Rational design of conformationally restricted quinazolinone inhibitors of poly(ADP-ribose)polymerase

Article abstract of DOI:10.1016/j.bmcl.2007.07.091

A successful design of conformationally restricted novel quinazolinone derivatives linked via a cyclopentene moiety as potent poly(ADP-ribose)polymerase-1 (PARP-1) inhibitors has been developed. One selected member of the new series, 8-chloro-2-[(3S)-3-(4

Full text of DOI:10.1016/j.bmcl.2007.07.091

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5577–5581
Rational design of conformationally restricted
quinazolinone inhibitors of poly(ADP-ribose)polymerase
Kouji Hattori,a,* Yoshiyuki Kido,a Hirofumi Yamamoto,a Junya Ishida,a
Akinori Iwashitab and Kayoko Miharab
aChemistry Research Laboratories, Astellas Pharma Inc., 21, Miyukigaoka, Tsukuba-shi, Ibaraki 305-8585, Japan
bPharmacological Research Laboratories, Astellas Pharma Inc., 21, Miyukigaoka, Tsukuba-shi, Ibaraki 305-8585, Japan
Received 29 March 2007; revised 27 July 2007; accepted 30 July 2007
Available online 22 August 2007
Abstract—A successful design of conformationally restricted novel quinazolinone derivatives linked via a cyclopentene moiety as
potent poly(ADP-ribose)polymerase-1 (PARP-1) inhibitors has been developed. One selected member of the new series, 8-
chloro-2-[(3S)-3-(4-phenylpiperidin-1-yl)cyclopent-1-en-1-yl]quinazolin-4(3H)-one (S-16d), was found to be highly potent with
IC50 = 8.7 nM and good brain penetration.
Ó 2007 Elsevier Ltd. All rights reserved.
Poly(ADP-ribose)polymerase-1 (PARP-1: EC 2.4.2.30)
is a chromatin-bound nuclear enzyme involved in a vari-
ety of physiological functions related to genomic repair,
including DNA replication and repair, cellular prolifer-
ation and differentiation, and apoptosis.1 PARP-1 func-
tions as a DNA damage sensor and signaling molecule.
Upon binding to DNA breaks, activated PARP cleaves
NAD+ into nicotinamide and ADP-ribose and polymer-
izes the latter onto nuclear acceptor proteins including
histones, transcription factors, and PARP itself. A cellu-
lar suicide mechanism of both necrosis and apoptosis by
PARP activation has been implicated in the pathogene-
sis of brain injury and neurodegenerative disorders and
PARP inhibitors have been shown to be effective in ani-
mal models of stroke, traumatic brain injury, and Par-
kinson’s disease.2,3 And also, the use of PARP
inhibitors as potential adjuvant cancer therapy with
alkylating agents and/or radiation has been recently an
area of primary interest in this field.4 Therefore, inhibi-
tion of PARP by pharmacological agents may prove
useful for the therapy of neurodegenerative disorders
and several other diseases involved in PARP activation.
for PARP inhibitors based on the prototypical PARP
inhibitor 3-aminobenzamide have been developed. Our
laboratory has previously reported the structure-based
design and synthesis of the series of quinazolinones
shown in Scheme 1.5 We designed quinazolinone 3,
which is a structural hybrid of 1 and 2 based on molec-
ular modeling of the human active site of PARP-1, and
found that quinazolinone 4e shows strong potency
(IC50 = 13 nM) and good oral bioavailability with a high
brain/plasma concentration ratio of approximately five.
Having identified a general template for the design of
PARP-1 inhibitors, we set out to further explore this ser-
ies by examining modification to the ubiquitous propyl-
ene side chain present in 4e. Accordingly, we have
designed a new series of analogues, in which the propyl-
ene linker unit has been replaced with a conformational-
ly restricted cyclic structure to reduce the number of
rotatable bonds. In this paper, we describe the SAR re-
sults of this novel series of modified cyclic linker of qui-
nazolinones against PARP-1.
Table 1 shows the SAR result of various cyclic struc-
tures in place of linear linker.6 Cyclopentene ring com-
pound 6a displayed the same potency as the original
compound 4a, while cyclopentane ring analogues 7a
and 8a and pyrrolidine ring analogue 14a were appropri-
ately 20 times less potent than 6a. Cyclohexene ring 9a
and cyclohexane ring analogues 10a, 11a, and 12a were
10-fold and 100-fold less active, respectively. As these
results, we have selected the restricted cyclopentene lin-
ker as a new lead compound of quinazolinone analogue.
Over the last two decades, extensive investigations have
been conducted in the identification of novel PARP-1
inhibitors. Various approaches to design the scaffolds
Keywords: Poly(ADP-ribose)polymerase PARP-1; SBDD; X-ray;
Quinazolinone; Conformationally restricted.
*
Corresponding author. Tel.: +81 29 863 7179; fax: +81 29 852
2971; e-mail: kouji.hattori@jp.astellas.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.07.091

