5-benzamidoisoquinolin-1-ones and 5-(ω-Carboxyalkyl)isoquinolin-1- ones as isoform-selective inhibitors of poly(ADP-ribose) polymerase 2 (PARP-2)

Article abstract of DOI:10.1021/jm1010918

PARP-2 is a member of the poly(ADP-ribose) polymerase family, with some activities similar to those of PARP-1 but with other distinct roles. Two series of isoquinolin-1-ones were designed, synthesized, and evaluated as selective inhibitors of PARP-2, usin

Full text of DOI:10.1021/jm1010918

ARTICLE
pubs.acs.org/jmc
5-Benzamidoisoquinolin-1-ones and 5-(ω-Carboxyalkyl)isoquinolin-
1-ones as Isoform-Selective Inhibitors of Poly(ADP-ribose) Polymerase
2 (PARP-2)
||
Peter T. Sunderland,† Esther C. Y. Woon,†, Archana Dhami,† Aoife B. Bergin,†,^ Mary F. Mahon,‡
Pauline J. Wood,† Louise A. Jones,§ Sophie R. Tully,§ Matthew D. Lloyd,† Andrew S. Thompson,†
Hashim Javaid,§ Niall M. B. Martin,§ and Michael D. Threadgill*,†
†Medicinal Chemistry, Department of Pharmacy and Pharmacology, University of Bath, Claverton Down, Bath BA2 7AY, U.K.
‡Crystallography Unit, Department of Chemistry, University of Bath, Claverton Down, Bath BA2 7AY, U.K.
§KuDOS Pharmaceuticals Ltd., 410 Cambridge Science Park, Milton Road, Cambridge CB4 0PE, U.K.
S Supporting Information
b
ABSTRACT: PARP-2 is a member of the poly(ADP-ribose) polymerase family, with some
activities similar to those of PARP-1 but with other distinct roles. Two series of
isoquinolin-1-ones were designed, synthesized, and evaluated as selective inhibitors of
PARP-2, using the structures of the catalytic sites of the isoforms. A new efficient synthesis
of 5-aminoisoquinolin-1-one was developed, and acylation with acyl chlorides gave
5-acylaminoisoquinolin-1-ones. By examination of isoquinolin-1-ones with carboxylates
tethered to the 5-position, Heck coupling of 5-iodoisoquinolin-1-one furnished the
5-CHdCHCO2H compound for reduction to the 5-propanoic acid. Alkylation of
5-aminoisoquinolin-1-one under mildly basic conditions, followed by hydrolysis, gave 5-(carboxymethylamino)isoquinolin-1-
one, whereas it was alkylated at 2-N with methyl propenoate and strong base. Compounds were assayed in vitro for inhibition of
PARP-1 and PARP-2, using FlashPlate and solution-phase assays, respectively. The 5-benzamidoisoquinolin-1-ones were more
selective for inhibition of PARP-2, whereas the 5-(ω-carboxyalkyl)isoquinolin-1-ones were less so. 5-Benzamidoisoquinolin-1-one
is the most PARP-2-selective compound (IC50(PARP-1)/IC50(PARP-2) = 9.3) to date, in a comparative study.
’ INTRODUCTION
molecules downstream therefrom. Thus, inhibitors of the cata-
lytic activity of PARP-1 have shown interesting activity in vivo in
models of several disease states, including hemorrhagic shock,
myocardial infarction, stroke, and other ischemiaꢀreperfusion
injuries, as well as in inflammatory disorders and reproductive
health.15ꢀ20 Recently, inhibitors of PARP-1 have also been
shown to inhibit angiogenesis and metastasis in cancer.21ꢀ23
PARP-2 is a 62 kDa protein with three functional domains.24
The N-terminal DNA-binding domain lacks zinc fingers but is
rich in basic amino acids, which account for the binding to DNA.
The differences in structure of the DNA-binding domains of
PARPs 1 and 2 may also reflect the differences in the DNA
structures which the two enzymes recognize and the fact that
PARP-2 binds to single-strand breaks (SSBs) less efficiently than
does PARP-1. The automodification domain is also responsible
for the proteinꢀprotein interactions that PARP-2 shares with
various partners, including PARP-1, XRCC1, and DNA ligase 3.25
The catalytic domains of mouse and human PARP-2 are highly
conserved, and both show high homology with the human
PARP-1 catalytic domain (69% in the case of human PARP-2).24
Poly(ADP-ribosyl)ation of proteins, first reported in 1963,1
involves the transfer of multiple ADP-ribose units from substrate
NADþ onto glutamate residues of the target proteins, resulting in
the formation of polyanionic poly(ADP-ribose) (PAR) poly-
mers. There has been much research effort since to characterize
the structures and functions of the poly(ADP-ribose) poly-
merases (PARPs), the family of enzymes responsible for catalysis
of this reaction.2ꢀ5
PARP-1 is responsible for most of the poly(ADP-ribosyl)ation
activity in the cell and has a major role in regulating the repair of
damaged DNA. It is a 116 kDa protein with three major domains:
an N-terminal DNA-binding domain carrying three zinc fingers,6
a central automodification domain (containing also a nuclear
localization signal7 and sites for cleavage by caspase-38), and a
C-terminal NADþ-binding catalytic domain. The catalytic activ-
ity of the PARP-1 is activated by detection of a damaged site, and
PAR is built up on histone-1 and on PARP-1 itself. Inhibition of
PARP-1 therefore inhibits repair of damaged DNA, and several
inhibitors of this enzyme are currently in clinical trial as sensi-
tizers to the effects of cytotoxic DNA-damaging drugs in several
tumors and as single-agent therapies in BRCA-mutant breast
cancer.9ꢀ14 In addition to its role in regulating repair of damaged
DNA, PARP-1 also regulates NF-κB and processes and
Received: August 23, 2010
Published: March 18, 2011
r
2011 American Chemical Society
2049
dx.doi.org/10.1021/jm1010918 J. Med. Chem. 2011, 54, 2049–2059
|

Journal of Medicinal Chemistry
ARTICLE
Figure 1. Structures of reported potent non-isoform-selective inhibitors 1ꢀ7, PARP-1-selective inhibitor 8, and PARP-2-selective inhibitors 9ꢀ11.
This isoform accounts for approximately 5ꢀ10% of the total
cellular PARP activity.25 It can heterodimerize with PARP-1, and
it shares many functions and targets. However, PARP-2 has some
different protein targets and potential functions. Schreiber et al.25
showed that PARP-2ꢀ/ꢀ mouse embryonic fibroblasts have
delayed repair of alkylated bases in DNA following treatment
with N-methyl-N-nitrosourea (MNU) and suggest that the
PARP-1/PARP-2 heterodimer is important for efficient base-
excision repair. By contrast, Fisher et al.26 found that a reduction
in PARP-1 significantly reduced the rate of repair of single-strand
breaks in DNA but reduction in PARP-2 had only a minor effect.
PARP-2 interacts with TRF2,27 a telomere-binding regulatory
protein, and affects its ability to bind to DNA both through a
noncovalent interaction of PAR with the DNA-binding domain
of TRF2 and through a covalent modification at the dimerization
domain. PARP-2 may have a functional role in the maintenance
of telomeres, but further work using selective inhibitors of this
isoform is required to gain insights into the precise molecular
mechanisms at work. Antisense knockdown of PARP-2 in a
mouse model of colitis resulted in improvement in inflammation
and normalization of colonic function.28 Deletion of PARP-2 but
not of PARP-1 leads to a significant diminution in CD4þCD8þ
double-positive thymocytes,29 suggesting that PARP-2 has a role
in survival of T-cells during thymopoiesis and may regulate the
apoptosis of thymocytes. PARP-2 is widely expressed in the
seminiferous epithelium, in contrast to limitation of PARP-1 to
the peripheral cell layer,25,30 pointing to distinct roles in sperma-
togenesis and a more prominent role for PARP-2. PARP-2
knockout mice display lipodystrophy,31 probably due to regula-
tion of the expression of peroxisome proliferator-activated
receptor-γ by PARP-2. These roles for PARP-2, which are
distinct from those of the major isoform PARP-1,32 point to
the need for selective inhibitors initially as pharmacological tools
to help resolve and distinguish the roles of PARP-2 but also, in
the longer term, possibly as drugs.
Most attempts to identify PARP-2-selective inhibitors have
relied on screening libraries of inhibitors of PARP-1 against both
isoforms. Interestingly, most of the clinical candidate PARP-1
inhibitors and of the widely used pharmacological tool inhibitors
display little or no selectivity, and thus, any pharmacological
effect thereof must be regarded as the effect of pan-PARP
inhibition or, at minimum, of PARP-1 and PARP-2 together.
As shown in this paper, our lead PARP-1 inhibitor 5-aminoiso-
quinolin-1-one 1 (5-AIQ, Figure 1), which is potently active
in vivo,15,17ꢀ20,23 is devoid of isoform selectivity in vitro.
Similarly, two other standard PARP-1 inhibitors, DPQ 2 and
PJ34 3, are reported to have little isoform-selectivity between
PARP-1 and PARP-2.33 The latest clinical candidates veliparib
434 and olaparib 5,36 the biochemical tools BYK49187 6a,35
BYK236864 6b35 and the thiopyranopyrimidinone 737 also
inhibit the two isoforms potently but approximately equally.
