2-acylimino-3H-thiazoline derivatives: A novel template for platelet GPIIb/IIIa receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(01)00123-8

In the course of our research for the low-molecular weight RGD peptide mimics, we have found that a rigid 2-acylimino-3H-thiazoline structure is suitable for the peptide backbone mimics. Introduction of amidinophenyl and β-alanine moiety as arginine and aspartic acid side-chain surrogates to this backbone mimic resulted in a highly potent fibrinogen receptor antagonist 2-(4amidinobenzoylimino)-3,4-dimethyl-N-(2-carboxyethyl) -3H-thiazoline-5-carboxamide (7c), namely PS-028 (K_i = 46.5± 5.8 pM).

Full text of DOI:10.1016/S0960-894X(01)00123-8

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1031–1035
2-Acylimino-3H-thiazoline Derivatives: A Novel Template for
Platelet GPIIb/IIIa Receptor Antagonists
Akira Manaka, Masakazu Sato,* Mari Aoki, Makoto Tanaka, Tomotake Ikeda,
Yoshihisa Toda, Yuuko Yamane and Shiro Nakaike
Taisho Pharmaceutical Co. Ltd., Medicinal Research Laboratories, 1-403 Yoshinocho, Omiya-shi, Saitama, 330-8530, Japan
Received 22 December 2000; accepted 17 February 2001
Abstract—In the course of our research for the low-molecular weight RGD peptide mimics, we have found that a rigid 2-acylimino-
3H-thiazoline structure is suitable for the peptide backbone mimics. Introduction of amidinophenyl and b-alanine moiety as argi-
nine and aspartic acid side-chain surrogates to this backbone mimic resulted in a highly potent fibrinogen receptor antagonist 2-(4-
amidinobenzoylimino)-3,4-dimethyl-N-(2-carboxyethyl)-3H-thiazoline-5-carboxamide (7c), namely PS-028 (Ki=46.5ꢀ5.8 pM). # 2001
Elsevier Science Ltd. All rights reserved.
The aggregation of platelets is known to play a critical
role in hemostatis and arterial thrombosis.1 Thus,
uncontrolled platelet aggregation and platelet adhesion
to the subendothelium of damaged blood vessels may
cause life-threatening diseases such as myocardial
infarction, transient ischemic attack, and unstable
angina pectoris.2 The aggregation of platelet is invoked
by various ways, such as stimulation with chemical
inducers (ADP, collagen, TXA2, etc.),3 or sheer stress.4
However, the crosslinking of the platelets by binding of
dimeric fibrinogen to its receptor GPIIb/IIIa on the
activated platelet membrane is the common pathway.5
Therefore, inhibition of the binding of fibrinogen to
GPIIb/IIIa has been regarded as an attractive strategy
for a potent novel class of platelet aggregation inhibi-
tors. In fact, a Fab fragment of GPIIb/IIIa monoclonal
antibody Reo-Pro6 can suppress the platelet aggregation
invoked by various chemical inducers, and has now
been launched as a therapeutic for the treatment of
patients undergoing percutaneous transluminal cor-
onary angioplasty (PTCA).
almost all of them are not systemically available when
administered orally, and evaluated as parenteral agents
for acute clinical settings. Therefore, development of
orally active low-molecular weight GPIIb/IIIa antago-
nists are now desired for the treatment of chronic
thrombotic disorders.
We now report the synthesis and pharmacological eva-
luations of a novel potent low molecular weight RGD
mimic named PS-028. In the course of the investigation,
we have utilized the benzamidino groupas an arginine
isostere9 and b-alanine as an aspartate surrogate, and
then evaluated the linkage structure of these two moi-
eties. To avoid the difficulty in commercial large-scale
synthesis, we have tended to evaluate achiral linkage
systems.
Chemistry
The synthesis of target thiazole derivatives was accom-
plished as shown in Scheme 1. 2-Amino (3a) and
methylamino (3b) thiazole derivatives were obtained
from the corresponding 2-amino-4-methylthiazole-5-
carboxylates in 3 steps. 3-Alkyl-4-methyl-2-imino-3H-
thiazoline derivatives 5c–f were synthesized from 4-cya-
nobenzoylamino intermediate (4) by selective alkylation
in basic conditions (Scheme 2). Thus, treatment of 4
with sodium hydride in DMF and alkyl halides at 0 ꢁC
yielded 3-alkyl-3H-thiazoline 5c–f as a single product.
In an NMR study regarding 5c,10 the chemical shift of
The minimum binding epitope of fibrinogen to its
receptor, GPIIb/IIIa, is known to be the tripeptide
7
sequence, Arg-Gly-Asp(RGD),
and, a number of
RGD-including cyclic peptides,8 peptide mimetics have
been evaluated as GPIIb/IIIa antagonists. However,
*Corresponding author. fax:+81-48-652-7254; e-mail: s11400@ccm.
taisho.co.jp
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00123-8