5578
K. Hattori et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5577–5581
F
O
Lead
Generation
Cl
O
N
Cl
O
N
NH
NH
Me
NH
3
N
O
N
N
1
2
Lead
Optimization I
R2
R2
Lead Optimization II
Conformationally restricted
new PARP-1 Inhibitors
O
N
O
N
NH
X
NH
N
N
R1
R1
4e: R1 = Cl, R2 = F
4f R1 = Me, R2 = F
Scheme 1. Design of quinazolinone PARP-1 inhibitors.
Table 1. SAR of cyclic-linker analoguesa
in spite of a 2-fold improvement in activity for the linear
linker 4 and 5. The phenylpiperidine (S)-16d in particu-
lar showed 12-fold improved potency relative to the cor-
responding linear linker compound 5d and 4-fold more
potency than the fluorophenylpiperidine (S)-16e. Figure
1 displays an overlay of (S)-16e (green color) and 4f
(pink color). The conformation of 4f was based on the
structure in a X-ray co-crystal structure with the 4f/
PARP-1 complex (accession code of the PDB: 1
UK0).7 Indeed, docking study suggested that both the
quinazolinone core and the terminal phenylpyridine unit
of (S)-16e could be occupied the similar space as an
important pharmacophore for PARP inhibitory activity.
O
R1: A=
N
N
Ph
Ph
NH
R1
B=
N
L
Compound
L
R1
A
PARP-1 IC50 (nM)
16
4a
5a
B
46
16
6ab
A
(cis or trans)-7ab
(trans or cis)-8ab
9ab
B
B
A
435
555
160
Table 3 shows the brain/plasma concentrations after
oral administration in mice.8 In general, the compounds
linked with cyclopentene showed improved plasma or
brain concentrations compared to the corresponding
compounds with a linear linker, which would be ex-
plained by metabolic stability in Table 4.9 We have al-
ready confirmed the metabolic pathway of 3 and
identified the main metabolites which were produced
by oxidation of the linker unit in both rat and human
liver microsomes shown in Scheme 3.10 Accordingly,
the cyclopentene linker was more efficacious to block
this primary site of metabolism. The tetrahydropyridine
type (S)-6d and 6e, and the piperazine type 17c, had de-
creased brain penetration. The phenylpiperidine type
(S)-16d, however, showed good brain/plasma ratios over
two hours, an important prerequisite for treatment of
brain-related neurodegenerative disorders.
10ab,c
A
1800
(cis)-11a
A
A
1600
8400
(trans)-12a
13a
14ab
15a
A
A
A
504
300
131
N
N
N
a Values are averages from at least two independent dose–response
curves, and standard errors for PARP inhibition assay is typically
5% of the mean or less.
b Racemic compounds.
c Compound was mixture of cis and trans.
Compound (S)-16d was synthesized using the proce-
dures outlined in Scheme 2. Briefly, the racemate com-
pound was synthesized using the same procedure as in
our previous publication.5 Bromination of 18 with N-
bromosuccinimide gave 19 and its isomer in a 2:1 ratio.
After removal of isomer 20 by column chromatography,
hydrolysis of the methyl ester group gave the
corresponding carboxylic acid. Treatment with SOCl2,
followed by condensation with chloroanthranilamide,
gave amide derivative 21. Amination of 21 with
Table 2 shows the results of quinazolinone analogues
linked via a cyclopentene linker. Regarding the chirality
of the cyclopentene ring, the (S)-enantiomer showed
more potent activity than the (R)-enantiomer, which
agrees with predictions based on the result of molecular
modeling of the human active site of PARP-1. It is note-
worthy that substitution of a fluoro group on the phenyl
ring did not improve activity for the cyclopentene linker

