Design, synthesis and biological activity of novel non-amidine factor Xa inhibitors. Part 1: P1 structure-activity relationships of the substituted 1-(2-Naphthyl)-1H-pyrazole-5-carboxylamides

Article abstract of DOI:10.1016/S0960-894X(02)00239-1

Based on DuPont Pharmaceuticals' monobenzamidine lead structure SN429, we have designed the biphenyl 1-(2-naphthyl)-1H-pyrazole-5-carboxylamides as a novel series of non-basic factor Xa inhibitors. We have discovered that the displacement of the benzamidine moiety with substituted 2-naphthyl structures not only results in highly potent factor Xa inhibitors, but also significantly increases their enzyme specificity and oral bioavailability.

Full text of DOI:10.1016/S0960-894X(02)00239-1

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1651–1655
Design, Synthesis and Biological Activity of Novel Non-Amidine
Factor Xa Inhibitors. Part 1:P Structure–Activity Relationships
of the Substituted 1-(2-Naphthyl)-1H-pyrazole-5-carboxylamides
1
Zhaozhong J. Jia,a,* Yanhong Wu,a Wenrong Huang,a Erick Goldman,a Penglie Zhang,a
John Woolfrey,a Paul Wong,b Brian Huang,b Uma Sinha,b Gary Park,b Andrea Reed,b
Robert M. Scarborougha and Bing-Yan Zhua,*
aMedicinal Chemistry Department, Millennium Pharmaceuticals, Inc., 256 East Grand Ave, South San Francisco, CA 94080, USA
bBiology Department, Millennium Pharmaceuticals, Inc., 256 East Grand Ave, South San Francisco, CA 94080, USA
Received 29 November 2001; accepted 29 March 2002
Abstract—Based on DuPont Pharmaceuticals’ monobenzamidine lead structure SN429, we have designed the biphenyl 1-(2-naph-
thyl)-1H-pyrazole-5-carboxylamides as a novel series of non-basic factor Xa inhibitors. We have discovered that the displacement
of the benzamidine moiety with substituted 2-naphthyl structures not only results in highly potent factor Xa inhibitors, but also
significantly increases their enzyme specificity and oral bioavailability. # 2002 Elsevier Science Ltd. All rights reserved.
Cardiovascular diseases involving acute occlusive
thrombosis are a leading cause of death worldwide.
Warfarin, the only orally bioavailable anticoagulant on
the market for thrombotic disorders, suffers from a
number of shortcomings which limit its application.
With a narrow therapeutic index, it requires careful and
costly monitoring of clotting time to achieve desired
efficacy and dose titration to minimize excessive bleed-
ing and potential drug–drug interactions.1 Blood co-
agulation factor Xa (fXa), a trypsin-like serine protease
that plays a pivotal role in the blood coagulation cas-
cade, has emerged as a very attractive target for the
design of new therapeutic agents with potential for the
treatment of arterial and venous thrombosis.2 At the
final convergence point of both the intrinsic and extrin-
sic coagulation pathways, fXa forms a prothrombinase
complex with factor Va, Ca2+, and phospholipid.3 FXa
is the active enzyme in the prothrombinase complex,
converting inactive prothrombin to active thrombin.
This activation is a highly amplified process.4 One
molecule of fXa in the prothrombinase complex can
generate 138 molecules of thrombin per minute. There-
fore, inhibition of fXa not only interrupts both the
coagulation pathways of thrombin production, but also
is theoretically more therapeutically effective to prevent
thrombosis than direct thrombin inhibition. As sug-
gested by preclinical studies, fXa inhibitors have less
potential for increasing the risk of abnormal bleeding
and have a wider therapeutic index. Representing an
important advance in the management of both arterial
and venous thrombosis, fXa inhibitors have been
actively pursued as new orally active antithrombotic
agents. A promising fXa inhibitor must demonstrate
adequate potency, enzyme specificity, oral bioavail-
ability, duration of action (half-life, t1/2) and efficacy. It
should be highly specific for fXa versus thrombin, tryp-
sin and other enzymes such as tissue plasminogen acti-
vator (t-PA), activated protein C (aPC), plasmin and
kallikrein.
