Optimization of 2-phenyl-pyrimidine-4-carboxamides towards potent, orally bioavailable and selective P2Y12 antagonists for inhibition of platelet aggregation

Article abstract of DOI:10.1016/j.bmcl.2014.06.070

2-Phenyl-pyrimidine-4-carboxamide analogs were identified as P2Y ₁₂ antagonists. Optimization of the carbon-linked or nitrogen-linked substituent at the 6-position of the pyrimidine ring provided compounds with excellent ex vivo potency in the

Full text of DOI:10.1016/j.bmcl.2014.06.070

Bioorganic & Medicinal Chemistry Letters 24 (2014) 4323–4331
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Optimization of 2-phenyl-pyrimidine-4-carboxamides towards
potent, orally bioavailable and selective P2Y12 antagonists for
inhibition of platelet aggregation
⇑
Eva Caroff , Emmanuel Meyer, Alexander Treiber, Kurt Hilpert, Markus A. Riederer
Actelion Pharmaceuticals Ltd, Drug Discovery and Preclinical Research & Development, Gewerbestrasse 16, CH-4123 Allschwil, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
2-Phenyl-pyrimidine-4-carboxamide analogs were identified as P2Y12 antagonists. Optimization of the
carbon-linked or nitrogen-linked substituent at the 6-position of the pyrimidine ring provided
compounds with excellent ex vivo potency in the platelet aggregation assay in human plasma. Compound
23u met the objectives for activity, selectivity and ADMET properties.
Received 24 March 2014
Revised 20 June 2014
Accepted 23 June 2014
Available online 1 July 2014
Ó 2014 Elsevier Ltd. All rights reserved.
Keywords:
P2Y12 receptor
P2Y12 antagonist
GPCR antagonist
Antiplatelet
Antithrombotic
When vascular integrity is compromised, signals are generated
that cause circulating platelets to adhere to the vessel wall, become
activated, and aggregate, forming a plug that seals off the site of
injury to prevent blood loss. Autopsy studies demonstrated that
the same processes occur upon rupture of an atherosclerotic pla-
que, and can lead to uncontrolled platelet thrombus formation
and vessel occlusion.1 Inhibition of platelet activation and aggrega-
tion is recognized as an effective strategy for prevention of ather-
othrombotic events in patients with atherosclerotic disease in
the coronary, peripheral, and cerebrovascular circulation.2
The ADP receptor P2Y12 plays a critical role in platelet activation
and aggregation.3 It is required to amplify and sustain the initial
response to a vascular damage thereby leading to stable thrombus
formation. The P2Y12 receptor is a validated target for prevention
of major adverse vascular events in patients with acute coronary
syndromes (ACS, comprising unstable angina [UA], non-ST eleva-
tion myocardial infarction [MI] and ST elevation MI), as demon-
strated by the thienopyridine class of drugs, including clopidogrel
and prasugrel.4 Thienopyridines are converted to active metabo-
lites that irreversibly bind and inactivate the P2Y12 receptor for
the life of the platelet.
treatment option.4 Clopidogrel is an irreversible P2Y12 receptor
antagonist and its efficacy is limited by slow and variable conver-
sion of the prodrug to the active metabolite, modest and variable
inhibition of platelet aggregation,5 and increased risk of bleeding
at higher doses.6 Prasugrel, achieving more pronounced inhibition
of platelet aggregation, showed superior efficacy versus clopidogrel
in moderate- to high-risk patients with ACS undergoing percutane-
ous coronary intervention (PCI), but the improved efficacy was
associated with an increased bleeding risk.7
Novel principles of P2Y12 antagonism may have the potential to
improve efficacy without adversely affecting safety. Firstly, antag-
onists that bind reversibly to the receptor are less likely to cause
bleeding than those that cause irreversible blockade, possibly
because receptors remain responsive to high local concentrations
of ADP.8 Secondly, comparison of clopidogrel and prasugrel with
genetic ablation of the P2Y12 receptor in mice suggests that some
of the excess bleeding associated with thienopyridines may be
due to off-target effects.9 Therefore, a reversible and selective
P2Y12 antagonist has the potential to provide an improved thera-
peutic window as compared with the thienopyridines.
Berlex,10 Portola (Elinogrel11) and Astra Zeneca (Ticagrelor,12
BrilintaÒ) were major players in the field of reversible P2Y12 antag-
onists when we started our research. Ticagrelor and the Berlex
molecules (I) were considered as starting points for knowledge-
based programs (Fig. 1). One approach was to simplify the struc-
ture of Ticagrelor by cutting the triazolo-pyrimidine scaffold down
to pyrimidine and by removing the sugar-like moiety. A set of
In patients with ACS, dual antiplatelet therapy with acetylsali-
cylic acid and clopidogrel is recommended today as the preferred
⇑
Corresponding author. Tel.: +41 615656231.