However, in 2006, Ishida et al. reported the identification of a
series of quinazolin-4-ones as selective inhibitors of PARP-1, of
which 8 was the most selective, with IC50(PARP-1) = 13 nM and
IC50(PARP-2) = 500 nM (selectivity ratio of 39); minor modifica-
tions to the substitution on the fluorophenyl ring were tolerated
in the structureꢀactivity relationship (SAR), but the long
pendant 3-(4-aryl-1,2,5,6-tetrahydropyridin-1-yl)propyl group
at the 3-position of the quinazolinone was shown to be essential
in this series, as 5-chloro-2-methylquinazolin-4-one was non-
isoform-selective and was 92-fold less active against PARP-1 than
was 8.38 The same paper also reports that 9a was some 12-fold
more potent as an inhibitor of PARP-2 than it was of PARP-1;
other varied substituents in the 4-position of the phenyl were
tolerated with little loss of selectivity.38 The observed selectivities
were rationalized by examining X-ray structures of cocrystals of
2050
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
Scheme 1. New Efficient Synthesis of 1 and Synthesis of 5-(Acylamino)isoquinolin-1-ones 15aꢀna
a Reagents andconditions:(i) HNO3, H2SO4, 0ꢀC, 92%; (ii) AcOH, 100ꢀC, 82%; (iii) H2, Pd/C, EtOH, aq HCl, 66%; (iv) RCOCl, pyridine, 90 ꢀC, 86% (15a),
82% (15b), 71% (15c), 72% (15d), 68% (15e), 77% (15f), 81% (15g), 76% (15h), 63% (15i), 61% (15j), 51% (15k), 68% (15l), 68% (15m), 59% (15n).
PARP-1 with inhibitors and of homology models of PARP-2.
Separately, this group also reported that a 2-arylquinoxaline-5-
carboxamide 9b had 5-fold selectivity for inhibition of PARP-2.39
By far the most selective inhibitors of PARP-2 claimed to date
are a series of 5-benzoyloxyisoquinolin-1-ones, 5-phenacyl-
oxyisoquinolin-1-ones, and 3,4-dihydroisoquinolin-1-ones,
of which 10 and 11 appear to show 60-fold and 16-fold
selectivity, respectively.33 These compounds, which pick up
the usual PARP-binding contacts of the classical (3,4-dihydro)
isoquinolin-1-one pharmacophore, achieve their selectivity by
relatively weak micromolar binding to PARP-1 rather than great
potency against PARP-2. This poor binding to PARP-1 is
probably due to the 5-benzoyloxy and 5-phenacyloxy substitu-
ents being too large to be accommodated in the slightly smaller
hydrophobic binding pocket of PARP-1, compared to that of
PARP-2 (see below).
needed with requirements for high yield and avoidance of scale-
limiting steps such as column chromatography. Scheme 1 shows
our development of this important new route. In this route, the
5-nitrogen substituent is introduced by nitration. However, the
most nucleophilic position of isoquinolin-1-ones is 4-C,42 so
direct nitration of isoquinolin-1-one was not feasible. The
activating enamide function of isoquinolin-1-one was masked
as the iminochloride in 12. Commercially available 1-chloroiso-
quinoline 12 was nitrated selectively in the 5-position in 92%
yield. To achieve this regioselectivity, it was necessary to predis-
solve 12 in concentrated sulfuric acid to ensure that the hetero-
cyclic ring was fully protonated and deactivated before addition
of the nitrating reagents. The iminochloride in 13 was hydro-
lyzed by heating in acetic acid to provide 14 in high yield, and
catalytic hydrogenation of the nitro group, under conditions
previously reported,42 then furnished 1. This new sequence is
reliable, high-yielding (54% overall), and highly reproducible and
involves only two intermediate recrystallizations and no chro-
matography for purification, giving the potential for it to be used
on larger scales.
5-Benzamidoisoquinolin-1-one 15a was synthesized in good
yield by acylation of 1 with benzoyl chloride in hot pyridine
(Scheme 1); the relatively forcing conditions were required to
overcome the poor nucleophilicity of the exocyclic amine in 1
and to ensure good solubility. To explore the structureꢀactivity
relationships around this phenyl ring, a series of 5-(substituted
benzamido)isoquinolin-1-ones 15bꢀj were prepared similarly
from 1 and substituted benzoyl chlorides. Following a classical
medicinal chemical replacement of the benzene aromatic ring
with thiophene led to design of 5-(thiophen-2-ylcarboxami-
do)isoquinolin-1-one 15k, which was prepared from 1 with
thiophene-2-carbonyl chloride. The need for aromaticity was
tested by replacement with cyclohexyl in 15l and by 3D bulky
aliphatic groups in the pivalamido analogue 15m and the
adamantylcarboxamido compound 15n.
In the present work, we report on two different approaches to
the discovery of PARP-2-selective inhibitors, both based on the
isoquinolin-1-one core which gives good binding to both iso-
forms through the hydrogen-bonding network to the lactam
oxygen and NꢀH and through π-stacking of the aromatic ring to
the adjacent Tyr residues (Tyr907/Tyr449).
’ CHEMICAL SYNTHESIS
Noting the remarkable selectivity for inhibition of PARP-2 by
10 claimed by Pellicciari et al.,33 we sought to use this core to
develop further selective inhibitors of this isoform. Inhibitors
10 and 11 both contain O-aryl esters, which are likely to be labile
to hydrolysis in physiological media. To obviate this potential
problem, the ester was replaced by an amide in target compound
15a (Scheme 1); this amide should be isosteric but may restrict
rotation in this part of the molecule.
Synthesis of 15a and its analogues 15bꢀn required quantities
of 1. This compound is commercially available but is prohibi-
tively expensive for applications in synthesis. There are three
previous published syntheses of 1 (where the critical steps are a
Polonowski rearrangement 5-nitroisoquinoline N-oxide,40 re-
ductive cyclization of methyl 2-cyanomethyl-3-nitrobenzoate,41
and condensation of methyl 2-methyl-3-nitrobenzoate with
dimethylformamide dimethylacetal15), but all have severe limita-
tions. Thus, for large-scale preparation of 1, a new synthesis was
Compound 22 (Scheme 2) extends the structure of 15a by
introducing a methyl group at the 3-position of the isoquinolin-1-
one ring system. The synthesis of this target was analogous to
that of 15a, thus requiring 3-methylisoquinolin-1-one 21 as a
starting material. As for 1, this educt is commercially available but
prohibitively expensive (>£5000 gꢀ1). None of the synthetic
sequences to 1 could be adapted to the 3-methyl analogue 21;
2051
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
Scheme 2. Synthesis of 5-Benzamido-3-methylisoquinolin-1-one 22a
a Reagents and conditions: (i) pentane-2,4-dione, KOtBu, Cu, tBuOH, reflux, 23%; (ii) NH3, MeO(CH2)2OH, reflux, 68%; (iii) SnCl2, EtOH, 70ꢀC,
59%; (iv) PhCOCl, pyridine, 90 ꢀC, 72%.
reaction of methyl 2-methyl-3-nitrobenzoate with dimethylace-
tamide dimethylacetal gives 3-dimethylamino-1-methoxy-5-ni-
tronaphthalene rather than 19,43 and 1-chloro-3-methyl-5-
nitroisoquinoline is only available by a low-yielding Polonowski
rearrangement of 3-methyl-5-nitroisoquinoline N-oxide.44 A new
route was required, starting from a suitable 1,2,3-trisubstituted
benzene. The Hurtley reaction involves Cu-catalyzed displace-
ment of the bromine of o-bromobenzoic acids by the enolates of
β-diketones.45,46 However, we have previously shown that the
reaction fails with 3-bromo-2-nitrothiophene-4-carboxylic acid.47
Despite this adverse omen, 2-bromo-3-nitrobenzoic acid 16 did
react with the potassium enolate of pentane-2,4-dione in the
presence of Cu powder. Under the reaction conditions (boiling
tert-butanol), the immediate Hurtley coupling product 17 was
not observed; rather it was deacetylated to the monoketone 18,
which cyclized to give the isocoumarin 19, achieving several
synthetic steps in one pot. The isocoumarin 19 was converted to
the isoquinolin-1-one 20, and reduction of the nitro group with
tin(II) chloride furnished the required intermediate 21 for
acylation to provide the target amide 22.
In 2004, the crystal structure of the catalytic fragment of
murine PARP-2 was solved48 at 2.8 Å resolution, 7 years after the
data for the chicken PARP-1 catalytic fragment were made
available.49 The high degree of homology of the PARP catalytic
domain between different species means that PARP inhibitors
are unlikely to show a wide species difference in their binding of
inhibitors between chicken, mouse, and human PARPs, for a
specific isoform. The binding of inhibitors to PARP-1 was
deduced by Ruf et al.50 to be due to hydrogen bonding
from the carbonyl of lactams/amides in the inhibitors to
Gly863-NH and Ser904-OH and from the lactam/amide NꢀH
to Gly863-O and hydrophobic or π-stacking interactions with
Tyr907. PARP-2 is a much smaller protein than is PARP-1, but the
hydrogen-bonding and π-stacking motifs are retained in the
nicotinamide-binding site, with Gly405 (PARP-2) corresponding
to Gly863 (PARP-1), Ser446 (PARP-2) corresponding to Ser904
(PARP-1), and Tyr449 (PARP-2) corresponding to Tyr907
(PARP-1). Our inspection of the binding sites of the isoforms
noted that Gln763 (neutral, polar) in PARP-1 is replaced by
Lys308 (basic, polar); these residues are located slightly above
the plane of the nicotinamide mimic and at the edge of the
hydrophobic pocket, i.e. “south” of the 5-position of potentially
bound isoquinoline-1-ones. By use of molecular modeling to
design further potentially selective inhibitors, the crystal struc-
tures of chicken PARP-1 with 8-hydroxy-2-methylquinazolin-4-
one bound50 and murine PARP-248 were used as starting
structures, from which the binding pockets were established
and compared. The binding pocket of PARP-1 was measured and
mapped, together with key interactions between the bound
inhibitor and the pocket. Comparisons were also made to
PARP-1 without inhibitor to confirm that the receptor binding
pocket is relatively rigid and does not change conformation upon
binding the substrate or an inhibitor ligand. The distances
observed in PARP-1 were then used to dock a minimized and
charged (Gastieger/H€uckel) structure of 1 into the binding
pocket of PARP-2. Once docked, restraints were added (as
observed in the PARP-1 model) and the ligand was subjected
to molecular dynamics (300 K for 5 ps) and then reminimized.