1032
A. Manaka et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1031–1035
Scheme 1.
Scheme 2.
the 3-methyl proton was d 3.83 (400 MHz, CDCl3), and
HMBC coupling was observed with the 4-position of
the thiazoline carbon atom (d 146.16), which was also
observed coupling with the 4-methyl proton (d 2.70).
This indicates that the methylation of 4 proceeded at the
3-position of the thiazole ring. The X-ray crystal-
lographic analysis11 of crystalline intermediate (5e)
revealed that the configuration of the acylimino moiety
was syn (Scheme 3). Moreover, this indicates that the 2-
benzoylimino-3-alkyl-3H-thiazoline template is an
extremely constrained, planar conjugated system. Prob-
ably due to this conjugated structure, the acylimino
moiety of these compounds is significantly stable for
hydrolysis conditions, namely, they are tolerable even in
refluxing 23% hydrobromic acid and 5% NaOH in
aqueous methanol for hours. N-Methyl-b-alanine deri-
vative 3g was synthesized by methylation of 3c in good
yield (Scheme 2). Bicyclic thiazoline derivative 3h was
synthesized from 4 in 4 steps (Scheme 4). Thus, the 4-
methyl groupof the thiazoline ring in 4 was brominated
by N-bromosuccinimide in the presence of a catalytic
amount of 2,20-azobisisobutyronitrile (AIBN), and sub-
sequent introduction of the b-alanine side chain gave 6
in good yield. Hydrolysis of 6 with 1% aqueous NaOH
and intramolecular cyclization using dehydrating
reagent afforded the desired bicyclic derivative 3h. The
nitrile moiety in compounds 3a–h was transformed into
an amidine using conventional procedures12 to afford
desired compounds 7a–h (Scheme 5).
Results and Discussion
Evaluation of fibrinogen receptor binding ability of the
compounds was performed using purified activated
GPIIb/IIIa protein (ELISA)13 and washed platelet (125I-
labelled fibrinogen).14 First, according to the report by
Alig et al., which indicated that the 1,3-disubstituted
phenyl ring was suitable for this purpose,15 we intro-
duced a 2,5-disubstituted thiazole ring for a linkage of
the terminal two moieties instead of phenyl ring. As
shown in Table 1, the IC50 value of compound 7a
against Fbg-GPIIb/IIIa binding assay (0.57 nM) was
almost equal to that of 1,3-disubstituted phenyl deriva-
tive 8 (0.55 nM), and about 2 times more potent in 125I-
labelled Fbg-WP binding assay (60.5 vs 105 nM). These
data indicate that thiazole ring system is suitable for the
linkage groupfor low-molecular weight GPIIb/IIIa
antagonist. Introduction of methyl groupon the nitro-
gen of the benzamide moiety reduced the activity by
about half (7b: IC50=1.0 nM). This might be because of
a loss of the ability for hydrogen bond donor to the
binding site on the receptor, however, it could be due to
the loss of rigidity of the linkage group. Thus, com-
pound 7c which possesses a methyl group at the nitro-
gen atom on the thiazoline ring and consequently lost
Scheme 3. ORTEP diagram of 5e.
Scheme 4.

