Synthesis and evaluation of novel and potent protease activated receptor 4 (PAR4) antagonists based on a quinazolin-4(3H)-one scaffold

Article abstract of DOI:10.1016/j.ejmech.2021.113764

Protease activated receptor 4 (PAR4) is an important target in antiplatelet therapy to reduce the risk of heart attack and thrombotic complications in stroke. PAR4 antagonists can prevent harmful and stable thrombus growth, while retaining initial thrombus formation, by acting on the late diffusion stage of platelet aggregation, and may provide a safer alternative to other antiplatelet agents. To date, only two PAR4 antagonists, BMS-986120 and BMS-986141 have entered clinical trials for thrombosis. Thus, the development of a potent and selective PAR4 antagonist with a novel chemotype is highly desirable. In this study, we explored the activity of quinazolin-4(3H)-one-based PAR4 antagonists, beginning with their IDT analogues. By repeated structural optimisation, we developed a series of highly selective PAR4 antagonists with nanomolar potency on human platelets. Of these, 13 and 30g, with an 8-benzo[d]thiazol-2-yl-substituted quinazolin-4(3H)-one structure, showed optimal activity (h. PAR4-AP PRP IC₅₀ = 19.6 nM and 6.59 nM, respectively) on human platelets. Furthermore, 13 and 30g showed excellent selectivity for PAR4 versus PAR1 and other receptors (IC₅₀s > 10 μM) on human platelets. And 13 and 30g were lack of cross-reactivity for PAR1 or PAR2 (PAR1 AP FLIPR IC₅₀ > 3162 nM, PAR2 AP FLIPR IC₅₀ > 1000 nM) in the calcium mobilization assays. Metabolic stability assays and cytotoxicity tests of 13 and 30g indicated that these compounds could sever as promising drug candidates for the development of novel PAR4 antagonists. In summary, the quinazolin-4(3H)-one-based analogues are the first reported chemotypes with excellent activity and selectivity against PAR4, and, in the current study, we expanded the structural diversity of PAR4 antagonists. The two compounds, 13 and 30g, found in our study could be promising starting points with great potential for further research in antiplatelet therapy.

Full text of DOI:10.1016/j.ejmech.2021.113764

European Journal of Medicinal Chemistry 225 (2021) 113764
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Synthesis and evaluation of novel and potent protease activated
receptor 4 (PAR4) antagonists based on a quinazolin-4(3H)-one
scaffold
,
, *
Shangde Liu a, Duo Yuan a, Shanshan Li a, Roujie Xie a, Yi Kong b **, Xiong Zhu a
a Institute of Medicinal & Chemistry, China Pharmaceutical University, Nanjing, 210009, PR China
b School of Life & Technology, China Pharmaceutical University, Nanjing, 210009, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Protease activated receptor 4 (PAR4) is an important target in antiplatelet therapy to reduce the risk of
heart attack and thrombotic complications in stroke. PAR4 antagonists can prevent harmful and stable
thrombus growth, while retaining initial thrombus formation, by acting on the late diffusion stage of
platelet aggregation, and may provide a safer alternative to other antiplatelet agents. To date, only two
PAR4 antagonists, BMS-986120 and BMS-986141 have entered clinical trials for thrombosis. Thus, the
development of a potent and selective PAR4 antagonist with a novel chemotype is highly desirable. In
this study, we explored the activity of quinazolin-4(3H)-one-based PAR4 antagonists, beginning with
their IDT analogues. By repeated structural optimisation, we developed a series of highly selective PAR4
antagonists with nanomolar potency on human platelets. Of these, 13 and 30g, with an 8-benzo[d]
thiazol-2-yl-substituted quinazolin-4(3H)-one structure, showed optimal activity (h. PAR4-AP PRP
IC50 ¼ 19.6 nM and 6.59 nM, respectively) on human platelets. Furthermore, 13 and 30g showed excellent
Received 30 May 2021
Received in revised form
17 July 2021
Accepted 6 August 2021
Available online xxx
Keywords:
Protease activated receptor 4
Small-molecule antagonists
Structural optimisation
Quinazolin-4(3H)-one
Antiplatelet agents
selectivity for PAR4 versus PAR1 and other receptors (IC50s > 10 mM) on human platelets. And 13 and 30g
were lack of cross-reactivity for PAR1 or PAR2 (PAR1 AP FLIPR IC50 > 3162 nM, PAR2 AP FLIPR
IC50 > 1000 nM) in the calcium mobilization assays. Metabolic stability assays and cytotoxicity tests of 13
and 30g indicated that these compounds could sever as promising drug candidates for the development
of novel PAR4 antagonists. In summary, the quinazolin-4(3H)-one-based analogues are the first reported
chemotypes with excellent activity and selectivity against PAR4, and, in the current study, we expanded
the structural diversity of PAR4 antagonists. The two compounds, 13 and 30g, found in our study could be
promising starting points with great potential for further research in antiplatelet therapy.
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
Abbreviations: PAR, Protease activated receptor; COX, Cyclooxygenase; ASA,
Aspirin; ADP, Adenosine 50-diphosphate; GP, Glycoprotein; GPCRs, G-protein-
coupled receptors; IDT, Imidazole[2,1-b][1,3,4]thiadiazole; PRP, Platelet-rich
plasma; AP, Agonist peptide; BMS, Bristol-Myers Squibb Company (NY, USA); DDQ,
2,3-Dichloro-5,6-dicyano-1,4-benzoquinone; TEA, Triethylamine; DMF, N,N0-
Dimethylformamide; DCM, Dichloromethane; ACN, Acetonitrile; Pd(dppf)Cl2, [1,10-
Bis(diphenylphosphino)ferrocene]dichloropalladium(II); EtOH, Ethanol; SAR,
Structure activity relationships; CRCs, Concentration-response curves; CDI, N,N0-
Carbonyldiimidazole; THF, Tetrahydrofuran; NMP, N-Methylpyrrolidone; Pd(PPh3)4,
Tetrakis(triphenylphosphine)palladium; MsCl, Methanesulfonyl chloride; MeOH,
Methanol; TMSOTf, Trimethylsilyl trifluoromethanesulfonate; DIPEA, N,N0-Diiso-
propylethylamine; FLIPR, fluorescent imaging plate reader; HEK, human embryonic
kidney; LM, liver microsomes; T1/2, half-life; CCK-8, cell counting kit; ICR, Institute
of Cancer Research; PPP, Platelet-poor plasma.
In pathological conditions, abnormal activation of platelets may
cause unexpected thrombosis [1], leading to myocardial infarction,
ischaemic stroke, or other cardiovascular diseases. Antiplatelet
agents are commonly used in clinical practice to prevent and treat
ischemic cardiovascular diseases [2e4]. For example, a COX-1 in-
hibitor, aspirin (ASA), inhibits the thromboxane pathway. Clopi-
dogrel, a P2Y12 inhibitor, blocks the binding of the ADP agonist to
the P2Y12 receptor to inhibit platelet activation. Tirofiban, a GP IIb/
IIIa receptor antagonist, inhibits the common final pathway of
platelet aggregation [4e8]. In recent years, various propitious ad-
vances in antiplatelet therapies have significantly enhanced the
antiplatelet efficacy and reduced the side effects, such as bleeding
risk or thrombocytopenia, of therapeutic agents [4]. One of the
* Corresponding author.
** Corresponding author.
E-mail addresses: liushangde0707@163.com (S. Liu), yikong@cpu.edu.cn
(Y. Kong), mcicpuzx@hotmail.com (X. Zhu).
https://doi.org/10.1016/j.ejmech.2021.113764
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
most remarkable antiplatelet agents is the protease activated re-
ceptor (PAR) antagonist.
Moreover, both BMS-986120 and BMS-986141 have been in phase
II studies since 2017, and no further information has been disclosed
to date [8]. In our laboratory, after a failure in structural optimisa-
tion of IDT analogues, we decided to think out of the box and
develop novel chemotypes for small-molecule PAR4 antagonists. In
this study, we aimed to develop selective and efficacious PAR4
antagonists with improved drug-like properties as promising
leading compounds for antiplatelet therapy.
The effects of thrombin on human platelets are mediated pri-
marily by two PARs, PAR1 and PAR4 [9]. In addition to activating
PARs, thrombin triggers fibrin formation and protein C activation
[10,11]. PAR antagonists have the potential to block platelet acti-
vation by thrombin while sparing the other functions of thrombin
[11]. Thus, PAR antagonists might provide a larger therapeutic in-
dex than that achieved with coagulation inhibitors, especially for
chronic applications [12]. Although both belong to a family of G-
protein-coupled receptors (GPCRs) [13], PAR4 has some unique
properties compared to those of PAR1, which is likely due to
structural differences between the two receptors. Specifically, PAR1
contains a hirudin-like thrombin-binding site that is absent in
PAR4, whereas PAR4 contains an anionic sequence downstream of
the thrombin cleavage site [14e16]. The lack of the hirudin-like site
may decrease the sensitivity of PAR4 to thrombin, whereas the
anionic sequence seems to be important for a sustained thrombin
signal from PAR4 [17,18]. This results in distinct signalling kinetics
between these two receptors e PAR1 activation drives a rapid initial
signal, whereas PAR4 activation induces a slower but more pro-
longed response [16,19] e and explained that PAR4 antagonists
prevent thrombosis by inhibiting the late-stage activation events of
platelet procoagulant function. This suggests that PAR4 represents
a safe and effective target for antiplatelet therapy compared to
other targets of antiplatelet agents in current use, including PAR1
[20,21].
2. Results and discussion
In our laboratory, PAR4 antagonist activity was evaluated
through mouse or human platelet-rich plasma (PRP) aggregation
upon PAR4 agonist peptide (PAR4 AP) AYPGKF-NH2 stimulation
using a phototurbidimetry assay [29,33,34]. For purposes of com-
parison, we also synthesised several IDT-based compounds that
were disclosed by Bristol-Myers Squibb, as shown in Fig. 2. HTS Hit
(BMS) was one of the most potent hits identified via high-
throughput screening (HTS) by Bristol-Myers Squibb, and BMS-1
was chosen based on the article published in 2019 [35]. The syn-
thesis of BMS-986120, BMS-986141, HTS Hit (BMS), and BMS-1 was
carried out as described previously [27,35e37], except for the last
two steps of BMS-986120 synthesis for which Williamson ether
synthesis was used instead of the Mitsunobu reaction (Supporting
Information). Among the comparison compounds, BMS-986120
and BMS-986141 showed IC50s of 0.093
respectively, whereas HTS Hit (BMS) and BMS-1 showed that of
0.94 M and 1.82 M, respectively, on mouse PRP. More details
mM and 0.045 mM,
Vorapaxar is the first and the only FDA-approved (2014) small-
molecule PAR1 antagonist used for the prevention of myocardial
infarction and for the treatment of peripheral arterial disease [22].
However, clinical trials revealed an increase in severe bleeding
rates in patients receiving vorapaxar, most notably, intracranial
haemorrhage in patients with a prior stroke, which has substan-
tially limited the clinical use of vorapaxar [23,24]. So far, no PAR4
antagonist has been approved for marketing and only two antag-
onists, BMS-986120 and BMS-986141 from the Bristol-Myers
Squibb Company (NY, USA) have entered clinical trials for throm-
bosis [25e28], as shown in Fig. 1. In general, small-molecule PAR4
antagonists that have been reported include indazole, indole,
imidazole[2,1-b] [1,3,4]thiadiazole (IDT), quinoline, and quinoxa-
line analogues. In recent years, due to the excellent performance of
BMS-986120 and BMS-986141, most institutions have focused on
the structural optimisation of BMS-986120 and IDT derivatives
[29e32]. For example, Chen et al. [29] synthesised a series of
deuterated derivatives of BMS-986120, and in vitro metabolic sta-
bility studies showed that deuterium of several sites might not be
suitable for improving metabolic stability. In 2019, several patents
[30e32] were published, which reported new structures and bio-
logical evaluation results of IDT analogues for PAR4 antagonists.
However, there have been hardly any breakthrough innovations
either in structure or pharmacology over the past few years.
m
m
about the chemical synthesis and structural characterization can be
found in Supporting Information.
2.1. Design and development of IDT-substituted analogues
Our initial optimisation strategy is shown in Fig. 2, after trans-
posing and overturning the 2-IDT moiety of BMS-1 to 7-position,
several 7-IDT-substituted benzofuran analogues, 1a-e, were ob-
tained (Fig. 3). Briefly, the 1-series benzofuran analogues were
obtained by Suzuki coupling reactions between the corresponding
benzofuran-7-ylboronic acids and IDT bromides. Of these, com-
pounds 1b and 1c, with a 5-methyl substituent, showed weak
antagonistic activity on mouse PRP aggregation induced by PAR4
AP.
Initially, we held the IDT moiety and introduced other benzo-
hetercyclic rings to replace the benzofuran group, and hoped to see
an increase in PAR4 activity. Therefore, three structural skeletons
were designed to replace the benzofuran moiety, as shown in Fig. 4.
However, the replacement of benzofuran was unsatisfactory. There
were hardly any breakthrough improvements in activity against
PAR4 AP; 2a, 3b, and 4a-b displayed maintained, but still, micro-
molar activity, compared with that of the initial IDT analogues 1. In
addition, we found it was extremely difficult to acquire satisfactory
yields of the above compounds, including analogues 1a-e, due to
their final Suzuki coupling with IDT bromides, which prohibits
efficient synthesis of the compounds. After several unsuccessful
attempts at synthesis, we focused on finding other groups to
replace the IDT moiety.
2.2. Design and synthesis of benzo[d]thiazole -substituted
analogues
Interestingly, from 2018 to 2019, Bristol-Myers Squibb disclosed
a novel quinoline- or azaquinoline-based series of PAR4 antago-
nists, many of which possessed a benzo[d]thiazole bicyclic system
[38e41]. As an opportunity, we synthesised three novel com-
pounds (2b, 3c, and 4c) involving the 4-chloro-6-methoxybenzo[d]
Fig. 1. Protease activated receptor 4 (PAR4) antagonists in clinical stage.
2