E-mail address: eva.caroff@actelion.com (E. Caroff).
http://dx.doi.org/10.1016/j.bmcl.2014.06.070
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

4324
E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
R3
F
F
HN
N
O
R6
N
(R4)n
N
N
N
(R5)m
N
N
N
R2
O
S
O
N
O
R1
O
I
HO
Berlex
OH
HO
Ticagrelor
R'
R3
N
HN
O
N
H
N
N
N
R
N
(R4)n
O
N
O
R2
O
R1
II
III
Figure 1. Implementation of a pyrimidine core structure (II) into Berlex molecules (I) leading to the 2-phenyl-pyrimidine-4-carboxamide series (III).
derivatives was prepared and few compounds of structure (II)
caught our attention: they showed weak activity in the binding
assay (5–10 lM, data not shown) and the analogy to the Berlex
quinoline structure (I) was compelling. Gratifyingly, the Berlex
quinoline was successfully replaced by scaffold (II) resulting in
the 2-phenyl-pyrimidine-4-carboxamide series (III).13
Suzuki conditions with the boronic acid 17, mostly using Pd(PPh3)4
as catalyst and K3PO4 as base in refluxing dioxane. In case an
alkene boronic ester 20 was used, such as for compound 23m,
the coupling was performed in presence of Pd2(dba)3/PPh3 in a
mixture of EtOH, toluene and Na2CO3 at 100 °C. An iron-catalyzed
cross-coupling reaction19 provided intermediate 22 using alkyl-
magnesium bromide 19 in presence of Fe(acac)3 in THF at room
temperature. The Stille protocol was performed using stannane
derivative 2020 and Pd(PPh3)4 as catalyst in toluene at 110 °C.
Finally, a Sonogashira reaction was carried out using alkyne 21 in
presence of Pd(PPh3)2(OAc)2, CuI and NEt3 in DMF at room temper-
ature. Treatment of intermediates 22 with TFA provided the set of
final compounds 23. In case of analogs 23g and 23l–n, the interme-
diates 22 were submitted to additional chemical transformations
before TFA treatment.21
All residues R1 to R4 were thoroughly explored. We investigated
the carbamate moiety of (III), R1–O–CO–N–, and came to similar
conclusions as in previously published work by Pfizer.14 Attempts
to modify the piperazine core led to substantial or complete loss of
biological activity. Substitution of the phenyl ring by various R4
groups did not lead to improved biological activities. Therefore in
the present paper R1 represents ethyl or n-butyl and R4 represents
hydrogen. For R2, the side chain of
L-glutamic acid was kept con-
stant throughout the study. The SAR of the R3 group whereby R3
is linked to the pyrimidine core by a nitrogen or a carbon atom is
reported in this Letter.
Another synthetic access to 6-(C-linked-R3)-2-phenyl-pyrimi-
dine-4-carboxamide analogs 23 is described in Scheme 3. The acid
chloride 24 was reacted with acetylene 25 in presence of catalytic
amounts of Pd(PPh3)2(OAc)2 and CuI in NEt3 at room tempera-
ture.22 The resulting keto-alkyne 26 was coupled to benzamidine
2 with Na2CO3 and a catalytic amount of water in refluxing
MeCN.23 The acid derivative 28 was obtained after cleavage of
the acetyl group of 27 in basic conditions (K2CO3 in MeOH/water)
and oxidation of the resulting alcohol with KMnO4 as previously
described in Scheme 1. Finally intermediate 28 was coupled with
compound 6 using EDCI and HOBt in DCM and the tert-butyl group
was cleaved (TFA in DCM) to provide the final compounds 23d–e.
The synthetic pathway to access 6-(C-linked-R3)-2-phenyl-
pyrimidine-4-carboxamide analogs 33 wherein R3 is CH2–NRR0 is
described in Scheme 4. Compound 29 was submitted to the Suzuki
conditions described previously in Scheme 2 using phenylboronic
acid. The resulting intermediate was oxidized to the aldehyde in
presence of SeO2 in refluxing dioxane. After reduction of the alde-
hyde function with NaBH4 in MeOH/DCM at 0 °C, the resulting
alcohol was converted to the chloro derivative by refluxing in
POCl3. After saponification, intermediate 30 was further coupled
with compound 6 using EDCI and HOBt. The chlorine atom of inter-
mediate 31 was displaced by the amine 8 in THF at a temperature
from 100 °C to 150 °C. Cleavage of the tert-butyl group provided
the set of final compounds 33.
The synthesis of the molecules was designed and carried out as
shown hereafter. The synthetic pathway to access 6-(N-linked-R3)-
2-phenyl-pyrimidine-4-carboxamide analogs 11, 14 and 16 is
described in Scheme 1. The pyrimidine ring was built by condensa-
tion of ethyl 4-methoxy-3-oxobutanoate 1 with benzamidine 2 to
provide hydroxypyrimidine 3.15 Subsequent chlorination in reflux-
ing POCl3 afforded the desired chloro derivative 4. The methoxy
group was cleaved using BBr3 in DCM at 0 °C and the resulting pri-
mary alcohol was oxidized to the acid with KMnO4/NaOH in water/
dioxane at room temperature.16 The resulting intermediate 5 was
coupled to the amine 617 to give intermediate 7, ready for deriva-
tization. Primary and secondary amines 8 were reacted with 7 in
THF at 40 °C to give intermediate 9. Heteroaromatic amines 10,
such as imidazole, were coupled to 7 using NaH in THF at room
temperature. Final treatment of 9 with TFA in DCM provided the
first set of final analogs 11. The intermediate 12 was obtained by
reacting 7 with NH3 in MeOH and heating in a sealed tube at
90 °C in a microwave oven. Compound 12 was further treated with
an acyl chloride 13 in pyridine heating at 70 °C. Final cleavage of
the tert-butyl ester with TFA provided the second set of final com-
pounds 14. Reaction of 12 with sulfonyl chloride derivatives 15
using NaH in THF at 70 °C followed by deprotection with TFA affor-
ded the third set of final molecules 16.