In early 1999, the recurring structural feature of many
fXa inhibitors was the presence of a benzamidine moi-
ety. One of the most potent P1 monobenzamidine fXa
inhibitors was DuPont’s SN429.5 Although it indicated
a very low in vitro human fXa Ki at 0.013 nM, SN429
was not a promising orally active antithrombotic agent
due to its poor oral bioavailability (4% in dog) and
short half-life (0.82 h in dog). The amidino group is
likely a major contributor to the low oral bioavailability
due to its high basicity. In order to improve oral bio-
availability, clearly the amidino group should be
replaced with less basic moieties. It should be pointed
out that SN429 also inhibits thrombin (Ki 300 nM),
trypsin (Ki 16 nM) and kallikrein.
*Corresponding authors. Fax.: +1-650-244-9287; e-mail: zhaozhong.
jia@mpi.com (Z.J. Jia); bing-yan.zhu@mpi.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00239-1

1652
Z. J. Jia et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1651–1655
Our approach toward non-benzamidine fXa inhibitors
was to substitute the benzamidine moiety with a less
basic or totally neutral replacement. We envisioned the
neutral 2-nathphyl group as a novel replacement for the
P1 benzamidine in the fXa S1 pocket. SN429 presented
to us a useful template to exercise this hypothesis. For
the proof of principal, compound 1 was designed. The
aromaticity and planarity of the 2-naphthyl group mat-
ched the predicted shape of the S1 pocket very well.
While a reduction in fXa affinity was predicted relative
to SN429 due to the elimination of an important
hydrogen bond interaction of the benzamidine with fXa
Asp189 carboxylate residue, the potentially improved
oral bioavailability and pharmacokinetic profile were
enticing. Compound 1’s enzyme specificity profile
should be better than that of SN429 by eliminating the
additional hydrogen bond interaction of the P1 benza-
midine with the trypsin Ser190 hydroxyl group. The
corresponding fXa Ala190 residue contributed to the
hydrophobic character of the S1 pocket, which should
favorably affect binding of the P1 naphthyl moiety.6 The
2-naphthyl ring system also presented multiple potential
sites for substitution and derivatization in order to
increase the fXa affinity.
To increase fXa potency, we carried out a systematic
structure–activity relationship (SAR) exploration of the
substitution on the 2-naphthyl moiety. The proximity of
the C-3 position to the available Ser195 hydroxyl group
suggested potential hydrogen bond interactions could
be gained by appropriate substitutions. This interaction
might be direct or mediated through a water molecule.
A variety of groups with hydrogen bond donating
and/or accepting ability were considered for Z3, includ-
ing halides, –CN, –OH, –NH2, –NO2, –OMe, –SO2Me,
–SO2NH2, –CO2H, –CO2R, –CONH2, –C(NH)NH2,
–C(NH)NR2, –CH2OH and –CH2NH2. The depth of
the fXa S1 pocket, generally defined by the location of
Tyr129, provided impetus for additional SAR explora-
tion by substituting at position C-6. Favorable van der
Waals and potential hydrogen bond interactions were
considered in choosing halides, –OH and –OMe as Z6
substituents.
The 3-fluoronaphthyl target 7 was prepared according
to Scheme 2. Schiemann reaction was used to convert
the amine to fluoride in over 80% yield. Curtius rear-
rangement cleanly converted the carboxylic acid to
amine. The hydrazine salt was produced by SnCl2-
reduction of the diazonium salt and was used without
purification. Pyrazole 6 was isolated in >70% yield.
Compound 1 was synthesized according to the route
shown in Scheme 1. Biphenylamine 2 was prepared in
70 to 85% yield by Suzuki coupling.5 The condensation
of 3 with the 2-naphthylhydrazine afforded the desired
pyrazole product in excellent regioselectivity. Weinreb
method7 was employed to build the amide connection.
The fXa IC50 value of compound 1 is 104 nM. Though
much less potent than SN429, compound 1 exhibited
significantly improved enzyme specificity and pharma-
cokinetic profile. Its IC50 values against thrombin,
trypsin, t-PA, aPC, plasmin, and kallikrein are all
higher than an 11 mM threshold. By orally dosing Spra-
gue–Dawley rats, it displayed 24.5% oral bioavailability
and 2.4 h half-life.