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Fig. 2. Initial optimisation strategy from BMS-986120.
Fig. 3. Protease activated receptor 4 (PAR4) activity for 7-IDT-substituted benzofuran analogues 1.
Fig. 4. Protease activated receptor 4 (PAR4) activity for IDT-substituted benzimidazole analogue 2a, 3,4-dihydroisoquinolin-1(2H)-one analogues 3a-b, and isoquinolin-1(2H)-one
analogues 4a-b.
thiazole group, as shown in Fig. 5. The final Suzuki coupling with
benzo[d]thiazole bromides resulted in a great improvement in
yields. For purposes of comparison, we also synthesised BMS-2, a
quinoline-based compound that was disclosed by BMS in 2019 (see
3

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Fig. 5. Protease activated receptor 4 (PAR4) activity for BMS-2 (comparison compound, chosen from BMS patent published in 2019 [38]), benzo[d]thiazole-substituted benz-
imidazole analogue 2b, 3,4-dihydroisoquinolin-1(2H)-one analogues 3c, and isoquinolin-1(2H)-one analogues 4c.
Supporting Information). We were pleased to observe that the
(CRCs) on mouse PRP to determine PAR4 AP IC50s followed by a
single-point screen at 0.2 M and assessing the maximum per-
centage of inhibition induced by PAR4 AP on human PRP to
reconfirm the activity of compound 13 on human platelet. We
found that compound 13, with a 3,6-dimethylquinazolin-4(3H)-one
group, displayed dramatic activity (m. PAR4 AP PRP
three compounds displayed modest activity (1e2
AP on mouse PRP, comparable with BMS-2 (0.79
m
m
M) against PAR4
M), highlighting
m
that benzo[d]thiazole might be a potent group in maintaining PAR4
activity. The most potent analogue in this initial optimisation was
4c, with a PAR4 AP IC50 of 0.98
comparable to that of HTS-hit (BMS) (m. PAR4 AP PRP
IC50 ¼ 0.94 M). However, compared to that of BMS-986120 or
mM on mouse PRP, which was
IC50 ¼ 0.084
mM) and 94.5% inhibition at 0.2 mM chemical con-
m
centration on human PRP, which was comparable to that of BMS-
986120. Having identified the optimal quinazolin-4(3H)-one moi-
ety in compound 13, our next optimisation approach is summa-
rized in Fig. 6.
BMS-986141, compound 4c did not possess enough activity. In
addition, as shown in Scheme 1, the synthesis of 4c was rather
complicated. Beckmann rearrangement of intermediate 7 produced
8 in 20% of the yield and an unwanted by-product, 9, was found in
45% of the yield. Other synthetic routes such as Knoevenagel
condensation of 5 and DDQ dehydrogenation of 10 also resulted in
unsatisfactory yields. As a result, we shifted our focus to quinazolin-
4(3H)-one analogues, primarily to simplify the synthesis of com-
pound 4c, which produced unexpected results.
2.3.2. Impact of substitution at the 8-position of quinazolin-4(3H)-
one (R3)
First, we replaced the 4-chloro-6-methoxybenzo[d]thiazole
group of compound 13 with substituted benzene rings (20a-i), and
monocyclic (20j-k), bicyclic (20l-p and 20r-s), or tricyclic (20q)
heteroaryl groups. As shown in Scheme 2, 20a-r were obtained by
Suzuki coupling between corresponding boronic acids and bro-
mides of two moieties in good yield. However, 20s was obtained via
Stille coupling between 8-bromo-3,6-dimethylquinazolin-4(3H)-
one and organotin intermediate, 2-(tributylstannyl)-4,5,6,7-
tetrahydrobenzo[d]thiazole (synthesis shown in Supporting In-
formation). As shown in Table 1, the benzo[d]thiazole moieties are
identified as the important functional groups contributing to the
PAR4 inhibitory activity, probably due to the thiazole ring, and their
replacement leads to significant loss in activity. Moreover, most of
the analogues (20a-s) couldn't afford better or even comparable
activity against PAR4 AP, compared to 13 (m. PAR4 AP PRP
2.3. Design, synthesis and SAR of quinazolin-4(3H)-one
-substituted analogues
2.3.1. Compound 13 is a potent PAR4 antagonist
By replacing the isoquinolin-1(2H)-one core of 4c with quina-
zolin-4(3H)-one, compound 13 was obtained by a simple four-step
synthesis as shown in Scheme 2. Condensation of commercially
available 2-amino-3-bromo-5-methylbenzoic acid with corre-
sponding amino-congeners and
a subsequent condensation/
cyclised with methyl orthoformate delivered 8-bromo-3,6-
dimethylquinazolin-4(3H)-one. This was followed by Miyaura
borylation and Suzuki coupling to provide a satisfactory yield of 13.
Our screening strategy relied on concentration-response curves
IC50 ¼ 0.084
mM, 61.8% inhibition at 0.2 mM on human PRP). Among
the benzene rings (20a-i) or monocyclic heteroaryl groups (20j-k),
Scheme 1. Synthetic routes of compound 4c. Reagents and conditions: (a) Meldrum's acid, TEA/HCOOH, 100 ꢀC, 2 h, 29%; (b) oxalyl chloride, DMF, DCM, 0 ꢀC, 2 h; (c) AlCl3, DCM,
50 ꢀC, 8 h, 45% (two steps); (d) methanesulfonic acid, sodium azide, DCM, 0 ꢀC, 16 h, 20% of 8, 45% of 9; (e) DDQ, 1,4-dioxane, 100 ꢀC, 24 h, 32%; (f) Cs2CO3, CH3I, DMF/ACN, 35 ꢀC, 3 h,
85%; (g) bis(pinacolato)diboron, Pd(dppf)Cl2, KOAc, 1,4-Dioxane, 120 ꢀC, 2 h; (h) 2-bromo-4-chloro-6-methoxybenzo[d]thiazole, Pd(dppf)Cl2, Na2CO3 (2 M), toluene/EtOH, 120 ꢀC,
1.5 h, 87%.
4

