Discovery of diarylurea P2Y1 antagonists with improved aqueous solubility

Article abstract of DOI:10.1016/j.bmcl.2013.03.125

Preclinical data suggests that P2Y₁ antagonists, such as diarylurea compound 1, may provide antithrombotic efficacy similar to P2Y ₁₂ antagonists and may have the potential of providing reduced bleeding liabilities. This manuscript d

Full text of DOI:10.1016/j.bmcl.2013.03.125

Bioorganic & Medicinal Chemistry Letters 23 (2013) 3239–3243
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of diarylurea P2Y1 antagonists with improved aqueous
solubility
⇑
Tammy C. Wang, Jennifer X. Qiao , Charles G. Clark, Ji Jua, Laura A. Price, Qimin Wu, Ming Chang,
Joanna Zheng, Christine S. Huang, Gerry Everlof, William A. Schumacher, Pancras C. Wong,
Dietmar A. Seiffert, Anne B. Stewart, Jeffrey S. Bostwick, Earl J. Crain, Carol A. Watson, Robert Rehfuss,
Ruth R. Wexler , Patrick Y. S. Lam
Medicinal Chemistry, Molecular Sciences and Candidate Optimization, Bristol-Myers Squibb, 311 Pennington-Rocky Hill Road, Pennington, NJ 08534, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Preclinical data suggests that P2Y1 antagonists, such as diarylurea compound 1, may provide antithrom-
botic efficacy similar to P2Y12 antagonists and may have the potential of providing reduced bleeding lia-
bilities. This manuscript describes a series of diarylureas bearing solublizing amine side chains as potent
P2Y1 antagonists. Among them, compounds 2l and 3h had improved aqueous solubility and maintained
antiplatelet activity compared with compound 1. Compound 2l was moderately efficacious in both rat
and rabbit thrombosis models and had a moderate prolongation of bleeding time in rats similar to that
of compound 1.
Received 20 February 2013
Revised 22 March 2013
Accepted 27 March 2013
Available online 6 April 2013
Keywords:
P2Y1 antagonists
Diarylurea
Ó 2013 Elsevier Ltd. All rights reserved.
Structure–activity relationship (SAR)
Aqueous solubility
In vivo efficacy/bleeding model
Recent advancement in platelet biology reveals that adenosine-
50-diphosphate (ADP) plays a crucial role in platelet function. There
are two distinct G protein-coupled P2 purinergic ADP receptors:
P2Y1 and P2Y12. P2Y12, through a Gi-dependent pathway, is respon-
sible for mediating the inhibition of adenylyl cyclase, which sus-
tains platelet aggregation; and P2Y1, through a Gq-dependent
pathway, is responsible for calcium activation, which causes plate-
let shape change and initiates rapidly reversible aggregation. Syn-
ergistic activation of both P2Y1 and P2Y12 is essential for a
complete platelet response and blockade of either receptor signif-
icantly decreases both in vitro and ex vivo ADP-induced platelet
aggregation and thrombosis formation.1
The irreversible P2Y12 antagonist, thienopyridine analog clopi-
dogrel (PlavixÒ) is in world-wide clinical use to treat peripheral ar-
tery disease and acute coronary syndrome, and is also used for the
secondary prevention of ischemic stroke, vascular death, and myo-
cardial infarction. Combination therapy of clopidogrel with aspirin
is the current standard of care for most patients. Several new P2Y12
antagonists, such as prasugrel and ticagrelor, were recently
approved.2
On the other hand, preclinical studies using P2Y1-deficient
(P2Y1ꢀ/ꢀ) mice and selective nucleotide-derived small molecule
P2Y1 antagonists (such as MRS-2500) demonstrated a complete
blockade of ADP-induced platelet aggregation and an effective
reduction of arterial thrombosis with a moderate prolongation of
the bleeding time.3,4 Our in-house preclinical data also suggests
that P2Y1 and P2Y12 inhibition provides equivalent antithrombotic
efficacy, while targeting P2Y1 has the potential of providing a re-
duced bleeding liability.5a–d P2Y1 is therefore a promising target
for new antithrombotic drugs. In addition, recent studies also sug-
gested P2Y1 to be a target for atherothrombosis,6,7 showing that
the P2Y1 receptor may be involved in atherosclerosis and may play
a role in inflammatory events such as leukocyte recruitment in in-
flamed mouse arteries.