The first synthetic pathway to access 6-(C-linked-R3)-2-phenyl-
pyrimidine-4-carboxamide analogs 23 is described in Scheme 2.
Intermediates 22 were prepared using various protocols for the
formation of a CAC bond.18 Intermediate 7 was reacted under
The project aimed at the discovery of potent P2Y12 antagonists
in vivo, that is, active at the site of action (platelets) in human
blood. The first goal was to get affinity in the P2Y12 receptor bind-
ing assay.24 Since the inhibition of platelet aggregation in human

E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
4325
OH
Cl
N
NH
O
O
i
ii
N
N
O
+
O
O
H2N
O
N
2
1
3
4
iii, iv
N
Cl
Cl
O
RR'NH
or
8
NH2
v
O
N
N
H
N
+
+
HO
10
Heteroaryl-NH
O
N
R1
N
N
N
O
N
O
O
R1
O
COOtBu
6
5
O
COOtBu
7
ix
vi or vii
O
R3
R
NH
NH2
x, viii
O
N
O
N
O
N
H
N
H
H
RCOCl
+
N
N
N
N
N
N
N
N
13
O
N
O
O
O
N
O
O
N
R1
R1
R1
O
COOH
14, R3 = -NHCOR
O
COOtBu
O
COOtBu
9
12
+
RSO2Cl
15
xi, viii
viii
O
O
S
R3
N
R
NH
O
N
O
N
H
N
H
N
N
N
N
O
O
N
O
O
N
R1
R1
O
COOH
O
COOH
16, R3 = -NHSO2R
11, R3 = -NHRR'
or -N-heteroaryl
Scheme 1. Synthesis of 6-(N-linked-R3)-2-phenyl-pyrimidine-4-carboxamide analogs 11, 14 and 16. Reagents and conditions: (i) NaOMe, EtOH, 80 °C; (ii) POCl3, 110 °C; (iii)
BBr3, DCM, 0 °C; (iv) KMnO4, NaOH, H2O, dioxane, rt; (v) (COCl)2, MeCN, 80 °C, then 1 equiv 6, NEt3, MeCN, 0 °C–rt; (vi) excess 8 or 1 equiv 8 and Et3N, THF, 40 °C; (vii) NaH,
1 equiv 10, THF, rt; (viii) TFA/DCM, rt; (ix) excess NH3, MeOH, 90 °C microwave; (x) 1 equiv 13, pyridine, 70 °C; (xi) 1 equiv 15, NaH, THF, 70 °C.
17
18
19
20
21
R-B(OH)2
or
R-B(OR')2
R
N
R3
N
Cl
N
or
i, ii, iii, iv or v
vi
O
N
O
N
O
N
H
N
H
N
H
N
R-MgBr
or
R-SnBu3
+
N
N
N
O
O
O
N
O
N
O
O
N
R1
R1
R1
or
O
COOtBu
O
COOH
O
COOtBu
R-C CH
23
7
22
Scheme 2. Synthesis of 6-(C-linked-R3)-2-phenyl-pyrimidine-4-carboxamide analogs 23. Reagents and conditions: (i) 1 equiv 17, Pd(PPh3)4, K3PO4, dioxane, 100 °C or
Pd(PPh3)4, 2 M K2CO3, EtOH/DME, 90 °C; (ii) 1 equiv 18, Pd2(dba)3, PPh3, 1 M Na2CO3, EtOH/toluene, 100 °C; (iii) 2 equiv 19, Fe(Acac)3, THF, rt; (iv) 1.2 equiv 20, Pd(PPh3)4,
toluene, 110 °C; (v) 2 equiv 21, Pd(PPh3)2Cl2, CuI, Et3N, DMF, rt; (vi) TFA/DCM, rt.
R3
N
R3
N
R3
N
R3
O
O
iii, iv
i
ii
v, vi
O
O
N
O
N
N
H
N
+
2
+
+
6
R3
Cl
O
HO
O
N
O
O
N
O
O
O
R1
O
COOH
24
25
26
27
23
28
Scheme 3. Alternative synthesis of 6-(C-linked-R3)-2-phenyl-pyrimidine-4-carboxamide analogs 23. Reagents and conditions: (i) Pd(PPh3)2(OAc)2, CuI, NEt3, rt; (ii) 2 equiv 2,
Na2CO3, MeCN, 1 drop H2O, 90 °C; (iii) K2CO3, MeOH/H2O, rt; (iv) KMnO4, NaOH, H2O, dioxane, rt; (v) 1 equiv 6, HOBt, EDCI, DCM, rt; (vi) TFA/DCM, rt.