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Scheme 2. Synthetic routes of quinazolin-4(3H)-one analogues 13 and 20a-aq. Reagents and conditions: (a) ReNH2, CDI, THF, 70 ꢀC, 2e5 h; (b) methyl orthoformate, HCl/1,4-
dioxane, NMP, 110 ꢀC, 1e6 h; (c) bis(pinacolato)diboron, Pd(dppf)Cl2, KOAc, 1,4-Dioxane, 120 ꢀC, 1.5e20 h; (d) Pd(dppf)Cl2, Na2CO3 (2 M), toluene/EtOH, 120 ꢀC, 0.5e4 h; (e)
Pd(PPh3)4, KOAc, 1,4-dioxane, 120 ꢀC, 0.5e10 h; (f) formamide, 100 ꢀC, 24 h; (g) HCHO/H2O, 1,4-dioxane, 80e110 ꢀC, 24e48 h; (h) MsCl, TEA, DCM, 0e25 ꢀC, 1 h; (i) NaOR1, R1OH,
25 ꢀC, 1e1.5 h.
only 20g, with a 2,4-diclorophenyl group, displayed passable ac-
M, 61.8% inhibition at 0.2 M on
displayed micromolar activity, with IC50s of 2.62
2.07 M, respectively. However, 20s, with
tetrahydrobenzo[d]thiazolyl group, displayed modest activity (m.
m
M, 5.57
m
M and
tivity (m. PAR4 AP PRP IC50 ¼ 1.22
m
m
m
a
4,5,6,7-
human PRP). Among the bicyclic groups (20l-p, 20r-s), 1-
naphthalene analogue 20l and 2-quinoline analogue 20m were
inactive, as was a 2-benzofuran analogue 20n. Replacing the 40-Cl
or 60-methoxyl groups (20o-r) of 13 decreased its activity signifi-
cantly: 20o, replacing the 40-Cl of 13 with F atom displayed sub-
PAR4 AP PRP IC50 ¼ 0.11
mM, 92.4% inhibition at 0.2 mM on human
PRP).
2.3.3. Influence of the 2-, 3- and 6-position substitution of
quinazolin-4(3H)-one
As a result, the three substituted bicyclic thiazolyl groups, used
in analogues 13, 20o and 20s, were selected for the optimisation of
micromolar activity (m. PAR4 AP PRP IC50 ¼ 0.83
mM); replacing the
60-methoxyl group of 13 with 60-Cl (20p), 60,70-diydro [1,4]dioxine
group (20q) or 6’-((tert-butoxycarbonyl)amino)ethoxy (20r)
5

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
the quinazolin-4(3H)-one structural skeleton. As shown in Scheme
2, 2-unsubstituted quinazolin-4(3H)-one analogues (20t-aq) could
be obtained via a similar approach as that used in 13 or 20s. The
synthesis of 23a-c and 28 are shown in Scheme 3. The activities of
these compounds are shown in Table 2. The structure and activity
relationships (SAR) indicated that: (1) compounds with a 4-chloro-
6-methoxybenzo[d]thiazole group displayed better activity
compared to that of compounds with two other thiazolyl groups;
(2) 2-methyl or 2-isopropyl substituted analogues (23a-c, 28)
showed significant loss in activity against PAR4 AP; (3) in many
cases, replacement of 6-methyl with H, ethyl, or n-butyl groups
leads to sharply loss in activity, except for 20aa, with a 6-ethyl
group, which displayed maintained activity (m. PAR4 AP PRP
Fig. 6. Optimisation strategy for 13 to improve protease activated receptor 4 (PAR4)
IC50 ¼ 0.16
mM; 82.8% inhibition at 0.2 mM on human PRP); (4) the
replacement of the 3-methyl moiety with benzyl or alkyl groups
antagonist activity.
decreased the activity of the various analogues sharply, only 20y
Table 1
Structures and activities of quinazolin-4(3H)-one analogues (8- position structure-activity relationships).
Compd. R3
m. PAR4 AP PRP IC50 % inhibition at 0.2
M)a
PRP)
m
M (h. PAR4 AP Compd. R3
m. PAR4 AP PRP IC50 % inhibition at 0.2
M)a
(m PRP)
mM (h. PAR4 AP
(
m
13
0.084
94.5
20a
-Ph
>10
ND
20b
20d
20f
20h
20j
-Ph-4-CH3
-Ph-2-OCH3
-Ph-2-OH-4-F
-Ph-2,3-Cl
>10
7.32
>10
3.70
>10
ND
17.4
6.5
34.3
1.8
20c
20e
20g
20i
-Ph-2-CH3
-Ph-2-F
-Ph-2,4-Cl
-Ph-2-CN
>10
5.48
1.22
>10
3.64
ND
16.0
61.8
ND
20k
7.1
20l
>10
ND
20m
20o
>10
ND
20n
4.26
44.4
0.83
89.9
20p
20r
2.62
2.07
21.3
54.1
20q
20s
5.57
0.11
35.7
92.4
a
In vitro inhibition of protease-activated receptor 4 agonist peptide (PAR4-AP) (100
typically the mean of at least two independent experiments.
mM)-induced platelet-rich plasma (PRP) platelet aggregation; reported IC50 values are
6