The ligand and binding pocket (5.0 Å from ligand) were then
subjected to molecular dynamics (300 K for 5 ps), and
the complete complex (enzyme and 1) was minimized to give
the final model. In this model, the basic Lys308 sat close (3.1 Å) to
the 5-amine of 1. Two short series of compounds were designed
to exploit this opportunity for isoform-selectivity by linking an
anionic carboxylate to the 5-position of the isoquinolin-1-one.
In the first three compounds, the carboxylic acid was attached
to the 5-position of the isoquinolin-1-one through all-carbon
links (Scheme 3). Compound 24 was prepared by hydrolysis of
5-cyanoisoquinolin-1-one 23.51,52 The targets 26 and 27 have a
C2 chain between the carboxylic acid and the isoquinoline, the
former with a rigid link and the latter with a flexible link. These
were approached by a Heck coupling of 5-iodoisoquinolin-1-one
25 to give 26.52 Selective hydrogenation of the exocyclic double
bond under acidic conditions led to 27.
Exploiting the now ready availability of 1 as a starting material,
it was planned to attach carboxylic acids through linkers to the
exocyclic amine to provide target compounds such as 29
(Scheme 4). Heating 1 with ethyl bromoacetate in DMF in the
presence of tertiary amine base alkylated the weakly nucleophilic
amine, giving the ester 28 in modest yield. Hydrolysis then
furnished the carboxylic acid 29. In an attempt to access the
homologous ester 30, 1 was treated with sodium hydride and
2052
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
Scheme 3. Synthesis of C-Linked 1-Oxoisoquinoline-5-car-
boxylic Acids 24, 26, and 27a
Figure 2. X-ray crystal structure of methyl ester 33, showing the
hydrogen-bonded pairs in the crystal. Ellipsoids are represented at
30% probability, and O10 is related to O1 by the 1 ꢀ x, 1 ꢀ y, 3 ꢀ z
symmetry operation.
a Reagents and conditions: (i) KOH, EtOH, reflux, then aq HCl, 83%;
(ii) HO2CCHdCH2, Pd(OAc)2, Et3N, EtCN, reflux, 97%; (iii) H2,
Pd/C, EtOH, 66%.
1-position under Mitsunobu conditions.52 Hydrolysis of the ester
of 31 with aqueous acid gave the corresponding carboxylic
acid 32.
Scheme 4. Synthesis of 5-(ω-Carboxyalkylamino)-
isoquinolin-1-ones 29 and 32a
In view of the unexpected regioselectivity of the reaction of 1
with methyl propenoate under strongly basic conditions, it was
necessary to prove the structure of 28 or 29 through an X-ray
crystal structure determination. Careful recrystallization of the
carboxylic acid 29 from methanol provided diffraction-quality
crystals of the corresponding methyl ester 33. The crystal
structure (Figure 2) confirmed the expected structure, with the
CH2CO2Me attached to the exocyclic amine. The isoquinolin-1-
ones formed hydrogen-bonded pairs in the crystal, with the
lactam NꢀH bonded to the carbonyl oxygen of the partner
molecule. The 5-NHCH2CO2Me unit adopted an extended
conformation but appears not to be hydrogen-bonded; the
exocyclic nitrogen was planar, as expected, and the methylene
was orientated away from the peri 4-H.
’ BIOCHEMICAL AND CELL BIOLOGICAL EVALUATION
The two series of 5-substituted isoquinolin-1-ones were
evaluated in vitro for inhibition of the catalytic activities of full-
length active human PARP-1 isolated from HeLa cell nuclear
extract (using a FlashPlate assay previously developed by us55)
and full-length mouse PARP-2 (using a solution-phase assay);
the results are presented in Table 1 for 15aꢀn, 22, 24, 26, 27, 29,
and 32. 1 and the two claimed PARP-2-selective inhibitors 9b38
and 1033 from the literature were assayed as controls.
As expected, 1 showed no selectivity toward either isoform,
with IC50 ≈ 1 μM for each. The 5-benzamido compound 15a
was approximately equipotent with 1 against PARP-2, but the
increased bulk of the benzoyl group and loss of the basicity of the
amine diminished the inhibition of PARP-1 dramatically, leading
to a 9.3-fold selectivity toward inhibition of PARP-2. Exploration
of the SAR around the phenyl ring indicated that introducing a
para-substituent reduced the selectivity, mostly through decreas-
ing the potency against PARP-2 (15b,dꢀf) rather than by
increasing the inhibition of PARP-1; the 4-nitrobenzamido and
4-iodobenzamido analogues 15c,h were the exception, where the
compounds were more potent against PARP-1 than was 15a.
Moving the methyl group to the ortho-position of the benza-
mide, in 15i, reduced inhibition against both isoforms, but
the corresponding ortho-iodo compound 15j was nonselective
but had fairly good potency against PARP-1 and PARP-2.
a Reagents and conditions: (i) EtO2CH2Br, iPr2NEt, NaI, DMF, 80ꢀC,
19%; (ii) aq HCl, reflux, 87% (29), 85% (32); (iii) NaH, THF,
MeO2CCHdCH2, 67%.
methyl acrylate. Surprisingly, the sole isolable product was 31,
where the CH2CH2CO2Me unit is attached to the lactam 2-N,
rather than the exocyclic amine. The location of this unit was
confirmed by the 1H NMR chemical shift of the NCH2 protons
(δ 4.09, more typical of CH2 attached to the lactam rather than
CH2 attached to an aniline nitrogen) and by HMBC correlations
between these CH2 protons and the lactam carbonyl-13C at δ
161.2 and between these CH2 protons and 3-C at δ 130.6;
these correlations are incompatible with the alternatives 30 or
1-(MeO2CH2CH2CO)-5-NH2-isoquinoline (from alkylation at
oxygen). Sodium hydride is a much stronger base than the
diisopropylethylamine used in the synthesis of 28 and is capable
of removing the lactam NꢀH proton. In the anion so generated,
the Nꢀ of the lactam is more nucleophilic than the neutral
exocyclic NH2 and reacts with the electrophile. Anions derived
from other 5-substituted isoquinolin-1-ones usually react
through the 2-N with alkylating electrophiles,42,53,54 although
there are some examples of reaction at the exocyclic oxygen at the
2053
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
Table 1. Inhibition of the Activities of PARP-1 and PARP-2 by 5-Amidoisoquinolin-1-ones 15aꢀn and 22, by 5-(ω-Carboxyalkyl)-
and 5-(ω-Carboxyalkylamino)isoquinolin-1-ones 24, 26, and 29, and by 2-(2-Carboxyethyl)isoquinolin-1-one 32a
isoquinolinone
5-substituent
IC50 (μM)
compd
3-substituent
PARP-1
PARP-2
observed selectivity (IC50(PARP-1)/(IC50(PARP-2))
1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H2N-
0.94 ( 0.10
13.9 ( 1.2
13.4 ( 1.2
3.0 ( 0.4
10.7 ( 1.5
18.0 ( 1.2
11.2 ( 1.3
NDb
1.05 ( 0.11
1.5 ( 0.2
6.5 ( 0.8
1.6 ( 0.2
3.3 ( 1.0
3.6 ( 0.4
3.9 ( 0.3
NDb
0.9
9.3
2.1
1.9
3.2
5.0
2.9
15a
15b
15c
15d
15e
15f
15g
15h
15i
15j
15k
15l
15m
15n
22
PhCONH-
4-MePhCONH-
4-O2NPhCONH-
4-F3CPhCONH-
4-FPhCONH-
4-ClPhCONH-
4-BrPhCONH-
4-IPhCONH-
2-MePhCONH-
2-IPhCONH-
(thiophen-2-yl)CONH-
cHexCONH-
tBuCONH-
7.6c
1.3c
5.8
5.6
31.6 ( 3.4
4.5 ( 1.2
22.4 ( 3.4
>80
5.6 ( 0.4
3.2 ( 1.0
7.0 ( 1.2
27.9 ( 4.5
29 ( 3.4
19.9 ( 3.0
6.3 ( 0.3
>25
1.4
3.2
>2.9
>3.4
>2.5
2.6
>100
(adamantan-1-yl)CONH-
PhCONH-
>50
16.6 ( 1.2
>25
24
HO2C-
26
HO2CHCdHC-
HO2CCH2CH2-
HO2CCH2NH-
H2N-
6.6 ( 1.0
8.6 ( 1.1
1.6 ( 0.1
0.55 ( 0.1
4.1 ( 0.5
0.03 ( 0.006
4.7 ( 0.4
3.0 ( 0.2
0.55 ( 0.1
1.6 ( 0.3
1.5 ( 0.4
0.09 ( 0.02
1.4
27
2.9
29
32d
2.9
0.4
10
PhCO2-
2.75
0.33
9b
[2-(4-ClPh)quinoxaline-5-CONH2]
a Data for 1, 5-benzoyloxyisoquinolin-1-one 10, and 2-(4-chlorophenyl)quinoxaline-5-carboxamide 9 are shown for comparison. b Not determined
owing to limited solubility. c Measured once only. d Also carries -CH2CH2CO2H at the 2-position.