4326
E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
R'
N
O
Cl
Cl
R
O
i, ii
iii, iv, v
vii, viii
vi
N
N
O
N
N
O
N
R'R''NH
+
6
+
H
N
H
N
O
O
HO
N
Cl
N
8
N
N
N
N
N
O
O
O
N
O
O
O
N
R1
R1
O
COOtBu
O
COOH
33, R3 = -CH2-NRR'
29
30
31
32
Scheme 4. Synthesis of 6-(C-linked-R3)-2-phenyl-pyrimidine-4-carboxamide analogs 33. Reagents and conditions: (i) 1 equiv phenylboronic acid, Pd(PPh3)4, K3PO4, dioxane,
100 °C; (ii) SeO2, dioxane, 100 °C; (iii) NaBH4, MeOH, DCM, 0 °C; (iv) POCl3, 100 °C; (v) 1 M NaOH, MeOH, rt; (vi) 1 equiv 6, HOBt, EDCI, DCM, rt; (vii) excess 8, THF, 100–
150 °C; (viii) TFA/DCM, rt.
platelet-rich plasma (PRP) was shown to correlate with anti-
thrombotic efficacy in human clinical trials, the second key step
was to optimize the potency in PRP. A first set of 6-(N/C-linked-
R3)-2-phenyl-pyrimidine-4-carboxamide analogs 11, 14, 16, 23
and 33 were prepared by the above synthetic pathways. The most
active compounds in the binding assay (generally IC50 value below
200 nM) were further profiled by measuring their inhibitory action
in the ex vivo light transmission aggregometry (LTA) assay per-
formed in human PRP.25 Remaining activity (RA) was measured
at 1000 nM, 500 nM or 100 nM concentration of the compounds,
followed by determination of the IC50 for the most promising mol-
ecules. A panel of examples shown in Table 1 illustrates the first
optimization round. Regarding 6-(N-linked-R3)-2-phenyl-pyrimi-
dine-4-carboxamide analogs 11, alkyl groups on the 6-nitrogen
atom generally provided antagonists in the binding assay that
translated into low to moderate potency in the LTA assay (see
examples 11a–b). Aryl and benzylic groups showed low ex vivo
activities (see examples 11c–d). The introduction of a hydroxyl
group onto the alkyl chain (11e) provided for the first time prom-
ising activities with no plasma shift between in vitro and ex vivo
assays. Replacing the hydroxyl of 11e by an acid (11f) retained
the in vitro activity but the ex vivo activity was lost. Replacement
of the hydroxyl of 11e by a dimethylamino function (11g) led to a
significant loss in binding affinity. Finally, replacement of the
hydroxyl of 11e by the weakly basic morpholino group (11h) led
only to moderate activities in both assays. Derivatives with the
6-nitrogen atom embedded into a cycloalkyl ring showed promises
in the LTA assay, especially when a second heteroatom was
included into the ring (11j, 11k). Including the 6-nitrogen atom
into a heteroaryl ring (see example 11l) was found to increase
the binding activity but the plasma shift remained high.
theless, reaching potency in the LTA assay remained challenging,
as experienced by other groups.26 Several compounds such as
11e, 11j, 11k, 23e and 23g, that combine high activities in vitro
and ex vivo, showed us the path to move forward. We decided to
put emphasis on compounds extending an oxygen or nitrogen
atom one to five atoms away from the position 6 of the pyrimidine
core. A second set of compounds 11 and 23 was subsequently pre-
pared following synthetic procedures described previously. Table 2
displays representative analogs illustrating the findings discussed
hereafter. Firstly, O/N heteroatoms introduced as alcohol, ether
or amine were tolerated by the receptor. Secondly, three parame-
ters seemed to play a role in modulating the binding affinity: (a)
the rigidity of the R3 substituent which could impact on the entro-
pic loss upon antagonist binding, (b) the lipophilic volume close to
the position 6 of the pyrimidine and finally (c) the specific position
of the heteroatom for additional favorable interactions in the
pocket. For example, analogs 11m and 23l show moderate binding
IC50 values of 279 nM and 218 nM, respectively. Both possess linear
linkers that place a hydroxyl function four and three atoms away
from position 6 of the pyrimidine, respectively. The flexibility of
the linkers may allow to reach an acceptable position for the
hydroxyl group but comes at an entropic cost. The more con-
strained analogs 11n and 11w show medium binding activities
as well. This may be due to an unfavorable presentation of the het-
eroatoms. One could also speculate that analog 11u is more active
than 11p because, although being structurally very close, the cyclic
ether function is more favorable than the acyclic ether with regard
to entropy. Finally, 11t and 23r were the most potent compounds
in the binding assay with IC50 values of 16 nM and 15 nM, respec-
tively, closely followed by 11s, 11u and 23o (IC50 values of 21, 25
and 23 nM, respectively). All five compounds combine the struc-
tural requirements observed in the second optimization round:
indeed the R3 substituents are relatively rigid, contain some lipo-
philic volume close to the position 6 of the pyrimidine and might
orient the oxygen atom favorably within the receptor pocket. As
a conclusion, the second optimization round successfully produced
highly potent antagonists in the binding and LTA assays, all analogs
of Table 2 with the exception of compound 23q exhibiting plasma
shifts below 6. In addition, the structural requirements of R3 for
efficient binding to the P2Y12 receptor were refined.
Finally, incorporation of the 6-nitrogen atom into an amide (see
examples 14a–b) or sulfonamide (see examples 16a–b) function
led to moderate in vitro biological activities. Therefore amides
and sulfonamides 14 and 16 were not explored further.
Similar SAR trends were observed for 6-(C-linked-R3)-2-phenyl-
pyrimidine-4-carboxamide analogs 23. Alkyl groups led to active
compounds in the binding assay, but devoid of ex vivo activity as
measured in the LTA assay (see examples 23a–d). Introduction of a
hydroxyl group two carbon atoms away from the pyrimidine core
provided potency in both assays, as exemplified by 23e. Aryl and
heteroaryl groups (see examples 23f, 23h, 23i–k) also showed good
binding IC50 values, but moderate or low activity in the LTA assay.