Scheme 2. (a) (1) NaNO2 (1.2 equiv), concd HCl, À10 ꢀC, 30 m; (2)
HBF4 (48% aq, 2 equiv), À10 ꢀC, 1 h; filtration; (b) xylene, reflux, 4 h;
(c) (COCl)2 (3 equiv), DMF (cat.), DCM, 3 h; evaporation; (d) NaN3
(2 equiv), dioxane, À5 ꢀC, 2 h; evaporation; workup; (e) DMF/water
(2:1), reflux, 4 h; (f) (1) NaNO2 (1.2 equiv), concd HCl, À5 ꢀC, 30 m;
ꢀ
.
(2) SnCl2 2H2O (2.5 equiv), concd HCl, À5 C, 1 h; filtration; (g) 3
(1.5 equiv), THF/HOAc (1:2), reflux, 2 h; (h) 2 (1 equiv), AlMe3 (2 M
in hexane, 5 equiv), DCM, overnight; (i) TFA.
The 3-bromo and 3-hydroxy targets (8 and 9) were pre-
pared in a manner similar to that described in Scheme 2
using the corresponding commercial 2-naphthylamines.
Scheme 1. (a) (1) BuLi (1.6 M in hexane, 2 equiv), THF, 0 ꢀC; (2)
(iPrO)3B (1.4 equiv), 10 ꢀC, 1 h; (3) HCl (1 N), 0 ꢀC, overnight; extrac-
tion; (b) 4-bromoaniline (1 equiv), Pd(Pꢀh3P)4 (0.1 equiv), Cs2CO3
.
(3 equiv), n-BuOH/H2O/PhMe (1:2:4), 80 C, 5 h; (c) H2NOMe HCl
(1.1 equiv), EtOH, reflux, 2 h; (d) 2-naphthylhydrazine.HCl (1 equiv),
THF/HOAc (1:2), reflux, 3 h; (e) 2 (1 equiv), AlMe3 (2 M in hexane,
5 equiv), DCM, overnight; (f) TFA.

Z. J. Jia et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1651–1655
1653
Compounds derived from Z3 carboxylic acid and nitrile
were synthesized according to the routes shown in
Scheme 3. Direct methylation and reduction of car-
boxylic acid 10 yielded 11 and 12, respectively. To
accomplish the 3-cyano substitution, the 3-carboxylic
acid was first converted to the corresponding amide,
which was then dehydrated using triphosgene. The
3-cyano function in 13 however was labile in strong
acid. Removal of the t-butyl by TFA afforded the cor-
responding amide 15 along with 14, even at room tem-
perature and in lower concentration. Clean deprotection
.
of 13 was achieved using BF3 OEt2. The 3-aminomethyl
target 16 was prepared via reduction of nitrile 13. Ami-
dines 18, 19, and 20 were prepared via the methyl imi-
date 17, freshly produced by saturating a methanolic
solution of nitrile 13 with dry HCl gas.
The 3-methylsulfonyl and 3-aminosulfonyl targets (23
and 24) were prepared from 3-fluoro intermediate 6
according to Scheme 4. A one-pot reaction employing
Huang’s method8 implemented the conversion of
methylsulfone 21 to sulfonamide 22 in high yield. The
couplings of the carboxylic acids and biphenylamine 2
were promoted by POCl3.
The preparation of 6-bromo, 6-chloro, 6-methoxy and
6-hydroxy targets (25, 26, 27, and 28) is shown in Scheme
5. Boron tribromide was used to unmask the 6-hydroxy.
Scheme 3. (a) (1) NaNO2 (1.2 equiv), concd HCl, À5 ꢀC, 30 m; (2)
The fXa inhibition data for the substituted 2-naphthyl
inhibitors are shown in Table 1. For C-3 substitution,
fluoro (7), cyano (14), amide (15), amidine (18), amino-
methyl (16), methylsulfone (23) and sulfonamide (24)
increase fXa potency 3- to 15-fold; conversely, hydroxy
(9), hydoxymethyl (12), carboxylic acid (10), carboxy-
late (11), and substituted amidines (19 and 20) decrease
fXa potency. For C-6 substitution, chloro (26) and
bromo (25) increase fXa potency 6- to 10-fold; methoxy
(27) and hydroxy (28) decrease fXa potency. It should
be noted that none of the compounds display an IC50
value lower than 11 mM against thrombin, trypsin, t-PA,
SnCl2 2H2O (3 equiv), concd HCl, À5 ꢀC, 1 h; filtration; (b)
3
.
