Discovery of potent and selective PARP-1 and PARP-2 inhibitors: SBDD analysis via a combination of X-ray structural study and homology modeling

Article abstract of DOI:10.1016/j.bmc.2005.09.061

We disclose herein our efforts aimed at discovery of selective PARP-1 and PARP-2 inhibitors. We have recently discovered several novel classes of quinazolinones, quinazolidinones, and quinoxalines as potent PARP-1 inhibitors, which may represent attractive therapeutic candidates. In PARP enzyme assays using recombinant PARP-1 and PARP-2, the quinazolinone derivatives displayed relatively high selectivity for PARP-1 and quinoxaline derivatives showed superior selectivity for PARP-2, and the quinazolidinone derivatives did not have selectivity for PARP-1/2. Structure-based drug design analysis via a combination of X-ray structural study utilizing the complexes of inhibitors and human PARP-1 catalytic domain, and homology modeling using murine PARP-2 suggested distinct interactions of inhibitors with PARP-1 and PARP-2. These findings provide a new structural framework for the design of selective inhibitors for PARP-1 and PARP-2.

Full text of DOI:10.1016/j.bmc.2005.09.061

Bioorganic & Medicinal Chemistry 14 (2006) 1378–1390
Discovery of potent and selective PARP-1 and PARP-2 inhibitors:
SBDD analysis via a combination of X-ray structural
study and homology modeling
Junya Ishida,a Hirofumi Yamamoto,a Yoshiyuki Kido,a Kazunori Kamijo,a
Kenji Murano,a Hiroshi Miyake,a Mitsuru Ohkubo,b Takayoshi Kinoshita,b
Masaichi Warizaya,b Akinori Iwashita,c Kayoko Mihara,c
Nobuya Matsuokac and Kouji Hattoria,*
aMedicinal Chemistry Research Laboratories, Fujisawa Pharmaceutical Co. Ltd, 2-1-6 Kashima,
Yodogawa-ku, Osaka 532-8514, Japan
bExploratory Research Laboratories, Fujisawa Pharmaceutical Co. Ltd, 5-2-3 Tokodai, Tsukuba, Ibaraki 300-2698, Japan
cMedicinal Biology Research Laboratories, Fujisawa Pharmaceutical Co. Ltd, 2-1-6 Kashima, Yodogawa-ku, Osaka 532-8514, Japan
Received 16 May 2005; revised 26 September 2005; accepted 27 September 2005
Available online 8 November 2005
Abstract—We disclose herein our efforts aimed at discovery of selective PARP-1 and PARP-2 inhibitors. We have recently discov-
ered several novel classes of quinazolinones, quinazolidinones, and quinoxalines as potent PARP-1 inhibitors, which may represent
attractive therapeutic candidates. In PARP enzyme assays using recombinant PARP-1 and PARP-2, the quinazolinone derivatives
displayed relatively high selectivity for PARP-1 and quinoxaline derivatives showed superior selectivity for PARP-2, and the qui-
nazolidinone derivatives did not have selectivity for PARP-1/2. Structure-based drug design analysis via a combination of X-ray
structural study utilizing the complexes of inhibitors and human PARP-1 catalytic domain, and homology modeling using murine
PARP-2 suggested distinct interactions of inhibitors with PARP-1 and PARP-2. These findings provide a new structural framework
for the design of selective inhibitors for PARP-1 and PARP-2.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
and consequently ATP, culminating in cell dysfunction
or necrosis. Furthermore, PARP has also been implicat-
ed in a caspase-independent apoptosis pathway medi-
ated by apoptosis-inducing factor.1 PARP inhibitors
provided remarkable protection from tissue damage in
various forms of reperfusion injury, inflammation, and
neurotoxicity in animal models.2 Thus, inhibition of
PARP by pharmacological agents could be useful in
the treatment of inflammatory disease, neurodegenera-
tive disease, and several other diseases involved in
PARP activation.
Poly(ADP-ribose) polymerase (PARP) is an abundant
nuclear enzyme in eukaryotic cells that has been impli-
cated to become activated in response to DNA damage.
Activated PARP catalyzes the transfer of ADP-ribose
units from nicotinamide adenine dinucleotide (NAD+)
to nuclear acceptor proteins such as histones, topo-
isomerases, DNA polymerases, DNA ligases, and PARP
itself. Excessive activation of PARP consumes NAD+
Keywords: PARP-1, poly(ADP-ribose)polymerase-1; PARP-2, poly
(ADP-ribose) polymerase-2 (PARP-2); NAD+, nicotinamide adenine
dinucleotide; PAR, poly(ADP-ribose); SBDD, structure based drug
design; NI site, nicotinamide-ribose binding site; AD site, adenine-
ribose binding site.
Characterized family members of PARP currently in-
clude the proteins PARP-1, PARP-2, PARP-3, Tankyr-
ase-1, Tankyrase-2, TiPARP, and vPARP. Although
PARP-1 has been believed to be responsible for all the
DNA-damage-dependent poly(ADP-ribose) (PAR) syn-
thesis in mammalian cells, a novel DNA-damage-depen-
dent PARP-2 was subsequently discovered as a result of
the presence of residual DNA-dependent PARP activity
*
Corresponding author at present address: Astellas Pharmaceutical
Co. Ltd., 21 Miyukigaoka, Tsukuba 305-8585, Japan. Tel.: +81 29
863 7179; fax: +81 29 852 2971; e-mail: kouji.hattori@jp.
astellas.com
0968-0896/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2005.09.061

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1379
in embryonic fibroblasts derived from PARP-1-deficient
mice.3,4 Of the PARP family currently identified, PARP-
2 bears the strongest resemblance to PARP-1 and might
be functionally similar. PARP-2 binds to DNA and is
activated by DNAse-treated DNA. It appears that
PARP-2 contributes to the residual PARP activity ob-
served in PARP-1 (ꢀ/ꢀ) cells after treatment with
DNA-damaging agents, suggesting a biological role for
PARP-2 in response to DNA damage. Interestingly,
the DNA-binding domain of PARP-2 is distinct from
PARP-1 and PARP-2 lacks the central automodification
domain. The mouse PARP-2, mPARP-2, lacks the zinc-
finger-motif, despite DNA-binding activity.
between PARP-1 and PARP-2.6 Furthermore, biologi-
cal assay data and SAR study on specific PARP inhibi-
tors are rare, thus, the discovery of new PARP-isoform
selective ligands presents an exciting challenge and new
opportunities to the field.
We have recently discovered some classes of quinazoli-
nones, quinazolidinones, and quinoxalines as potent
PARP-1 inhibitors which can represent attractive thera-
peutic candidates for neurodegenerative disorders such
as cerebral ischemia or ParkinsonÕs disease in Figure
111. In this paper, we describe our research on PARP-
1 and PARP-2 selective inhibitors using the SBDD anal-
ysis by a combination of X-ray analysis and homology
modeling.
In view of the significant structural information, the first
crystal structure available was that of the catalytic do-
main of chicken PARP-1, which informed us to identify
the topology of the donor and acceptor sites harboring
the substrate NAD+ and the ADP-ribose moiety of the
poly(ADP-ribose) chain of the target protein, respective-
ly. Furthermore, it helped us to elucidate inhibitor–en-
zyme interactions for a number of chemically different
compounds that showed competition to NAD+ in its
binding fold for the binding mode of inhibitors to the
NAD+-binding site.5 Recently, the second crystal struc-
ture of murine PARP-2 has been solved. Pearl and col-
leagues have disclosed the catalytic fragment of murine
2. Chemistry
Our initial procedure to prepare quinazolinone ana-
logues linked via appropriate substitution is outlined
in Scheme 1. Solid-phase synthesis with Rink amide res-
in produced rapidly and efficiently the desired quinazoli-
none derivatives on a 10 mg scale in five steps. Loading
commercially available anthranilic acid derivatives onto
the support-bound secondary amine employing DIC
resulted in amide bond formation. Amide compounds
on the resins were subsequently treated with 4-bromo-
butyryl chloride, followed by N-alkylation with several
substituted amines to give the cycloamine derivatives.
After cleavage from the resin with a mixture of TFA
and CH2Cl2 (1:1), treatment with 1 N NaOH gave the
desired quinazolinone derivatives in a total yield of
40–80%. Re-synthesis and 1f were performed using li-
quid synthesis and in a similar method instead of using
resin.
˚
PARP-2, at 2.8 A resolution, and compared this to the
catalytic fragment of PARP-1, with an emphasis on pro-
viding a possible framework for rational drug design to
develop isoform-specific inhibitors.6 Furthermore, it has
been reported that the PARP catalytic domain shows
the highest degree of homology between different spe-
cies,7,8 suggesting that PARP inhibitors may have no
species-difference in terms of inhibitory activity among
human, rat, and mouse PARPs.9
Over the last two decades, a large number of PARP-1
inhibitors have been developed,10 the majority of which
mimic to some degree the nicotinamide moiety of NAD+
and bind to the donor site of the protein. However, the
residues in the donor site of PARP-1 that provide
hydrogen-bonding interactions with these inhibitors
are completely conserved in PARP-2, suggesting that
these existing PARP-1 inhibitors would not discriminate
Synthesis of the quinazolidinone derivatives linked with
corresponding substitution is outlined in Scheme 2.
Treatment of 4 with 4-phenyl-tetrahydropyridine under
basic conditions leads to the corresponding ester 5.
Hydrolysis of ester 5 with KOH, which was converted
to the corresponding amide by DIC and aq NH3, fol-
lowed by KOH under reflux resulted in formation of
the desired quinazolidinone ring 2a and 2b.
Cl
F
O
O
N
NH
N
NH
N
N
N
R
1d: R = Cl
1e: R = Me
1g
R
O
N
R
O
NH2
N
NH
O
N
N
2b: R = F
3b: R = Cl
Figure 1.