K. Hattori et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5577–5581
Table 2. SAR of cyclopentene-linker analoguesa
5579
Compound
PARP-1 IC50 (nM)
R1
R2
6a
6b
H
H
F
16
18
R2
O
N
H
NH
(S)-6d
(R)-6d
6e
Cl
Cl
Cl
Me
H
H
F
15
N
286
36
R1
6
6f
F
25
(S)-16d
(S)-16e
(R)-16e
16f
Cl
Cl
H
F
8.7
R2
O
N
35
257
30
NH
Cl
F
N
N
Me
Me
F
Cl
R1
16g
46
16
17c
17e
17g
H
Cl
F
25
25
12
R2
O
N
Cl
Me
NH
N
Cl
R1
17
4a
4b
4c
4d
4e
4f
H
H
F
16
H
8.9
R2
O
N
H
Cl
H
F
23
26
13
NH
Cl
Cl
Me
Me
N
N
R1
F
Cl
8.0
4
5
4g
16
R2
5a
5d
5e
H
H
H
F
46
103
47
O
N
Cl
Cl
NH
R1
a Values are the averages from at least two independent dose–response curves, and standard error for PARP inhibition assay is typically 5% of the
mean or less.
Table 3. Brain/plasma concentration in micea
Brain (lg/g tissue)
Plasma (lg/mL)
0.5 h 2.0 h
0.5 h
2.0 h
(S)-6d
6e
2.0
0.4
3.0
0.9
17.1
15.2
8.7
18.1
17.7
4.4
(S)-16d
(S)-16e
17c
19.8
6.7
18.1
12.5
2.2
13.6
>20.1
1.3
11.3
>20.4
1.2
1.3
8.0
4d
4e
8.4
2.8
8.4
11.6
1.9
12.5
1.2
9.4
5d
3.9
a The concentration is determined by ex vivo assay after po adminis-
tration, see Ref. 8.
Table 4. Metabolism by liver microsomesa
Clint (mL/min/kg)
Mouse
Human
6e
1333
1164
686
54
22
(S)-16d
(S)-16e
4e
17
156
3910
Figure 1. Overlay of (S)-16e (green color) and 4f (pink color).
a Results are average of two experiments, see Ref. 9.