The common medicinal chemical replacement of a benzene ring
with a thiophene, in 15k, reduced potency against both isoforms,
and replacement with bulky aliphatic groups, in 15lꢀn, contin-
ued the trend of diminishing activity without gain of selectivity.
Adding a methyl group at the 3-position, in 22, caused loss of
potency against PARP-2 relative to that of 15a but retention of
activity against PARP-1, leading to a loss of selectivity toward the
former enzyme.
Table 2. Inhibition of Growth of HT29 Human Colon
Carcinoma Cells, MDA-MB-231 Human Breast Carcinoma
Cells, LNCaP Human Prostate Carcinoma Cells, and
FEK4 Human Dermal Fibroblast Cells by 1 and Selected
5-Amidoisoquinolin-1-ones 15a,lꢀna
IC50 (μM)
compd
HT29
MDA-MB-231
LNCaP
FEK4
Curiously, the benzoate ester 10, which is claimed to be the
most selective for inhibition of PARP-2 with a reported ratio of
IC50 values of ∼60,33 had activity against PARP-2 very similar to
that of the close analogue, the benzamide 15a, as expected, but
proved to be potent in our assay of inhibition of PARP-1 catalytic
activity, giving a selectivity ratio of only 2.75. The quinoxaline-5-
carboxamide 9b was remarkably potent against both isoforms,
with IC50 = 30 nM vs PARP-1 and IC50 = 90 nM vs PARP-2.
These values give an approximate 3-fold selectivity for inhibition
of PARP-1, in contrast to the claimed 5-fold selectivity for PARP-2.39
These marked differences in apparent selectivities for the
different isoforms probably arise from the different assays used
and different sources of the enzymes used. Pellicciari et al.33
compared bovine PARP-1 with murine PARP-2 using the
incorporation of radioactivity from 3H-NADþ into trichloroace-
tic acid-insoluble material as an assay of enzymic activity, whereas
Iwashita et al.39 compared human PARP-1 with murine PARP-2
using the same method.60 We compared human PARP-1 with
murine PARP-2 using different assay methods. Thus, it is highly
1
>200
>200
>200
>200
>200
170
139
>200
98
>200
>200
>200
>200
>200
15a
15l
15m
15n
>200
>200
>200
>200
>200
42
a Data are from single experiments using duplicates for each data point.
likely that our comparative assays are a much more stringent test
of selectivity for inhibition of PARP-2 than the assay pairs
previously used or, alternatively, tend to bias toward reporting
selectivity for inhibition of PARP-1.
Table 1 also shows the evaluation of the isoquinolin-1-ones
bearing a carboxy-terminated function at the 5-position. These
carboxylates had been designed to interact with the positively
charged side chain of Lys308 near the PARP-2 NADþ-binding
site. By use of the current assays for PARP-1 and for PARP-2,
isoquinolinone-5-carboxylic acid 24 showed little inhibition of
2054
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Modeling of the non-isoform-selective inhibitor 1 and the PARP-2-selective inhibitors 15a and 29 into the NADþ-binding sites of PARP-1
and PARP-2: (A) 1 in human PARP-1; (B) 1 in murine PARP-2; (C) 15a in murine PARP-2; (D) 29 in murine PARP-2.
either isoform, a surprising result in view of the observation of
IC50 < 13 μM in an earlier assay using broken cell nuclei as the
source of PARP activity.51 The isoquinolin-1-ones 26, 27, and 29
carrying the carboxy group attached to the 5-position through a
more or less flexible linker were active against both isoforms but
with modest selectivity. The isoquinolinone-5-propenoic acid
had previously shown IC50 < 11 μM in the broken nuclei assay
against mixed PARPs, broadly consistent with the presently
observed values in the purified enzyme assays.51 Curiously, 32,
which bears the carboxy group attached through a tether to the
2-N of the ring system, was potent against both isoforms, with
∼3-fold selectivity for inhibition of PARP-1 over PARP-2. This
compound appears to be slightly more potent against PARP-1
than is the lead PARP inhibitor 1 but lacks the NꢀH generally
taken3 to be part of the essential pharmacophore for binding to
the PARPs (through an H-bond to Gly863-O in PARP-1 and to
Gly405-O in PARP-2), but Eltze et al. have recently reported
some inhibitors of PARP lacking this NꢀH (fused 2,3-dihydro-
quinolin-4-ones 5, thus with CH2 replacing the usual NH).35 The
precise binding mode of 32 to the PARPs will be the subject of
later studies.
The most selective inhibitor of PARP-2 identified in our study,
5-benzamidoisoquinolin-1-one 15a, together with the non-iso-
form-selective lead inhibitor 1 and examples 15lꢀn of 5-(bulky
alkyl)amidoisoquinolin-1-ones, which are poorly active against
both isoforms, were evaluated for cytotoxicity toward a panel of
three diverse human carcinoma cell lines and one human
fibroblast cell line in vitro using the MTS colorimetric assay.
None of the isoquinolin-1-ones were active against HT29 human
colon carcinoma cells or the FEK4 normal human fibroblasts
(Table 2). Two compounds, 1 and the most selective PARP-2
inhibitor 15a, were very weakly cytotoxic toward the MDA-MB-
231 human breast carcinoma cells, and two, 15a and 15n, were
weakly cytotoxic toward the LNCaP human prostate carcinoma
cells. These results are consistent with our previous observation
that 1 caused only 40% inhibition of the proliferation of CT26
murine colon carcinoma cells at the very high concentration of
1 mM.57 None of the cell lines used here have mutant BRCA, so
the lack of cytotoxicity is unsurprising.
’ POST FACTO STRUCTURAL STUDIES
Post facto molecular modeling studies were undertaken to
rationalize the strong selectivity for inhibition of PARP-2 ob-
served for the 5-benzamidoisoquinolin-1-one 15a and for the
smaller but significant selectivity observed for the carboxylic acid
29. These represent the most isoform-selective examples from
each of the two series originally designed. These post facto
studies were aided by the disclosure in 2009 (after the current
experimental work had been completed) of a crystal structure of
human PARP-158 and in 2010 of a crystal structure of human
PARP-2 complexed with ABT888 4.59 Starting structures were
taken from published X-ray crystallographic structures of human
PARP-1 complexed with a quinoxalinone inhibitor (A861696)58
and of murine PARP-2 without an inhibitor ligand.48 The latter
was compared with the more recent data for human PARP-2
obtained with 3-aminobenzamide bound;59 this comparison was
used to drive docking and to refine the nicotinamide/inhibitor-
binding pocket of PARP-2. Nonselective inhibitor 1 and PARP-
2-selective inhibitors 15a and 29 were then docked into the
models using the existing bound inhibitor (for PARP-1) and
the docked inhibitor (for PARP-2) as templates. Once docked,
the inhibitors were subjected to molecular mechanics and
dynamics calculations to establish optimal docking conforma-
tions. During these calculations, the receptor was restrained to its
original conformation. Lastly, both the inhibitors and binding
pockets (radius 10 Å) were subjected to molecular dynamics and
finally molecular mechanics calculations to give the final struc-
tures (Figure 3). The structures of 15a and of 29 made the
expected hydrogen-bonding contacts of the lactam with the Ser
and Gly residues and π-stacks of the isoquinoline-1-one cores
with Tyr907 (PARP-1) and Tyr449 (PARP-2).
Differences between the binding pockets of PARP-1 and
PARP-2 can be observed from the models. The binding pocket
2055
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
in PARP-1 is smaller than that of PARP-2. Furthermore, the
cavity that accepts groups attached to the 5-position of isoqui-
nolin-1-ones is arranged essentially in plane with the bound 1 in
PARP-2 but lies significantly above the plane of the bound ligand
in PARP-1. This means that any bulky substituent in this position
should lead to selectivity for binding to PARP-2 and inhibit
binding to PARP-1. Compound 1 has a very small amino group
in this location, which can easily be accommodated in both
structures. By contrast, the 5-benzamido group in 15a can only
be accommodated by the larger and in-plane void of the PARP-2
structure and is sterically excluded by PARP-1 (Figure 3).
Indeed, it was not possible to generate a corresponding model
for binding of 15a to PARP-1, owing to steric clash of the
benzamide unit with the walls of the pocket. Now, the apparent
SAR requirement for a benzamide in the 5-position, rather than
an equivalent aliphatic amide, can be rationalized by the estab-
lishment of an additional aromatic π-stacking interaction onto a
Tyr residue (Tyr438) in the wall of the pocket in PARP-2. The
structure of human PARP-2 contains a captive water molecule
near the 3-amino group of the bound inhibitor 3-aminobenza-
mide, and it has been proposed that this mediates hydrogen
bonding to the amine.59 A similar water-mediated hydrogen
bond to the 5-NH2 could also be responsible, in part, for the good
binding of 1 to both PARP-1 and PARP-2. The facility for
hydrogen bonding (now to the secondary amide NꢀH) is
retained in bound 15a.
PARP-1, were disappointing. Reasonable potency against PARP-2
was observed for 26, 27, and 29, but inhibition of PARP-1 activity
was also strong. The modest selectivity was rationalized by post
facto modeling studies.