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Scheme 3. Synthesis of 2-substituted quinazolin-4(3H)-one analogues, 23 and 28. Reagents and conditions: (a) Ac2O, toluene, 140 ꢀC, 3 h; (b) NH3.H2O, 100 ꢀC, 2 h; (c) Cs2CO3, CH3I,
DMF, 25 ꢀC, 3 h; (d) bis(pinacolato)diboron, Pd(dppf)Cl2, KOAc, 1,4-Dioxane, 120 ꢀC, 1.5 h; (e) R3-Br, Pd(dppf)Cl2, Na2CO3 (2 M), toluene/EtOH, 120 ꢀC, 1e2 h; (f) R3-Sn(C4H9)3,
Pd(PPh3)4, KOAc, 1,4-dioxane, 120 ꢀC, 1.5 h; (g) c.H2SO4, MeOH, 85 ꢀC, 18 h; (h) isobutyryl chloride, pyridine, 0e25 ꢀC, 12 h; (i) NH3.H2O, sealed tube, 120 ꢀC, 12 h; (j) K2CO3, MeOH,
reflux, 12 h.
(m. PAR4 AP PRP IC50 ¼ 0.95
m
M; 70.2% inhibition at 0.2
m
M on
2.3.4. Impact of 60-substituents on the benzo[d]thiazole moiety in
20r
human PRP) displayed submicromolar activity. In general, few of
them were activity-competitive with 13.
Except for 4-chloro-6-methoxybenzo[d]thiazole moiety, we
Table 2
Structures and activities of quinazolin-4(3H)-one analogues (2-, 3-, 6- position structure-activity relationships).
Compd.
R
R1
R2
m. PAR4 AP PRP IC50
(
m
M)a
% inhibition at 0.2 mM (h. PAR4 AP PRP)
20t
20u
20v
20w
20x
20y
H
-CH3
-CH3
-CH3
-CH3
-CH3
-CH2CH3
-CH(CH3)2
-CH(CH2)2
-CH2OCH3
-CH2OCH2CH3
-Bn
H
H
H
H
H
H
H
H
6.78
1.85
9.77
4.26
1.00
0.65
>10
0.16
1.38
2.09
>10
5.34
9.95
6.74
4.20
5.45
2.61
1.38
3.74
>10
3.37
2.32
4.77
3.54
>10
7.04
6.28
8.21
25.6
36.5
15.2
29.2
45.3
91.2
ND
92.8
47.6
39.8
ND
18.7
18.8
10.2
27.7
20.1
19.8
41.6
26.8
ND
17.4
35.7
15.1
25.7
ND
-CH3
-CH3
20z
20aa
20 ab
23a
-CH2CH3
-CH2CH2CH2CH3
-CH3
-CH3
-CH3
-CH3
-CH3
H
-CH3
-CH(CH3)2
H
H
H
H
-CH3
H
28
-CH3
H
-CH3
-CH3
-CH3
-CH3
H
-CH3
-CH3
-CH3
-CH3
-CH3
20ac
20ad
20ae
20af
23b
20 ag
20ah
20ai
20aj
20ak
20 al
20am
20an
20ao
20ap
20aq
23c
-CH3
-CH2CH3
-CH(CH3)2
-CH(CH2)2
-CH3
-CH3
H
-CH2CH3
-CH(CH3)2
-CH(CH2)2
-CH2OCH3
-CH2OCH2CH3
-CH3
-CH2CH3
-CH3
-CH2CH3
-CH3
H
H
H
H
H
H
H
H
-CH3
-CH2CH3
-CH2CH3
-CH2CH2CH2CH3
-CH2CH2CH2CH3
-CH3
H
H
-CH3
25.7
9.3
17.0
a
In vitro inhibition of protease-activated receptor 4-agonist peptide (PAR4-AP) (100
typically the mean of at least two independent experiments.
mM)-induced platelet-rich plasma (PRP) platelet aggregation; reported IC50 values are
7

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
found that many of the quinoline-based series of PAR4 antagonists
disclosed in BMS patents [38,39] possessed sulfnamido moieties, as
shown in Fig. 7. As an opportunity, we elected to synthesis several
quinazolin-4(3H)-one derivatives with sulfonamido moieties,
hoping to see an increase in activity against PAR4 AP. As shown in
Scheme 4, compound 29 was obtained by deprotection of 20r, and
followed by condensation to afford 30a-i in good yields. Here,
several compounds with benzenesulfonamido moieties (30b, 30c,
30f, and 30g) afforded submicromolar PAR4 inhibition on mouse
PRP, as shown in Table 3. However, the most potent analogue in this
of several quinazolin-4(3H)-one analogues on PAR4, a calcium
mobilization assay (PAR fluorescent imaging plate reader (FLIPR))
was performed using PAR-expressing human embryonic kidney
(HEK) 293 cell lines and PAR agonist peptides [42,43]. As shown in
Table 4, these quinazolin-4(3H)-one analogues (13, 20o, 20s, 20aa
and 30g) inhibited PAR4 AP-induced cellular calcium mobilization
in HEK293 cells with high potency, and compound 13 showed a
comparable, even better activity (PAR4 FLIPR IC50 ¼ 0.53 nM)
against PAR4 AP, compared with BMS-986120 (PAR4 FLIPR
IC50 ¼ 0.70 nM). Moreover, all of these compounds were lack of
cross-reactivity for PAR1 or PAR2, because these compounds were
ineffective against stimulation with human PAR1 AP or human
PAR2 AP (PAR1 FLIPR IC50 > 3162 nM, PAR2 FLIPR IC50 > 1000 nM),
which indicated that the tested compounds were both high po-
tency and selectivity PAR4 antagonists.
route was 30g, with an IC50 of 0.17 mM in mouse PRP, and <0.2 mM
on human PRP. Herein, the results indicated that the benzene-
sulfonamido moieties on the benzo[d]thiazole group might be
beneficial to inhibit PAR4 activation.
2.4. Reconfirmation of PAR4 AP activity of quinazolin-4(3H)-one
analogues on human PRP
2.5.2. In vitro selectivity of 13 and 30g on human PRP
After several attempts to create structural optimisation, several
quinazolin-4(3H)-one analogues with potential activity against
PAR4 AP have been disclosed. Therefore, we next focused on the
examination of the in vitro antiplatelet aggregation activity of these
analogues on human platelets. As shown in Fig. 8, BMS-986120 and
BMS-986141 were used as positive control compounds, and seven
optimised compounds were determined at human PRP levels
The selectivity of 13 and 30g on human PRP was tested by a
single-point screen at 10
mM and assessing the maximum per-
centage of inhibition induced by a variety of platelet agonists on
human PRP [35,44,45] as shown in Fig. 9. 13 and 30g showed potent
inhibition of platelet aggregation triggered by PAR4 AP (AYPGKF-
NH2) or thrombin, but no measurable inhibition of the effects of a
PAR1 AP (SFLLRN-NH2), collagen, adenosine 50-diphosphate (ADP),
or the thromboxane A2 receptor agonist U46619 on human plate-
lets (IC50s > 10 mM), which reconfirmed that the quinazolin-4(3H)-
one analogues 13 and 30g are selective and potent anti-PAR4
agents.
induced by PAR4 AP (75 mM). We found that 20g showed decent
activity against PAR4 on human PRP, with an IC50 of 174 nM, which
once again indicated that the absence of the thiazole ring caused
loss in activity. Among these optimised compounds, 20o, 20s, 20y,
and 20aa displayed decent activity on human PRP, with IC50s of
77.8, 28.5, 106.7, and 40.0 nM, respectively, which matched the
activity data measured on mouse PRP, as shown in Table 3. How-
ever, the two most potent compounds, 13 and 30g, showed
2.6. Metabolic stability assays of 13 and 30g
reversed activity on human PRP, i.e., 13 showed an IC50 of 0.084
on mouse PRP and 19.6 nM on human PRP, whereas 30g showed an
IC50 of 0.17 M on mouse PRP and 6.59 nM on human PRP. Herein,
mM
In vitro metabolic study using liver microsomes (LM) is the most
common approach to early estimation and prediction of in vivo
metabolic stability of a compound. To evaluate the metabolic sta-
bility of 13 and 30g, liver microsomes of three species (mouse, rat
and human) were employed. In vitro metabolic assay data of 13 and
30g are summarized in Table 5. The results showed that in com-
parison with BMS-986120 (T1/2 ¼ 11.1 min), there was a significant
improvement of in vitro metabolic stability for 13 (T1/2 ¼ 42.6 min)
and 30g (T1/2 ¼ 444 min) in human liver microsomes. Moreover,
compared to 13, 30g showed an excellent metabolic stability in
human, rat and mouse liver microsomes, indicated that the ben-
zenesulfonamido moiety of 30g might be beneficial to improve
drug-like properties. This finding demonstrated that 13 and 30g
exhibited good metabolic stability, suggesting that these com-
pounds could sever as promising drug candidates for the devel-
opment of novel PAR4 antagonists.
m
compound 13 showed a little slightly potent relative to that of BMS-
986120, and the most potent analogue in the above compounds
was 30g with a human PAR4-AP PRP IC50 of 6.59 nM, which was
comparable to that of BMS-986120 (h. PAR4-AP PRP IC50 ¼ 9.70 nM)
and BMS-986141 (h. PAR4-AP PRP IC50 ¼ 3.35 nM). In general, the
results indicated that some of the quinazolin-4(3H)-one analogues
are highly potent PAR4 antagonists.
2.5. Specificity of quinazolin-4(3H)-one analogues on PAR4
2.5.1. Calcium mobilization assays
PAR1, PAR2 and PAR4 couple to the G-protein Gq, which upon
receptor activation induces intracellular-signalling events resulting
in an increase in cytosolic calcium [42]. To conform the selectivity
Fig. 7. Representative quinoline-based compounds with sulfnamido moieties in BMS patents [38,39].
8