1380
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
O
O
N
a
b
R1
H
NH2
N
HN
R1
H
H2N
O
O
O
N
Br
N
HN
R1
H
R1
c
d, e
HN
O
R3R2N
1
R3R2N
Scheme 1. Reagents: (a) anthranilic acid derivatives, DIC, DMF; (b) 4-bromobutyryl chloride diisopropylethylamine, CH2Cl2; (c) 1,2,3,
6-tetrahydro-4-phenylpyridine derivatives HCl, diisopropylethylamine, NMP; (d) TFA, CH2Cl2; (e) 1 N NaOH, dioxane.
R
R
R
O
N
CO2Me
a
CO2Me
Ph
b,c,d
NH
Cl
N
Ph
N
N
O
N
O
OEt
O
OEt
2a: R = H
2b: R = F
4a: R = H
4b: R = F
5a: R = H
5b: R = F
Scheme 2. Reagents: (a) 1,2,3,6-tetrahydro-4-phenylpyridine HCl, K2CO3, NaI, DMF; (b) KOH, H2O–MeOH; (c) DlC, HOBT, THF, then NH3 aq;
(d) KOH, EtOH.
Synthesis of the phthalazinone derivatives is outlined in
Scheme 3. Wittig reaction of 4-benzyloxybutanal 6 and
phosphonium bromide 7 with triethylamine at room
temperature, followed by hydrazine monohydrate, gave
the desired phthalazinone ring 8. Treatment of 8
with BBr3 gives the corresponding bromide, followed
O
O
O
R
a
b,c
NH
N
NH
N
OHC
OBn
+
O
-P+Ph
N
Br
3
OBn
6
7
8
9a: R = H
9b: R = F
Scheme 3. Reagents: (a) Et3N, THF, then hydrazine monohydrate, EtOH; (b) BBr3, CH2Cl2; (c) 1,2,3,6-tetrahydro-4-phenylpyridine derivatives
HCl, Et3N, DMF.
EtO2CHN
NHCO2Et
EtO2CHN
Br
a
b
Br
HO
Br
O
O
10
12
11
NH
NH
c,d
e
Br
N
Ph
13
14
Scheme 4. Reagents: (a) Br2, AcONa, AcOH; (b) phenylboronic acid, Pd(PPh3)4, 2 M Na2CO3, DME; (c) P2O5, POCl3; (d) 4 N HCl, EtOH;
(e) 1,2,3,6-tetrahydro-4-phenylpyridine HCl, Et3N, DMF.

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1381
2HCl
NH2
H2NOC
O
NH2
N
O
O
NH2
N
R
a
Br
+
NH2
R
N
N
3
R
17
16
Scheme 5. Reagents: (a) Et3N, MeOH.
15
by reaction with the corresponding pyridine derivatives
to obtain phthalazinone derivatives 9a and 9b.
inhibitory potency of these derivatives was found to be
largely dependent on the unique linker of the quinazoli-
none ring. Quinazolinone linked with a 4-phenyl-tetra-
hyropyridine moiety (1a) exhibited strong potency
against PARP-1 (IC50: 21 nM) and exhibited appropri-
ately 30-fold less potency against PARP-2 (IC50:
608 nM). The addition of a chloro group at the 8-posi-
tion (1b), which is important for bioavailability and
brain penetration, resulted in almost the same activity
and selectivity, as previously reported.11a Substitution
of the p-cyano group at the terminal phenyl ring was
the most advantageous, with IC50 = 3 nM, but the same
selectivity was observed. The addition of a methyl group
at the 8-position (1e) and a chloro group at the 5-posi-
tion (1f) showed three times lower selectivity than 1a.
N-Phenylpiperazine derivative (1g) also had good activ-
ity (IC50: 11 nM) and selectivity (PARP-2/1: 27). On the
other hand, well-known inhibitors such as 2-methyl-
4(3H)quinazolinone (18) and 3-AB were less potent
and selective, with PARP-2/1: 0.6 and 0.9, respectively.
Synthesis of the phenanthridinone 14 is outlined in
Scheme 4. Bromination of the commercially available
carbonate 10 with bromine and sodium acetate in acetic
acid at the room temperature gives dibromide 11. After
Suzuki coupling with phenylboric acid and Pd(PPh3)4,
diphenyl derivative 12 was treated with P2O5 in phospho-
rus oxychloride at the room temperature, followed by
4 N HCl solution under reflux for 30 min to form the
phenanthridinone ring 13. 13 was treated with the same
amine to give the corresponding phenanthridinone 14.
Finally, synthesis of the quinoxaline derivatives is out-
lined in Scheme 5. Treatment of 2,3-diaminobenzamide
dihydrochloride 15 with the corresponding phenacyl
bromide derivatives 16 in the presence of the triethyl-
amine at the room temperature overnight produced a
mixture of 3 and minor regioisomer 17. The mixture
was easily purified by column chromatography to give
desired quinoxaline derivatives 3.
˚
We performed X-ray crystallography refined to 3.0 A of
the catalytic domain of human PARP-1 complexed with
1e, as shown in Figure 2.8 The results indicated that the
quinazolinone part of the inhibitor 1e tightly binds to
the nicotinamide-ribose binding site (NI site) and the
4-phenyl-tetrahydropyridine moiety provides secondary
contacts to the adenine-ribose binding site (AD site),
resulting in improvement of potency. Figure 2B shows
the NI site as a critical binding mode for a common
3. Results and discussion
3.1. Quinazolinone structure
PARP-1/2 inhibitory activity of quinazolinone deriva-
tives12 in vitro is outlined in Table 1.13 In general, the
Table 1. PARP-1/2 inhibitory activities of quinazolinone analogues13
R3
R1
O
N
NH
N
R2
1
Compound
R1
R2
R3
PARP-1 IC50 (nM)a
PARP-2 IC50 (nM)a
Selectivity PARP-2/1
1a
1b
H
H
H
H
H
Cl
H
H
H
CN
F
21 1.9
23 1.8
608 12.9
610 18.1
87 3.6
29
27
29
39
10
9
Cl
Cl
Cl
Me
H
1c
1d
3.0 0.6
13 1.3
16 1.1
500 15.2
167 25.7
630 28.5
300 12
1e
F
1f
1g
18b
3-ABc
H
68 2.4
11 0.4
27
0.6
0.9
1200 80
11200 810
660 24.7
9810 770
a The values are presented as means SE of three independent experiments (each experiment was triplicate).
b 5-Chloro-2-methyl-4(3H)-quinazolinone.
c 3-Aminobenzamide.