The 5-amidoisoquinolin-1-ones were investigated as more
hydrolytically stable analogues of the benzoate ester 10, for
which Pellicciari et al. claimed 60-fold selectivity for inhibition of
PARP-2.33,56 Of this series, the simple benzamide 15a, the closest
analogue of 10, proved to be the most selective (IC50(PARP-1)
/
IC50(PARP-2) = 9.3). In our comparative assays, 15a had activity
equal to that of 10 against PARP-2 but 10 was much more potent
than was 15a against PARP-1, leading to an observed selectivity
of only 2.75-fold for 10. Similarly, the quinoxaline-5-carboxamide
9b showed a 3-fold selectivity for PARP-1 in our comparative
assays, contrasting with a claimed 5-fold selectivity for PARP-2.39
In light of these results, it is now evident from our direct
comparisons that 5-benzamidoisoquinolin-1-one 15a is the most
isoform-selective inhibitor of PARP-2 reported to date. Post
facto modeling studies rationalized the structural basis of the
observed selectivity.
’ EXPERIMENTAL SECTION
General. NMR spectra were recorded on JEOL Delta 270 and
Varian Mercury 400 spectrometers. Mass spectra were obtained using
VG7070E and Bruker microTOF spectrometers. IR spectra were
measured on a Perkin-Elmer RXI FTIR spectrometer. The stationary
phase for chromatography was silica gel. All reactions were carried out at
ambient temperature unless otherwise stated. Solvents were evaporated
under reduced pressure. Melting points were determined using a
Reichert-Jung Thermo Galen instrument and are uncorrected. Target
compounds for biochemical evaluation were >95% pure, as shown by
CHN combustion microanalyses (carried out at the School of Pharmacy,
University of London) and TLC for novel compounds and by identity of
mp and TLC for known compounds. Compound 12 was obtained from
the Aldrich Chemical Co.
Similar docking/modeling of the less selective inhibitor 29
into the active site of murine PARP-2 was carried out. The above
putative water-mediated hydrogen bond is also available in this
structure (from the 5-NH). The side chain is located within the
“in plane” pocket of the PARP-2 protein, although the π-stacking
is, of course, not possible. The terminal carboxylate can also
access the ammonium side chain of Lys308. The smaller size and
the flexibility of the side chain of 29 permit access to both the
larger pocket of PARP-2 and the smaller pocket of PARP-1,
leading to more modest isoform-selectivity than for 15a.
5-Aminoisoquinolin-1-one Hydrochloride (1). Compound
14 (1.6 g, 8.4 mmol) was stirred with Pd/C (10%, 1.0 g) in EtOH
(120 mL) and aqueous HCl (9 M, 4 mL) under H2 for 2 h. The
suspension was then filtered through Celite. The Celite pad and residue
were suspended in water (1000 mL) and heated. The hot suspension
was filtered through a second Celite pad. Concentration of the filtrate
and drying gave 1 (1.2 g, 71%) as white crystals: mp 248ꢀ252 ꢀC (dec)
(lit.15 250ꢀ260 ꢀC (dec)).
’ CONCLUSIONS
In this paper, we report the design, synthesis, and biochemical
evaluation of new 5-substituted isoquinolin-1-ones as selective
inhibitors of PARP-2. Compounds were designed partly by
modeling using the reported crystal structures of the catalytic
NADþ-binding sites of PARP-1 and PARP-2.48,49 A new high-
yielding synthesis of our lead PARP inhibitor 1 has been
developed, which has the potential for large-scale preparations
of this important compound to make it more economically
available for biological and chemical studies. A Hurtley coupling
to 2-methyl-3-nitrobenzoic acid 16 led directly to the isocou-
marin 19; conversion of 19 to the isoquinolin-1-one 20 and
reduction of the nitro group provided a good route to 21.
Acylation of the exocyclic amines of 1 and 21 gave the corre-
sponding 5-amidoisoquinolin-1-ones 15aꢀn and 22 in good
yields. Four isoquinolin-1-ones carrying carboxylates at the
5-position (linked directly to the ring or through tethers) were
synthesized by hydrolysis of 5-cyanoisoquinolin-1-one 23, by a
Heck coupling to 5-iodoisoquinolin-1-one 25 or by chain-
extension from the amine of 1. Under strongly basic conditions,
1 was alkylated at the ring 2-N rather than at the exocyclic amine.
The isoform-selectivities of compounds 24, 26, 27, and 29,
carrying carboxylates designed to bind electrostatically to basic
Lys308 in the PARP-2 structure, which replaces neutral Gln763 in
1-Chloro-5-nitroisoquinoline (13). Aqueous HNO3 (70%, 850
mg, 13.4 mmol) in concentrated H2SO4 (5 mL) was added dropwise to
12 (2.00 g, 12.2 mmol) in concentrated H2SO4 (10 mL) at 0ꢀ5 ꢀC. The
mixture was stirred at 0 ꢀC for 2 h, then poured onto ice. The precipitate
was collected, washed (H2O), dried, and recrystallized (EtOAc/
hexanes) to give 13 (2.34 g, 92%) as pale yellow crystals: mp
181ꢀ183 ꢀC (lit.61 mp 183ꢀ184 ꢀC).
5-Nitroisoquinolin-1-one (14). Compound 13 (5.00 g, 24.0
mmol) was stirred at 100 ꢀC in AcOH (100 mL) for 40 h. The cooled
suspension was then poured onto ice. The solid was collected, washed
(H2O), and recrystallized (EtOH) to give 14 (3.74 g, 82%) as pale
yellow crystals: mp 247ꢀ249 ꢀC (dec) (lit.15 mp 247ꢀ249 ꢀC).
5-Benzamidoisoquinolin-1-one (15a). Compound 1 (50 mg,
0.25 mmol) was stirred with PhCOCl (39 mg, 0.28 mmol) in pyridine
(2.0 mL) at 90 ꢀC for 16 h. Evaporation and recrystallization (EtOAc)
gave 15a (57 mg, 86%) as an off-white solid: mp >310 ꢀC (dec).
3-Methyl-5-nitroisocoumarin (19). 2-Bromo-3-nitrobenzoic
acid 1662 (2.5 g, 10 mmol) and Cu powder (67 mg, 1.1 mmol) were
added to pentane-2,4-dione (5.3 g, 53 mmol) and KOtBu (2.3 g,
2056
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
20 mmol) in tBuOH (50 mL). The mixture was boiled under reflux for
16 h, then poured into H2O (350 mL) and acidified with aqueous HCl
(2 M). Extraction (Et2O), evaporation, and chromatography (hexane/
EtOAc 3:2) gave 19 (470 mg, 23%) as yellow crystals: mp 199ꢀ200 ꢀC.
3-Methyl-5-nitroisoquinolin-1(2H)-one (20). A solution of 19
(470 mg, 2.3 mmol) in MeO(CH2)2OH (100 mL) was saturated with
NH3, boiled under reflux for 4 h, then evaporated until 10 mL remained.
The concentrate was stored at 4 ꢀC for 16 h. The solid was collected,
washed (H2O, EtOH), and recrystallized (MeOH) to give 20 (320 mg,
68%) as bright yellow crystals: mp 231ꢀ232 ꢀC (dec).
using a logarithmic concentration scale to a dose-response curve using
SigmaPlot 11; IC50 values were measured usually in two or three
independent experiments, and the mean values are reported.
PARP-2 Inhibition Assay. Compounds were assayed for inhibi-
tion of the catalytic activity of PARP-2 using a method in which
recombinant full-length murine PARP-2 protein (Alexis) (∼50 ng)
was bound by a PARP-2-specific antibody in a 96-well white-walled
3
plate. PARP-2 activity was measured following addition of H-NADþ
3
and DNA.36 After washing, scintillant was added to measure the H
incorporated. Compounds were evaluated at eight different concentra-
tions (0.75 nM to 1.5 μM þ further concentrations to 10 μM if
necessary) in triplicate. Compounds were dissolved in DMSO prior to
addition to give a final concentration of 2% in the assay mixture, a
concentration that had been shown to have no effect on the activity of
the enzyme. Oligonucleotides 50-ACTTGATTAGTTACGTAACGT-
TATGATTGA-30/50-TCAATCATAACGTTACGTAACTAATCA-
AGT-30 were used as the DNA ligand. The concentration of NADþ was
2.5 μM. The reaction buffer was HEPES (1.0 M, 12.5 mL), MgCl2
(1.0 M, 6.25 mL), KCl (3.0 M, 8.3 mL), dithiothreitol (77 mg), glycerol
(propane-1,2,3-triol, 50 mL), NP-40 (50 μL) made up to 500 mL with
Milli-Q water, then adjusted to pH 7.6 with aqueous KOH (3.0 M).
Compounds were incubated with the enzyme for 10 min. Then the other
components were added, and the reaction was allowed to proceed for
45 min. Data were fitted using a logarithmic concentration scale to a
dose-response curve using SigmaPlot 11. IC50 values were measured
usually in two or three independent experiments, and the mean values
are reported.
5-Amino-3-methylisoquinolin-1(2H)-one (21). Compound
20 (320 mg, 1.6 mmol) was heated with SnCl2 (900 mg, 4.7 mmol)
in EtOH (20 mL) at 70 ꢀC for 4 h, then carefully poured into iceꢀwater
(200 mL). The suspension was made alkaline with aqueous NaOH, and
the precipitate was filtered. Extraction of the filtrate (EtOAc), evapora-
tion, and recrystallization (hexane, EtOAc) gave 21 (160 mg, 59%) as
yellow crystals: mp 183ꢀ184 ꢀC.
