Discovery of a potent and selective αvβ3 integrin antagonist with strong inhibitory activity against neointima formation in rat balloon injury model

Article abstract of DOI:10.1016/j.bmcl.2004.02.075

A new series of phenylpiperazine-based derivatives with strong antagonistic activity for α_vβ₃ integrin were synthesized. Of these derivatives, the fluorine-substituted compound 8 showed strong inhibitory activity and high selectivity for α_vβ ₃ integrin receptor (IC₅₀=0.055nM). In vivo evaluation of the antistenotic effects of 8 indicated that this compound significantly inhibits neointima formation in rat balloon injury model.

Full text of DOI:10.1016/j.bmcl.2004.02.075

Bioorganic & Medicinal Chemistry Letters 14 (2004) 2567–2570
Discovery of a potent and selective avb3 integrin antagonist with
strong inhibitory activity against neointima formation in rat balloon
injury model
Seiji Iwama,a,* Tomoko Kitano,b Fumiyo Fukuya,b Yayoi Honda,b Yuji Sato,b
Mitsue Notakeb and Toshiya Moriea
aChemistry Research Laboratories, Dainippon Pharmaceutical Co., Ltd, Enoki 33-94, Suita, Osaka 564-0053, Japan
bPharmacology & Microbiology Research Laboratories, Dainippon Pharmaceutical Co., Ltd, Enoki 33-94, Suita,
Osaka 564-0053, Japan
Received 22 January 2004; revised 23 February 2004; accepted 23 February 2004
Abstract—A new series of phenylpiperazine-based derivatives with strong antagonistic activity for avb3 integrin was synthesized. Of
these derivatives, the fluorine-substituted compound 8 showed strong inhibitory activity and high selectivity for avb3 integrin
receptor (IC50 ¼ 0.055 nM). In vivo evaluation of the antistenotic effects of 8 indicated that this compound significantly inhibits
neointima formation in rat balloon injury model.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
clasts and various tumour cells, and is known to mediate
a variety of biological events, including migration and
proliferation of SMCs during restenosis,5 adhesion of
osteoclasts to bone matrix and angiogenesis.6 The inte-
grin avb3 binds to extracellular matrix proteins, such as
vitronectin, fibronectin, osteopontin, fibrinogen and von
Willebrand factor through interaction with the tripep-
tide Arg-Gly-Asp (RGD) sequence.
Long-term success of percutaneous coronary interven-
tions (PCIs) remains limited due to restenosis, which
occurs within 6 months in 30–50% of patients who
undergo balloon angioplasty and in 20–40% of patients
who receive bare stent.1 Therefore, reduction in the
incidence of restenosis after angioplasty or stenting has
been a research theme and clinical aim for a number of
years. To date, numerous pharmacological agents have
been evaluated in an attempt to reduce the rate of
restenosis in patients who underwent PCIs. However,
only a few compounds have been shown to influence
restenosis after balloon angioplasty or stenting.2 Path-
ologically, experimental evidence has suggested that
restenosis is characterized by vascular smooth muscle
cells (SMCs) migration, proliferation and extracellular
matrix production in disease states.3
A vast number of cyclic peptide and nonpeptide antag-
onists with high affinity for avb3 integrin have been
reported.7 Nevertheless, only a few avb3 integrin
antagonists have been shown to inhibit neointima for-
mation in the arterial injury model.5 Our interest in this
field was prompted by the finding that avb3 antagonists
inhibit migration of SMC into neointima and/or pro-
liferation of SMC in neointima after arterial injury.
In early studies, Kessler and co-workers disclosed cyclic
RGD peptides as selective avb3 integrin antagonists, and
proposed an avb3 ligand-binding model. Their findings
suggest that conformation of the RGD-sequence with
The avb3 integrin, also known as the vitronectin recep-
tor, is expressed on a variety of cell types,4 including
endothelial cells, SMCs, macrophage, platelets, osteo-
ArgCa=AspCa-distance
of
about
500 pm
and
ArgCb=AspCb-distance of about 700 pm fits better in the
avb3 receptor than in the aIIbb3 receptor.8 In our search
for a new scaffold based on KesslerÕs model, we found
Keywords: Integrin rat balloon model.
* Corresponding author. Tel.: +81-663375901; fax: +81-663387656;
e-mail: seiji-iwama@dainippon-pharm.co.jp
3-aminophenyl-piperazine moiety as
a constrained
scaffold for the nonpeptide avb3 integrin antagonists.
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.02.075