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Scheme 4. Synthesis of 29 and 30a-i. Reagents and conditions: (a) 2,6-lutidine, TMSOTf, DCM, 25 ꢀC, 2 h; (b) R4-Cl, DIPEA, DMF, 25 ꢀC, 20e60 min.
Table 3
Structures and activities of quinazolin-4(3H)-one analogues 29 and 30a-i.
Compd. R4
m. PAR4 AP PRP IC50
M)a
% inhibition at 0.2
PRP)
m
M (h. PAR4 AP Compd. R4
m. PAR4 AP PRP IC50
(m
M)a
% inhibition at 0.2
PRP)
mM (h. PAR4 AP
(
m
29
H
3.02
33.2
30a
>10
ND
30b
30d
30f
2.52
17.9
32.3
47.1
ND
30c
30e
30g
30i
0.52
30.5
8.7
1.04
0.95
>10
8.22
0.17
>10
94.5
ND
30h
a
In vitro inhibition of protease activated receptor 4-agonist peptide (PAR4-AP) (100
typically the mean of at least two independent experiments.
m
M)-induced platelet-rich plasma (PRP) platelet aggregation; reported IC50 values are
2.7. Cytotoxicity test of 13 and 30g against HEK293T cells
human, rat and mouse liver microsomes. And cytotoxicity test
indicated that 13 and 30g displayed low cytotoxic activities
To acquire insights about cytotoxicity of this new series of qui-
nazolin-4(3H)-one analogues, the most potent derivatives 13 and
30g have been tested against HEK293T cell. As shown in Table 6,
both 13 and 30g displayed low cytotoxic activities, as indicated by
(IC50s > 20 mM) against HEK293T cell. All of these results sug-
gesting that both 13 and 30g could sever as promising drug can-
didates for the development of novel antagonists. Moreover, It was
the first reported PAR4 antagonist with quinazolin-4(3H)-one
chemotype. However, the in-depth investigation on the biological
profile of our compounds is still insufficient. Collectively, our
findings suggest that quinazolin-4(3H)-one analogues are a good
choice for the design and development of PAR4 antagonists and
deserve further research. Future investigation on the structural
modifications of these antagonists and in vitro/in vivo studies are
ongoing in our laboratory and will be reported in due course.
the high cell viability % at the examined doses of 1e20 mM.
3. Conclusions
Starting from structure modification of BMS-986120, a novel
chemotype with quinazolin-4(3H)-one structural skeleton for PAR4
antagonist has been developed after several routes of optimisation.
Based on this scaffold, we have obtained the compounds 13 and
30g; the results of antiplatelet aggregation assays showed that both
these compounds were potent PAR4 antagonists on mouse and
human platelets, comparable to the activity of BMS-986120 or
BMS-986141. In vitro selectivity of 13 and 30g in the calcium
mobilization assays or on human PRP indicated that both 13 and
4. Experimental section
4.1. Preparation of platelet-rich plasma
Institute of Cancer Research (ICR) mice (male, 18e22 g) were
purchased from Nanjing Qinglongshan Animal Center (Nanjing,
Jiangsu, China). The animals were acclimatised for at least 1 d prior
to use. Blood was collected from the abdominal aorta of healthy
mice with 3.8% sodium citrate as anticoagulant (1:9 citrate/blood,
30g were lack of cross-reactivity for PAR1 or PAR2 (IC50s > 1
and 13 and 30g were inactive against PAR1 or other receptors
(IC50s > 10 M) on human platelets. Metabolic stability assays
showed that 13 and 30g exhibited good metabolic stability in
mM),
m
9

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Fig. 8. The IC50 of positive control compounds: BMS-986120 and BMS-986141, and several optimised compounds were determined at human platelet-rich plasma (PRP) levels
induced by protease activated receptor 4-agonist peptide (PAR4-AP) (75
mM). Data are presented as the mean SD (n ¼ 3).
aggregometer (AG400, Beijing Techlink Medical Technology Co.,
China) according to the phototurbidimetry method. In brief, 270
of PRP were pipetted into aggregometer cuvettes and were pre-
incubated at 37 ꢀC for 5 min with 20
L sample solution at appro-
priate concentrations. Saline served as vehicle control. Platelet
aggregation was stimulated with 10 L PAR4 AP (AYPGKF-NH2) in
the presence of magnetic beads, and the maximum platelet ag-
gregation rate was determined by a lasting measurement of light
transmittance for 5 min. The selectivity of the compounds was
tested following the same protocol, except that platelets were
Table 4
In vitro selectivity of quinazolin-4(3H)-one analogues and BMS-986120 in the cal-
cium mobilization assays.
mL
Compd.
PARs FLIPR IC50 values (nM)
m
PAR4a
PAR1b
PAR2c
m
BMS-986120
0.70 0.08
0.53 0.02
3.2 0.61
5.7 0.33
6.8 0.17
1.2 0.26
>3162
>3162
>3162
>3162
>3162
>3162
>1000
>1000
>1000
>1000
>1000
>1000
13
20o
20s
20aa
30g
stimulated by a PAR1-agonist peptide SFFLLRN-NH2 (5
m
M),
M), or the
M). For single concentrations
a
PAR4 antagonist activity was measured using HEK293/ga15/PAR4 cell line and a
PAR4 AP (AYPGKF-NH2).
thrombin (0.1 U/mL), collagen (10
thromboxane A2 mimetic U46619 (3
m
m
g/mL), ADP (10 m
b
PAR1 antagonist activity was measured using HEK293/ga15/PAR1 cell line and a
of each compound tested, all data are an average of at least two
independent determinations. For IC50 of each compound tested, at
least five concentrations were chosen, and the curve of inhibition
rate and concentration was derived, then the IC50 value of this
compound was obtained by using Prism 7.0 (GraphPad, San Diego,
CA) [29,33,34].
PAR1 AP (SFFLRR-NH2).
c
PAR2 antagonist activity was measured using HEK293/ga15/PAR2 cell line and a
PAR2 AP (2-furoyl-LIGRLO-NH2). Reported IC50 values are typically the mean of at
least two experiments.
v/v). Human blood was purchased from the Nanjing Red Cross
Blood Center, China. Both mouse and human PRP was obtained by
centrifugation at 160ꢁg for 15 min. Platelet-poor plasma (PPP) was
prepared from the precipitated fraction of PRP by centrifugation at
2000ꢁg for 15 min. PRP was diluted to 3 ꢁ 108/mL with PPP [29].
4.3Calcium mobilization assays
Fluorometric Imaging Plate Reader (FLIPR) calcium mobilization
assays were preformed using HEK293 cells. HEK293 Aequorin cells
that stably express human PARs were used to screen for PARs
antagonist activity. Cells were preincubated with a fluorescent
calcium indicator (Codex Biosolutions) according to manufacturer's
instructions and then treated with test compounds. The cells were
4.2. Platelet aggregation activity assay
Platelet aggregation was measured using
a
4-channel
10