3-(1-Oxoisoquinolin-5-yl)propanoic Acid (27). Compound
26 (160 mg, 8.4 mmol) in EtOH (25 mL) and aqueous HCl (34%,
4 mL) was stirred vigorously with Pd/C (10%, 100 mg) under H2 for 2 h.
Filtration (Celite) and evaporation of the solvent from the filtrate gave
27 (1.1 g, 66%) as white crystals: mp 260ꢀ263 ꢀC.
Ethyl 2-(1-Oxoisoquinolin-5-ylamino)acetate (28). Com-
i
pound 1 (1.0 g, 4.4 mmol), Pr2NEt (1.4 g, 11 mmol), ethyl bromoa-
cetate (885 mg, 5.3 mmol), and NaI (100 mg, 0.7 mmol) were stirred at
80 ꢀC in DMF (60 mL) for 16 h. Evaporation and recrystallization
(MeOH) gave 28 (121 mg, 19%) as pale buff crystals: mp 199ꢀ201 ꢀC.
5-(Carboxymethylamino)isoquinolin-1-one Hydrochlor-
ide (29). Ester 28 (94.0 mg, 0.39 mmol) was boiled under reflux in
aqueous HCl (6.0 M, 4.0 mL) for 3 h. Evaporation gave 29 (86 mg, 87%)
as a pale amber solid: mp 275ꢀ280 ꢀC (dec).
’ ASSOCIATED CONTENT
S
Supporting Information. Synthetic procedures and
b
spectroscopic data for 15bꢀn, 22, 24, 26, 31, and 32; spectro-
scopic data for 1, 13, 14, 15a, 19, 20, 21, and 27ꢀ29; biochem-
ical and cell biological evaluation procedures; dose-response
curves; and elemental combustion analytical data. This material
is available free of charge via the Internet at http://pubs.acs.org.
Crystal Data for 33. All data were collected at 150 K on a Nonius
kappaCCD diffractometer. The structure was solved using SHELXS-
9763 and refined using full-matrix least-squares in SHELXL-97.
C12H12N2O3, M = 232.24, λ = 0.710 73 Å, triclinic, space group = P1
(No. 2), a = 7.7290(3) Å, b = 7.7930(3) Å, c = 9.2330(4) Å, R =
103.760(2)ꢀ, β = 93.294(2)ꢀ, γ = 94.054(2)o, U = 537.23(4) Å3, Z = 2,
Dc = 1.436 g cmꢀ3, μ = 0.105 mmꢀ1, F(000) = 244. Crystal size = 0.25
mm ꢁ 0.25 mm ꢁ 0.10 mm, unique reflections = 2405 [Rint = 0.0494],
observed (I > 2σ(I)) = 1503, data/restraints/parameters = 2405/1/160.
Final R indices [I > 2σ(I)], R1 = 0.0547, wR2 = 0.1355; R indices (all
data) = R1 = 0.1010, wR2 = 0.1589. Max peak/hole, 0.718 and ꢀ0.309
e Åꢀ3. Crystallographic data for 33 have been deposited at the Cam-
bridge Crystallographic Data Centre: CCDC 781071. Requests for data
should be addressed to CCDC, 12 Union Road, Cambridge CB2
1EZ, U.K.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: þ44 1225 386840. Fax: þ44 1225 386114. E-mail: m.d.
threadgill@bath.ac.uk.
Present Addresses
Chemistry Research Laboratory, University of Oxford, 12
Mansfield Road, Oxford OX1 3TA, U.K.
^Department of Pharmacy and Pharmaceutical Sciences, Trinity
College Dublin, College Green, Dublin 2, Ireland.
PARP-1 Inhibition Assay. Compounds were assayed for inhibi-
tion of the catalytic activity of PARP-1 using the FlashPlate scintillation
proximity assay previously developed at KuDOS.55 Compounds were
evaluated at eight different concentrations (0.75 nM to 1.5 μM þ further
concentrations to 10 μM if necessary) in triplicate, using full-length
human PARP-1 (∼50 ng) isolated from HeLa cell nuclear extract.
Compounds were dissolved in DMSO prior to addition to give a final
concentration of 2% in the assay mixture, a concentration that had been
shown to have no effect on the activity of the enzyme. Oligonucleotides
50-ACTTGATTAGTTACGTAACGTTATGATTGA-30/50-TCAAT-
CATAACGTTACGTAACTAATCAAGT-30 were used as the DNA
ligand. The concentration of NADþ was 5.0 μM. The reaction buffer was
HEPES (1.0 M, 12.5 mL), MgCl2 (1.0 M, 6.25 mL), KCl (3.0 M,
8.3 mL), dithiothreitol (77 mg), glycerol (propane-1,2,3-triol, 50 mL),
NP-40 (50 μL) made up to 500 mL with Milli-Q water, then adjusted to
pH 7.6 with aqueous KOH (3.0 M). Compounds were incubated with
the enzyme for 10 min. Then the other components were added and
the reaction was allowed to proceed for 45 min. Data were fitted
’ ACKNOWLEDGMENT
We thank Dr. Timothy J. Woodman (University of Bath) for
many of the NMR spectra, Dr. Anneke Lubben (University of
Bath) for the mass spectra, and Professor Rex M. Tyrrell
(University of Bath, U.K.) for the kind gift of the FEK4 cells.
We are very grateful to KuDOS Pharmaceuticals Ltd., Cancer
Research UK, and the University of Bath for financial support. M.
D.T., M.D.L., and A.S.T. are members of Cancer Research at
Bath (CR@B).
’ ABBREVIATIONS USED
PARP-2, poly(ADP-ribose) polymerase 2; PARP-1, poly(ADP-
ribose) polymerase 1; 5-AIQ, 5-aminoisoquinolin-1-one; ADP,
2057
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
adenosine diphosphate ribose; NADþ, nicotinamide adenine
dinucleotide; PAR, poly(ADP-ribose); BRCA, breast cancer type
1 susceptibility protein; SSB, single-strand break; TRF2, telo-
meric repeat-binding factor 2; MNU, N-methyl-N-nitrosourea
of 5-aminoisoquinolinone, a water-soluble, potent inhibitor of the activity
of poly(ADP-ribose)polymerase on the organ injury and dysfunction
caused by haemorrhagic shock. Br. J. Pharmacol. 2000, 130, 843–850.
(16) Ferraris, D.; Pargas Ficco, R.; 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 imidazobenzodiazepines as potent PARP-1
inhibitors for treatment of ischemic injuries. Bioorg. Med. Chem. 2003,
11, 3695–3707.
(17) Chatterjee, P. K.; Chatterjee, B. E.; Pedersen, H.; Sivarajah, A.;
McDonald, M. C.; Mota-Filipe, H.; Brown, P. A. J.; Stewart, K. N.;
Cuzzocrea, S.; Threadgill, M. D.; Thiemermann, C. 5-Aminoisoquino-
linone reduces renal injury and dysfunction caused by ischaemia/
reperfusion. Kidney Int. 2004, 65, 499–509.
’ REFERENCES
(1) Chambon, P.; Weill, J. D.; Mandel, P. Nicotinamide mononu-
cleotide activation of new DNA-dependent polyadenylic acid synthesiz-
ing nuclear enzyme. Biochem. Biophys. Res. Commun. 1963, 11, 39–43.
(2) Rouleau, M.; Patel, A.; Hendzel, M. J.; Kaufmann, S. H.; Poirier,
G. G. PARP inhibition: PARP1 and beyond. Nat. Rev. Cancer 2010,
10, 293–301.
(3) Woon, E. C. Y.; Threadgill, M. D. Poly(ADP-ribose)polymerase
inhibition—where now? Curr. Med. Chem. 2005, 12, 2373–2392.
(4) Smith, S. The world according to PARP. Trends Biochem. Sci.
2001, 26, 174–179.
(18) Hendryk, S.; Czuba, Z. P.; Je-drzejowska-Szypuzka, H.; Szliszka, E.;
Phillips, V. A.; Threadgill, M. D.; Krꢀol, W. Influence of 5-aminoisoqui-
nolin-1-one (5-AIQ) on neutrophil chemiluminescence in rats with
transient and prolonged focal cerebral ischaemia and after reperfusion.
J. Physiol. Pharmacol. 2008, 59, 811–822.
(5) De Murcia, G.; Mꢀenissier-de Murcia, J. Poly(ADP-ribose) poly-
merase—a molecular nick-sensor. Trends Biochem. Sci. 1994, 19,
172–176.
(19) Cuzzocrea, S.; McDonald, M. C.; Mazzon, E.; Dugo, L.;
Serraino, I.; Threadgill, M. D.; Caputi, A. P.; Thiemermann, C. Effects
of 5-aminoisoquinolinone, a water-soluble potent inhibitor of the
activity of poly(ADP-ribose) polymerase, in a rodent model of lung
injury. Biochem. Pharmacol. 2002, 63, 293–304.
(20) Cuzzocrea, S.; Mazzon, E.; Di Paola, R.; Genovese, T.; Patel,
N. S. A.; Muiꢁa, C.; Threadgill, M. D.; De Sarro, A.; Thiemermann, C.
5-Aminoisoquinolinone reduces colon injury by experimental colitis.
Naunyn-Schmiedeberg's Arch. Pharmacol. 2004, 370, 464–473.
(21) Rajesh, M.; Mukhopadhyay, P.; Godlewski, G.; Bꢀatkai, S.;
Haskꢀo, G.; Liaudet, L.; Pacher, P. Poly(ADP-ribose)polymerase inhibi-
tion decreases angiogenesis. Biochem. Biophys. Res. Commun. 2006,
350, 1056–1062.