1382
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
A
B
D
C
Figure 2. 8,13 (A) X-ray structure (3.0 A resolution) of compound 1e in complex with human PARP-1 catalytic domain. 1e is yellow with heteroatoms
being blue for nitrogen, red for oxygen, and green for fluoride. (B) Close-up views of 1e in the NI site of PARP-1. Important residues in the NI site
are shown in ball and stick style. The quinazolinone part was tightly bound by three hydrogen bonds (C@O to Ser904Oc and Gly863NH, NH to
Gly863C@O) and by a sandwiched hydrophobic interaction of Tyr907 and Tyr869. (C) Close-up views of the spacer part of 1e. Methylene part is
˚
˚
twisted and the distance between nitrogen atom and a-CH is 2.6 A. (D) Close-up views of the terminal site of 1e in the AD site of PARP-1. The
terminal phenyl ring lies in a deep hydrophobic pocket which consists of the side chains of Leu769, Ile879, Pro881, and the methylene chain of
Arg878 in PARP-1.
structure motif: three hydrogen bonds (C@O to Ser904-
Oc and Gly863NH, NH to Gly863C@O) and a sand-
wiched hydrophobic interaction consisting of a p–p
interaction with the phenyl ring of Tyr907 and a CH–
p interaction with Cb of Tyr869.14 4-Phenyl-tetrahydro-
pyridine moiety of the inhibitor lies in the deep pocket,
which consists of the side chains of Leu769, Ile879,
Pro881, and the methylene chain of Arg878, by van
der Waals interaction in Figure 2D. The spacer part of
the inhibitors, which we designed as a three carbon unit
to allow a maximum fix of the terminal 4-phenyl-tetra-
hydropyridine moiety in the AD site, the a-CH2 is locat-
these findings, it is clearly suggested that replacement
of Leu769 by Gly314 leads to loss of a hydrophobic
pocket and leads to a loss of potency against PARP-2.
The change of only one amino acid is therefore suffi-
cient to allow discrimination between PARP-1 and
PARP-2.
3.2. Quinazolidinone and related structures
Table 2 shows the SAR results of quinazolidinone and
related structures. Quinazolidinone structure 2 was dis-
covered from the Fujisawa sample collection and phtha-
lazinone 9 and phenanthridinone 14 have already served
as substrates for NAD+.16 We designed the related ana-
logues 9 and 14, which are linked by a unique 4-phenyl-
tetrahydropyridine moiety extended with a three- or
four-carbon unit by modeling to introduce it to the
AD site. In general, these derivatives had similar
potency for PARP-1 as the quinazolinone derivatives.
However, they did not have selectivity for PARP-1
and PARP-2 in spite of overlapping with the terminal
4-phenyl-tetrahydropyridine moiety of all compounds
in AD site. PARP-2/1 selectivities for 2a and 9a, and
14 were 1.4 and 0.8, and 1, respectively.
˚
ed 2.6 A from the nitrogen in Figure 2C. As we
previously reported, transformation from nitrogen of
the pyridine moiety into carbon led to a complete loss
of potency of PARP-111a due to the steric hindrance
with the a-CH2 and SP3 carbon.
Next we performed homology modeling of the catalytic
domain of human PARP-2 and 1e by FAMS15 based on
the solved X-ray structure of murine PARP-2.6 As
expected, there are no differences in the nicotinamide-ri-
bose binding site (NI site) between PARP-1 and PARP-
2. The quinazolinone 1e binds to the NI site of PARP-2
by three hydrogen bonds (C@O to Ser446Oc and
Gly405NH; NH to Gly405C@O) and by a sandwiched
hydrophobic interaction with the phenyl ring of
Tyr438 and Tyr449 (Fig. 3A). However, the critical dif-
ference between PARP-1 and PARP-2 in this hydropho-
bic pocket in the AD site is the replacement of Leu769 in
PARP-1, with Gly314 in PARP-2 (Fig. 3B). Based on
In order to validate our results, we have compared by X-
ray crystallography the catalytic domain of human
PARP complexes with 2b and 1e. Figure 4A shows that
quinazolidinone 2b binds to both the NI and the AD
subsites of the donor site in a similar binding mode as
1e. The quinazolidinone core binds to the NI site by

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1383
A
B
Gly405
Ser446
Tyr449
Gly314
Tyr438
Figure 3. Refs. 13. (A) Homology modeled structure of compound1e in complex with human PARP-2 catalytic domain based on the solved X-ray
˚
structure (2.8 A resolution) of murine PARP-2. Important residues in the NI site are shown in ball and stick style. (B) Close-up views of the terminal
site of 1e in the AD site of PARP-2. The replacement of Leu769 in PARP-1with Gly314 in PARP-2.
Table 2. PARP-1/2 inhibitory activities of quinazolidinone and related structures
Compound
R
H
X
N
PARP-1 IC50 (nM)a
PARP-2 IC50 (nM)a
Selectivity PARP-2/1
R
O
N
NH
2a
120 3.2
70 3.6
0.6
1.4
Ph
O
X
N
2b
2c
F
F
N
C
60 2.7
>1000
83 2.5
NT
O
R
NH
N
9a
H
F
120 4.3
90 4.7
0.8
9b
14
49 3.1
82 5.1
84 4.2
84 2.8
1.7
1.0
H
N
O
Ph
N
(CH2)3
a The values are presented as means SE of three independent experiments (each experiment was triplicate).
A
B
Figure 4. (A) Compound 2b bound to the human PARP-1 active site. Compound 2b is yellow with heteroatoms being blue for nitrogen, red for
oxygen, and green for fluoride. Important residues are shown in stick style. The nitrogen atom of the terahydropyridine moiety of 2b directly binds to
CO2H of Asp766 located 3.6 A from the nitrogen. See Ref. 11a (B) Overlay of co-crystal structures of 1e (yellow for carbon) and 2b (gray for carbon)
˚
in the PARP-1 active site.
three hydrogen bonds and by a sandwiched hydropho-
bic interaction. The terminal phenyl ring lies in the deep
pocket by van der Waals interaction in the AD site.
The different binding mode is the nitrogen atom of the
tetrahydropyridine moiety of 2b, which directly binds
˚
to CO2H of Asp766 located 3.6 A from the nitrogen.

1384
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
Figure 4B displays an overlay of cocrystal structures of
1e (yellow for carbon) and 2b (gray for carbon) in the
PARP-1 active site, showing that they occupy the same
general space in the NI site and AD site. However, the
location of the nitrogen atom of the tetrahydropyridine
moiety is in a different location. Transformation from
nitrogen of pyridine moiety into carbon (2c) also led
to a complete loss of potency of PARP-1, suggesting
that this nitrogen is a very important factor for binding
mode between ligand and PARP-1. 2b and 9, and 14
overlapped in the same position of the nitrogen atom
of 4-phenyltetrahydropyridine by our modeling and
show the same potency for PARP-1 and PARP-2. On
the other hand, the replacement of Asp766 in PARP-1
with Glu311 in PARP-2 would be expected to lead to
a more strong interaction of the CO2H of Glu311 with
the nitrogen atom of side chain to recover the loss of
the hydrophobic pocket in PARP-2.
ylic acid of Glu988, suggesting that this binding mode
was conserved in PARP-2 (Fig. 5B). However, the main
differences in the binding site of the terminal phenyl
group of quinoxaline analogues within the groove could
be used to introduce discrimination between PARP-1
and PARP-2. The Cl-phenyl group of 3b provides sec-
ondary contacts to the side chain of Asp766 formed by
a hydrogen bond with Tyr889, which replaces the hydro-
gen bond between Glu311 and Tyr431 in PARP-2. The
second source of selectivity results from the replacement
of Glu763 in PARP-1 with Gln308 in PARP-2. X-ray
analysis shows that the side chain of Glu763 in PARP-
1 is restricted by a hydrogen bond with neighbor
Gln759 and Ala760 (Fig. 5A),18 while the residue of
Gln308 in PARP-2 is mobile and can make a more favor-
able interaction with the Cl-phenyl group, since the pre-
vious study indicated that the side chain of Gln763 in
chicken PARP-1 appeared to be mobile and movement
of this residue nicely accommodated the p-benzylamine
substitution of inhibitors.17,19 It is considered that the
Cl-phenyl group extends into this groove containing
Glu311 and Gln308 in PARP-2 to make a more favor-
able interaction and may account for the increase in
potency observed (Fig. 5C vs. D).
3.3. Quinoxaline structure
The SAR results for quinoxaline derivatives are outlined
in Table 313. The activity of these inhibitors against
PARP-2 is about 10-fold more potent than against
PARP-1. Substitution at the para position of the termi-
nal phenyl ring was more advantageous to activity than
either meta or ortho substitution (data not shown). A 2-
fold improvement in activity against PARP-2 was
obtained by p-chloro (3b) with IC50 = 7.0 nM and
selectivity: 0.21. Addition of p-cyano also gave good
potency for PARP-2 (IC50 = 8 nM) and good selectivity
(selectivity: >10-fold).
4. Conclusions
We have discovered the first PARP-1 and PARP-2 selec-
tive inhibitors. The quinazolinone derivatives are
PARP-1 selective and the quinoxaline derivatives are
PARP-2 selective. SBDD study by a combination of
X-ray analysis and homology modeling has revealed
the distinct binding mode for the discrimination between
ligands and PARP-1/2. The information from this study
should be useful for the further design of novel and
selective PARP inhibitors. It is not known at present
how PARP inhibition therapy affects the function of
PARP isoforms or how inhibition of these isoforms con-
tributes to possible reduction of the side effects. Com-
pounds 1d and 1g, and compounds 3b were found to
have a good pharmacokinetic profile with brain penetra-
tion. These inhibitors provide important tools for the
development of new conceptual medication for neurode-
To validate our results, we performed structural analysis
by X-ray crystallography of human PARP-1 complexes
with quinoxaline ligand and molecular modeling of
PARP-213. Figure 5A shows the structure of the human
catalytic domain of PARP-1 complex with 3b, refined to
˚
3.0 A. The binding mode of 3b with human PARP-1 is al-
most the same as those of the previously disclosed benz-
imidazole analogues.17 The carboxamide part of the
inhibitor tightly binds to the same NI site by the three
critical hydrogen bonds with Ser904 and Gly863, and
the nitrogen of the quinoxaline ring binds to the carbox-
Table 3. PARP-1/2 inhibitory activities of quinoxaline analogues13
O
NH2
R
N
N
3
Compound
R
PARP-1 IC50 (nM)a
PARP-2 IC50(nM)a
Selectivity PARP-2/1
3a
3b
3c
3d
3e
3f
H
Cl
131 2.6
33 0.9
101 1.8
118 3.1
71 2.2
87 3.8
14 1.3
7.0 2.0
8.0 1.0
11 0.9
8.0 0.8
9.0 1.0
0.11
0.21
0.08
0.09
0.11
0.10
CN
CF3
OMe
NH2
a The values are presented as means SE of three independent experiments (each experiment was triplicate).