(1.5 equiv), THF/HOAc (1:2), reflux, overnight; (c) 2 (1 equiv), AlMe3
(2 M in hexane, 5 equiv), DCM, overnight; (d) TFA; (e) TMSCH2N2
(2 M in hexane, 2 equiv), THF, 2 h; (f) BH3 (1 M in THF, 4 equiv),
THF, 0 ꢀC, 1 h; (g) NH3 (g), BOP (1.2 equiv), DMF, 10 m; (h)
(Cl3CO)2CO (1.2 equiv), Et3N (2 equiv), DCM, 1 h; (i) BF3OEt2
(30 equiv), DCM, 2 days; (j) NaBH4 (8 equiv), CoCl2 (2 equiv), DMF,
overnight; (k) HCl (g), MeOH, 0 ꢀC to rt, overnight; (l) NH4OAc
(5 equiv), MeOH, reflux, 1 h; (m) HNMe2 (2M in MeOH, 5 equiv),
MeOH, reflux, 1 h; (n) MeNHCH2CH2NH2 (5 equiv), MeOH, reflux,
1 h.
Scheme 5. (a) (COCl)2 (3 equiv), DMF (cat), DCM, 3 h; evaporation;
(b) NaN3 (2 equiv), dioxane, À5 ꢀC, 2 h; evaporation; workup; (c)
DMF/water (2:1), reflux, 4 h; (d) (1) NaNO2 (1.2 equivꢀ), concd HCl,
Scheme 4. (a) NaSMe (4 equiv), DMSO, 100 ꢀC, 1 h; (b) MCPBA
(2.5 equiv), DCM, 1 h; (c) (1) LiHMDS(1 M in hexane, 4 equiv), THF,
0 ꢀC, 30 m; (2) Bu3B (5 equiv), rt, 2 h; then 90 ꢀC, 2 h; cool to rt; (3)
NaOAc (10 equiv), water, H2NOSO3H (8 equiv), 18 h; removal of
THF; concd HCl (till pH 1); extraction; (d) 2 (1 equiv), POCl3
(3 equiv), pyr, 0 ꢀC, 1 h; (e) TFA.
ꢀ
.
À5 C, 30 m; (2) SnCl2 2H2O (3 equiv), concd HCl, À5 C, 1 h; filtra-
tion; (e)
3 (1.5 equiv), THF/HOAc (1:2), reflux, 4 h; (f) CuCl
(10 equiv), CuI (1 equiv), DMF, reflux, 1 h; (g) 2 (1 equiv), AlMe3 (2 M
in hexane, 5 equiv), DCM, overnight; (h) TFA; (i) BBr3 (1 M in DCM,
1.5 equiv), DCM, overnight.

1654
Z. J. Jia et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1651–1655
Table 2. 3-Pyrazole substitution on fXa potency
Table 1. 2-Naphthyl substitution on fXa potency
Compd
Y
Z3
Z6
IC50 (nM)
Compd
Z3
Z6
IC50 (nM)
26
30
7
CH3
CF3
CH3
CF3
H
H
F
Cl
Cl
H
11
4
9
27
1
7
8
14
9
H
F
Br
CN
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Br
OMe
OH
104
9
143
6
31
F
H
OH
>11,000
483
38
12
16
10
11
15
23
24
18
19
20
26
25
27
28
CH2OH
CH2NH2
CO2H
CO2Me
CONH2
SO2Me
SO2NH2
C(NH)NH2
C(NH)NMe2
2-Imidazolinyl
H
376
545
24
45
24
pyrazole formation. The furyl ring acted as a protected
carboxylic acid, and was conveniently oxidized to the
free acid using KMnO4 in hot acetone and water.
21
The fXa potency results for 3-methyl versus 3-tri-
fluoromethyl substitution in the pyrazole nucleus are
listed in Table 2. The trifluoromethyl group does not
compromise the inhibitors’ excellent enzyme specificity
toward fXa.
1650
7820
11
18
3300
3510
H
H
H
The pharmacokinetic profile of this class of novel fXa
inhibitors was explored by examining the representative
analogue 7 (fXa Ki 5.7 nM). By orally dosing conscious
Sprague–Dawley rats at a nominal dose of 4 mg/kg in
25% PEG-300, the oral bioavailability and half-life of
compound 7 are 22% and 6.0 h, respectively. Intrave-
nous administration to the conscious rats at a dose of
0.2 mg/kg in 50% PEG-300 demonstrated half-life,
volume distribution and clearance values of 1.7 h, 2.6 L/kg
and 17.7 mL/min/kg, respectively.