A. Manaka et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1031–1035
1033
Scheme 5.
Scheme 6.
Table 1. Antiplatelet activity of compounds 7a–c, 8 and Echistatin
Compds
X
Y
Fbg-GPIIb/IIIa ELISAa IC50 (nM)
0.57
125I-Fg-WPb IC50 (nM)
60.5
7a
NH
7b
NMe
N
1.00
0.090ꢀ0.008
0.55
NTc
11.3
105
7c (PS-028)
815
NH
Echistatin
0.20
NTc
aBinding assay was performed with human fibrinogen (Fbg) versus purified activated GPIIb/IIIa16 using ELISA method.
bBinding assay was performed with 125I-labelled human Fbg versus washed platelet.
cNT=Not tested.
hydrogen bond donor at the benzamide moiety, still
maintains the activity, and possesses 6-fold more potent
(0.09 nM) activity than 7a. Introduction of alkyl groups
(7d,e) other than the methyl groupon the 3-position of
the thiazoline ring also increased activity, and even the
introduction of the rather bulky benzyl group( 7f) still
maintained high potency (0.09 nM, Table 2). The sub-
stituents on the 3-position of the thiazole ring seemed to
be widely tolerable, thus, it might act only to fix the exo-
imino resonance form (Scheme 6).
the activity (7g, 0.89 nM). Substitution of the secondary
amide groupby a flexible tertiary amide groupshould
result in the loss of the desired ‘exo’ amide conforma-
tion. In fact, the ‘endo’ amide derivative (7h), which had
confirmed the amide moiety by ring-closure, showed
dramatically reduced activity by about one-fifty magni-
tude (4.76 nM) as reported in the other templates.17 We
also investigated the relationshipbetween IC
value
(ELISA) and concentration of fibrinogen. As shown in
50
Figure 1, 7c (PS-028) appeared to be a competitive
inhibitor
of
fibrinogen-GPIIb/IIIa
binding
Moreover, the rigid 2-acylimino thiazoline template of
7c should fix two receptor-binding residues (an amidine
and carboxyl group) to the suitable position to give a
highly potent low-molecular weight GPIIb/IIIa antago-
nist. As in the case of a benzamide moiety at the 2-
position of the thiazole ring, introduction of a methyl
groupon the nitrogen atom to the amide moiety at the
5-position of the thiazoline template of 7c diminished
(Ki=46.5ꢀ5.8 pM).18 IC50 value of 7c for 125I-fibrino-
gen-washed human platelet binding was 11.3 nM which
was almost the theoretical optimal value due to a
large distribution of GPIIb/IIIa or the platelet mem-
brane (upto 10 ꢂ8 M in plasma),19 and in the same
concentration, 7c strongly inhibits the ADP-induced
platelet aggregation (IC50 for human PRP: 16.8 nM,
Table 3).