(22) Li, M.; Threadgill, M. D.; Wang, Y.; Cai, L.; Lin, X. Poly(ADP-
ribose) polymerase inhibition down-regulates expression of metastasis-
related genes in CT26 colon carcinoma cells. Pathobiology 2009,
76, 108–116.
(23) Qin, Y.; Wang, Y.; Li, Y.-Y. Effect of poly(ADP-ribose)-
polymerase inhibitor on the liver metastasis of mouse colorectal
carcinoma CT-26 cell line in vivo. J. Third Mil. Med. Univ. 2008, 30,
1330–1333.
(24) Amꢀe, J.-C.; Rolli, V.; Schreiber, V.; Niedergang, C.; Apiou, F.;
Decker, P.; Muller, S.; H€oger, T.; Mꢀenissier-de Murcia, J.; de Murcia, G.
PARP-2, A novel mammalian DNA damage-dependent poly(ADP-
ribose) polymerase. J. Biol. Chem. 1999, 274, 17860–17868.
(25) Schreiber, V.; Amꢀe, J.-C.; Dollꢀe, P.; Schultz, I.; Rinaldi, B.;
Fraulob, V.; Mꢀenissier-de Murcia, J.; de Murcia, G. Poly(ADP-ribose)
polymerase-2 (PARP-2) is required for efficient base excision DNA
repair in association with PARP-1 and XRCC1. J. Biol. Chem. 2002,
277, 23028–23036.
(26) Fisher, A. E.; Hochegger, H.; Takeda, S.; Caldecott, K. W.
Poly(ADP-ribose) polymerase 1 accelerates single-strand break repair in
concert with poly(ADP-ribose) glycohydrolase. Mol. Cell. Biol. 2007,
27, 5597–5605.
(27) Dantzer, F.; Giraud-Panis, M. J.; Jaco, I.; Amꢀe, J.-C.; Schultz, I.;
Blasco, M.; Koering, C. E.; Gilson, E.; Mꢀenissier-deMurcia, J.;deMurcia, G.;
Schreiber, V. Functional interaction between poly(ADP-Ribose) poly-
merase 2 (PARP-2) and TRF2: PARP activity negatively regulates TRF2.
Mol. Cell. Biol. 2004, 24, 1595–1607.
(28) Popoff, I.; Jijon, H.; Monia, B.; Tavernini, M.; Ma, M.; McKay,
R.; Madsen, K. Antisense oligonucleotides to poly(ADP-ribose) poly-
merase-2 ameliorate colitis in interleukin-10-deficient mice. J. Pharma-
col. Exp. Ther. 2002, 303, 1145–1154.
(6) Langelier, M. F.; Servent, K. M.; Rogers, E. E.; Pascal, J. M. A
third zinc-binding domain of human poly(ADP-ribose) polymerase-1
coordinates DNA-dependent enzyme activation. J. Biol. Chem. 2008,
283, 4105–4114.
(7) Schreiber, V.; Molinete, M.; Boeuf, H.; de Murcia, G.; Mꢀenissier-de
Murcia, J. The human poly(ADP-ribose) polymerase nuclear localization
signal is a bipartite element functionally separate from DNA binding and
catalytic activity. EMBO J. 1992, 11, 3263–3269.
(8) Boulares, A. H.; Yakovlev, A. G.; Ivanova, V.; Stoica, B. A.; Wang,
G.; Iyer, S.; Smulson, M. Role of poly(ADP-ribose) polymerase (PARP)
cleavage in apoptosis. Caspase 3-resistant PARP mutant increases rates
of apoptosis in transfected cells. J. Biol. Chem. 1999, 274, 22932–22940.
(9) Plummer, R.; Jones, C.; Middleton, M.; Wilson, R.; Evans, J.;
Olsen, A.; Curtin, N.; Boddy, A.; McHugh, P.; Newell, D.; Harris, A.;
Johnson, P.; Steinfeldt, H.; Dewji, R.; Wang, D.; Robson, L.; Calvert, H.
Phase I study of the poly(ADP-ribose) polymerase inhibitor, AG014699,
in combination with temozolomide in patients with advanced solid
tumors. Clin. Cancer Res. 2008, 14, 7917–7923.
(10) Kummar, S.; Kinders, R.; Gutierrez, M. E.; Rubinstein, L.;
Parchment, R. E.; Phillips, L. R.; Ji, J.; Monks, A.; Low, J. A.; Chen, A.;
Murgo, A. J.; Collins, J.; Steinberg, S. M.; Eliopoulos, H.; Giranda, V. L.;
Gordon, G.; Helman, L.; Wiltrout, R.; Tomaszewski, J. E.; Doroshow,
J. H. Phase 0 clinical trial of the poly (ADP-ribose) polymerase inhibitor
ABT-888 in patients with advanced malignancies. J. Clin. Oncol. 2009,
27, 2705–2711.
(11) Jones, P.; Altamura, S.; Boueres, J.; Ferrigno, F.; Fonsi, M.;
Giomini, C.; Lamartina, S.; Monteagudo, E.; Ontoria, J. M.; Orsale,
M. V.; Palumbi, M. C.; Pesci, S.; Roscilli, G.; Scarpelli, R.; Schultz-
Fademrecht, C.; Toniatti, C.; Rowley, M. Discovery of 2-{4-[(3S)-
piperidin-3-yl]phenyl}-2H-indazole-7-carboxamide (MK-4827): a no-
vel oral poly(ADP-ribose)polymerase (PARP) inhibitor efficacious in
BRCA-1 and-2 mutant tumors. J. Med. Chem. 2009, 52, 7170–7185.
(12) Gartner, E. M.; Burger, A. M.; LoRusso, P. M. Poly(ADP-
ribose) polymerase inhibitors a novel drug class with a promising future.
Cancer J. 2010, 16, 83–90.
(13) Fong, P. C.; Boss, D. S.; Yap, T. A.; Tutt, A.; Wu, P.; Mergui-
Roelvink, M.; Mortimer, P.; Swaisland, H.; Lau, A.; O’Connor, M. J.;
Ashworth, A.; Carmichael, J.; Kaye, S. B.; Schellens, J. H.; de Bono, J. S.
Inhibition of poly(ADP-ribose) polymerase in tumors from BRCA
mutation carriers. N. Engl. J. Med. 2009, 361, 123–134.
(14) Hay, T.; Matthews, J. R.; Pietzka, L.; Lau, A.; Cranston, A.;
Nygren, A. O. H.; Douglas-Jones, A.; Smith, G. C. M.; Martin, N. M. B.;
O’Connor, M.; Clarke, A. R. Poly(ADP-ribose) polymerase-1 inhibitor
treatment regresses autochthonous Brca2/p53-mutant mammary tumors
in vivo and delays tumor relapse in combination with carboplatin. Cancer Res.
2009, 69, 3850–3855.
(29) Yꢀelamos, J.; Monreal, Y.; Saenz, L.; Aguado, E.; Schreiber, V.;
Mota, R.; Fuente, T.; Minguela, A.; Parrilla, P.; de Murcia, G.; Almarza, E.;
Aparicio, P.; Mꢀenissier-de Murcia, J. PARP-2 deficiency affects the
survival of CD4þCD8þ double-positive thymocytes. EMBO J. 2006, 25,
4350–4360.
(15) McDonald, M. C.; Mota-Filipe, H.; Wright, J. A.; Abdelrahman,
M.; Threadgill, M. D.; Thompson, A. S.; Thiemermann, C. Effects
2058
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059

Journal of Medicinal Chemistry
ARTICLE
(30) Quꢀeneta, D.; Mark, M.; Govin, J.; van Dorsseleard, A.; Schreiber,
V.; Khochbin, S.; Dantzer, F. Parp2 is required for the differentiation
of post-meiotic germ cells: identification of a spermatid-specific complex
containing Parp1, Parp2, TP2 and HSPA2. Exp. Cell Res. 2009, 315,
2824–2834.
(31) Bai, P.; Houten, S. M.; Huber, A.; Schreiber, V.; Watanabe, M.;
Kiss, B.; de Murcia, G.; Auwerx, J.; Mꢀenissier-de Murcia, J. Peroxisome
proliferator-activated receptor (PPAR)-2 controls adipocyte differentia-
tion and adipose tissue function through the regulation of the activity of
the retinoid X receptor/PPARg heterodimer. J. Biol. Chem. 2007,
282, 37738–37746.
(32) Yꢀelamos, J.; Schreiber, V.; Dantzer, F. Toward specific func-
tions of poly(ADP-ribose) polymerase-2. Trends Mol. Med. 2008,
14, 169–178.
(33) Pellicciari, R.; Camaioni, E.; Costantino, G.; Formentini, L.;
Sabbatini, P.; Venturoni, F.; Eren, G.; Bellocchi, D.; Chiarugi, A.;
Moroni, F. On the way to selective PARP-2 inhibitors. Design, synthesis,
and preliminary evaluation of a series of isoquinolinone derivatives.
ChemMedChem 2008, 3, 914–923.
nitro-2-methylbenzoate esters with dimethylacetamide dimethyl acetal.
Tetrahedron Lett. 2002, 43, 2299–2302.
(44) Serban, A. 3-Chloro-5-acetamidoisoquinoline as a herbicide. U.S.
Patent 3930837, 1976
(45) Hurtley, W. R. H. Replacement of halogen in orthobromo-
benzoic acid. J. Chem. Soc. 1929, 1870–1873.