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
Fig. 9. Effects of 13 and 30g on platelet aggregation of human PRP induced by AYPGKF peptide (75 mM), thrombin (0.1 U/mL), SFLLRN peptide (5 mM), collagen (10 mg/mL), ADP
(10
mM), and U46619 (3
mM). Percentages of inhibition are presented as mean SD (n ¼ 3).
Table 5
The metabolic stability of 13 and 30g in liver microsomes.
Compd.
Human Liver Microsomal Stability
Rat Liver Microsomal Stability
Mice Liver Microsomal Stability
T1/2 (min) CLint (mL/min/g)
282
428 29
e
T1/2 (min)a
CLint (mL/min/g)b
T1/2 (min)
CLint (mL/min/g)
13
30g
BMS-986120
42.6 5.5
444 15
11.1 0.32
36.6 1.8
3.51 0.23
125 3.74
31.9 3.6
323 18
e
78.1 2.8
7.73 0.21
e
6
19.3 1.4
12.8 4.0
e
a
LM t1/2 ¼ 0.693/k (k is the rate constant from a plot of In [concentration] vs. incubation time).
b
CLint (intrinsic clearance) ¼ 0.693 ꢁ liver weight ꢁ P450/(0.5 ꢁ t1/2). Data are presented as the mean SD (n ¼ 3).
Table 6
4.4Metabolic stability of compounds in liver microsomes
Cytotoxicity evaluation of 13 and 30g against HEK293T cell line.
Metabolic stability was assessed in the presence of human,
mouse and rat liver microsomes. Compound stock solutions were
initially prepared in 100% DMSO and subsequently diluted in
acetonitrile for the assay. The pH of the reactions was kept at ~7.4
with potassium phosphate buffer. The reaction wells were prepared
by adding microsomes to a well and allowed to warm to 37 ꢀC. Then
compound was added to each well. The reactions were started by
adding NADPH to the reaction well containing microsomes and
compounds. Negative controls received buffer only (instead of
NADPH). Immediately after reactions started, 0 min aliquots were
promptly collected and mixed in a separate well with ice cold
acetonitrile (spiked with internal standards) to quench the re-
actions. The remainder of the reaction volume was incubated at
37 ꢀC with shaking. An additional aliquot was collected at 60 min
after the start of the reaction and promptly quenched with ice cold
acetonitrile (spiked with an internal standard). Samples were vor-
texed and centrifuged at 3700 rpm for 10 min. The amount of
compound in the supernatant was determined by LC/MS/MS and
the percent of parent compound remaining after 60 min was cal-
uated by the following formula: CLint (intrinsic clearance) ¼ 0.693/
(T1/2 [microsomal protein]); T1/2 (half-life) ¼ 0.693/-k; k ¼ slope.
All reactions were run in triplicate, except negative controls (no
NADPH) which were performed as single reactions. Results re-
ported are the mean of each reaction triplicate, normalized to the
internal standard, and expressed as the percent of compound
remaining after the incubation time.
Compd.
Tested concentrations
Cell viability (%)a
13
1
m
M
91.26 5.49
73.68 5.09
59.33 2.73
87.71 6.01
76.80 11.87
61.80 3.72
10
20
m
m
M
M
30g
1
mM
10
20
m
m
M
M
a
Data are presented as the mean SD (n ¼ 3).
then incubated for 30 min at room temperature followed by addi-
tion of AP for measuring antagonist activity. A PAR4 AP, AYPGKF-
NH2 was used for activating PAR4. Compound potency was
derived from 10-point concentration-response curves (top dose
10 mM, half-log dilution). Selectivity against PAR1 was tested using
HEK293 cells expressing human PAR1 (HEK293/ga15/PAR1) and a
PAR1 AP (SFFLRR), and selectivity against PAR2 was tested using
HEK293 cells expressing human PAR2 (HEK293/ga15/PAR2) and a
PAR2 AP (furoyl-LIGRLO-NH2). Add 30
mL/well of the cultured
matrix fluid to a 384-well transparent. After 30 min, centrifuge for
1 min to remove the matrix fluid. After digesting HEK293/Ga15/PAR
cells with cell dissociation fluid, using 45,000 cells/well plate,
incubate at 37 ꢀC, 5% CO2 for 24 h. After the incubation is
completed, remove the culture medium, quickly add the prepared
buffer at a volume of 100 mL/well, and continue to incubate for
60 min at 37 ꢀC, 5% CO2 in the dark. Dissolve the sample com-
pounds in DMSO, dilute these concentration to 4% with HBSS
detection buffer, add 25 mL diluent to each well, and incubate at
room temperature in the dark for 15 min. After the incubation is
4.5Cytotoxicity assay of compounds on HEK293 cells
complete, place the 384-well plate on the fluorometric imaging
Cell Counting Kit-8 (CCK-8) assays were carried out on
HEK293 cells to primarily determine the cytotoxicity of test com-
pounds. Briefly, HEK293 cells with exponential growth were
seeded into 96-well culture plates at a density of 4000 cells/well
and allowed to adhere overnight. The cells were then treated with a
plate reader, and add 25
mL PARs-AP to each well. Measure the
relative fluorescence intensity of each well, and calculate the in-
hibition of the sample compounds to PAR1, PAR2 or PAR4.
11

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
series of 3 concentrations of the test compounds for 48 h. Cells were
treated with an equal volume of DMSO as the control. Thereafter,
10 mL CCK-8 solutions were added to each well and the plates were
organic phase was washed with brine, dried with anhydrous
Na2SO4, and evaporated at reduced pressure. Purification was car-
ried out via silica chromatography (5% EA/PE - 10% EA/PE) to afford
title compound as a yellow solid in 50% yield. 1H NMR (300 MHz,
incubated at 37 ꢀC for 2 h in the dark. The UV absorbance was
measured at 450 nm with an Epoch plate reader (BioTek). Cell
viability was calculated as (optical density of experimental sample/
optical density of control) ꢁ 100%.
Chloroform-d)
3H).
d
6.86 (t, J ¼ 3.0 Hz, 1H), 6.76e6.65 (m, 2H), 3.74 (s,
4-chloro-6-methoxybenzo[d]thiazol-2-amine.
Ammonium
thiocyanate (170 mg, 2.41 mmol) was added to an ACN solution
(5 mL) of 2-chloro-4-methoxyaniline (200 mg, 1.27 mmol). The
solution was stirred at room temperature for 10 min, after which an
ACN solution (2 mL) of benzyltrimethylammonium tribromide
(700 mg, 1.65 mmol) was added dropwise. The solution was stirred
at room temperature for 18 h. On completion, the mixture was
quenched with saturated NaHCO3 aqueous (10 mL), and extracted
with EA/THF (1/1, 3 ꢁ 10 mL). The combined organic layer was
washed with brine, dried with anhydrous Na2SO4, and evaporated
at reduced pressure to afford title compound as an off-white solid
used for next step without further purification.
4.6. Procedure for the synthesis of the compounds used in the study
The general chemistry, experimental information, and synthesis
of all compounds, including positive control BMS-1, HTS Hit (BMS),
BMS-986120, BMS-986141 and BMS-2, are supplied in Supporting
Information. The synthesis of compound 13 was provided below as
an example.
8-(4-chloro-6-methoxybenzo[d]thiazol-2-yl)-3,6-
dimethylquinazolin-4(3H)-one (13):
2-amino-3-bromo-N,5-dimethylbenzamide. In an argon at-
mosphere, N,N0-carbonyldiimidazole (500 mg, 3.14 mmol) was
added to an anhydrous THF solution (20 mL) of 2-amino-3-bromo-
5-methylbenzoic acid (480 mg, 2.09 mmol) in small protons. The
mixture was refluxed for 2 h, cooled to room temperature, and a
THF solution of methylamine (2 M, 4 mL) was added to it. The
mixture was refluxed for another 2 h. On completion, the mixture
was cooled to room temperature, and evaporated under reduced
pressure. The residue was diluted with EA (20 mL), washed with
NaOH aqueous (1 M, 10 mL) and brine, dried with anhydrous
Na2SO4, and evaporated at reduced pressure. Purification was car-
ried out by silica chromatography (20% EA/PE - 100% EA/PE) to
obtain title product as an off-white solid in 85% yield. 1H NMR
2-bromo-4-chloro-6-methoxybenzo[d]thiazole. To an ACN
solution (10 mL) of CuBr2 (210 mg, 1.40 mmol) was added tert-butyl
nitrite (0.2 mL) at 40 ꢀC. The mixture was stirred for 10 min, and an
ACN solution (8 mL) of 4-chloro-6-methoxybenzo[d]thiazol-2-
amine (250 mg, 1.17 mmol) was added dropwise to it over
10 min. The mixture was stirred at 40 ꢀC for 2 h. On completion, the
mixture was quenched with HCl aqueous (0.5 M, 30 mL) and
extracted with EA (3 ꢁ 20 mL). The combined organic layer was
washed with saturated NaHCO3 aqueous (40 mL) and brine, dried
with anhydrous Na2SO4, and evaporated at reduced pressure. Pu-
rification was carried out by silica chromatography (5% EA/PE - 20%
EA/PE) to obtain title compound as a light yellow solid in 74% yield.
(300 MHz, Chloroform-d)
d
7.37 (s, 1H), 7.09 (s, 1H), 6.04 (broad s,
1H NMR (300 MHz, Chloroform-d)
1H), 3.90 (s, 3H).
d
7.30 (s, 1H), 7.18 (d, J ¼ 8.5 Hz,
1H), 5.84 (broad s, 2H), 2.98 (d, J ¼ 3.9 Hz, 3H), 2.24 (s, 3H).
8-bromo-3,6-dimethylquinazolin-4(3H)-one. Methyl ortho-
formate (350 mg, 3.3 mmol) was added to an N-methylpyrrolidone
solution (10 mL) of 2-amino-3-bromo-N,5-dimethylbenzamide
(400 mg, 1.65 mmol), followed by a DCM solution of HCl (4 M,
0.5 mL) dropwise. The mixture was refluxed for 1.5 h. On comple-
tion, the reaction mixture was cooled to room temperature, poured
into ice-water, quenched with saturated NaHCO3 aqueous (50 mL)
carefully, and extracted with EA (3 ꢁ 50 mL). The combined organic
layer was washed with brine, dried with anhydrous Na2SO4, and
evaporated at reduced pressure. The residue was triturated with
isopropyl ether (5 mL), and the solid was collected and washed
with isopropyl ether (3 mL) to give title compound as a white solid
8-(4-chloro-6-methoxybenzo[d]thiazol-2-yl)-3,6-
dimethylquinazolin-4(3H)-one (13). In an argon atmosphere,
(3,6-dimethyl-4-oxo-3,4-dihydroquinazolin-8-yl)boronic
acid
(500 mg, 2.29 mmol), 2-bromo-4-chloro-6-methoxybenzo[d]thia-
zole (300 mg, 1.08 mmol), and Pd(dppf)Cl2 (3 mg, 1% w/w) were
added to a mixed solution of toluene (21 mL) and EtOH (7 mL). To
this mixture, Na2CO3 aqueous (2 M, 0.5 mL) was added dropwise.
The mixture was stirred at 130 ꢀC for 2.5 h. On completion, the
reaction mixture was cooled to room temperature, diluted with
saturated NaHCO3 aqueous (30 mL), and extracted with EA
(3 ꢁ 20 mL). The combined organic layer was washed with brine,
dried with anhydrous Na2SO4, and evaporated at reduced pressure.
Purification by silica chromatography (10% EA/PE - 30% EA/PE) to
afford crude product. The crude product was triturated with iso-
propyl ether (4 mL), the solid collected, and washed with isopropyl
ether (1 mL) to give title compound as a white solid in 76% yield. 1H
in 99% yield. 1H NMR (300 MHz, Chloroform-d)
d 8.12 (s, 1H), 8.08
(dd, J ¼ 2.0, 1.0 Hz, 1H), 7.88 (d, J ¼ 1.9 Hz, 1H), 3.61 (s, 3H), 2.49 (s,
3H).
(3,6-dimethyl-4-oxo-3,4-dihydroquinazolin-8-yl)boronic
acid. In an argon atmosphere, 8-bromo-3,6-dimethylquinazolin-
4(3H)-one (340 mg, 1.34 mmol), bis(pinacolato)diboron (800 mg,
3.32 mmol), and anhydrous AcOK (560 mg, 4.18 mmol) were added
to 1,4-dioxane (10 mL), where Pd(dppf)Cl2 (6 mg, 2% w/w) was
added in one portion. The mixture was stirred at 120 ꢀC for 2 h,
cooled to room temperature, diluted with H2O (20 mL), and
extracted with EA (3 ꢁ 20 mL). The combined organic layer was
washed with brine, dried with anhydrous Na2SO4, and evaporated
at reduced pressure to give the title compound as brown oil, which
was used in the next step directly without further purification.
2-chloro-4-methoxyaniline. Zinc dust (3.5 g, 53.3 mmol) was
added to a MeOH solution (25 mL) of commercially available 2-
chloro-4-methoxy-1-nitrobenzene (1.0 g, 5.33 mmol), followed by
NH4Cl (5.72 g, 106.6 mmol) in small portions. The mixture was
stirred at room temperature for 1 h. On completion, the solvents
were removed at reduced pressure. The residue was diluted with
water (50 mL) and extracted with EA (3 ꢁ 20 mL). The combined
NMR (300 MHz, DMSO‑d6)
7.74e7.63 (m, 1H), 7.31e7.22 (m, 1H), 3.90 (s, 3H), 3.57 (s, 3H), 2.58
(s, 3H). 13C NMR (126 MHz, DMSO‑d6)
160.92, 160.66, 157.87,
d 8.72 (s, 1H), 8.55 (s, 1H), 8.12 (s, 1H),
d
148.45, 143.21, 140.40, 137.14, 134.04, 129.03, 128.56, 127.32, 122.46,
116.05, 104.13, 56.61, 34.13, 21.42. MS (ESI): m/z 336.1 [M ꢂ Cl]þ.
ESI-HRMS: calcd. for C18H14ClN3O2Sþ [MþH]þ 372.0495, found:
372.0571.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
Data availability statement
All data associated with the study are provided with the
manuscript and its supplementary information.
12