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1385
A
B
Ser904 Gly863
Ser446
Gly405
Tyr907
Tyr449
Gln308
Glu763
Asp766
Glu988
Tyr896
Glu534
Tyr438
Glu311
Ala760
Gln759
Tyr889
Tyr431
C
D
Glu763
Gln3
Asp766
Tyr431
Tyr889
˚
Figure 5. Ref. 13. (A) X-ray structure (3.0 A resolution) of compound 3b in complex with human PARP-1 catalytic domain. Compound 3b is yellow
with heteroatoms being blue for nitrogen, red for oxygen, and green for chloride. Important residues in the NI site are shown in ball and stick style.
˚
(B) Homology modeled structure of compound3b in complex with human PARP-2 catalytic domain based on the solved X-ray structure (2.8 A
resolution) of murine PARP-2. Important residues in the NI site are shown in ball and stick style. (C) Close-up views of the binding mode of the
Cl-Phenyl site of 3b in PARP-1. (D) Close-up views of the binding mode of the Cl-Phenyl site of 3b in PARP-2.
generative disorders such as cerebral ischemia or Parkin-
sonÕs diseases.
bound secondary amine, 1.0 g, 0.59 mmol/g, loading).
After the vessel was shaken for 12 h at ambient temper-
ature, the resin was washed with dichloromethane, and
THF, DMF, and dichloromethane.
5. Experimental
5.1. Chemistry
2. Diisopropylethylamine(10 equiv),
4-bromobutyl
chloride (10 equiv), and DMAP (1 equiv) were add-
ed to a mixture of the obtained resin in dichloro-
methane (10 mL). After the vessel was shaken for
12 h at ambient temperature, the resin was washed
with dichloromethane, and THF, DMF, and
dichloromethane.
5.1.1. General information. 1H NMR spectra were
recorded on a Varian EM-390 NMRspectrometer or a
Bruker AC200P spectrometer. Chemical shifts are
1
reported downfield from tetramethylsilane (=0) for H
3. Cyclic amine derivatives (10 equiv), diisopropyleth-
ylamine (10 equiv), and DMAP (1 equiv) were
added to a mixture of the obtained resin in N-meth-
yl-2-pyrrolidinone (10 mL). After the vessel was
shaken for 12 h at ambient temperature, the resin
was washed with dichloromethane, and THF,
DMF, and dichloromethane.
4. Cleavage from resin was performed with 50% trifluo-
roacetic acid in dichloromethane (10 mL) for 30 min
at ambient temperature. After the filtered solvent
was evaporated under pressure, the residue was dis-
solved in dioxane (5 mL). An aqueous solution of
sodium hydroxide (1 M, 5 mL) was added to the solu-
tion at room temperature, and the mixture was stirred
at that temperature for 15 h. The organic materials
were extracted with chloroform, and the organic layer
was washed with water and dried over sodium sulfate.
Appropriate purification of the crude product by col-
umn chromatography on silica gel or HPLC gave the
desired product.
NMR. Mass spectra (MS) were determined on a Hitachi
Model M-80 mass spectrometer. Elementary analyses
were performed on a Perkin-Elmer 2400 CHN elemental
analyzer. Reagent and solvent were used as obtained
from commercial suppliers without further purification.
Chromatographic purification of the compounds was
performed on silica gel 60 (63–200 lm) purchased from
Merck Co. Recombinant human PARP enzyme was
purchased from Trevigen, Inc. (Gaithersburg, MD)
and recombinant mouse PARP-1 and PARP-2 enzymes
were purchased from Alexis Biochemicals (San Diego,
CA).
5.1.2. Solid-phase synthesis of quinazolinone derivatives
1. A solution of anthranilic acid derivatives (5 equiv), pyr-
idine (5 equiv), 1,3-diisopropylcarbodiimide (5 equiv),
and DMAP (1 equiv) in DMF (5 mL) was added to a
reaction vessel containing Rink amide resin (support-