2568
S. Iwama et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2567–2570
Herein, we disclose our initial effort for the synthesis and
structure–activity relationship of a new series of phenyl-
piperazine based avb3 integrin antagonists, which have
strong inhibitory activity of neointima formation in rat
balloon injury model.
Likewise, the amide and urea analogues (13–15) showed
low inhibitory activity for avb3 integrin, although their
selectivity for aIIbb3 integrin was modest. However, the
benzylcarbamate analogue 2 showed high inhibitory
activity for avb3 integrin (IC50 ¼ 0.058 nM) and high
selectivity for aIIbb3 integrin. As for the effect of the
guanidine moiety, the ring size [five-membered (1), six-
membered (2) and seven-membered (5)] had no signifi-
cant influence on both the inhibitory activity for avb3
integrin and inhibition of HVSMC proliferation. How-
ever, introduction of a fluorine atom on the six-mem-
bered cyclic guanidine moiety of 2 (yielding 3) resulted
in good improvement in inhibition of HVSMC prolif-
eration. Likewise, the inhibitory activity of 8 for both
avb3 integrin and HVSMC proliferation increased
compared with that of 7, 9 or 10.
2. Chemistry
Synthesis of the avb3 integrin antagonists 1–15 was
accomplished by attaching the groups bearing the acidic
and basic moieties in 1–15 to 3-aminophenyl-piperazine.
As shown in Scheme 1, reaction of 3-fluoronitrobenzene
with piperazine in DMSO at 100 °C,9 followed by cou-
pling with Na-substituted glutamic acid tert-butyl ester
(17) using EDCI and HOBt in DMF provided the
coupled product 18. Reduction of the nitro group in 18
with Fe and NH4Cl in EtOH and H2O gave the aniline
19. Introduction of a cyclic guanidine moiety with
HgCl2 and diBoc thiourea derivatives (20a–f) using
KimÕs protocol,10 followed by deprotection of both
diBoc groups and the tert-butyl ester group provided the
desired compounds 1–15.
Among the compounds synthesized in this study, the
fluorine-substituted analogues 3 and 8 having both high
inhibitory activity for avb3 integrin (3: IC50 ¼ 0.063 nM;
8: IC50 ¼ 0.055 nM) and potent inhibition of HVSMC
proliferation (3: IC50 ¼ 0.26 lM, 8: IC50 ¼ 0.18 lM) were
selected as candidates for inhibition of neointima for-
mation in rat balloon injury model.
3.2. Rat balloon injury model
3. Results and discussion
3.1. Binding study
To evaluate the antistenotic effects of 3 and 8, their
ability to inhibit neointima formation in the carotid
artery of rat balloon injury model was examined.14
For in vitro screening, we examined the inhibitory
activity of 1–15 for avb3 and aIIbb3 integrins binding
using human vitronectin and fibrinogen as ligands,
respectively,11;12 and for human vascular smooth muscle
cells (HVSMC) proliferation.13 As shown in Table 1,
among the nonsubstituted six-membered guanidine
analogues (2, 6, 7, 11–15), the sulfonamide analogues
(11, 12) showed low selectivity for aIIbb3 integrin.
For a period of 15 days starting 1 day before injury,
compound 3 or 8 was infused to rats subcutaneously at a
constant rate of 1.3 lmol/rat/day using an Alzete os-
motic pump. Measurements of plasma levels of 8 indi-
cated a steady state of plasma concentration exceeding
the in vitro IC50 value for inhibition of HVSMC pro-
liferation (8: ca. 0.35 lM, IC50 ¼ 0.18 lM).
R
HN
NO2
(17)
HOOC
COOBut
NO2
NO2
EDCI, HOBt, iso-Pr2NEt /DMF
1) piperazine / DMSO
2) 4.0 M HCl / AcOEt
N
NH HCl
NH2
F
16
Fe, NH4Cl
EtOH, H2O
R
R
HN
N
HN
N
OBut
OBut
N
N
O
O
O
O
19
18
S
S
S
BocN
NBoc
NBoc
BocN
BocN NBoc
1) 20a- f, HgCl2
Et3N / DMF
R3 R4
1- 15
20f
2) TFA / CH2Cl2
20b: R3= R4= H
20a
20c: R3= F, R4= H
20d: R3= OBoc,R4= H
20e: R3= R4= CH3
Scheme 1. Synthesis of avb3 integrin antagonists 1–15.