X
R
OCF3
O
O
N
N
N
H
N
H
N
H
N
H
O
O
1
R = amine-containing solublizing groups
X = H, F
BMS researchers recently reported a series of non-nucleotide small
molecule P2Y1 antagonists, represented by the diphenyl urea analog
1.5d,8 Compound 1 showed good binding affinity in washed platelets
⇑
Corresponding author. Tel.: +1 609 818 5298; fax: +1 609 818 6810.
E-mail address: jennifer.qiao@bms.com (J.X. Qiao).
Present address: Lam Drug Discovery Consulting LLC, PA, USA.
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.03.125

3240
T. C. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3239–3243
(P2Y1 Ki = 6 nM) and moderate in vitro antiplatelet activity in the
ADP-induced platelet aggregation assay in platelet-enriched plasma
(PA IC50 = 2.1 lM using 2.5 lM of ADP). Compound 1 was the first in-
One of the ‘‘sweet spots’’ is the trifluoromethoxy portion of the
urea phenyl ring. This paper describes our efforts to modify the tri-
fluoromethoxy portion of the phenyl ring in compound 1, resulting
in a series of novel P2Y1 antagonists with much increased aqueous
solubility, some of which with maintained binding affinity and
antiplatelet functional activity. The in vivo efficacy and bleeding
profile of a representative soluble diarylurea compound (2l) will
also be presented.
Table 1 shows that replacing the trifluoromethyl group with
hydrophobic groups, such as cyclopropyl ether 2a and cyclopropyl
ester 2b, maintained the binding affinity and the functional po-
tency. The neutral polar groups, such as alcohols in 2c and 2e, on
the other hand, led to significant loss of the P2Y1 binding affinity.
In spite of a great improvement of aqueous solubility and HPLC
logP, the pyrrolidinyl analog 2d showed more than 200-fold de-
crease of the binding affinity. Basic piperazines 2f (NH) and 2g
(NMe) demonstrated decreased binding affinities consistent with
aforementioned general SAR trends. However, incorporation of an
additional lipophilic phenyl group in 2h (NBn) regained both the
binding and the functional potency but decreased solubility. Inter-
estingly, compounds bearing some of the polar functional groups,
such as piperidine 2i and morpholine 2j, showed only a slight de-
crease of binding affinity and platelet aggregation activity com-
house molecule to demonstrate a reduction in thrombus weight and
less of a bleeding liability than a P2Y12 antagonist in the rat models of
thrombosis and bleeding. However, lack of solubility of compound 1
in vehicles compatible with rabbit models precluded it from being
studied in rabbits. We decided to improve the physical properties
of compound 1, in particular, to increase aqueous solubility and re-
duce lipophilicity (HPLC logP) while maintaining the antiplatelet
activity as measured in the in vitro platelet aggregation assay.
Diarylurea moieties like that in compound 1 are usually associ-
ated with poor aqueous solubility, partly due to ease of forming
intermolecular hydrogen bonding and the co-planarity of the two
aryl and urea groups in crystal packing. Replacing the urea linker
of compound 1 with other two to four atom linkers resulted in sig-
nificant loss of potency. Therefore, we searched for solubilizing
side chains to append to compound 1 to improve its physical prop-
erties. In general, adding polar or basic amine groups to different
parts of compound 1 resulted in significant loss of binding affinity
(data not shown). After comprehensive SAR studies, we identified a
couple of ‘‘sweet spots’’ where polar groups, in particular, basic
amines, may be tolerated.