1386
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
5.1.3. Liquid-phase synthesis for quinazolinone derivatives
1a
DMSO-d6): d 1.8–2.1 (2H, m), 2.3–2.8 (8H, m), 3.05
(2H, br s), 6.07 (1H, m), 7.2–7.5 (5H, m), 7.86 (1H,
dd, J = 8.0, 1.4 Hz), 7.97 (1H, dd, J = 8.0,1.4 Hz).
API-ESMS: 380 (M++H). Anal. Calcd for
C22H22ClN3O + 0.29H2O: C, 68.61; H, 5.91; N, 10.91.
Found: C, 68.60; H, 5.79; N, 10.91.
1. Under a nitrogen atmosphere, a solution of 4-bromo-
butyryl chloride (4.9 g, 26.4 mmol) in dichlorometh-
ane (10 mL) was added dropwise to a solution of 2-
aminobenzamide (3.0 g, 22 mmol) in pyridine
(18 mL, 220 mmol) and dichloromethane (15 mL) at
0 °C. The mixture was stirred for 1.5 h at 0 °C. The
reaction mixture was poured into ice-cooled 1 N
hydrochloric acid, and the product was extracted
with chloroform. The organic layer was washed with
1 N hydrochloric acid and water, and dried over sodi-
um sulfate. The crude product was triturated with tol-
uene to give 2-[(4-bromobutanoyl)amino]benzamide
(5.11 g, 81.3%) as a powder. 1H NMR (200 MHz,
CDCl3) d 2.29 (2H, quint., J = 6.8 Hz), 2.61 (2H, t,
J = 7.2 Hz), 3.52 (2H, t, J = 6.4 Hz), 5.5–6.5 (2H,
br), 7.09 (1H, dt, J = 7.6, 1.1 Hz), 7.51 (1H, t,
J = 7.6 Hz), 7.53 (1H, d, J = 7.6 Hz), 8.62 (1H, d,
J = 8.5 Hz), 11.25 (1H, s). API-ESMS: 307 (M++Na).
5.1.3.2. 8-Chloro-2-{3-[4-(4-cyanophenyl)-3,6-dihydro-
1(2H)-pyridinyl]propyl}-4(3H)-quinazolinone (1c). 1H
NMR (200 MHz, DMSO-d6) d 1.98 (2H, quint.,
J = 6.9 Hz), 2.3–2.8 (8H, m), 3.11 (2H, d, J = 2.9 Hz),
6.29 (1H, br s), 7.36 (1H, t, J = 7.9 Hz), 7.53(2H, d,
J = 8.5 Hz), 7.77 (2H, d, J = 8.4 Hz), 7.90 (1H, d,
J = 7.8 Hz), 7.97 (1H, d, J = 7.9 Hz), 12.49 (1H, br).
API-ESMS:405 (M++H). HRMS (EI) 405.1485 [calcd
for C23H21ClN4 405.1482].
5.1.3.3. 8-Chloro-2-{3-[4-(4-fluorophenyl)-3,6-dihydro-
1(2H)-pyridinyl]propyl} -4(3H)-quinazolinone (1d). 1H
NMR (200 MHz, DMSO-d6): d 1.8–2.1 (2H, m), 2.2–
2.8 (8H, m), 3.3 (2H, br s), 6.03 (1H, m), 7.0–7.2 (2H,
m), 7.3–7.6 (2H, m), 7.42 (1H, t, J = 8.0 Hz), 7.90 (1H,
dd, J = 8.0, 1.4 Hz), 7.99 (1H, dd, J = 8.0, 1.4 Hz).
API-ESMS: 398 (M++H). Anal. Calcd for
C22H21FClN3O + 1.19H2O: C, 63.02; H, 5.42; N,
10.02. Found: C, 63.01; H, 5.10; N, 9.92.
2. Under
(0.73 mL, 5.26 mmol) was added to a solution of
2-[(4-bromobutanoyl)amino]benzamide (500 mg,
a
nitrogen atmosphere, triethylamine
1.75 mmol) and 4-phenyl-1,2,3,6-tetrahydropyridine
hydrochloride (412 mg, 2.10 mmol) in N,N-dimethyl-
formamide (5 mL) at 0 °C. The mixture was allowed
to warm to room temperature and stirred for 24 h.
The reaction was quenched with water, and the prod-
uct was extracted with chloroform. The organic layer
was washed with water and dried over sodium sulfate.
Purification over silica gel chromatography gave 2-
{[4-(4-phenyl-3,6-dihydro-1(2H)-pyridinyl)butanoyl]-
amino}benzamide (477 mg, 74.8%) as a pale-yellow
powder. 1H NMR (200 MHz, CDCl3) d 2.01 (2H,
quint., J = 7.3 Hz), 2.41–2.56 (4H, m), 2.72 (2H, t,
J = 5.4 Hz), 3.76 (2H, d, J = 5.7 Hz), 5.4–6.3 (2H,
br), 6.05 (1H, m), 7.05 (1H, t, J = 7.0 Hz), 7.21–7.37
(6H, m), 7.45–7.51 (2H, m), 8.64 (1H, d,
J = 8.6 Hz). API-ESMS: 364 (M++H).
5.1.3.4. 8-Methyl-2-{3-[4-(4-fluorophenyl)-3,6-dihydro-
1(2H)-pyridinyl]propyl}- 4(3H)- quinazolinone (1e). 1H
NMR (200 MHz, DMSO-d6): d 1.8–2.0 (2H, m), 2.59
(3H, s), 2.4–2.8 (8H, m), 3.0–3.2 (2H, m), 6.05 (1H,
m), 7.0–7.5 (5H, m), 7.80 (1H, dd, J = 7.6, 1.4 Hz),
7.95 (1H, dd, J = 7.6, 1.4 Hz). API-ESMS: 378
(M++H). HRMS (EI) 378.1979 [calcd for C23H25FN3O1
378.1982].
5.1.3.5.
5-Chloro-2-{3-[4-phenyl-3,6-dihydro-1(2H)-
pyridinyl]propyl}-4(3H)-quinazolinone (1f). 1H NMR
(200 MHz, DMSO-d6): d 1.8–2.0 (2H, m), 2.2–2.8 (8H,
m), 2.9–3.1 (2H, m), 6.05 (1H, m), 7.1–7.8 (7H, m),
8.01 (1H, d, 1.6 Hz). API-ESMS: 380 (M++H). Anal.
Calcd for C22H22ClN3O: C, 69.56; H, 5.84; N, 11.06.
Found: C, 69.54; H, 5.88; N, 11.03.
3. 2-{[4-(4-Phenyl-3,6-dihydro-1(2H)-pyridinyl)butano-
yl]amino}benzamide (475 mg, 1.31 mmol) was dis-
solved in dioxane (5 mL). An aqueous solution of
sodium hydroxide (1 M, 3.92 mL) was added to the
solution at room temperature, and the mixture was
stirred at that temperature for 15 h. The organic
materials were extracted with chloroform, and the
organic layer was washed with water and dried over
sodium sulfate. Recrystallization of the crude product
from chloroform–methanol gave 2-{3-[4-phenyl-3,
6-dihydro-1(2H)-pyridinyl]propyl}-4(3H)-quinazoli-
none 1a (329 mg, 72.9%). 1H NMR (200 MHz,
CDCl3) d 2.05 (2H, quint., J = 6.0 Hz), 2.66 (2H, t,
J = 6.0 Hz), 2.81–2.94 (4H, m), 3.31 (2H, d,
J = 3.2 Hz), 6.12 (1H, t, J = 2.9 Hz), 7.21–7.49 (7H,
m), 7.61–7.72 (2H, m), 8.23 (1H, d, J = 6.6 Hz).
API-ESMS: 346 (M++H). Anal. Calcd for
C22H23N3O: C, 76.49; H, 6.71; N, 12.16. Found: C,
76.15; H, 6.75; N, 12.12.
5.1.3.6.
2-[3-(4-(4-Chlorophenyl)-1-piperazinyl)pro-
pyl]-4(3H)-quinazolinone (1g). 1H NMR (200 MHz,
DMSO-d6): d 1.7–2.0 (2H, m), 2.3–2.7 (8H, m), 3.1–3.3
(4H, m), 6.88 (2H, d, J = 9.2 Hz), 7.35 (2H, d,
J = 9.2 Hz), 7.38 (1H, t, J = 9.0 Hz), 7.71 (1H, d,
J = 9.0 Hz), 7.78 (1H, td, J = 9.0, 1.2 Hz), 8.05 (1H,
dd J = 9.0, 1.2 Hz). API-ESMS: 383 (M++H). Anal.
Calcd for C21H23ClN4O + 1.61 H2O: C, 61.24; H, 6.42;
N, 13.80. Found: C, 61.23; H, 6.22; N, 14.04.
5.1.3.7.
Methyl-3-fluoro-2-{(ethoxycarbonyl)[4-(4-
phenyl-3,6-dihydro-1(2H)-pyridinyl) butyl]amino}benzo-
ate (5b). Sodium iodide (225 mg, 1.50 mmol), potassium
carbonate (622 mg, 4.5 mmol), and 4-phenyl-1,2,3,6-tet-
rahydropyridine hydrochloride (323 mg, 1.65 mmol)
were added to a solution of 4b (543 mg, 1.50 mmol) in
N,N-dimethylformamide (10 mL) at room temperature.
The mixture was heated at 80 °C for 3 h. The reaction
was quenched by addition of saturated aqueous
5.1.3.1. 8-Chloro-2-{3-[4-phenyl-3,6-dihydro-1(2H)-
pyridinyl]propyl}-4(3H)-quinazolinone (1b): 1b–g were
prepared in a similar method. 1H NMR (200 MHz,