S. Iwama et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2567–2570
Table 1. Inhibitory activity of phenyl-piperazine antagonists 1–15
2569
R1
R2
N
N
N
N
H
HN
HN
N
H
HN
N
H
O
O
O
R
N
N
N
HN
O
HN
HN
O
OH
N
N
N
OH
OH
O
O
O
O
O
5
1
2- 4, 6-15
Compound
R
R1
R2
avb3 IC50 (nM)
aIIbb3 IC50 (nM) avb3/aIIbb3 ratio HVSMCa proliferation
assay IC50 (lM)
1
––
––
H
F
––
H
0.16
0.058
0.063
0.15
0.056
0.59
0.32
0.055
0.30
0.38
1.4
720
660
4500
11,000
8600
3.0
1.1
2
PhCH2OCO–
PhCH2OCO–
PhCH2OCO–
––
3
H
540
340
0.26
3.5
4
OH
––
H
H
2300
5
––
H
>1000
>1000
>1000
>1000
>1000
>1000
88
>18,000
>1700
>3100
>18,000
>33,000
>2600
63
1.7
2.1
6
EtOCO–
7
iso-BuOCO–
iso-BuOCO–
iso-BuOCO–
iso-BuOCO–
PhSO2–
H
H
0.91
0.18
2.0
8
F
H
9
OH
CH3
H
H
10
11
12
13
14
15
CH3
H
3.4
>10
Mes–b
H
H
0.22
6.8
30
140
>150
>10
>10
>10
>10
PhCH2CO–
EtNHCO–
PhCH2NHCO–
H
H
>1000
>1000
>1000
H
H
3.6
1.4
>280
>710
H
H
a HVSMC: human vascular smooth muscle cell.
b 2,4,6-Trimethylbenzenesulfonyl.
Table 2. Results of morphometrical examination of carotid cross sections
Compound
Number of animals
Intima areaa
(103 lm2)
%Inhibition
from control
Media areaa
(103 lm2)
I=M ratioa;b %Inhibition
from control
Experiment no. 1
Control
7
6
181 ꢀ 16
156 ꢀ 4
149 ꢀ 6
1.16 ꢀ 0.09
3
113 ꢀ 13ꢁꢁ
38
0.76 ꢀ 0.09ꢁꢁ 34
Experiment no. 2
Control
7
7
178 ꢀ 17
81 ꢀ 9ꢁꢁꢁ
171 ꢀ 8
162 ꢀ 7
1.06 ꢀ 0.10
8
54
0.50 ꢀ 0.05ꢁꢁꢁ 53
ꢁꢁp < 0:01 and ꢁꢁꢁp < 0:001 compared with control.
a Values represent the mean ꢀSE.
b I=M ratio, intima area to media area ratio.
Rats carotid arteries were isolated 14 days after injury,
cross-sectioned and morphometrically analyzed. As
shown in Table 2, compounds 3 and 8 inhibited neoin-
tima formation by 38% (p < 0:01) and 54% (p < 0:001),
respectively, compared to each control group, whereas
no difference in media area was observed between
compounds treated and untreated control rats. The
ratios of intima area to media area (I=M) in the case of 8
were higher than those in the case of 3 (3: 34%, 8: 53%).
From these results, compound 8 was selected as a better
inhibitor of neointima formation in rat balloon injury
model. Representative cross sections of the arteries are
shown in Figure 1. I/M ratios in the sections were close
to the mean values in each group (control rat: 1.16,
8-treated rat: 0.51). Following injury, the intima area in
the control rat (Fig. 1b,e) remarkably increased com-
pared with that in the uninjured normal rat (Fig. 1a,d).
However, treatment with 8 inhibited neointima forma-
tion following injury (Fig. 1c,f). In addition, compound
8 exhibited a dose-dependent inhibition; I=M ratios of
21% and 38% at doses of 0.14 and 0.43 lmol/rat/day,
respectively.
4. Conclusion
In summary, we synthesized highly selective avb3 inte-
grin antagonists based on a phenyl-piperazine RGD
peptidemimetic. Among the synthesized compounds, the
fluorine-substituted 8 showed the strongest inhibitory
activity with high selectivity for avb3 integrin. In vivo
evaluation of the antistenotic effects of 8 indicated that