One exception was analog 23g with an oxygen atom attached onto
the aromatic ring as an N-oxide, displaying in the LTA assay an
IC50 value of 126 nM and a plasma shift of 0.98. Finally it was also
decided to investigate compounds with a nitrogen atom attached
onto the carbon atom on position 6 of the pyrimidine ring. In partic-
ular, piperidine and morpholine rings were introduced to compare
with analogs 11i and 11k. The corresponding analogs 33a and 33b
were not potent enough in vitro to be further profiled.
We postulated that the heteroatom in R3 as hydrogen bond donor
or acceptor would decrease plasma protein binding (PPB) thereby
improving the IC50 value in theLTA assay. Table 3 displays a selection
of protein binding data in human plasma and plasma shifts between
in vitro IC50 and ex vivo IC50 values measured in human PRP. The
compounds are listed according to decreasing plasma shift. High
plasma shifts (>20) were indeed associated with high PPB (>98.5%,
analogs 23b, 11a, 23a) and low shifts (<2) characterized less plasma
protein bound compounds (PPB <91.5%, analogs 11u, 23l, 11e). Ana-
log 23r showed a plasma shift and PPB in the medium range. Within
this set of compounds 11 and 23, R1 being ethyl, our hypothesis that
the heteroatom in R3 is required to decrease PPB thereby decreasing
plasma shift seems to hold true.
Preparation of active compounds in the binding assay was suc-
cessfully achieved during the first round of optimization. Never-

E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
4327
Table 1
First optimization round
a
c
Compound
R3
Binding IC50 (nM)
LTA activityb
LTA IC50 (nM)
Plasma shiftd
11a
11b
11c
11d
11e
11f
–NH-nPr
–NH-iPr
–NH-Ph
–NH-Bn
–NH–CH2–CH2–OH
–NH–CH2–CH2–COOH
–NH–CH2–CH2–N(Me)2
162
95.5
177
122
169
198
2780
3820
724
24
7.6
(ꢀ)
(ꢀ)
(++)
(ꢀ)
114
0.67
11g
O
N
11h
11i
257
201
(+)
(+)
HN
N
N
11j
46.0
(++)
N
O
11k
11l
100
(++)
(+)
N
N
N
27.0
318
12
14a
14b
16a
16b
23a
23b
23c
23d
–NH–CO–cyclohexyl
–NH–CO–Ph
–NH–SO2–Et
–NH–SO2–Ph
–nBu
–Cyclopropyl
–iPr
–tBu
219
178
142
266
98.0
59.5
52.0
36.0
(ꢀ)
(ꢀ)
(ꢀ)
(ꢀ)
1930
3230
1140
2050
20
54
22
57
(ꢀ)
(ꢀ)
OH
23e
21.0
(++)
121
5.8
N
23f
53.0
(ꢀ)
O
N
23g
23h
129
(++)
126
0.98
64.0
(ꢀ)
O
23i
31.0
(+)
523
17
S
23j
11.0
73.0
533
428
(ꢀ)
(ꢀ)
HN
N
23k
33a
33b
N
N
O
R1 is ethyl. Binding IC50, LTA activity and IC50 and plasma shift of analogs 11, 14, 16, 23 and 33 analogs.
a
CHO cells expressing recombinant P2Y12 receptors incubated with tritium-labeled 2-methyl-thio-adenosine 50-diphosphate and compound.
b
c
Incubation in human PRP with 3
l
M ADP. (ꢀ): RA at 1000 nM >30%, or RA at 500 nM >60%, (+): 30% <RA at 500 nM <60%; (++): RA at 500 nM <30% or RA at 100 nM <70%.
M ADP.
Plasma shift = LTA IC50/binding IC50
Incubation in human PRP with 3
l
d
.
Finally, extending R1 from ethyl to n-butyl improved the biolog-
ical activity in the binding assay while slightly increasing the
plasma shift, as seen with examples 11t/11x, 23p/23s, 23r/23u
(Table 4).

4328
E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
Table 2
Second optimization round
a
c
Compound
R3
Binding IC50 (nM)
279
LTA activityb
(+)
LTA IC50 (nM)
Plasma shiftd
11m
–NH–CH2–CH2–CH2–OH
OH
11n
223
(+)
HN
OH
HN
11o
11p
56
97
(++)
(++)
55
0.98
1.4
O
139
HN
OH
(R)
11q
52
(++)
99
1.9
N
OH
(S)
11r
11s
46
21
(++)
(++)
68
73
1.5
3.5
N
OH
N
O
(S)
11t
16
(++)
32
2.0
N
O
11u
11v
11w
23l
25
76
(++)
(++)
(+)
58
86
2.3
1.1
HN
(R)
NH2
N
N
N
291
218
OH
(++)
264
171
1.2
4.3
OH
23m
40
(++)
OH
23n
23o
35
23
(++)
(++)
118
71
3.4
3.1
OH
(rac)
O
23p
23q
23r
39
43
15
(++)
(+)
192
589
87
4.9
14
(rac)
O
(R,R)
O
(++)
5.8
(S,S)
R1 is ethyl. Binding IC50, LTA activity and IC50 and plasma shift of 11 and 23 analogs.
a
CHO cells expressing recombinant P2Y12 receptors incubated with tritium-labeled 2-methyl-thio-adenosine 50-diphosphate and compound.
b
c
Incubation in human PRP with 3
l
M ADP. (ꢀ): RA at 1000 nM >30%, or RA at 500 nM >60%, (+): 30% <RA at 500 nM <60%; (++): RA at 500 nM <30% or RA at 100 nM <70%.