Table 1
In vitro P2Y1 binding, functional platelet aggregation (PA) activity, aqueous solubility and experimental HPLC logP9 of compounds 2a–n
X
R
O
N
N
H
N
H
O
Entry
X
R
P2Y1 Ki (nM)a
PA IC50
2.1
(l
M)b
Solubility (
l
g/mL)c
HPLC logP9
1
H
OCF3
6
<1
5.7
2a
2b
H
H
18
7
3.2
3.1
<1⁄⁄
1⁄⁄
5.2
5.1
O
O
O
2c
H
H
176
–
–
5⁄⁄⁄
4.6
3.7
OH
N
2d
1396
355⁄⁄
OH
2e
F
54
91
2⁄
5.5
N
N
NH
2f
H
H
8041
592
–
–
62⁄
3.2
3.9
N
2g
51⁄⁄
Ph
2h
H
7
3.3
5⁄
5.8
N
N
N
N
2i
H
F
11
4.4
3.0
16
9⁄⁄
5.5
4.5
4.2
5⁄⁄⁄
160⁄
O
2j
25
N
2k
F
322
N
2l
F
17
2.9
173⁄⁄, 38⁄
5.3
N
2m
2n
F
F
34
37
10.9
2.3
17⁄
7⁄
5.5
5.4
N
a
b
c
Ki values are reported as the mean of at least three individual determinations. Variability around the mean was <50%.
Platelet aggregation (PA) was tested with 2.5 M of ADP, n P 3.
Aqueous solubility was tested at pH 6.5 using 25
M phosphate buffer with either crystalline (⁄), partially crystalline (⁄⁄), or amorphous materials (⁄⁄⁄).
l
l

T. C. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3239–3243
3241
Table 2
In vitro P2Y1 binding, functional platelet aggregation (PA) activity, aqueous solubility and experimental HPLC logP9 of compounds 3a–h
X
R
O
N
N
H
N
H
O
Entry
1
X
R
P2Y1 Ki (nM)a
PA IC50
2.1
(l
M)b
Solubility (
l
g/ml)c
HPLC logP9
H
H
OCF3
6
<1⁄
103⁄
5.7
5.1
3a
137
4.6
O
N
N
3b t-Bu to Si(CH3)3
H
F
163
52
5.5
4.2
173⁄⁄
204⁄
4.7
5.2
O
O
N
3c
N
Ph
3d
3e
F
F
11
4.5
–
2⁄
5.8
3.8
O
232⁄
N
989
O
N
3f
F
121
–
87⁄
5.5
O
O
O
N
N
3g
3h
H
F
87
30
5.6
2.9
27⁄⁄
4.5
5.2
102⁄⁄
a
b
c
Ki values are reported as the mean of at least three individual determinations. Variability around the mean was <50%.
Platelet aggregation (PA) was tested with 2.5 M of ADP, n P 3.
Aqueous solubility was tested at pH 6.5 using 25
M phosphate buffer with either crystalline (⁄), partially crystalline (⁄⁄), or amorphous materials (⁄⁄⁄).
l
l
Table 3
In vitro and rat PK profile of diarylureas 1, 2l, and 3h
Cpd
number
P2Y1 Ki
(nM)
PA IC50
M)
Human PB %
bound
Rat PB %
bound
RLMd Caco-2 Pc (A and B)
PAMPA CL (mL/min/
Vss (L/
kg)
t1/2
(h)
F
(%)
(
l
(nm/s)
kg)
1a
6
17
30
2.1
2.9
2.9
99.3
99.85
99.7
>99.8
99.56
–
86
24
67
21
–
16
–
–
103
16
158
22
5.6
8.7
5
15
1.5
3.4
7
–
40
2lb
3hc
a
b
c
Dose: 1 mg/kg (IV/PO), vehicle: 10% DMAC/10% EtOH/10% cremophor/70% H2O.
Dose: 1 mg/kg (IV), vehicle: 10% EtOH/10% cremophor/80% H2O.
Dose: 2 mg/kg (IV), 5 mg/kg (PO), vehicle: 10%EtOH/10% cremophor/80% H2O.