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
mobilization, J. Biol. Chem. 282 (2007) 5478e5487, https://doi.org/10.1074/
jbc.M609881200.
Author contributions
[20] S.L. French, J.F. Arthur, H. Lee, W.S. Nesbitt, R.K. Andrews, E.E. Gardiner,
J.R. Hamilton, Inhibition of protease-activated receptor 4 impairs platelet
procoagulant activity during thrombus formation in human blood, J. Thromb.
Haemostasis 14 (2016) 1642e1654, https://doi.org/10.1111/jth.13293.
[21] K. Hosokawa, T. Ohnishi, N. Miura, H. Sameshima, T. Koide, K.A. Tanaka,
I. Maruyama, Antithrombotic effects of PAR1 and PAR4 antagonists evaluated
under flow and static conditions, Thromb. Res. 133 (2014) 66e72, https://
doi.org/10.1016/j.thromres.2013.10.037.
All authors contributed to the study conception and design. XZ is
the supervisor for the project. The synthesis was done by SDL. The
assays were performed by YK, SSL and RJX. The article was written
by SDL and DY. All authors read and approved the final manuscript.
Declaration of competing interest
[22] R.M. Poole, S. Elkinson, Vorapaxar: first global approval, Drugs 74 (2014)
1153e1163, https://doi.org/10.1007/s40265-014-0252-2.
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
[23] P. Vranckx, H.D. White, Z. Huang, K.W. Mahaffey, P.W. Armstrong, F.V. De
Werf, D.J. Moliterno, L. Wallentin, C. Held, P.E. Aylward, J.H. Cornel, C. Bode,
K. Huber, J.C. Nicolau, W. Ruzyllo, R.A. Harrington, P. Tricoci, Validation of
BARC bleeding criteria in patients with acute coronary syndromes: the
TRACER Trial, J. Am. Coll. Cardiol. 67 (2016) 2135e2144, https://doi.org/
10.1016/j.jacc.2016.02.056.
[24] ClinicalTrials gov, Vorapaxar in the Human Endotoxemia Model, 2016. https://
clinicaltrials.gov/ct2/show/NCT02875028. (Accessed 13 March 2021).
accessed.
[25] ClinicalTrials gov, CV004-007 Thrombosis Chamber Study, 2015. https://
clinicaltrials.gov/ct2/show/NCT02439190. (Accessed 13 March 2021).
accessed.
[26] ClinicalTrials gov, Multiple Ascending Dose Study to Evaluate the Safety,
Tolerability and Pharmacokinetics of BMS-986120 in Healthy Subjects and the
Effects of Co-adiministration of Midazolam and BMS-986120, 2015. https://
clinicaltrials.gov/ct2/show/NCT02208882. (Accessed 13 March 2021).
accessed.
[27] P.C. Wong, C.A. Watson, J. Bostwick, J. Banville, R.R. Wexler, E.S. Priestley,
A. Marinier, M. Bouvier, D. Gordon, W. Schumacher, J. Yang, An orally-active
small-molecule antagonist of the platelet protease-activated receptor-4,
BMS-986141, prevents arterial thrombosis with low bleeding liability in
Cynomolgus monkeys, Circulation 136 (2017).
[28] S.J. Wilson, F.A. Ismat, Z. Wang, M. Cerra, H. Narayan, J. Raftis, T.J. Gray,
S. Connell, S. Garonzik, X. Ma, J. Yang, D.E. Newby, PAR4 (Protease-Activated
Receptor 4) antagonism with BMS-986120 inhibits human ex vivo thrombus
formation, Arterioscler. Thromb. Vasc. Biol. 38 (2017) 448e456, https://
doi.org/10.1161/ATVBAHA.117.310104.
[29] P. Chen, S. Ren, H. Song, C. Chen, F. Chen, Q. Xu, Y. Kong, H. Sun, Synthesis and
biological evaluation of BMS-986120 and its deuterated derivatives as PAR4
antagonists, Bioorg. Med. Chem. 27 (2019) 116e124, https://doi.org/10.1016/
j.bmc.2018.11.024.
[30] F.L. Yang, L. Zhang, J.H. Han, F. He, W.K. Tao, Benzoheteroaryl Derivative and
Preparation Method and Medical Application Thereof, WO, 2019, 2019/
149205 A1.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2021.113764.
References
[1] G. Davi, C. Patrono, Platelet activation and atherothrombosis, N. Engl. J. Med.
357 (2007) 2482e2494, https://doi.org/10.1056/NEJMra071014.
[2] GBD 2016 Causes of Death Collaborators, Global, regional, and national age-
sex specific mortality for 264 causes of death, A systematic analysis for the
global burden of disease study 2016, Lancet 390 (2017) 1151e1210, https://
doi.org/10.1016/S0140-6736(17)32152-9, 1980e2016.
[3] D. Zhao, J. Liu, M. Wang, X. Zhang, M. Zhou, Epidemiology of cardiovascular
disease in China: current features and implications, Nat. Rev. Cardiol. 16
(2019) 203e212, https://doi.org/10.1038/s41569-018-0119-4.
[4] A.D. Michelson, Antiplatelet therapies for the treatment of cardiovascular
disease, Nat. Rev. Drug Discov. 9 (2010) 154e169, https://doi.org/10.1038/
nrd2957.
[5] A.O. Maree, D.J. Fitzgerald, Aspirin and coronary artery disease, J. Thromb.
Haemostasis 6 (2004) 1175e1181, https://doi.org/10.1160/TH04-02-0127.
[6] P.A. Gurbel, C.M. O'Connor, C.C. Cummings, V.L. Serebruany, Clopidogrel: the
future choice for preventing platelet activation during coronary stenting?
Pharm. Res. (N. Y.) 65 (1999) 109e123, https://doi.org/10.1006/
phrs.1999.0478.
[7] K. Bledzka, S.S. Smyth, E.F. Plow, Integrin aIIbb3: from discovery to efficacious
therapeutic target, Circ. Res. 112 (2013) 1189e1200, https://doi.org/10.1161/
circresaha.112.300570.
[8] F. Franchi, D.J. Angiolillo, Novel antiplatelet agents in acute coronary syn-
drome, Nat. Rev. Cardiol. 12 (2015) 30e47, https://doi.org/10.1038/
nrcardio.2014.156.
[9] M.L. Kahn, Y.W. Zheng, W. Huang, V. Bigornia, D. Zeng, S. Moff, R.V. Farese Jr.,
C. Tam, S.R. Coughlin, A dual thrombin receptor system for platelet activation,
Nature 394 (1998) 690e694, https://doi.org/10.1038/29325.
[10] M.D. Rand, J.B. Lock, C.V. Veer, D.P. Gaffney, K.G. Mann, Blood clotting in
minimally altered whole blood, Blood 88 (1996) 3432e3445, https://doi.org/
10.1182/blood.V88.9.3432.bloodjournal8893432.
[11] S.R. Coughlin, Protease-activated receptors in hemostasis, thrombosis and
vascular biology, J. Thromb. Haemostasis 3 (2005) 1800e1814, https://doi.org/
10.1111/j.1538-7836.2005.01377.x.
[12] Y. Bulani, S.S. Sharma, Therapeutic potential of targeting protease activated
receptors in cardiovascular diseases, Curr. Pharmaceut. Des. 21 (2015)
4392e4399, https://doi.org/10.2174/138161282130151007144725.
[13] P.J. O’brien, M. Molino, M. Kahn, L.F. Brass, Protease activated receptors:
theme and variations, Oncogene 20 (2001) 1570e1581, https://doi.org/
10.1038/sj.onc.1204194.
[31] J.J. Wu, S. Li, X.M. Wen, H. Yang, G.P. Wei, Y.J. Hu, S. Qian, Compound as
Protease Activated Receptor 4 (PAR4) Inhibitor for Treating Platelet Aggre-
gation, WO, 2019, 2019/218955 A1.
[32] J.J. Wu, S. Li, X.M. Wen, H. Yang, G.P. Wei, Y.J. Hu, S. Qian, Compound Acting as
Protease-Activated Receptor 4 (PAR4) Inhibitor for Treating Platelet Aggre-
gation, WO, 2019, 2019/218956 A1.
[33] X. Ye, M. Chen, Y. Chen, X. Su, Y. Wang, W. Su, Y. Kong, Isolation and char-
acterization of a novel antithrombotic peptide from enzymatic hydrolysate of
Agkistrodon acutus venom, Int. J. Pept. Res. Therapeut. 21 (2015) 343e351,
https://doi.org/10.1007/s10989-015-9463-y.
[34] X. Su, W. Su, Z. He, X. Ming, Y. Kong, Tripeptide SQL inhibits platelet aggre-
gation and thrombus formation by affecting PI3K/Akt signaling, J. Cardiovasc.
Pharmacol.
66
(2015)
254e260,
https://doi.org/10.1097/
FJC.0000000000000269.
[35] M.M. Miller, J. Banville, T.J. Friends, M. Gagnon, J.J. Hangeland, J.F. Lavallee,
ꢀ
ꢀ
A. Martel, H. O'Grady, R. Remillard, E. Ruediger, F. Tremblay, S.L. Posy,
N.J. Allegretto, V.R. Guarino, D.G. Harden, T.W. Harper, K. Hartl, J. Josephs,
S. Malmstrom, C. Watson, Y. Yang, G. Zhang, P. Wong, J. Yang, M. Bouvier,
D.A. Seiffert, R.R. Wexler, R.M. Lawrence, E.S. Priestley, A. Marinier, Discovery
of potent protease-activated receptor
4 antagonists with in vivo antith-
[14] L. Covic, A.L. Gresser, A. Kuliopulos, Biphasic kinetics of activation and
signaling for PAR1 and PAR4 thrombin receptors in platelets, Biochemistry 39
(2000) 5458e5467, https://doi.org/10.1021/bi9927078.
rombotic efficacy, J. Med. Chem. 62 (2019) 7400e7416, https://doi.org/
10.1021/acs.jmedchem.9b00186.
[36] E.H. Ruediger, D.H. Deon, M. Gagnon, S.L. Posy, Benzothiazole and Benzo-
thiophene Compounds, WO, 2016, 2016/138199 A1.
[15] L. Covic, C. Singh, H. Smith, A. Kuliopulos, Role of the PAR4 thrombin receptor
in stabilizing platelet-platelet aggregates as revealed by
a patient with
[37] National Center for Biotechnology Information, PubChem compound sum-
mary for CID 72188743. https://pubchem.ncbi.nlm.nih.gov/compound/
72188743, 2021. (Accessed 13 March 2021) accessed.
[38] X.J. Zhang, E.S. Priestley, J.A. Bates, O.S. Halpern, S.K. Reznik, J.M. Richter,
Bicyclic Heteroaryl Substituted Compounds, US, 2019, 2019/0292176 A1.
[39] X.J. Zhang, E.S. Priestley, O.S. Halpern, W. Jiang, S.K. Reznik, J.M. Richter, Tri-
cylic Heteroaryl-Substituted Quinoline and Azaquinoline Compounds as PAR4
Inhibitors, US, 2019, 2019/0315774 A1.
Hermansky-Pudlak syndrome, J. Thromb. Haemostasis 87 (2002) 722e727,
https://doi.org/10.1055/s-0037-1613071.
[16] T. Sidhu, S.L. French, J.R. Hamilton, Differential signaling by protease-activated
receptors: implications for therapeutic targeting, Int. J. Mol. Sci. 15 (2014)
6169e6183, https://doi.org/10.3390/ijms15046169.
[17] S.L. Jacques, A. Kuliopulos, Protease-activated receptor-4 uses dual prolines
and an anionic retention motif for thrombin recognition and cleavage, Bio-
chem. J. 376 (2003) 733e740, https://doi.org/10.1042/BJ20030954.
[18] L. Covic, A.L. Gresser, J. Talavera, S. Swift, A. Kuliopulos, Activation and inhi-
bition of G protein-coupled receptors by cell-penetrating membrane-tethered
peptides, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 643e648, https://doi.org/
10.1073/pnas.022460899.
[19] A. Mazharian, S. Roger, E. Berrou, F. Adam, A. Kauskot, P. Nurden, M. Jandrot-
perrus, M. Bryckaert, Protease-activating receptor-4 induces full platelet
spreading on a fibrinogen matrix: involvement of ERK2 and p38 and Ca2þ
[40] X.J. Zhang, X.F. Zhang, L.S. Chupak, Aryl Substituted Bicyclic Heteroaryl
Compounds, WO, 2017, 2017/019828 A1.
[41] Q.H. Fu, X.J. Zhang, E.S. Priestley, O.S. Halpern, S.K. Reznik, J.M. Richter, Bicyclic
Heteroaryl Substituted Compounds, WO, 2018, 2018/013770 A1.
[42] P.C. Wong, D.A. Seiffert, J.E. Bird, C.A. Watson, J.S. Bostwick, M. Giancarli,
N.J. Allegretto, J. Hua, D.G. Harden, J. Guay, Blockade of protease-activated
receptor-4 (PAR4) provides robust antithrombotic activity with low
13