5580
K. Hattori et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5577–5581
O
NH2
NH
MeO
MeO
O
b, c, d
MeO
O
a
Br
+
O
Cl
Br
Br
20
O
18
19
21
O
O
NH2
f, g, h
NH
e
NH
N
N
Cl
N
O
Cl
22
(S)-16d
Scheme 2. Reagents and conditions: (a) N-bromosuccinimide, CCl4, reflux; (b) 1 N NaOH, THF–MeOH, rt; (c) SOCl2, CH2Cl2, rt; (d)
chloroanthranilamide, TEA, CH2Cl2, rt; (e) phenylpiperidine, TEA, DMF, rt; (f) 1 N NaOH, dioxane, rt; (g) (L)-di-p-toluoyl-tartaric acid, EtOH; (h)
1 N NaOH, rt.
Cl
O
N
NH
+N
M-5
Cl
O
N
Cl
O
OH
Cl
O
N
NH
NH
NH
N
N
HO
N
N
3
M-3
M-4
Main pathway for metabolites
Cl
O
N
NH
+
O
HN
M-2
M-1
Scheme 3. Metabolites pathway of 3.
phenylpiperidine in the presence of triethylamine gave
22, which was treated with 1 N NaOH in dioxane at
room temperature to yield 16d as a racemic mixture.
The optical resolution of 16d by crystallization with
(L)-di-p-toluoyl-tartaric acid from ethanol (three times)
produced the optically pure (S)-16d with > 98% ee.11
The preparation of the tetrahydropyridine type 6 and
the piperazine type 17 was synthesized using the similar
procedure in Scheme 2.
Acknowledgments
We express our thanks to Dr. David Barrett for his crit-
ical reading of the manuscript and Mr. Akihiro Yokota
at University of Kyoto for docking study of (S)-16e and
4f.
References and notes
In conclusion, we have refined our previously described
PARP-1 inhibitor templates, replacing the propylene
linker unit found in 4 with a conformationally restricted
cyclopentene linker. This has led to a new series of
PARP-1 inhibitors described herein. The phenylpiperi-
dine type (S)-16d was found to be very potent with good
bioavailability and high brain penetration. These find-
ings suggest (S)-16d could be an attractive therapeutic
candidate for neurodegenerative disorders such as Par-
kinson’s disease.
´
1. For recent reviews: (a) Jagtap, P.; Szabo, C. Nat. Rev.
Drug Disc. 2005, 4, 42; (b) Woon, E. C. Y.;
Threadgill, M. D. Curr. Med. Chem. 2005, 12, 2373;
(c) Peukerr, S.; Schwahn, U. Expert Opin. Ther. Pat.
2004, 14, 1531; (d) Southan, G. J.; Szabo, C. Curr.
Med. Chem. 2003, 10, 321; (e) Cosi, C. Expert Opin.
2002, 12, 1047.
´
2. (a) LaPlaca, M. C.; Zhang, J.; Raghupathi, R.; Li, J. H.;
Smith, F.; Bareyre, F. M.; Snyder, S. H.; Graham, D. I.;
Mclntosh, T. K. J. Neurotrauma 2001, 18, 369; (b)

K. Hattori et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5577–5581
5581
Abdelkarim, G. E.; Harms, C.; Katchanov, J.; Dirnagl,
´
Mihara, K.; Yamazaki, S.; Matsuoka, N.; Teramura, Y.;
Miyake, H. J. Med. Chem. 2004, 47, 4151.
U.; Szabo, C.; Endres, M. Int. J. Mol. Med. 2001, 7, 255;
(c) Takahashi, K.; Pieper, A. A.; Croul, S. E.; Zhang, J.;
Snyder, S. H.; Greenberg, J. H. Brain Res. 1999, 829, 46;
(d) Cosi, C.; Colpaert, F.; Koek, W.; Degryse, A.; Marien,
M. Brain Res. 1996, 729, 264.
6. Measurements of PARP-1 inhibitory activity in vitro were
carried out by a standard method using human recombi-
nant PARP-1. See Ref. 5.
7. Kinoshita, T.; Nakanishi, I.; Warizuka, M.; Iwashita, Y.;
Kido, Y.; Hattori, K.; Fujii, T. FEBS 2004, 556, 43.
8. Measurements of the concentration of inhibitors in brain
and plasma were performed in mice following po admin-
istration (32 mg/kg, n = 2) by ex vivo assay. See Ref. 5.
9. Compounds were incubated at 37 °C for 60 min with mice
and humans in the presence of the MADPH-generating
system.
3. Recent studies have demonstrated that dopaminergic
neurons lacking the gene coding PARP are spared from
the neurotoxic effects of MPTP: Mandir, A. S.; Przedbor-
ski, S.; Jackson-Lewis, V.; Wang, Z.; Simbulan-Rosenthal,
C. M.; Smulson, M. E.; Hoffman, B. E.; Guastella, D. B.;
Dawson, V. L.; Dawson, T. M. Proc. Natl. Acad. Aci.
U.S.A. 1999, 96, 5774.
4. Madhusudan, S.; Middleton, M. R. Cancer Treat. Rev.
2005, 31, 603.
5. Hattori, K.; Kido, Y.; Yamamoto, H.; Ishida, J.; Kamijo,
K.; Murano, K.; Ohkubo, M.; Kinoshita, T.; Iwashita, A.;
10. M-1 and M-2 were the main metabolites in both mouse
and human liver microsomes shown in Scheme 3.
11. The ee ratio was determined by HPLC analysis with a
chiral column.