M ADP.
Plasma shift = LTA IC50/binding IC50
Incubation in human PRP with 3
l
d
.
After the second optimization round, our criteria in terms of
CYP2C9) was seen.27 Overall, metabolic stability in human and
rat liver microsomes was high28 and none of the compounds
tested on the hERG channel showed any inhibitory effect.29
Selected potent P2Y12 antagonists were therefore profiled
in vivo in the rat. Oral administration was performed primarily,
followed by intravenous experiments, if deemed necessary.
in vitro biological activity and ex vivo potency were met and
further profiling was undertaken. Solubility was generally low
at acidic pHs for analogs 23, and increased for analogs 11. At
neutral and basic pHs, all compounds showed good solubility.
No inhibition of cytochrome P450 enzymes (CYP3A4, CYP2D6,

E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
4329
Table 3
Binding IC50, LTA IC50, plasma shift and PPB of selected analogs 11 and 23
Compound
R3
Heteroatom in R3
Plasma shifta
PPBb (%)
23b
11a
23a
–cyclopropyl
–NH-nPr
–nBu
No
No
No
54
24
20
99.7
98.5
99.3
O
23r
Yes
Yes
5.8
2.3
96.6
91.5
(S,S)
O
11u
HN
(R)
OH
23l
Yes
Yes
1.2
87.2
86.5
11e
–NH–CH2–CH2–OH
0.67
R1 is ethyl. The compounds are organized according to decreasing plasma shift.
a
Plasma shift = LTA IC50/binding IC50
Protein binding in human plasma.
.
b
Table 4
Binding IC50, LTA IC50 and plasma shift of 11 and 23 compounds wherein R1 is n-butyl, available corresponding ethyl analogs are shown in brackets for comparison
R1
R3
Binding IC50 (nM)
LTA IC50 (nM)
Plasma shiftc
a
b
Compound
O
(S)
11x
(11t)
Bu
(Et)
3
(16)
21
(32)
7.0
(2.0)
N
O
23s
Bu
3
21
7.0
(rac)
(23p)
(Et)
(39)
(192)
(4.9)
O
23t
Bu
9
287
32
(rac)
O
23u
(23r)
Bu
(Et)
4.8
(15)
31
(87)
6.4
(5.8)
(S,S)
a
b
c
CHO cells expressing recombinant P2Y12 receptors incubated with tritium-labeled 2-methyl-thio-adenosine 50-diphosphate and compound.
Incubation in human PRP with 3 M ADP.
Plasma shift = LTA IC50/binding IC50
l
.
Representative data are summarized in Table 5. Systemic plasma
clearance in the rat spanned over a large range, from 14 up to
100 mL/minkg. The volume of distribution at steady-state was
in the range of total body water, as expected for acidic com-
pounds with extensive binding to plasma proteins. All molecules
containing a hydroxyl group in R3 (11e, 23m, 23n, 23p) showed
low oral exposures. A methoxy group seemed to be better toler-
ated (see examples 11t, 11x, 23q, 23u). The N-oxide derivative
23g was not detected in rat plasma. The higher oral exposure
of 23u as compared to 23q was likely a consequence of the
lipophilicity increase resulting from the elongation of the ethyl
chain to a n-butyl group. The lipophilicity increase was indeed
reflected in the logD (pH 7.4) of the two compounds, being 0.8
for 23q and 1.9 for 23u. Overall, 23u offered an acceptable profile
with a low clearance and a medium bioavailability of 33%.
Further characterization of 23u was undertaken. 23u was
found to be a reversible P2Y12 antagonist (data not shown). No
antagonistic activity on genetically related P2Y receptors was
technology (DiscoveRx). In this assay, no significant agonistic or
antagonistic activity, other than on P2Y12, was detected (data
not shown). In conclusion, 23u is a selective and reversible
antagonist of P2Y12
.
In summary, we have identified within the 2-phenyl-pyrimi-
dine-4-carboxamide series (III) two sets of highly potent P2Y12
antagonists, the 6-(N-linked-R3)-2-phenyl-pyrimidine-4-carbox-
amide analogs 11 and the 6-(C-linked-R3)-2-phenyl-pyrimidine-
4-carboxamide analogs 23. The structural requirements for high
affinity to the P2Y12 receptor both in binding and in the LTA
assay were described and a correlation between the plasma shift
and protein binding could be established within this optimization
round. Extensive fine-tuning led to the discovery of two R3
groups (exemplified in 11t/11x and 23q/23u) which provided
excellent potency in the low nanomolar range both in binding
and in the ex vivo platelet aggregation assay performed in human
PRP. Furthermore, moving from R1 being ethyl to n-butyl
increased the lipophilicity of the compounds thereby improving
oral absorption in the rat. No inhibition of the main human cyto-
chrome P450 enzymes nor of the hERG channel was observed.
