CH2-(1-tetrazolyl)
–CH2-(2-tetrazolyl)
—
N-Pyrrolidine
N-Pyrrolidine
N-Pyrrolidine
—
20.1
53
6.5
62
a All final compounds were purified by prep HPLC and gave satisfactory spectral data.
b Ki values were obtained from purified human enzyme and were averaged from multiple determinations (n = 2), as described in Ref. 6a.
c The aPTT (activated partial thromboplastin time) and PT (prothrombin time) in vitro clotting assays were performed in human plasma as described
in Ref. 6a.

J. M. Fevig et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3755–3760
Table 3. Human enzyme selectivity profile
3759
primary amine 26 were 10- to 15-fold less potent than
20–24. The 1H-pyrazolo-[3,4-d]-pyridazin-7(6H)-one
core analog 32 was also found to have comparable
potency to compounds 20–24.
Enzyme Ki (nM) 39a
39b
40a
Razaxaban
fXa
1.1
0.74
1.7
0.19
540
Thrombin
Trypsin
aPC
1600 1100
>4200 >2500
>76,000 >76,000 48,710
1800
>2500
>10,000
19,700
Recently, inspired by the basic N,N-dimethylamino-
methyl moiety present in razaxaban 2, it was demon-
strated that the sulfonamide group on the terminal
aryl ring of the P4 biphenylsulfonamides could be effec-
tively replaced by a variety of cyclic and acyclic N,N-
dialkylaminomethyl, N-alkylaminomethyl, and related
residues.13 This strategy was also employed in this series,
with the expectation that the resulting compounds might
have greater potency, both in fXa binding and in the
in vitro clotting assays.
fIXa
fVIIa
>41,000 >15,000 >15,000 9000
>15,000 >15,000 >15,000 >15,000
>15,000 >15,000 >15,000 >15,000
>33,000 >33,000 >33,000 >33,000
>13,000 >13,000 >13,000 >13,000
Plasmin
tPA
Urokinase
Chymotrypsin
7930
530
2030
8500
All Kis were obtained from purified human enzymes and are averaged
from multiple determinations (n = 2). See Ref. 6a for more details.
orally (Table 4). The amides 39a and b compare very
favorably with razaxaban, with each having lower
clearance, longer half-life, higher bioavailability, and
comparable volume of distribution.
The SAR for these compounds are shown in Table 2.
In general, there was little difference in fXa potency
upon replacing the terminal ring sulfonamide with
N,N-dialkylaminomethyl residues. Compounds 38a,b,
39a–c, and 40a,b were all potent fXa inhibitors, com-
parable to the corresponding analogs from Table 1.
The alcohol 38c, an intermediate from Scheme 4 with
a non-basic sulfonamide replacement, was an order of
magnitude less potent than the corresponding N-pyr-
rolidine 38a. Compound 40c demonstrates that 5-methyl
substitution is tolerated on the bicyclic core. Despite
no improvement in potency, compounds 38a and
39a–c show slightly enhanced activity in the aPTT as-
say relative to similar sulfonamide analogs from Table
1, suggesting that the basic P4 substitution may be hav-
ing a favorable effect in lowering protein binding. The
aPTT and PT profiles of the 3-amido analogs 39a–c are
comparable to that of razaxaban. Additional 3-substit-
uents, represented by 43a–c, were explored for the pos-
sibility of further improving fXa potency. However,
these compounds offered no potency advantage relative
to the nitrile 40c. Furthermore, the 2-tetrazolylmethyl
analog 43c was less potent in the in vitro clotting as-
says relative to 39a–c, suggesting an unfavorable effect
on protein binding. Compound 46 was equipotent with
the corresponding 1H-pyrazolo-[3,4-d]-pyridazin-7(6H)-
one example 32, but had poor activity in the aPTT and
PT assays, again indicative of high protein binding,
which is likely highly influenced by the 3-trifluoromethyl
substituent.
The higher permeabilities for 39a,b in the Caco-2 assay
correlate with the observed higher bioavailabilities.
These results demonstrate that the 3-amido group is a
well-tolerated substituent on the 1H-pyrazolo[4,3-d]pyr-
imidin-7(6H)-one core. The fluoro substituent on the in-
ner phenyl ring has little effect on the PK profile or
Caco-2 permeability. The nitrile analog 40a has an unfa-
vorable PK profile, with very high clearance and volume
of distribution.