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1387
ammonium chloride, and the organic materials were
extracted with ethyl acetate (647 mg, 89.0%). API-
ESMS: 455 (M++H).
chromatographed on silica gel using toluene as an eluent
to give an oil, which was dissolved in ethanol and re-
fluxed in the presence of hydrazine monohydrate
(1.4 g) for 1 h. The mixture was concentrated, then
dichloromethane and water were added, and the organic
layer was separated. The aqueous layer was further
extracted with dichloromethane, then the combined
extracts were dried over magnesium sulfate and concen-
trated. The residue was triturated with dichloromethane
and diisopropylether, then the resulting powder was col-
lected, washed with diisopropylether, and dried in vacuo
to give the objective compound (2.78 g, 51%).
5.1.3.8.
5-Fluoro-1-(4-(4-phenyl-3,6-dihydro-1(2H)-
pyridinyl)butyl)-2,4(1H,3H)-quinazolinedione (2b). An
aqueous solution of potassium hydroxide (2 M,
5.3 mL) was added to a solution of 5b (647 mg,
1.33 mmol) in methanol. The mixture was heated at re-
flux for 14 h. Cooled to room temperature, methanol
was removed in vacuo. Neutralized with an aqueous
solution of hydrogen chloride, the organic materials
were extracted with ethyl acetate. The crude carboxylic
acid derivative was used for next step without any puri-
1H NMR (DMSO-d6, d): d 1.50–2.00 (4H, m), 2.94 (2H,
t, J = 7.2 Hz), 3.49 (2H, t, J = 6.1 Hz), 4.45 (2H, s),
7.10–7.50 (5H, m), 7.70–8.20 (3H, m), 8.26 (1H, dd,
J = 1.9, 7.1 Hz), 12.45 (1H, br s). API-ESMS: 309
(M++1).
fication.
N,N0-diisopropylcarbodiimide
(0.418 mL,
2.67 mmol) was added to a solution of carboxylic acid
derivative and 1H-1,2,3-benzotriazol-1-ol monohydrate
(361 mg, 2.67 mmol) in N,N-dimethylformamide
(10 mL) at room temperature, and the mixture was stir-
red for 2 h. This mixture was dropped into an excess
amount of aqueous solution of ammonia (28%,
8.1 mL), and it was stirred for 1 h at room temperature.
The crude product, which the residue was dissolved in
ethanol (30 mL). Potassium hydroxide (powder,
180 mg) was added to it, and the mixture was heated
at reflux for 2h. Compound 2b was obtained by treating
the crude product with methanol (165 mg, 30.2%) as a
white powder.
5.1.3.12. 4-{4-[4-(4-Fluorophenyl)-3,6-dihydro-1(2H)-
pyridinyl]butyl}-1(2H)-phthalazinone (9b). To a slurry
of
4-[4-(benzyloxy)butyl]-1(2H)-phthalazinone
in
dichloromethane (5 mL) was added dropwise 1 M bor-
on tribromide in dichloromethane (0.97 mL), and the
mixture was stirred at room temperature for 2 h. The
reaction was quenched with water and extracted with
dichloromethane twice. The combined extracts were
dried over magnesium sulfate and concentrated. The
residue was triturated with diisopropylether, and the
resulting powder was collected, washed with diisoprop-
ylether, and dried in vacuo. A mixture of 4-(4-Bromo-
butyl)-1(2H)-phthalazinone (100 mg), 4-fluorophenyl-1,
2,5,6-tetrahydropyridine hydrochloride (91 mg), and
triethylamine (0.149 mL) in N,N-dimethylformamide
(5 mL) was stirred at room temperature overnight.
The mixture was diluted with water and extracted with
ethyl acetate twice. The combined extracts were washed
with water three times, dried over magnesium sulfate,
and concentrated. The residue was purified by prepara-
tive thin layer chromatography (10% methanol in
dichloromethane) to give 9b (70 mg, 30%) as a colorless
powder.
1H NMR (200 MHz, DMSO-d6): d 1.5–1.8 (4H, m), 2.3–
2.7 (6H, m), 3.08 (2H, m), 4.04 (2H, m), 6.17 (1H, br s),
7.00 (1H, dd, J = 8.2, 12.0 Hz), 7.1–7.8 (7H, m). API-
ESMS:
394
(M++H).
Anal.
Calcd
for
C23H24FN3O + 0.13H2O: C, 69.80; H, 6.18; N, 10.62.
Found: C, 69.78; H, 6.18; N, 10.54.
5.1.3.9. 1-(4-(4-Phenyl-3,6-dihydro-1(2H)-pyridi-
nyl)butyl)-2,4(1H,3H)-quinazolinedione (2a): Compound
1
2a was obtained in a similar method as 2b. H NMR
(200 MHz, DMSO-d6): d 1.5–1.8 (4H, m), 2.3–2.7 (6H,
m), 3.08 (2H, m), 4.04 (2H, m), 6.20(1H, br s), 7.1–7.5
(7H, m), 7.74 (1H, m), 0.04 (1H, d, J = 8.2 Hz). API-
ESMS: 376 (M++H). Anal. Calcd for C23H25N3O2-
HCl: C, 67.06; H, 6.36; N, 10.20. Found: C, 67.10; H,
6.46; N, 10.11.
1H NMR (DMSO-d6): d 1.40–2.00 (4H, m), 2.30–3.30
(10H, m), 6.12 (1H, s), 7.00–7.60 (5H, m), 7.70–8.00
(2H, m), 8.04 (1H, d, J = 7.6 Hz), 8.26 (1H, d,
J = 7.6 Hz), 12.44 (1H, br s). API-ESMS: 378 (M++1).
Anal. Calcd for C23H24FN3O + 0.30H2O: C, 72.12; H,
6.48; N, 10.98. Found: C, 72.16; H, 6.61; N, 10.77.
5.1.3.10.
5-Fluoro-1-1-[4-(4-phenyl-3-cyclohexen-1-
yl)butyl]-2,4(1H,3H)-quinazolinedione (2c). 11H NMR
(200 MHz, DMSO-d6): d 1.1–1.7 (8H, m), 1.7–1.9 (2H,
m), 2.1–2.5 (3H, m), 4.04 (2H, t, J = 7.3 Hz), 6.12 (1H,
br s), 7.03 (1H, dd, J = 11.2, 8.2 Hz), 7.1–7.5 (6H, m),
7.73 (1H, dt, J = 8.4, 6.0 Hz), 11.51 (1H, br s). API-
ESMS: 393 (M++H). HRMS (EI) 393.1975 [calcd for
C24H26FN2O2 393.1978].
5.1.3.13.
4-[4-(4-Phenyl-3,6-dihydro-1(2H)-pyridi-
nyl)butyl]-1(2H)-phthalazinone (9a). Compound 9a was
obtained by the same method as 9b 1H NMR
(DMSO-d6): d 1.10–1.90 (4H, m), 2.30–3.00(8H, m),
3.07 (2H, d, J = 2.8 Hz), 6.15(1H, s), 7.10–8.40 (9H,
m), 12.45 (1H, br s). API-ESMS: 360 (M++1). HRMS
(EI) 360.2072 [calcd for C23H26N3O1 360.2076].
5.1.3.11. 4-[4-(Benzyloxy)butyl]-1(2H)-phthalazinone
(8). A slurry of 4-benzyloxybutanal, (3-oxo-1,3-dihy-
dro-2-benzofuran-1-yl)(triphenyl)phosphonium
bro-
mide (560 mg), and triethylamine (7.39 mL) in
tetrahydrofuran (50 mL) was stirred at room tempera-
ture overnight. The resulting precipitates were removed
by filtration and washed with ethyl acetate, and then the
combined filtrate was concentrated. The residue was
5.1.3.14. Ethyl 2-bromo-4-(3-hydroxypropyl)phenyl-
carbamate (11). Bromine (3.51 g) was added to a solu-
tion of ethyl 4-(3-hydroxypropyl)phenyl carbamate
(10) (4.46 g) and sodium acetate (3.28 g) in acetic acid
(50 mL). The mixture was stirred for 5 h and the solvent