Compound 23u showed the best profile amongst 11 and 23 ana-
logs, being an orally bioavailable, selective and reversible P2Y12
antagonist with an excellent inhibitory effect on platelet aggrega-
tion in human plasma. The cyclopropyl derivative 23u was
seen. Specifically, up to 10 lM of 23u did not interfere with ago-
nist binding to P2Y1, P2Y2, P2Y4, P2Y6, P2Y11 and P2Y13 (data not
shown). Furthermore, no antagonistic activity was detected on
P2X1, P2X2, P2X3, P2X4 and P2X7 receptors (data not shown). In
addition, 23u was tested at 10
lM for potential antagonistic or
agonistic activity on a panel of 161 GPCRs using the PathHunter

4330
E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
Table 5
Rat pharmacokinetic profiles of selected P2Y12 antagonists
Compound
R1
Et
R3
CL (mL/minkg)
100
Vss (L/kg)
T1/2 (h)
AUC (nMh)
8.44
Cmax (nM)
F (%)
11e
–NH–CH2–CH2–OH
1.0
0.20
17.5
3
N
N
11j
11t
Et
Et
49
36
0.8
1.8
0.30
0.63
164
638
192
683
3
8
O
(S)
N
O
(S)
11x
23g
23m
Bu
Et
Et
24
0.5
1.2
1170
BLQ
1690
BLQ
9
N
O
N
OH
78.2
46.3
OH
23n
23q
Et
Et
109
56.1
O
14
17
0.3
0.4
0.55
2.0
2310
2820
11
33
(R,R)
O
23u
Bu
6000
7620
(S,S)
Male Wistar rats.
Dose: iv infusion at 1 mg/kg (n = 2 rats); po at 10 mg/kg (n = 3 rats).
BLQ: below limit of quantification (ca. 1 ng/mL).
4. (a) Antman, E. M.; Anbe, D. T.; Armstrong, P. W.; Bates, E. R.; Green, L. A.; Hand,
M.; Hochman, J. S.; Krumholz, H. M.; Kushner, F. G.; Lamas, G. A., et al.
Circulation 2004, 110, 82; (b) Smith, S. C., Jr.; Feldman, T. E.; Hirshfeld, J. W., Jr.,
et al. Circulation 2006, 113, 166; (c) Anderson, J. L.; Adams, C. D.; Antman, E. M.,
et al. Circulation 2011, 123, 426.
selected for further in vivo antithrombotic and safety experi-
ments in preclinical species. This work will be reported in due
course.
5. (a) Jernberg, T.; Payne, C. D.; Winters, K. J., et al. Eur. Heart J. 2006, 27, 1166; (b)
Wallentin, L.; Varenhorst, C.; James, S., et al. Eur. Heart J. 2008, 29, 21.
6. (a) Yusuf, S.; Zhao, F.; Mehta, S. R., et al. N. Engl. J. Med. 2001, 345, 494; (b) Fox,
K. A.; Mehta, S. R.; Peters, R., et al. Circulation 2004, 110, 1202.
7. Wiviott, S. D.; Braunwald, E.; McCabe, C. H., et al. N. Engl. J. Med. 2001, 2007,
357.
8. (a) van Giezen, J. J.; Berntsson, P.; Zachrisson, H.; Bjorkman, J. A. Thromb. Res.
2009, 124, 565; (b) Becker, R. C.; Gurbel, P. A. Thromb. Haemost. 2010, 103, 535.
9. André, P.; Deguzman, F.; Haberstock-Debic, H.; Mills, S.; Pak, Y.; Inagaki, M.;
Pandey, A.; Hollenbach, S.; Phillips, D. R.; Conley, P. B. J. Pharmacol. Exp. Ther.
2011, 338, 22.
10. (a) Bryant, J. A.; Buckman, B. O.; Islam, I.; Mohan, R.; Morrissey, M. M.; Wei, G.
P.; Xu, W.; Yuang, S. WO 052366, 2004.; (b) Bryant, J. A.; Buckman, B. O.; Islam,
I.; Mohan, R.; Morrissey, M. M.; Wei, G. P.; Xu, W.; Yuang, S. WO 098856, 2002.;
(c) Bryant, J.; Post, J. M.; Alexander, S.; Wang, Y.-X.; Kent, L.; Schirm, S.; Tseng,
J.-L.; Subramanyam, B.; Buckman, B.; Islam, I.; Yuan, S.; Sullivan, M. E.; Snider,
M.; Morser, J. Thromb. Res. 2008, 122, 523; (d) Post, J. M.; Alexander, S.; Wang,
Y.-X.; Vincelette, J.; Vergona, R.; Kent, L.; Bryant, J.; Sullivan, M. E.; Dole, W. P.;
Morser, J.; Subramanyam, B. Thromb. Res. 2008, 122, 533; (e) Wang, Y.-X.;
Vincelette, J.; da Cunha, V.; Martin-McNulty, B.; Mallari, C.; Fitch, R. M.;
Alexander, S.; Islam, I.; Buckman, B. O.; Yuan, S.; Post, J. M.; Subramanyam, B.;
Vergona, R.; Sullivan, M. E.; Dole, W. P.; Morser, J.; Bryant, J. Thromb. Haemost.
2007, 97, 847.