S. Liu, D. Yuan, S. Li et al.
European Journal of Medicinal Chemistry 225 (2021) 113764
bleeding,
scitranslmed.aaf5294.
Sci.
Transl.
Med.
9
(2017),
https://doi.org/10.1126/
Synthesis of 1-benzyl-3-(5'-hydroxymethyl-2'-furyl)indazole analogues as
novel antiplatelet agents, J. Med. Chem. 44 (2001) 3746e3749, https://doi.org/
10.1021/jm010001h.
[43] A. Kawabata, M. Saifeddine, B. Al-Ani, L. Leblond, M.D. Hollenberg, Evaluation
of proteinase-activated receptor-1 (PAR1) agonists and antagonists using a
cultured cell receptor desensitization assay: activation of PAR2 by PAR1-
targeted legands, J. Pharmacol. Exp. Therapeut. 288 (1999) 358e370. PMID:
9862790.
[45] C. Wu, T. Hwang, C. Liao, S. Kuo, F. Lee, C. Lee, C. Teng, Selective inhibition of
protease-activated receptor 4-dependent platelet activation by YD-3,
J. Thromb. Haemostasis 87 (2002) 1026e1033, https://doi.org/10.1055/s-
0037-1613128.
[44] F. Lee, J. Lien, L. Huang, T. Huang, S. Tsai, C. Teng, C. Wu, F. Cheng, S. Kuo,
14