Percentage remaining after 10 min rat liver microsomal (RLM) incubation.
d
pared with compound 1. On the other hand, the more basic N,N-
dimethylaminomethyl analog 2k, having a significantly enhanced
were the two major metabolic pathways. As a result, clearance of
compound 2l was very high in rats. In addition, in the GPCR panel
aqueous solubility (160
lg/ml solubility for the crystalline mate-
testing, compound 2l showed ca. 90% inhibition (at 10 lM) against
rial), was 60-fold less potent in binding but showed only eight-fold
decrease in antiplatelet activity compared with 1. We then incor-
porated the N,N-dimethylaminomethyl side chain into the ortho,
meta, and para position of the biphenyl system (2l, 2m, 2n) to
try to further improve potency. Compound 2l was subsequently
identified as the lead compound with activities comparable to that
of compound 1 (human P2Y1 Ki of 17 nM and in vitro PA IC50 of
2.9
ubility (173
showed good antiplatelet activity in rats (PA IC50 of 0.8
10 M of ADP) and in rabbits (PA IC50 of 1.6 M at 10
ADP). It was selective against P2Y2 (Ki >30
l
M (at 2.5
lM of ADP) but significantly increased aqueous sol-
l
g/ml for partially crystalline material). Compound 2l
l
l
M at
M of
l
l
l
M) and P2Y12 (Ki
M), and blood-coagulation factors (fXa Ki >25 M), fVIIa Ki
M and fXIa Ki >55 M). At the same time, the P2Y1 antagonist
M) when using either thrombin receptor
>30
>15
l
l
l
l
2l was inactive (IC50 >20
l
activating peptide (TRAP), 5-HT2a/epinephrine (EPI), or collagen as
the agonist in the platelet aggregation assay, indicating that it spe-
cifically targets antagonism of ADP-activated platelet aggregation.
Further profiling revealed that compound 2l was metabolically
unstable in rat and human liver microsomal incubation (Table 3).
Metabolic ID studies revealed that N-demethylation of the N,N-
dimethylaminomethyl side chain and oxidative hydroxylation of
the tert-butyl group (C(CH3)3 to C(CH3)2CH2OH and C(CH3)2COOH)
compound 2l (mg/kg + mg/kg/h, IV)
Figure 1. Dose-dependent antithrombotic effect of compound 2l in the rabbit ECAT
model.

3242
T. C. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3239–3243
Vehicle (10% DMA/ 90% 5%-Dextrose; 1 mL/kg + 1 mL/kg/hr, i.v.; n=10)
1 + 2.5
1.3
3.6
Plasma
Concentration
(µMolar)
compound2l
1 + 5
*
(mg/kg + mg/kg/h, IV)
(n=6-8 per dose)
p<0.05 compared to Vehicle
=SEM
8.5
1 + 10
1 + 25
13.7
9
8
7
50
40
30
20
10
0
Maximum
response
7
Maximum
response
7
6
5
4
3
2
1
0
6
5
-39%
*
6
5
4
3
2
1
-49%
-56%
*
*
*
4
3
2.7x
2.6x
*
*
2.1x
1.9x
*
*
1.6x
*
2
1
*
*
Figure 2. Dose-dependent effect of compound 2l on arterial thrombosis, ex vivo platelet aggregation and provoked bleeding in rats.
several GPCRs, 5-HT2A, 5-HTT, alpha-2A adrenoreceptor, and al-
pha-1D adrenoreceptor. Compound 2l had superb solubility in
the rabbit electrolytic-mediated carotid arterial thrombosis (ECAT)
and bleeding model vehicle (>89 mg/mL in 10% DMA/90% 5%-dex-
trose). In spite of the less desirable pharmacokinetic profile and po-
tential selectivity issues against other GPCRs, compound 2l was
selected as the first program compound to test in the rabbit effi-
cacy and bleeding models.