Acknowledgments
The authors would like to thank Simon Buetikofer, Remy Castro,
Yannick Gallin and Gina Wintenberger for the synthetic work, Mar-
tine Baumann, Brigitte Butscha and Benoit Lack for biological test-
ing, Stephane Delahaye, Fabienne Drouet, Rolf Wuest, Isabelle
Weber, Carmela Gnerre, Kyle Landskroner, Aude Weigel and Nath-
alie Jaouen for generating DMPK data, Bruno Capeleto and Eric
Ertel for providing solubility and hERG data, Olivier Houille and
Heinz Fretz for their contribution at the beginning of the project,
Francis Hubler and Dorte Renneberg for reviewing the manuscript
and Beat Steiner for guidance and support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
06.070.
11. Gurbel, P. A.; Kereiakes, D.; Tantry, U. S. Drugs Fut. 2010, 35, 885. and references
herein.
References and notes
12. (a) Huber, K.; Hamad, B.; Kirkpatrick, P. Nat. Rev. Drug Disc. 2011, 10, 255; (b)
Springthorpe, B.; Bailey, A.; Barton, P.; Birkinshaw, T. N.; Bonnert, R. V.; Brown,
R. C.; Chapman, D.; Dixon, J.; Guile, S. D.; Humphries, R. G.; Hunt, S. F.; Ince, F.;
Ingall, A. H.; Kirk, I. P.; Leeson, P. D.; Leff, P.; Lewis, R. J.; Martin, B. P.;
McGinnity, D. F.; Mortimore, M. P.; Paine, S. W.; Pairaudeau, G.; Patel, A.; Rigby,
A. J.; Riley, R. J.; Teobald, B. J.; Tomlinson, W.; Webborn, P. J. H.; Willis, P. A.
Bioorg. Med. Chem. Lett. 2007, 17, 6013.
1. Davies, M. J.; Thomas, A. C.; Knapman, P. A.; Hangartner, J. R. Circulation 1986,
73, 418.
2. (a) Davì, G.; Patrono, C. N. Engl. J. Med. 2007, 357, 2482; (b) Michelson, A. D. Nat.
Rev. Drug Disc. 2010, 9, 154.
3. (a) Andre, P.; Delaney, S. M.; LaRocca, T.; Vincent, D.; DeGuzman, F.; Jurek, M.;
Koller, B.; Phillips, D. R.; Conley, P. B. J. Clin. Invest. 2003, 112, 398; (b) Dorsam,
R. T.; Kunapuli, S. P. J. Clin. Invest. 2004, 113, 340.
13. (a) Caroff, E.; Fretz, H.; Hilpert, K.; Houille, O.; Hubler, F.; Meyer, E. WO 114774,
2006.; (b) Caroff, E.; Hilpert, K.; Meyer, E. WO 050301, 2008.

E. Caroff et al. / Bioorg. Med. Chem. Lett. 24 (2014) 4323–4331
4331
14. Parlow, J. J.; Burney, M. W.; Case, B. L.; Girard, T. J.; Hall, K. A.; Hiebsch, R. R.;
Huff, R. M.; Lachance, R. M.; Mischke, D. A.; Rapp, S. R.; Woerndle, R. S.; Ennis,
M. D. Bioorg. Med. Chem. Lett. 2009, 19, 6148.
15. WO 114774, 2006, page 127, intermediate 1.3.
16. WO 114774, 2006, p146–147, intermediates 24.1, 24.2 and 24.3.
value or as mean of two measurements. Each data point of an IC50 curve was
determined in triplicate. In addition, reference compound A was profiled as
part of each compound-set analyzed. If the determined IC50 value of reference
compound A (geometric mean = 22 nM) was outside of the pre-determined
95% confidence interval of 19–25 nM, the entire compound-set was re-
analyzed.
17. Experimental details for the synthesis of amine
Supplementary material.
6 can be found in the
25. For the description of the in vitro Light Transmission Aggregometry (LTA)
assay, see the Supplementary material.
18. Reaction conditions leading to intermediates 22 can be found in the
Supplementary material.
19. (a) Fürstner, A.; Leitner, A.; Méndez, M.; Krause, H. J. Am. Chem. Soc. 2002, 124,
13856; (b) Scheiper, B.; Bonnekessel, M.; Krause, H.; Fürstner, A. J. Org. Chem.
2004, 69, 3943.
20. More details and literature references for the synthesis of reagents 20 leading
to analogs 23o–u can be found in the Supplementary material.
21. Reaction conditions leading to analogs 23g and 23l–n can be found in the
Supplementary material.
22. Van den Hoven, B. G.; El Ali, B.; Alper, H. J. Org. Chem. 2000, 65, 4131.
23. Dube, H.; Gommermann, N.; Knochel, P. Synthesis 2004, 12, 2015.
24. For the description of the P2Y12 radioligand displacement assay (binding
assay), see the Supplementary material. Binding IC50 values are given as single
26. (a) Zech, G.; Hessler, G.; Evers, A.; Weiss, T.; Florian, P.; Just, M.; Czech, J.;
Czechtizky, W.; Görlitzer, J.; Ruf, S.; Kohlmann, M.; Nazaré, M. J. Med. Chem.
2012, 55, 8615; (b) Kortum, S. W.; Lachance, R. M.; Schweitzer, B. A.;
Yalamanchili, G.; Rahman, H.; Ennis, M. D.; Huff, R. M.; TenBrink, R. E. Bioorg.
Med. Chem. Lett. 2009, 19, 5919.
27. For details of the P450 enzyme inhibition assay, see the Supplementary
material.
28. For details of the P450 enzyme inhibition assay, see the Supplementary
material.
29. For details of the hERG inhibition assay, see the Supplementary material.