Two of these compounds were also studied in the rab-
bit arterio-venous (A-V) shunt thrombosis model.5b
Upon intravenous dosing, compound 39a inhibited
thrombus formation with an IC50 of 1200 nM and
an ID50 of 5.5 lmol/kg/h (Table 5), which is about
3- to 4-fold less efficacious than razaxaban in this
model. Compound 39b is less potent, achieving only
33% inhibition of thrombus formation at 4.6 lmol/
kg/h. The reduced activity of 39a relative to 39b
may reflect the higher protein binding of 39b, since
the Ki and PT values are essentially equivalent. This
level of activity for 39a and 39b is somewhat surpris-
ing, given the potent fXa inhibition, low clearance in
the dog PK studies, and favorable protein binding in
the rabbit, where these analogs are expected to have
2- to 3-fold higher free fraction relative to razaxaban.
Apparently, their favorable PK and protein binding
are not sufficient to overcome their diminished fXa
potency compared to razaxaban. These results have
led to the conclusion that the fXa binding potency
Compounds with the 4-methoxyphenyl P1 residue have
previously been reported to show high selectivity
for fXa versus thrombin and trypsin.7 Additional
selectivity data are shown for selected compounds in
Table 3. Compounds 39a, 39b, and 40a all showed
>1000-fold selectivity relative to trypsin, aPC, factor
IXa, factor VIIa, plasmin, tPA, plasma kallikrein,
and urokinase. They are less selective relative to throm-
bin and chymotrypsin, but still show >600-fold selec-
tivity in all cases. The overall selectivity profile for
these compounds in most cases is comparable to that
of razaxaban.
Table 4. Dog pharmacokinetic profiles
Compound
Cl
Vdss
t1/2
F
Caco-2
(L/h/kg) (L/kg) (h) (%) (Papp · 10ꢀ6 cm/s)
39aa
39ba
40aa
0.55
0.76
6.8
3.9
4.9
5.4 94
5.1 96
8.3 77
3.4 84
12
15
75
5.3
3.1
5.6
Razaxabanb 1.1
a Compounds were dosed as the TFA salts in an N-in-1 format at
0.5 mg/kg i.v. and 0.2 mg/kg p.o. (n = 2).
b Ref. 6a.
The pharmacokinetic profiles of compounds 39a, b, and
40a were studied in dogs via a cassette dosing format,
with dosing at 0.5 mg/kg intravenously and 0.2 mg/kg

3760
J. M. Fevig et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3755–3760
Table 5. Anticoagulant activity in rabbits
Compound
fXa Ki (rabbit)
(nM)
PT IC2x (rabbit)
(uM)
Rabbit A-V shunt5b IC50
(nM)
Rabbit A-V shunt5b ID50
Protein binding14 (rabbit)
(% bound)
(lmol/kg/h)
39a
39b
Razaxabana
1.0
1.2
1.8
1.6
1.9
1200
>1000b
340
5.5
33% inh @ 4.6 lmol/kg/h
1.6
77
85
93
0.19
a Ref. 6a.
b Concentration at 4.6 lmol/kg/h.
M.; Fuster, V.; Badimon, J. J. Am. J. Cardiovasc. Drugs
2004, 4, 379.
of this series needs to be improved significantly in or-
der to achieve potent activity in the rabbit A-V shunt
thrombosis model.
5. (a) Pinto, D. J. P.; Orwat, M. J.; Wang, S.; Fevig, J. M.;
Quan, M. L.; Amparo, E.; Cacciola, J.; Rossi, K. A.;
Alexander, R. S.; Smallwood, A. M.; Luettgen, J. M.;
Liang, L.; Aungst, B. J.; Wright, M. R.; Knabb, R. M.;
Wong, P. C.; Wexler, R. R.; Wexler, R. R.; Lam, P. Y. S.
J. Med. Chem. 2001, 44, 566; (b) Wong, P. C.; Quan, M.
L.; Crain, E. J.; Watson, C. A.; Wexler, R. R.; Knabb, R.
M. J. Pharmacol. Exp. Ther. 2000, 292, 351.
6. (a) 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; (b) Lessen, M. R.; Davidson, B.
L.; Gallus, A.; Pineo, G.; Ansell, J.; Deitchman, D. Blood
2003, 102, 15a, Abstract 41.
In summary, the first examples of pyrazole-fused
bicyclic core fXa inhibitors have been described. The
1H-pyrazolo[4,3-d]pyrimidin-7(6H)-one core is especial-
ly promising as a means to maintain potent fXa inhibi-
tion, while also reducing the probability that in vivo
amide hydrolysis will liberate an aniline fragment.