In the rabbit ECAT model, compound 2l was efficacious and
showed a dose-dependent reduction of thrombosis weight with
an EC50 of 2 lM (Fig. 1). However, in the rabbit cuticle bleeding
model, hypotension was observed at the highest dose, 15 mg/
kg + 23.4 mg/kg/h IV with an average plasma concentration of
compound 2l and compound 15d were very similar in rats. It is
noteworthy that the hypotensive response of compound 2l in rab-
bits was not observed in rats.
We continued to investigate other amine-containing soluble
side chains in order to improve PK properties and GPCR selectivity.
Table 2 exemplifies several analogs bearing amine solublizing
groups via an ether linkage to the urea phenyl ring. One possibility
for the loss of translation between the binding affinity and the po-
tency of compound 1 was the high protein binding (99.3%) in hu-
man serum. Compound 3a, having a dimethylaminomethyl ether
side chain, was 20-fold less potent in binding but showed only
two-fold decrease in PA IC50 compared with compound 1, indicat-
ing a better translation presumably because of the significantly in-
creased free fraction of 3a in human serum (5% free for 3a vs <0.2%
free for 1). Adding an ortho-substituted fluorine atom on the urea
phenyl ring of 3a resulted in compound 3c with slightly improved
P2Y1 binding and similar platelet aggregation activity. However,
both 3a and 3c had low metabolic stability in human and rat liver
microsomal incubation. Metabolic ID study indicated extensive
oxidative hydroxylation on several parts of the molecule, including
the tert-butyl group and N-demethylation on the amino side chain.
5 lM and 94% inhibition of ex vivo platelet aggregation. This dose
level was five times higher than the highest efficacious dose tested
in the efficacy model. A lower dose of compound 2l (9 mg/
kg + 23.4 mg/kg/h) was then identified which still produced an
exaggerated effect (93% blockade) on ex vivo platelet function. This
dose produced a 3.2-fold increase in cuticle bleeding time. How-
ever on prolonged dosing the hypotension was again observed;
clouding the interpretation. The toxicity is likely due to the lack
of selectivity of compound 2l.
In previously established rat models,5c pretreatment with com-
pound 2l produced dose-dependent concurrent inhibition of FeCl2-
induced arterial thrombosis and ex vivo platelet aggregation
(Fig. 2). Compound 2l was administered as loading plus sustaining
IV infusions of 1 + 2.5, 1 + 5, 1 + 10, and 1 + 25 mg/kg + mg/kg/h,
which produced plasma concentrations of 1.3, 3.6, 8.5 and
N
F
N
O
NH2
a,b
c,d
N
N
O
NO2
F
N
H
N
H
5
F
O
13.7
lM, respectively. The 1 + 10 and 1 + 25 mg/kg + mg/kg/h dose
4
2l
levels inhibited ex vivo platelet aggregation to 10
lM ADP by 83%
and 96%, respectively. At these two dose levels, compound 2l de-
creased thrombus weight by 49% and 56%, respectively in the arte-
rial thrombosis model, and the antithrombotic EC50 was ca. 7 lM.
In separate rats, compound 2l produced dose-dependent increases
in bleeding time, but the prolongation of the bleeding time
achieved a plateau of less than three-fold at maximum antithrom-
botic doses in both the mesenteric bleeding time and the cuticle
bleeding time models. In the same bleeding models, clopidogrel
had produced maximum increases of at least six-fold.5c Overall,
the antithrombotic, hemostatic and platelet inhibitory profiles of
F
BocHN
Br
e,f,g
c,d
H2N
2n
N
(HO)2B
CHO
6
Scheme 1. Reagents and conditions: (a) 2-t-Butylphenol, K2CO3, DMF, 130 °C, 80%;
(b) H2, 10% Pd/C, MeOH/THF, 92%; (c) Phosgene, NaHCO3, CH2Cl2; (d) HCl salt of 20-
((dimethylamino)methyl)-3-fluorobiphenyl-4-amine (5), THF, 45–65 °C, 19 h, 78%
for 2l and 36% for 2n; (e) Pd(PPh3)4, THF, reflux, 40%; (f) NHMe2HCl, NaBH(OAc)3,
CH2Cl2, 61%; (g) TFA, CH2Cl2, 94%.