2570
S. Iwama et al. / Bioorg. Med. Chem. Lett. 14 (2004) 2567–2570
Figure 1. Typical cross sections of rats carotid arteries 14 days after balloon injury. (a) Normal, the right uninjured carotid artery ( ꢂ 44). (b) Control,
the left injured carotid artery in untreated rat ( ꢂ 44). (c) Treated rat, the left injured carotid artery in 8-treated rat ( ꢂ 44). (d) High magnification
( ꢂ 147) of panel (a). (e) High magnification ( ꢂ 147) of panel (b). (f) High magnification ( ꢂ 147) of panel (c).
Hutchinson, J. H. Exp. Opin. Ther. Patents 2000, 10, 1367;
€
this compound significantly inhibits neointima forma-
tion in rat balloon injury model. It is therefore suggested
that compound 8 may offer a new entry for the treat-
ment of restenosis after PCIs.
(c) Holzemann, G. IDrugs 2001, 4, 72; (d) Coleman, P. J.;
Duong, L. T. Exp. Opin. Ther. Patents 2002, 12, 1009.
8. (a) Haubner, R.; Gratias, R.; Diefenbach, B.; Goodman,
S. L.; Jonczyk, A.; Kessler, H. J. Am. Chem. Soc. 1996,
118, 7461; (b) Haubner, R.; Finsinger, D.; Kessler, H.
Angew. Chem., Int. Ed. 1997, 36, 1374.
9. Belfield, A. J.; Brown, G. R.; Foubister, A. J.; Ratcliffe,
P. D. Tetrahedron 1999, 55, 13285.
Acknowledgements
10. Kim, K. S.; Qian, L. Tetrahedron Lett. 1993, 34, 7677.
11. Inhibitory activity on human avb3 integrin binding was
determined with the use of biotinylated human vitronectin
as the ligand according to the methods described in
Mousa, S. A.; Forsythe, M.; Lorelli, W.; Bozarth, J.; Xue,
C.-B.; Wityak, J.; Sielecki, T. M.; Olson, R. E.; DeGrado,
W.; Kapil, R.; Hussain, M.; Wexler, R.; Thoolen, M. J.;
Reilly, T. M. Coron. Artery Dis. 1996, 7, 767.
12. Inhibitory activity on human aIIbb3 integrin binding was
determined with the use of biotinylated human fibrinogen
as the ligand. The binding assay was carried out essentially
in a similar manner as that for avb3 integrin binding,
although the modified antigen capture ELISA procedure
was used instead of immobilization of a purified aIIbb3
integrin protein; see: (a) Kosugi, S.; Tomiyama, Y.;
Shiraga, M.; Kashiwagi, H.; Mizutani, H.; Kanakura,
Y.; Kurata, Y.; Matsuzawa, Y. Thromb. Haemost. 1996,
75, 339; (b) Kashiwagi, H.; Tomiyama, Y.; Tadokoro, S.;
Honda, S.; Shiraga, M.; Mizutani, H.; Handa, M.;
Kurata, Y.; Matsuzawa, Y.; Shattil, S. J. Blood 1999, 93,
2559.
We thank Mr. Kenji Kojima and Mr. Naohiro
Nishimura for their work on the determinations of
compoundÕs concentration in plasma.
References and notes
1. (a) The STRESS Investigators. New Engl. J. Med. 1994,
331, 496; (b) The BENESTENT Study Group. New Engl.
J. Med. 1994, 331, 489; (c) Babapulle, M. N.; Eisenberg,
M. J. Circulation 2002, 106, 2734.
2. (a) Johnson, G. J.; Griggs, T. R.; Badimon, L. Thromb.
Haemost. 1999, 81, 835; (b) Gruberg, L.; Waksman, R.;
Satler, L. F.; Pichard, A. D.; Kent, K. M. Exp. Opin.
Invest. Drugs 2000, 9, 2555.
3. Ross, R. Nature 1993, 362, 801.
4. Byzova, T. V.; Rabbani, R.; DÕSouza, S. E.; Plow, E. F.
Thromb. Haemost. 1998, 80, 726.
5. Sajid, M.; Stouffer, G. A. Thromb. Haemost. 2002, 87, 187.
6. Brooks, P. C.; Clark, R. A. F.; Cheresh, D. A. Science
1994, 264, 569.
7. For recent reviews on avb3 integrin antagonist, see: (a)
Miller, W. H.; Keenan, R. M.; Willette, R. N.; Lark, M.
W. Drug Discovery Today 2000, 5, 397; (b) Duggan, M. E.;
13. Inhibitory activity on human VSMC proliferation was
determined using the WST-1 cell proliferation assay; see:
Ishiyama, M.; Shiga, M.; Sasamoto, K.; Mizoguchi, M.;
He, P.-G. Chem. Pharm. Bull. 1993, 41, 1118.
14. Clowes, A. W.; Reidy, M. A.; Clowes, M. M. Lab. Invest.
1983, 49, 327.