(46) Cirigottis, K. A.; Ritchie, E.; Taylor, W. C. Studies on the
Hurtley reaction. Aust. J. Chem. 1974, 27, 2209–2228.
(47) Shinkwin, A. E.; Whish, W. J. D.; Threadgill, M. D. Synthesis of
a series of thiophenecarboxamides, thieno[3,4-c]pyridin-4(5H)-ones
and thieno[3,4-d]pyrimidin-4(3H)-ones and preliminary evaluation as
inhibitors of poly(ADP-ribose)polymerase (PARP). Bioorg. Med. Chem.
1999, 7, 297–308.
(48) Oliver, A. W.; Amꢀe, J.-C.; Roe, S. M.; Good, V.; de Murcia, G.;
Pearl, L. H. Crystal structure of the catalytic fragment of murine
poly(ADP-ribose) polymerase-2. Nucleic Acids Res. 2004, 32, 456–464.
(49) Ruf, A.; Mꢀenissier de Murcia, J.; de Murcia, G.; Schulz, G. E.
Structure of the catalytic fragment of poly(AD-ribose) polymerase from
chicken. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7481–7485.
(50) Ruf, A.; de Murcia, G.; Schulz, G. E. Inhibitor and NADþ
binding to poly(ADP-ribose) polymerase as derived from crystal
structures and homology modeling. Biochemistry 1998, 37, 3893–3900.
(51) Watson, C. Y.; Whish, W. J. D.; Threadgill, M. D. Synthesis of
3-substituted benzamides and 5-substituted isoquinolin-1(2H)-ones
and preliminary evaluation as inhibitors of poly(ADP-ribose)polymerase
(PARP). Bioorg. Med. Chem. 1998, 6, 721–734.
(34) Penning, T. D.; Zhu, G.-D.; Gandhi, V. B.; Gong, J.; Liu, X.;
Shi, Y.; Klinghofer, V.; Johnson, E. F.; Donawho, C. K.; Frost, D. J.;
Bontcheva-Diaz, V.; Bouska, J. J.; Osterling, D. J.; Olson, A. M.; Marsh,
K. C.; Luo, Y.; Giranda, V. L. Discovery of the poly(ADP-ribose)
polymerase (PARP) inhibitor 2-[(R)-2-methylpyrrolidin-2-yl]-1H-ben-
zimidazole-4-carboxamide (ABT-888) for the treatment of cancer.
J. Med. Chem. 2009, 52, 514–523.
(35) Eltze, T.; Boer, R.; Wagner, T.; Weinbrenner, S.; McDonald,
M. C.; Thiemermann, C.; B€urkle, A.; Klein, T. Imidazoquinolinone,
imidazopyridine, and isoquinolindione derivatives as novel and potent
inhibitors of the poly(ADP-ribose) polymerase (PARP): a comparison
with standard PARP inhibitors. Mol. Pharmacol. 2008, 74, 1587–1598.
(36) Menear, K. A.; Adcock, C.; Boulter, R.; Cockcroft, X.-L.;
Copsey, L.; Cranston, A.; Dillon, K. J.; Drzewiecki, J.; Garman, S.;
Gomez, S.; Javaid, H.; Kerrigan, F.; Knights, C.; Lau, A.; Loh, V. M.;
Matthews, I. T. W.; Moore, S.; O’Connor, M. J.; Smith, G. C. M.; Martin,
N. M. B. 4-[3-(4-Cyclopropanecarbonylpiperazine-1-carbonyl)-4-
fluorobenzyl]-2H-phthalazin-1-one: a novel bioavailable inhibitor of
poly(ADP-ribose) polymerase-1. J. Med. Chem. 2008, 51, 6581–6591.
(37) Nakajima, H.; Kakui, N.; Ohkuma, K.; Ishikawa, M.; Hasegawa,
T. A newly synthesized poly(ADP-ribose) polymerase inhibitor,
DR2313 [2-methyl-3,5,7,8-tetrahydrothiopyrano[4,3-d]pyrimidine-4-
one]: pharmacological profiles, neuroprotective effects, and therapeutic
time window in cerebral ischemia in rats. J. Pharmacol. Exp. Ther. 2005,
312, 472–481.
(38) Ishida, J.; Yamamoto, H.; Kido, Y.; Kamijo, K.; Murano, K.;
Miyake, H.; Ohkubo, M.; Kinoshita, T.; Warizaya, M.; Iwashita, A.;
Mihara, K.; Matsuoka, N.; Hattori, K. Discovery of potent and selective
PARP-1 and PARP-2 inhibitors: SBDD analysis via a combination of
X-ray structural study and homology modeling. Bioorg. Med. Chem.
2006, 14, 1378–1390.
(39) 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-chlorophenyl)-5-
quinoxalinecarboxamide], ameliorates methamphetamine-induced do-
paminergic neurotoxicity in mice. J. Pharmacol. Exp. Ther. 2004,
310, 1114–1124.
(52) Ferrer, S.; Naughton, D. P.; Parveen, I.; Threadgill, M. D. N- and
O-Alkylation of isoquinolin-1-ones in the Mitsunobu reaction: develop-
ment of potential drug delivery systems. J. Chem. Soc., Perkin Trans. 1
2002, 335–340.
(53) Berry, J. M.; Threadgill, M. D. Labelled compounds of interest
as antitumour agents. V. Syntheses of [18O]-5-methylisoquinolinone
and 1-(furan-2-yl-[18O]-methoxy)-5-methylisoquinoline: Correction.
J. Labelled Compd. Radiopharm. 1997, 39, 453 (Correction).
(54) Berry, J. M.; Threadgill, M. D. Labelled compounds of interest
as antitumour agents. V. Syntheses of [18O]-5-methylisoquinolinone
and 1-(furan-2-yl-[18O]-methoxy)-5-methylisoquinoline. J. Labelled
Compd. Radiopharm. 1996, 38, 935–940.
(55) Dillon, K. J.; Smith, G. C. M.; Martin, N. M. B. A FlashPlate assay for
the identification of PARP-1 inhibitors. J. Biomol. Screening 2003, 8, 347–352.
(56) Moroni, F.; Formentini, L.; Gerace, E.; Camaioni, E.; Pellegrini-
Giampietro, D. E.; Chiarugi, A.; Pellicciari, R. Selective PARP-2 inhibitors
increase apoptosis in hippocampal slices but protect cortical cells in models
of post-ischaemic brain damage. Br. J. Pharmacol. 2009, 157, 854–862.
(57) Cai, L.; Threadgill, M. D.; Wang, Y.; Li, M. Effect of poly (ADP-
ribose) polymerase-1 inhibition on the proliferation of murine colon
carcinoma CT26 cells. Pathol. Oncol. Res. 2009, 15, 323–328.
(58) Miyashiro, J.; Woods, K. W.; Park, C. H.; Liu, X.; Shi, Y.;
Johnson, E. F.; Bouska, J. J.; Olson, A. M.; Luo, Y.; Fry, E. H.; Giranda,
V. L.; Penning, T. D. Synthesis and SAR of novel tricyclic quinoxalinone
inhibitors of poly(ADP-ribose)polymerase-1 (PARP-1). Bioorg. Med.
Chem. Lett. 2009, 19, 4050–4054.
(59) Karlberg, T.; Hammarstr€om, M.; Sch€utz, P.; Svensson, L.;
Sch€uler, H. Crystal structure of the catalytic domain of human PARP2 in
complex with PARP inhibitor ABT-888. Biochemistry 2010, 49, 1056–1058.
(60) Banasik, M.; Komura, H.; Shimoyama, M.; Ueda, K. Specific
inhibitors of poly(ADP-ribose) synthetase and mono(ADP-ribosyl)-
transferase. J. Biol. Chem. 1992, 267, 1569–1575.
(40) Wenkert, E.; Johnston, B. D. R.; Dave, K. G. Derivatives of
hemimellitic acid. A synthesis of erythrocentaurin. J. Org. Chem. 1964,
29, 2534–2542.
(61) Elpern, B.; Hamilton, C. S. Arsenicals in the isoquinoline series.
(41) Woon, E. C. Y.; Dhami, A.; Sunderland, P. T.; Chalkley, D. A.;
Threadgill, M. D. Reductive cyclisation of 2-cyanomethyl-3-nitrobenzo-
ates. Lett. Org. Chem. 2006, 3, 619–621.
(42) Berry, J. M.; Watson, C. Y.; Whish, W. J. D.; Threadgill, M. D. 5-
Nitrofuran-2-ylmethyl group as a potential bioreductively activated
prodrug system. J. Chem. Soc., Perkin Trans. 1 1997, 1147–1156.
(43) Wong, S.-M.; Shah, B.; Shah, P.; Butt, I. C.; Woon, E. C. Y.;
Wright, J. A.; Thompson, A. S.; Upton, C.; Threadgill, M. D. A
new synthesis of “pushꢀpull” naphthalenes by condensation of
J. Am. Chem. Soc. 1946, 68, 1436–1438.
(62) Sienkowska, M.; Benin, V.; Kaszynski, P. Preparation and NMR
analysis of 2,6-heterodifunctional halobenzenes as precursors for sub-
stituted biphenyls. Tetrahedron 2000, 56, 165–173.
(63) Sheldrick, G. M. SHELXL-97, a Computer Program for Crystal
Structure Refinement; University of G€ottingen: G€ottingen, Germany, 1997.
2059
dx.doi.org/10.1021/jm1010918 |J. Med. Chem. 2011, 54, 2049–2059