T. C. Wang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 3239–3243
3243
F
OH
dependent antithrombotic effect in both the rat and rabbit throm-
bosis models and a moderate prolongation of the bleeding time
similar to that of compound 1 in rat bleeding models. However,
toxicity of compound 2l in the rabbit model precluded it from
being further studied. Additional research which will be described
in due course, focused on improving both compound selectivity
and pharmacokinetic properties.
O
c or d
N
N
N
a, b
N
H
N
H
NH2
8 or 9
O
O
O
7
F
R
O
N
N
O
3c R =
3d R =
3h R =
N
H
N
H
Acknowledgments
Ph
O
O
e,f
The authors would like to thank Dr. David Nirschl, Mr. George
Morton, and Dr. Timothy Herpin for early exploratory work via par-
allel synthesis, Dr. Mary Grubb for metabolic ID study, and Dr. Dora
Schnur for CADD support.
N
Scheme 2. Reagents and conditions: (a) Phosgene, NaHCO3, CH2Cl2; (b) 4-Amino-5-
fluorophenol, THF, 47%; (c) 3-(Dimethylamino)-2,2-dimethylpropan-1-ol, PS-Ph3P,
di-tert-butyl azodicarboxylate (8), THF, 25%; (d) (1-Benzylpiperidin-4-yl)methanol
(9), Ph3P-PS, di-tert-butyl azodicarboxylate, THF, 58%; (e) 10% Pd/C, H2, MeOH, 98%;
(f) Isobutyraldehyde, NaBH3CN, conc. HCl, MeOH, 27%.
References and notes
1. (a) Hechler, B.; Gachet, C. Purinergic Signalling 2011, 7, 293; (b) Jacobson, K. A.;
Deflorian, F.; Mishra, S.; Costanzi, S. Purinergic Signalling 2011, 7, 305; (c)
Gachet, C. Thromb. Haemost. 2008, 99, 466; (d) Cattaneo, M. Expert Rev.
Cardiovasc. Ther. 2007, 5, 45; (e) Gachet, C.; Léon, C.; Hechler, B. Blood Cells Mol.
Dis. 2006, 36, 223; (f) Oury, C.; Toth-Zsamboki, E.; Vermylen, J.; Hoylaerts, M. F.
Curr. Pharm. Des. 2006, 12, 859. and references cited therein; (g) Gachet, C.
Annu. Rev. Pharmacol. Toxicol. 2006, 46, 277; (h) Cattaneo, M. Drug News
Perspect. 2006, 19, 253; (i) Murugappan, S.; Kunapuli, S. P. Front. Biosci. 2006,
11, 1977; (j) Hechler, B.; Cattaneo, M.; Gachet, C. Semin. Thromb. Hemost. 2005,
312, 150; (k) Herbert, J.-M. Expert Opin. Investig. Drugs 2004, 13, 457.
2. For a recent review on clopidogrel, prasugrel and ticagrelor, see: Giorgi, M. A.;
Arazi, H. C.; Gonzalez, C. D.; Di Girolamo, G. Expert Opin. Pharmacother. 2011,
12, 1285. and references cited therein.
3. (a) Hechler, B.; Nonne, C.; Roh, E. J.; Cattaneo, M.; Cazenave, J. P.; Lanza, F.;
Jacobson, K. A.; Gachet, C. J. Pharmacol. Exp. Ther. 2006, 316, 556 and references
cited therein; (b) Lenain, N.; Freund, M.; Léon, C.; Cazenave, J. P.; Gachet, C. J.
Thromb. Haemost. 2003, 1, 1144; (c) Léon, C.; Freund, M.; Ravanat, C.; Baurand,
A.; Cazenave, J. P.; Gachet, C. Circulation 2001, 103, 718; (d) Fabre, J. E.; Nguyen,
M.; Latour, A.; Keifer, J. A.; Audoly, L. P.; Coffman, T. M.; Koller, B. H. Nat. Med.