1034
A. Manaka et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1031–1035
Figure 1. Relationshipbetween IC 50 value (ELISA, 0.4 mg/mL of GPIIb/IIIa) and concentration of fibrinogen (0.31–50.4 mg/mL).
Table 2. Antiplatelet activity of compounds 7c–h Table 4. Selectivity for platelet fibrinogen receptor versus HUVEC
receptor
Compds IC50 for HUVEC adhesion IC50 (mM)a to plates coated with
Fibrinogen
Vitronectin
Laminin
Fibronectin
7c
echistatin
>100
>100
<0.001
>100
>100
0.007
0.01
0.1
aIC50 value for adhesion with immobilized HUVEC in the presence of
each extra cell matrix.
Compds
R
R0
Fbg-GPIIb/IIIa
ELISAa IC50 (nM)
125I-Fg-WPbb
IC50 (nM)
7c
7d
7e
7f
Me
Et
n-Bu
Bn
Me
H
H
H
H
Me
0.09
0.12
0.13
0.09
0.89
11.3
NTc
6.26
NTc
51.8
peptide backbone mimics, and demonstrated a potent
low molecular weight GPIIb/IIIa antagonist PS-028
(7c), which is suitable as a lead compound for orally
active potent platelet aggregation inhibitors. Moreover,
this unique template might be stable in chemical and
biological conditions, it is of significance to use as
the template for various kinds of pharmacological
compounds.
7g
7h
4.76
67.9
aꢂcSee Table 1.
Table 3. In vitro and ex vivo antiaggregation activities of compound
7c (PS-028)
Acknowledgements
In vitroa (IC50, nM)
Ex vivof (iv, ED50, mg/kg)
The authors would like to thank Keita Matsumoto for
X-ray crystallographic analysis.
Humanb Mousec Dogd Guinea pige
16.8
36.6
35.5
20.8
0.0063
aIC50 value for chemically induced aggregation of platelet-rich plasma
monitored by Aggrecoder.20
References and Notes
bADP (2 mM) was used as an inducer.
cADP (5 mM) was used as an inducer.
1. Philips, D. R.; Charo, I. F.; Parise, L. V.; Fitzgerald, L. A.
Blood 1988, 71, 831.
2. Fitzgerald, D. J.; Roy, L.; Catella, F.; Fitagerald, G. A. N.
Engl. J. Med. 1986, 315, 983.
3. Coller, B. S.; Anderson, K.; Weisman, H. F. Thromb.
Haemost. 1995, 74, 302.
dADP (10 mM) and epinephrine (10 mM) were used as inducers.
eADP (3 mM) was used as an inducer.
fED50 value for ex vivo ADP (3 mM)-induced platelet aggregation of
guinea-pig administered 7c intravenously 30 min prior to the prepara-
tion of platelet-rich plasma.
4. Leonard, E. F.; Turitto, V. T.; Vroman, L. Ann. N.Y. Acad.
Sci. 1987, 516, 1.
5. Kieffer, N.; Philips, D. R. Ann. Rev. Cell. Biol. 1990, 6, 329.
6. Drug Data Rep. 1995, 17, 446.
7. Ruoslhti, E.; Pierschbadher, M. D. Science 1987, 238, 491.
8. Tcheng, J. E.; Harrington, R. A.; Kottke-Marchant, K.;
Kleiman, N. S.; Ellis, S. G.; Kereiakes, D. J.; Mick, M. J.;
Navetta, F. I.; Smith, J. E.; Worley, S. J.; Miller, J. A.; Joseph,
D.; Sigmon, K. N.; Kitt, M. M.; Du Mee, C. P.; Califf, R. M.;
Topol, E. J. Circulation 1995, 91, 2151.
Compound 7c showed potent ex vivo antiaggregation
activity in guinea pig (Table 3). Compound 7c was
found to be chemically and biologically stable (data not
shown), and did not affect the RGD-mediated HUVEC
adhesion21 to extra cell matrix proteins (Table 4).
In conclusion, we have found a novel sterically con-
strained 2-imino-3H-thiazoline template for RGD

A. Manaka et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1031–1035
1035
9. Kaiser, B.; Markwardt, F. Thromb. Haemost. 1986, 55, 194.
10. 1H and 13C NMR spectra were obtained on a JEOL
GX400 spectrometer with TMS as an internal standard.
11. X-ray spectroscopy was obtained by Macscience MXC-18.
12. Wagner, G.; Voigt, B.; Vieweg, H. Pharmazie 1984, 39,
226.
16. Pytela, R.; Pierschbacher, M. D.; Ginsberg, M. H.; Plow,
E. F.; Ruoslahti, E. Science 1986, 231, 1559.
17. Askew, B. C.; McIntyre, C. J.; Hunt, D. A.; Claremon,
R. J.; Gould, R. J.; Lynch, R. J.; Armstrong, D. J. Bioorg.
Med. Chem. Lett. 1995, 5, 475.
18. Cheng, Y. C.; Prusoff, W. H. Biochem. Pharmacol. 1973,
22, 3099.
13. Kouns, W. C.; Kirchhofer, D.; Hadvary, P.; Edenhofer, T.
´
W.; Pfenninger, G.; Baumbartner, H. R.; Jennings, L. K.;
Steiner, B. Blood 1992, 80, 2539.
14. Kupinski, J. M.; Miller, J. L. Thromb. Res. 1986, 43, 335.
19. Cramer, E. M.; Savidge, G. F.; Vainchenker, W.; Berndt,
M. C.; Pidard, D.; Caen, J. P.; Masse, J. M.; Breton-Gorius, J.
Blood 1990, 75, 1220.
15. Alig, L.; Edenhofer, A.; Hadva
´
Knopp, D.; Muller, M.; Steiner, B.; Trzeciak, A.; Weller, T. J.
ry, P.; Hurzeler, M.;
20. Mebanix PAM-8C aggregometer was used.
21. Underwoods, P. A.; Bennett, F. A.; Kirkpatrick, A.; Bean,
P. A.; Moss, B. A. Biochem. J. 1995, 309, 765.
¨
¨
Med. Chem. 1992, 35, 4393.