Within this series, the 3-amido analogs 39a and 39b
are not only potent fXa inhibitors, but they are also
highly selective versus relevant serine protease inhibi-
tors, have excellent pharmacokinetic profiles, and have
relatively low protein binding. Unfortunately, these
compounds were not highly efficacious in the rabbit
A-V shunt thrombosis model, perhaps because they
are not potent enough versus fXa, both compounds
being about 5-fold less potent than razaxaban. Further
efforts to increase the potency of this series will be
described in due course.
7. Pruitt, J. R.; Pinto, D. J. P.; Galemmo, R. A., Jr.;
Alexander, R. S.; Rossi, K. A.; Wells, B. L.; Drummond,
S.; Bostrom, L. L.; Burdick, D.; Bruckner, R.; Chen, H.;
Smallwood, A.; Wong, P. C.; Wright, M. R.; Bai, S.;
Luettgen, J. M.; Knabb, R. M.; Lam, P. Y. S.; Wexler, R.
R. J. Med. Chem. 2003, 46, 5298.
8. Debnath, A. K.; Debnath, G.; Shusterman, A. J.; Hansch,
C. Environ. Mol. Mutagen 1992, 19, 37.
9. For a preliminary account of part of this work, see:
Fevig, J. M.; Cacciola, J.; Buriak, J., Jr.; Li, Y.-L.; Pinto,
D. J.; Orwat, M. J.; Galemmo, R. A., Jr.; Wells, B.; Li,
R.; Rossi, K. A.; Knabb, R. M.; Luettgen, J. M.; Wong,
P. C.; Bai, S.; Wexler, R. R.; Lam, P. Y. S. Abstracts of
Papers, 230th National Meeting of the American Chem-
ical Society, Washington, DC, Aug 28–Sept 1, 2005;
American Chemical Society: Washington, DC, 2005;
MEDI 279.
Acknowledgments
The authors thank Bruce Aungst, Frank Barbera, Tracy
Bozarth, Earl Crain, Andrew Leamy, Dale McCall, and
Carol Watson for technical assistance.
References and notes
10. Lam, P. Y. S.; Clark, C. G.; Li, R.; Pinto, D. J. P.; Orwat,
M. J.; Galemmo, R. A.; Fevig, J. M.; Teleha, C. A.;
Alexander, R. S.; Smallwood, A. M.; Rossi, K. A.;
Wright, M. R.; Bai, S. A.; He, K.; Luettgen, J. M.; Wong,
P. C.; Knabb, R. M.; Wexler, R. R. J. Med. Chem. 2003,
46, 4405.
1. (a) Hirsh, J.; O’Donell, M.; Weitz, J. I. Blood 2005, 105,
453; (b) Golino, P.; Loffredo, F.; Riegler, L.; Renzullo, E.;
Cocchia, R. Curr. Opin. Invest. Drugs 2005, 6, 298; (c)
Quan, M. L; Smallheer, J. J. Curr. Opin. In Drug Disc.
Develop 2004, 7, 460.
11. Tanaka, K.; Maeno, S.; Mitsuhashi, K. Chem. Lett. 1982,
543.
12. Compound 44 was prepared in 50% yield from 14 (Y = H)
with tin(II) chloride hydrate in warm isopropanol, which
prevented the concomitant conversion of the nitrile to an
ester.
13. Quan, M. L.; Han, Q.; Fevig, J. M.; Lam, P. Y. S.;
Bai, S.; Knabb, R. M.; Luettgen, J. M.; Wong, P. C.;
Wexler, R. R. Bioorg. Med. Chem. Lett. 2006, 16,
1795.
2. (a) Drout, L.; Bal dit Sollier, C. Eur. J. Clin. Invest 2005,
35(Suppl. 1), 21; (b) Mann, K. G.; Butenas, S.; Brummel,
K. Arterioscler. Thromb. Vasc. Biol. 2003, 23, 17; (c)
Leadley, R. J., Jr. Curr. Top. Med. Chem. 2001, 1, 151.
3. (a) 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;
(b) Wong, P. C.; Pinto, D. J.; Knabb, R. M. Cardiovasc.
Drug Rev. 2002, 20, 137.
4. (a) Rajagopal, V.; Bhatt, D. L. J. Thromb. Haem. 2005, 3,
436; (b) Viles-Gonzalez, J. F.; Gaztanaga, J.; Zafar, U.
14. Pacific, G. M.; Viani, A. Clin. Pharmacokin. 1992, 23, 449.