1999, 5, 1199; (e) Léon, C.; Hechler, B.; Freund, M.; Eckly, A.; Vial, C.; Ohlmann,
P.; Dierich, A.; LeMeur, M.; Cazenave, J. P.; Gachet, C. J. Clin. Invest. 1999, 104,
1731.
4. For recent publications on small molecule P2Y1 antagonists, see: (a)
Pfefferkorn, J. A.; Choi, C.; Winters, T.; Kennedy, R.; Chi, L.; Perrin, L. A.; Lu,
G.; Ping, Y.-W.; McClanahan, T.; Schroeder, R.; Leininger, M. T.; Geyer, A.;
Schefzick, S.; Atherton, J. Bioorg. Med. Chem. Lett. 2008, 18, 3338; (b) Morales-
Ramos, A. I.; Mecom, J. S.; Kiesow, T. J.; Graybill, T. L.; Brown, G. D.; Aiyar, N. V.;
Davenport, E. A.; Kallal, L. A.; Knapp-Reed, B. A.; Li, P.; Londregan, A. T.;
Morrow, D. M.; Senadhi, S.; Thalji, R. K.; Zhao, S.; Burns-Kurtis, C. L.; Marino, J.-
P. Bioorg. Med. Chem. Lett. 2008, 18, 6222; (c) Houston, D.; Costanzi, S.;
Jacobson, K. A.; Harden, T. K. Comb. Chem. High Throughput Screening 2008, 11,
410; (d) Thalji, R. K.; Aiyar, N.; Davenport, E. A.; Erhardt, J. A.; Kallal, L. A.;
Morrow, D. M.; Senadhi, S.; Burns-Kurtis, C. L.; Marino, J. P., Jr. Bioorg. Med.
Chem. Lett. 2010, 20, 4104.
5. (a) Bird, J. E.; Wang, X.; Smith, P. L.; Barbera, F.; Huang, C.; Schumacher, W. A. J.
Thromb. Thrombolysis 2012, 34, 199; (b) Wong, P. C.; Crain, E. J.; Jiang, X.;
Bostwick, J. S.; Thibault, C.; Ruel, R.; Rehfuss, R.; Schumacher, W. A.; Ogletree,
M. L. Circulation 2006, 114, 248; (c) Schumacher, W. A.; Bostwick, J. S.; Ogletree,
M. L.; Stewart, A.; Steinbacher, T. E.; Hua, J.; Price, L. A.; Wong, P. C.; Rehfuss, R.
J. Pharmacol. Exp. Ther. 2007, 322, 1; (d) Chao, H.; Turdi, H.; Herpin, T. F.;
Roberge, J. Y.; Liu, Y.; Schnur, D.; Poss, M. A.; Rehfuss, R.; Hua, J.; Wu, Q.; Price,
L. A.; Abell, L. M.; Schumacher, W. A.; Bostwick, J. S.; Steinbacher, T. E.; Stewart,
A. B.; Ogletree, M. L.; Huang, C. S.; Chang, M.; Cacace, A. M.; Arcuri, M. J.; Celani,
D.; Wexler, R. R.; Lawrence, R. M. J. Med. Chem. 2013, 56, 1704.
Thus, the tert-butyl group in 3a was replaced with a trimethylsilyl
group, a bioisostere that is believed to be less prone to oxidative
metabolism.10 But the corresponding silyl analog 3b showed only
slight improvement in metabolic stability (HLM 29% remaining
for 3b vs 12% remaining for 3a). A series of ether-linked piperidinyl
analogs 3d–3h were also investigated. Replacement of the benzyl
substitution in 3d with an isobutyl group led to compound 3h with
relatively good binding affinity and similar functional activity as
that of compound 1, but much increased aqueous solubility
(102 lg/mL). The two alkyl homologs 3e and 3f showed decreased
binding affinity compared with 3h. The 2-fluorophenyl analog 3h
was slightly more soluble than the corresponding des-fluoro com-
pound 3g similar to that of 3a and 3c. Compound 3h showed rela-
tively low clearance in rats similar to that of compound 1 due to a
much increased rat liver microsomal stability (67% for 3h vs 24%
for 2l). In addition, compound 3h showed improved bioavailability
(F: 40%) presumably at least in part as a result of increased aqueous
solubility (Table 3). However, compared with 2l, compound 3h had
similar in vitro hERG profile (patch clamp IC50 3.7
5.4 M for 2l) and GPCR selectivity profile (3h also showed more
than 90% inhibition against 5-HT2A, 5-HTT, and alpha-2A adreno-
receptor at 10 M). Consequently, 3h was not further studied in
lM for 3h vs
l
l
the in vivo efficacy/bleeding models.