In order to compare 2-naphthyl with other neutral aro-
matic bicyclic structures as P1 moieties in fXa inhibitor
design, compound 33 was prepared from commercial
ester 32 using the same chemistry as illustrated above.
The lack of inhibiting activity against fXa potency (IC50
>11 mM) reinforced our focus on the 2-naphthyl P1 fXa
inhibitors.
Scheme 6. (a) THF/HOAc (1:2), reflux, 4 h; (b) CuCl (10 equiv), CuI
(1 equiv), DMF, reflux, 1 h; (c) KMnO4 (5 equiv), acetone/water (2:1),
60 ꢀC, 3 h; (d) 2 (1 equiv), POCl3 (3 equiv), pyr, 0 ꢀC, 1 h; (e) TFA.
aPC, plasmin and kallikrein. Compared to SN429, the
enzyme specificity for fXa has been significantly
improved by incorporating substituted 2-naphthyls as
the P1 moiety.
In conclusion, we have designed a series of biphenyl 1-
(2-naphthyl)-1H-pyrazole-5-carboxylamides as novel
non-amidine fXa inhibitors possessing low nanomolar
fXa potency, excellent enzyme specificity and respect-
able oral bioavailability and half-life. Several C-3 and
C-6 2-naphthyl substitution patterns were discovered to
increase fXa potency. Further SAR study and optimi-
zation of the biphenyl/biaryl P4 and other novel P4
structures, leading to subnanomolar fXa inhibitors, will
be reported in future publications.
We also examined the 3-(trifluoromethyl)-1H-pyrazole
nucleus since it has been known as a potency-enhancing
replacement for 3-methyl-1H-pyrazole in fXa inhibitor
design.2 The 6-chloro and 3-fluoro targets (30 and 31)
were prepared according to Scheme 6. Commercial
diketone 29 was employed in the trifluoromethyl

Z. J. Jia et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1651–1655
1655
References and Notes
Pruitt, J. R.; Rossi, K. A.; Alexander, R. S.; Fevig, J. M.;
Cacciola, J.; Lam, P. Y. S.; Knabb, R. M.; Wong, P. C.;
Wexler, R. R. Book of Abstracts, 217th ACSNational Meet-
ing, Anaheim, CA, 1999; MEDI-006. (b) Pinto, D. J.; Orwat,
M. J.; Wang, S.; Fevig, J. M.; Quan, M. L.; Amparo, E. C.;
Cacciola, J.; Rossi, K. A.; Alexander, R. S.; Smallwood, A. S.;
Luettgen, J. M.; Liang, L.; Aungst, B. J.; Wright, M. R.;
Knabb, R. M.; Wong, P. C.; Wexler, R. R.; Lam, P. Y. S. J.
Med. Chem. 2001, 44, 566.
1. Bona, R. D.; Hicky, A. D.; Wallace, D. M. Thromb. Hae-
most. 1997, 78, 137 and references therein.
2. (a) Sanderson, P. E. J. Annu. Rep. Med. Chem. 2001, 36, 79.
(b) Zhu, B.-Y.; Scarborough, R. M. Annu. Rep. Med. Chem.
2000, 35, 83. (c) Zhu, B.-Y.; Scarborough, R. M. Curr. Opin.
Car. Pulmon. Renal Invest. Drugs 1999, 1, 63. (d) Leadley, R. J.
Curr. Top. Med. Chem. 2001, 1, 151. (e) Quan, M. L.; Wexler,
R. R. Curr. Top. Med. Chem. 2001, 1, 137 and references
therein.
6. Maignan, S.; Mikol, V. Curr. Top. Med. Chem. 2001, 1,
161.
3. Davre, E. W.; Fujikawa, K.; Kisiel, W. Biochemistry 1991,
30, 10363.
7. Basha, A.; Lipton, M.; Weinreb, S. M. Tetrahedron Lett.
1977, 18, 4171.
4. Elodi, S.; Varadi, K. Thromb. Res. 1979, 15, 617.
5. (a) Pinto, D. J.; Orwat, M. J.; Wang, S.; Amparo, E.;
8. Huang, H.-C.; Reinhard, E. J.; Reitz, D. B. Tetrahedron
Lett. 1994, 35, 7201.