1388
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
was evaporated. The residue was diluted with a mixture
of water and ethyl acetate and the aqueous layer was
separated. The organic layer was washed with saturated
aqueous sodium hydrogencarbonate, aqueous sodium
thiosulfate, brine, and dried over magnesium sulfate.
After evaporation of the solvent, the residue was puri-
fied by column chromatography on silica-gel eluting
with hexane–ethyl acetate to afford 11 (5.53 g, 92%).
5.1.3.17. 2-{3-[4-Phenyl-3,6-dihydropyridin-1(2H)-
yl]propyl}phenanthridin-6(5H)-one (14). 1,2,3,6-Tetrahy-
dro-4-phenylpyridine hydrochloride (152 mg) was add-
solution of 13 (150 mg) in N,N-
ed to
a
dimethylformamide (3 mL) at room temperature. The
mixture was cooled in an ice bath and triethylamine
(0.66 mL) was added. The whole mixture was stirred
for 1 h with ice cooling and stirred overnight at ambi-
ent temperature. The mixture was poured into a mix-
ture of water and ethyl acetate and the aqueous layer
was separated. The organic layer was washed with
brine and dried over magnesium sulfate. After evapo-
ration of the solvent, the residue was purified by col-
umn chromatography on silica gel eluting with
dichloromethane–acetone and chloroform–methanol
to afford 14 (75 mg, 41%).
1H NMR (DMSO-d6): d 1.32 (3H, t, J = 7.1 Hz), 1.8–2.0
(2H, m), 2.65 (2H, t, J = 7.2 Hz), 3.6–3.7 (2H, m), 4.23
(2H, q, J = 7.1 Hz), 7.02 (1H, br s), 7.13 (1H, dd,
J = 8.4, 2.0 Hz), 7.35 (1H, d, J = 2.0 Hz), 8.01 (1H, d,
J = 8.4 Hz). API-ESMS: 303 (M++1).
5.1.3.15. Ethyl 5-(3-bromopropyl)-1,10-biphenyl-2-yl-
carbamate (12). Under a nitrogen atmosphere, phen-
ylboronic acid (437 mg), 2 M aqueous sodium
dicarbonate (4.5 mL), tetrakis(triphenylphosphine)pal-
ladium (0) (173 mg) were added to a solution of ethyl
11 (1.1 g) in dimethoxyethane (13.5 mL) at room tem-
perature. The mixture was refluxed for 5 h, cooled to
room temperature, and poured into a mixture of water
and ethyl acetate. The aqueous layer was separated
and the organic layer was washed with brine, dried over
magnesium sulfate. After evaporation of the solvent, the
residue was purified by column chromatography on sil-
ica-gel eluting with toluene to afford 13 (1.1 g, 99%).
1H NMR (DMSO-d6): d 1.8–2.0 (2H, m), 2.4–2.5 (4H,
m), 2.6–2.8 (4H, m), 3.08 (2H, d, J = 2.8 Hz), 6.15
(1H, s), 7.1–7.5 (7H, m), 7.63 (1H, t, J = 7.2 Hz), 7.84
(1H, t, J = 7.2 Hz), 8.23 (1H, s), 8.32 (1H, d,
J = 8.0 Hz), 8.53 (1H, d, J = 8.0 Hz), 11.61 (1H, s).
API-ESMS:
395
(M++1).
Anal.
Calcd
for
C27H26N2O + 0.41H2O: C, 80.69; H, 6.73; N, 6.97.
Found: C, 80.698; H, 6.46; N, 6.88.
5.1.3.18. 3-Phenylquinoxaline-5-carboxyamide (3a).
To a suspension of 2,3-diaminobenzamide dihydrochlo-
ride (224 mg, 1 mmol) in methanol (10 mL) were added
triethylamine (1.4 mL, 10 mmol) and phenacyl bromide
(343 mg, 1.5 mmol) at room temperature. The mixture
was stirred overnight and poured into a mixture of water
and chloroform. The aqueous layer was separated and
the organic layer was washed with brine, dried over
magnesium sulfate. After evaporation of the solvent,
the residue was purified by column chromatography
on silica-gel eluting with dichloromethane–acetone to
afford 3a (128 mg, 35%).
1H NMR (CDCl3): d 1.24 (3H, t, J = 7.1 Hz), 2.0–2.4
(2H, m), 2.76 (2H, t, J = 7.0 Hz), 3.40 (2H, t,
J = 6.6 Hz), 4.16 (2H, r, J = 7.1 Hz), 6.55 (1H, br s),
7.0–7.5 (7H, m), 8.02 (1H, d, J = 8.3 Hz). Mass 384.1,
386.1 API-ESMS: 384 (M++Na).
5.1.3.16. 2-(3-Bromopropyl)phenanthridin-6(5H)-one
(13). Under a nitrogen atmosphere, phosphorus (V)
oxide (511 mg) was added to a solution of 12 (435 mg)
in phosphorus oxychloride (3 mL) at room temperature.
The mixture was refluxed for 2 h and the solvent was
evaporated. The residue was poured into a mixture of
ice water and ethyl acetate, and the solution was
brought to pH 9 with 10% aqueous potassium carbon-
ate. The aqueous layer was separated and the organic
layer was washed with brine, dried over magnesium sul-
fate. The solvent was evaporated in vacuo and the resi-
due was dissolved in a mixture of dioxane (6 mL) and
4 N aqueous hydrogen chloride (3 mL). The solution
was refluxed for 30 min, cooled to room temperature,
and poured into a mixture of water and ethyl acetate.
The mixture was neutralized with 10% aqueous potassi-
um carbonate and the aqueous layer was separated. The
organic layer was washed with brine, dried over magne-
sium sulfate. After evaporation of the solvent, the resi-
due was purified by column chromatography on silica-
gel eluting with dichloromethane–acetone to afford 13
(280 mg, 74%).
1H NMR (200 MHz, DMSO-d6): d 7.5–8.2 (8H, m),
9.37 (1H, m), 9.68 (1H, s). API-ESMS: 250 (M++1).
Anal. Calcd for C15H11N3O + 0.14H2O: C, 71.55;
H, 4.52; N, 16.69. Found: C, 71.56; H, 4.52; N,
16.69.
5.1.3.19. 3-(4-Chlorophenyl)quinxalin-5-carboxyamide
(3b). Compounds 3b–f were prepared in a similar meth-
1
od H NMR (200 MHz, DMSO-d6): d d 7.73 (2H, dd,
J = 8.6, 2.2 Hz), 7.95–8.00 (2H, m), 8.25–8.45 (4H, m),
9.14 (1H, br s), 9.67 (1H, s). API-ESMS: 284 (M++1).
HRMS (EI) 284.0584 [calcd for C15H11ClN3O
284.0591].
5.1.3.20.
3-(4-Cyanophenyl)quinoxaline-5-carboxy-
1
amide (3c). H NMR (200 MHz, DMSO-d6): d 7.8–8.6
(7H, m), 9.04 (1H, m), 9.72 (1H, s). API-ESMS: 275
(M++1). HRMS (EI) 275.0933 [calcd for C16H11lN4O
275.0937].
1H NMR (DMSO-d6): d 2.0–2.3 (2H, m), 2.83 (2H, t,
J = 7.0 Hz), 3.5–3.7 (2H, m), 7.25–7.4 (2H, m), 7.63
(1H, t, J = 7.1 Hz), 7.85 (1H, dt, J = 7.2, 1.5 Hz), 8.23
(1H, s), 8.32 (1H, dt, J = 7.9, 1.2 Hz), 8.52 (1H, d,
J = 8.1 Hz), 11.62 (1H, s). API-ESMS: 316 (M++1).
5.1.3.21.
3-(4-Trifluoromethylphenyl)quinoxaline-5-
1
carboxyamide (3d). H NMR (200 MHz, DMSO-d6): d
7.9–8.05 (4H, m), 8.30 (1H, dd, J = 8.3, 1.5 Hz), 8.45–
8.60 (3H, m), 8.59 (1H, br s), 9.72 (1H, s). API-ESMS:

J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
1389
340 (M++1). Anal. Calcd for C16H10F3N3O + 0.54H2O:
C, 58.77; H, 3.42; N, 12.85. Found: C, 58.76; H, 3.44; N,
12.73.
DNA damage-dependent poly(ADP-ribose)polymerase.
J. Biol. Chem. 1999, 274, 17860–17868.
5. (a) Ruf, A.; Murcia, G.; Schulz, G. E. Inhibitor and
NAD+ binding to poly(ADP-ribose) polymerase as
derived from crystal structures and homology modeling.
Biochemistry 1998, 37, 3893–3900; (b) Ruf, A.; Murcia, J.
M.; Murcia, G. M.; Schulz, G. E. Structure of the catalytic
fragment of poly(ADP-ribose) polymerase from chicken.
Proc. Natl. Acad. Aci. U.S.A. 1996, 93, 7481–7485.
6. Oliver, A. W.; Ame, J. C.; Roe, S. M.; Good, V.;
DeMurcia, G.; Pearl, L. H. Crystal structure of the
catalytic fragment of murine poly(ADP-ribose)polymer-
ase. Nucleic Acids Res. 2004, 32, 456–464.
5.1.3.22.
3-(4-Methoxyphenyl)quinoxaline-5-carb-
1
oxyamide (3e). H NMR (200 MHz, DMSO-d6): d 3.87
and 3.88 (total 3H, s), 7.15–7.22 (H, m), 7.85–7.95
(1H, m), 7.98 (1H, br s), 8.23–8.52 (4H, m), 9.44 (1H,
br s), 9.63 and 9.64 (total 1H, s). API-ESMS: 280
(M++1). Anal. Calcd for C16H13N3O2 + 0.04H2O: C,
68.63; H, 4.71; N, 15.01. Found: C, 68.61; H, 4.59; N,
14.84.
7. De Murcia, G.; Schreiber, V.; Molinate, M.; Saulier, B.;
Poch, O.; Masson, M.; Niedergang, C.; Menissier, J.
Structure and function of poly(ADP-ribose) polymerase.
Mol. Cell Biochem. 1994, 138, 15–24.
8. Kinoshita, T.; Nakanishi, I.; Warizaya, M.; Iwashita, A.;
Kido, Y.; Hattori, K.; Fujii, T. Inhibitor-induced struc-
tural change of the active site of human poly(ADP-ribose)
polymerase. FEBS Lett. 2004, 556, 43–46.
5.1.3.23.
3-(4-Aminophenyl)quinoxaline-5-carboxy-
1
amide (3f). H NMR (200 MHz, DMSO-d6): d 6.6–6.8
(2H, m), 7.7–8.4 (5H, m), 9.52 (1H, s). API-ESMS:
265 (M++1). HRMS (EI) 265.1087 [calcd for
C15H13lN4O 265.1089].
5.2. PARP assay
9. Iwashita, A.; Yamazaki, S.; Mihara, K.; Hattori, K.;
Yamamoto, H.; Ishida, J.; Matsuoka, N.; Mutoh, S.
Neuroprotective effects of a novel poly(ADP-ribose)poly-
merase-1 inhibitor, 2-{3-[4-(4-chlorophenyl)-1-piperazi-
nyl]propyl}-4(3H)- quinazolinone (FR255595) in an
in vitro model of cell death and in mouse 1-methyl-4-
phenyl-1236- tetrahydropyridine model of ParkinsonÕs
disease. J. Pharmacol. Exp. Ther. 2004, 309, 1067–1078.
10. Recent reviews on PARP inhibitors: (a) Peukert, S.;
Schwahn, U. New inhibitors of poly(ADP-ribose)poly-
merase (PARP). Expert Opin. Ther. Pat. 2004, 14, 1531–
To assess the inhibitory activity of novel inhibitors, the
PARP enzyme assay was carried out in a final volume
of 100 ll consisting of 50 mM Tris–HCl (pH 8.0),
25 mM MgCl2, 1 mM dithiothreitol, 10 lg activated
salmon testes DNA, 0.1 lCi of [adenylate-32P]-NAD,
0.2 units of recombinant human PARP-1, recombinant
mouse PARP-1 or PARP-2, and various concentrations
of PARP inhibitors. The reaction mixture was incubated
at room temperature (23 °C) for 15 min, and the reaction
was terminated by adding 200 ll of ice-cold 20% trichlo-
roacetic acid (TCA) and incubated at 4 °C for 10 min.
The precipitate was transferred onto GF/B filter (Pack-
ard Unifilter-GF/B) and washed three times with 10%
TCA solution and 70% ethanol. After the filter was
dried, the radioactivity was determined by liquid scintil-
lation counting. IC50 values were calculated from the
concentration dependence of the inhibition curves by
using computer-assisted non-linear regression analyses.
´
1551; (b) Southan, G. J.; Szabo, C. Poly(ADP-ribose)poly-
merase inhibitors. Curr. Med. Chem. 2003, 10, 321–340; (c)
Cosi, C. New inhibitors of poly(ADP-ribose)polymerase
and their potential therapeutic targets. Expert Opin Ther.
Pat. 2002, 12, 1047–1071.
11. (a) Hattori, K.; Kido, Y.; Yamamoto, H.; Ishida, J.;
Kamijo, K.; Murano, K.; Ohkubo, M.; Kinoshita, T.;
Iwashita, A.; Mihara, K.; Yamazaki, S.; Matsuoka, N.;
Teramura, Y.; Miyake, H. Rational approaches to
discovery of orally active and brain-penetrable quinazoli-
none inhibitors of poly(ADP-ribose)polymerase. J. Med.
Chem. 2004, 47, 4151–4154; (b) Iwashita, A.; Mihara, K.;
Yamazaki, S.; Matsuura, S.; Ishida, J.; Yamamoto, H.;
Hattori, K.; Matsuoka, N.; Mutoh, S. A new poly(ADP-
ribose) polymerase inhibitor, FR261529 [2-(4-chlorophe-
nyl)-5-quinoxalinecarboxamide], ameliorates metham-
phetamine-induced dopaminergic neurotoxicity in mice.
J. Pharmacol. Exp. Ther. 2004, 310, 1114–1124; (c)
Iwashita, A.; Tojo, N.; Matsuura, S.; Yamazaki, S.;
Kamijo, K.; Ishida, J.; Yamamoto, H.; Hattori, K.;
Matsuoka, N.; Mutoh, S. A. novel and potent poly(-
ADP-ribose)polymerase-1 inhibitor, FR247304 (5-chloro-
2-[3-(4-phenyl-36-dihydro-1(2H)-pyridinyl)propyl]-4(3H)-
quinazolinone attenuates neuronal damage in in vivo and
in vivo models of cerebral ischemia. J. Pharmacol. Exp.
Ther. 2004, 310, 425–436(d) Ref. 9.
12. Some quinazolinone-based PARP inhibitors have been
identified. (a) Griffin, R. J.; Srinivasan, S.; Bowman, K.;
Calvert, A. H.; Curtin, N. J.; Newell, D. R.; Pemberton, L.
C.; Golding, B. T. Resistance-modifying agents. 5. Syn-
thesis and biological properties of quinazolinone inhibi-
tors of the DNA repair enzyme poly (ADP-
ribose)polymerase (PARP). J. Med. Chem. 1998, 41,
5247–5256; (b) Banasik, M.; Komura, H.; Shimoyama,
M.; Ueda, K. Specific inhibitors of poly(ADP-ribose)syn-
thetase and mono(ADP-ribosyl)transferase. J. Biol. Chem.
1992, 267, 1569–1575.
Acknowledgment
We express our thanks to Dr. David Barrett for his crit-
ical reading of the manuscript.
References and notes
1. Yu, S. W.; Wang, H.; Poitras, M. F.; Coombs, C.; Bowers,
W. J.; Federoff, H. J.; Poirier, G. G.; Dawson, T. M.;
Dawson, V. L. Mediation of poly(ADP-ribosepolymerase-
1 dependent cell death by apoptosis-inducing factor.
Science 2002, 297, 259–263.
2. Szabo, C.; Dawson, V. L. Role of poly(ADP-ribose)syn-
thetase in inflammation and ischemia–reperfusion. Trends
Pharmacol. Sci. 1998, 19, 287–298.
3. Shieh, W. M.; Ame, J. C.; Wilson, M. V.; Wang, Z. Q.;
Koh, D. W. Poly(ADP-ribose)polymerase null mouse cells
synthesize ADP-ribose polymerase. J. Biol. Chem. 1998,
273, 30069–30072.
4. Ame, J. C.; Rolli, V.; Schreiber, V.; Niedergang, C.;
Apiou, F.; Decker, P.; Muller, S.; Hoger, T.; Menissier-de
Murcia, J.; De Murcia, G. PARP-2, a novel mammalian

1390
J. Ishida et al. / Bioorg. Med. Chem. 14 (2006) 1378–1390
13. Iwashita, A.; Hattori, K.; Yamamoto, H.; Ishida, J.; Kido,
Y.; Kamijo, K.; Murano, K.; Miyake, H.; Kinoshita, T.;
Warizaya, M.; Ohkubo, M.; Matsuoka, N.; Mutoh, S.
Discovery of quinazolinone and quinoxaline derivatives as
potent and selective poly(ADP-ribose)polymerase-1/2
inhibitors. FEBS Lett. 2005, 579, 1389–1393.
16. Some phthalazin-1(2H)-one derivatives and phenanthrid-
inone derivatives, PARP inhibitors have been identified:
see in Ref. 10b.
17. (a) Canan Koch, S. S.; Thoresen, L. H.; Tikhe, J. G.;
Maegley, K. A.; Almassy, R. J.; Li, J.; Yu, X.-H.; Zook,
S. E.; Kumpf, R. A.; Zhang, C.; Boritzki, T. J.;
Mansour, R. N.; Zhang, K. E.; Ekker, A.; Calabrese,
C. R.; Curtin, N. J.; Kyle, S.; Thomas, H. D.; Wang,
L.-Z.; Calvert, A. H.; Golding, B. T.; Griffin, R. J.;
Newell, D. R.; Webber, S.; Hostomsky, Z. Novel
tricyclic poly(ADP-ribose)polymerase-1 inhibitors with
potent anticancer chemopotentiating activity: design
synthesis and X-ray cocrystal structure. J. Med. Chem.
2002, 45, 4961–4974; (b) White, A. W.; Almassy, R.;
Calvert, A. H.; Curtin, N. J.; Griffin, R. J.; Hostomsky,
Z.; Maegley, K.; Newell, D. R.; Srinivasan, S.; Golding,
B. T. Resistance-modifying agents. 9. Synthesis and
biological properties of benzimidazole inhibitors of the
DNA repair enzyme poly(ADP-ribose) polymerase. J.
Med. Chem. 2000, 43, 4084–4097.
14. (a) Skalitzky, D. J.; Marakovits, J. T.; Maegley, K. A.;
Ekker, A.; Yu, X.-H.; Hostomsky, Z.; Webber, S. E.;
Eastman, B. W.; Almassy, R.; Li, J. Tricyclic benzimidaz-
oles as potent poly(ADP-ribose)polymerase-1 inhibitors.
J. Med. Chem. 2003, 46, 210–213; (b) Ferraris, D.; Ficco,
R. P.; Dain, D.; Ginski, M.; Lautar, S.; Lee-Wisdom, K.;
Liang, S.; Lin, Q.; Lu, M. X.-C.; Morgan, L.; Thomas, B.;
Williams, L. R.; Zhang, J.; Zhou, Y.; Kalish, V. J. Design
and synthesis of poly(ADP-ribose)polymerase-1 (PARP-1)
inhibitors. Part 4: biological evaluation of imi-
dazobenzodiazepines as potent PARP-1 inhibitors for
treatment of ischemic injuries. Bioorg. Med. Chem. 2003,
11, 3695–3707; (c) Costantino, G.; Macchiarulo, A.;
Camaioni, E.; Pellicciari, R. Modeling of poly(ADP-
ribose)polymerase (PARP) inhibitors. Docking of ligands
and quantitative structure–activity relationship analysis. J.
Med. Chem. 2001, 44, 3786–3794.
18. In X-ray crystallography of 1e and 2b, the side chain of
Glu763 in PARP-1 was also restricted by a hydrogen bond
with neighbor amino acid.
15. Ogata, K.; Umeyama, H. An automatic homology mod-
eling method consisting of database searches and simu-
lated annealing. J. Mol. Graphics Mod. 2000, 18, 258–272.
19. In murine PARP-2, the replacement of Gln308 by Lys308
was observed. The side chain of Lys308 is also not
restricted; see Ref. 6.