The synthesis of compounds 2l and 2n bearing the N,N-dimeth-
ylaminomethyl side chain is outlined in Scheme 1. SNAr displace-
ment of 2-fluoro-3-nitropyridine with 2-tert-butylphenol followed
by reduction of the nitro group provided the amine intermediate 4.
The 20-dimethylaminomethyl substituted biphenyl amine 5 was
prepared according to literature method,11 while the 40-dimethyl-
aminomethyl substituted biphenyl amine 6 was prepared from
commercially available starting materials via a three-step se-
quence: Suzuki coupling, reductive amination and then deprotec-
tion. The isocyanate was generated using phosgene and then
coupled directly with 5 or 6 to afford the desired compounds 2l
and 2n.
The ether-linked analogs 3c and 3h were prepared according to
Scheme 2. Mitsunobu reaction of the intermediate phenol 7 with
alcohol 8 or 9 afforded the ether-linked analog 3c or 3d using poly-
mer-bound triphenylphosphine and di-tert-butyl azodicarboxylate
in anhydrous THF. Compound 3h was then obtained via hydroge-
nation of 3d followed by reductive amination.
6. Hechler, B.; Freund, M.; Ravanat, C.; Magnenat, S.; Cazenave, J. P.; Gachet, C.
Circulation 2008, 118, 754.
7. Zerr, M.; Hechler, B.; Freund, M.; Magnenat, S.; Lanois, I.; Cazenave, J.-P.;
Catherine Léon, C.; Gachet, C. Circulation 2011, 123, 2404.
8. The diarylurea compounds reported in Ref. 5d and herein, were previously
disclosed in Chao, H.J.; Turdi, H.; Herpin, T.; Roberge, J.Y.; Liu, Y.; Lawrence,
R.M.; Rehfuss, R.P.; Clark, C.G.; Qiao, J.X.; Gungor, T.; Lam, P.Y.S.; Wang, T.C.;
Ruel, R.; L’Heureux, A.L.; Thibeault, C.; Bouthillier, G.; Schnur, D.M. U.S.
0,261,244A1, 2005.
9. Partition/distribution (LogP/LogD) coefficients are determined using reversed-
phase HPLC and eluted with isocratic methanol: 30 lM potassium phosphate
buffer (pH 7.0) mobile phase. A series of compounds with known literature
LogPs were used as standards. Details see: Zhao, Y.; Jona, J.; Chow, D. T.; Rong,
H.; Semin, D.; Xia, X.; Zanon, R.; Spancake, C.; Maliski, E. Rapid Commun. Mass
Spectrom. 2002, 16, 1548.
In summary, incorporating solublizing amine-bearing side
chains on the phenyl ring of the phenyl urea portion of compound
1 led to the discovery of several compounds, such as 2l and 3h, as
small molecule non-nucleotide P2Y1 antagonists with significantly
improved aqueous solubility and similar antiplatelet activity com-
pared with 1. Compound 2l was efficacious and showed a dose-
10. (a) Gately, S.; West, R. Drug Dev. Res. 2007, 68, 156; (b) Showell, G. A.; Mills, J. S.
Drug Discov. Today 2003, 8, 551.
11. Lam, P.Y.; Han, Q.; Pinto, D.J.P.; Quan, M.L.; Pruitt, J.R.; Li, R.; Fevig, J.M.; Clark,
C.G.; Dominguez, C. U.S. 6,339,099 B1, 2002.