Synthesis of 4-alkyl-, 4-aryl- and 4-arylamino-5-aminoisoquinolin-1-ones and identification of a new PARP-2 selective inhibitor

Article abstract of DOI:10.1039/c0ob00665c

The considerable interest in substituted isoquinolin-1-ones related to 5-aminoisoquinolin-1-one (5-AIQ) as drugs points to a need for an efficient and straightforward synthesis of the 4,5-disubstituted bicycles. Bromination of 5-nitroisoquinolin-1-one gave 4-bromo-5-nitroisoquinolin-1-one but neither this nor 5-amino-4-bromoisoquinolin-1-one would participate in Pd-catalysed couplings. Protection of the lactam as 1-methoxy- and 1-benzyloxy-4-bromo-5- nitroisoquinolines, however, permitted Stille, Suzuki and Buchwald-Hartwig couplings to take place in high yields, insensitive to electronic demands and severe steric bulk in the arylboronic acids. Lithiation of 4-bromo-1-methoxy-5- nitroisoquinoline and quench with iodomethane gave 1-methoxy-4-methyl-5- nitroisoquinoline in low yield. Demethylation of the 1-methoxy-4-substituted-5- nitroisoquinolines with hydrogen bromide gave 4-substituted-5-nitroisoquinolin- 1-ones, whereas hydrogenolytic debenzylation was achieved with simultaneous reduction of the 5-nitro group. 5-Amino-4-(4-trifluoromethylphenyl)isoquinolin- 1-one was identified as a new potent and selective inhibitor of poly(ADP-ribose)polymerase-2 (PARP-2).

Full text of DOI:10.1039/c0ob00665c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Synthesis of 4-alkyl-, 4-aryl- and 4-arylamino-5-aminoisoquinolin-1-ones and
identification of a new PARP-2 selective inhibitor†
Peter T. Sunderland,a Archana Dhami,a Mary F. Mahon,b Louise A. Jones,c Sophie R. Tully,c
Matthew D. Lloyd,a Andrew S. Thompson,a Hashim Javaid,c Niall M. B. Martinc and Michael D. Threadgill*a
Received 3rd September 2010, Accepted 22nd October 2010
DOI: 10.1039/c0ob00665c
The considerable interest in substituted isoquinolin-1-ones related to 5-aminoisoquinolin-1-one
(5-AIQ) as drugs points to a need for an efficient and straightforward synthesis of the 4,5-disubstituted
bicycles. Bromination of 5-nitroisoquinolin-1-one gave 4-bromo-5-nitroisoquinolin-1-one but neither
this nor 5-amino-4-bromoisoquinolin-1-one would participate in Pd-catalysed couplings. Protection of
the lactam as 1-methoxy- and 1-benzyloxy-4-bromo-5-nitroisoquinolines, however, permitted Stille,
Suzuki and Buchwald–Hartwig couplings to take place in high yields, insensitive to electronic demands
and severe steric bulk in the arylboronic acids. Lithiation of 4-bromo-1-methoxy-5-nitroisoquinoline
and quench with iodomethane gave 1-methoxy-4-methyl-5-nitroisoquinoline in low yield.
Demethylation of the 1-methoxy-4-substituted-5-nitroisoquinolines with hydrogen bromide gave
4-substituted-5-nitroisoquinolin-1-ones, whereas hydrogenolytic debenzylation was achieved with
simultaneous reduction of the 5-nitro group. 5-Amino-4-(4-trifluoromethylphenyl)isoquinolin-1-one
was identified as a new potent and selective inhibitor of poly(ADP-ribose)polymerase-2 (PARP-2).
Introduction
Isoquinolin-1-ones are of considerable interest as potential
drugs, particularly the 5-substituted analogues, which are
potent inhibitors of poly(ADP-ribose)polymerases (PARPs).
Simple 5-substituted isoquinolin-1-ones and 5-substituted 3,4-
dihydroisoquinolin-1-ones were reported as long ago as 1991
to be potent inhibitors of PARPs in vitro and in cells1,2 and 5-
hydroxyisoquinolin-1-one 1 (often mis-named as the tautomer
1,5-dihydroxyisoquinoline 2, Fig. 1) has shown good inhibition
of the enzyme in vitro and in models of inflammation and
other PARP-mediated diseases in vivo.3 However, of this series,
it is the 5-amino analogue 5-AIQ 3 which has shown most
promise as an inhibitor of PARPs, partly owing to the exceptional
solubility in water of its hydrochloride salt. 5-AIQ 3 is active
in models in vivo and in vitro of a wide range of disease states,
Fig. 1 Structures of isoquinolin-1-one inhibitors of PARP-1 and of
including colitis,4 ischaemic heart disease,5 haemorrhagic shock6
PARP-2. Compounds 4–6 are selective inhibitors of PARP-2.
and spinal cord trauma,7 and has recently been shown to have
strong antimetastatic effects in a murine model of cancer.8 We
have recently reported9 that 5-benzamidoisoquinolin-1-ones 4 and
paucity of reports of preparation of such compounds in the
chemical journal literature, shown in Scheme 1. We have dis-
closed recently that palladium-catalysed cyclisation of N-allyl-
2-iodo-3-nitrobenzamide 7a at high-temperature gives 4-methyl-
5-nitroisoquinolin-1-one 8a in low yield; similar reaction of N-
cinnamyl-2-iodo-3-nitrobenzamide 7b leads inefficiently to 4-
benzyl-5-nitroisoquinolin-1-one 8b.11 These products can be then
reduced readily to their 5-amino analogues 9a,b. Croisy-Delcey
et al. achieved the synthesis of 5-methoxy-4-methylisoquinolin-
1-one 11 in moderate yield by Curtius rearrangement/thermal
cyclisation of the acyl azide 10.12 Sercel et al. prepared
4-bromo-5-methylisoquinolin-1-one 13 by bromination of 5-
bromoisoquinolin-1-one 12 with pyridinium perbromide; this was
converted to the lactim 14 and from this they were then able
to introduce formyl, methyl and methylthio at the 4-position
(forming 15a–c) by lithiation and quench with an appropriate
one 3-substituted 5-benzamidoisoquinolin-1-one 5 are selective
inhibitors of the PARP-2 isoform; 5-benzoyloxyisoquinolin-1-one
6 also selectively inhibits PARP-2.10
In the light of these biological activities, we wished to ex-
plore 4,5-disubstituted isoquinolin-1-ones. There is a marked
aMedicinal Chemistry, Department of Pharmacy & Pharmacology, Univer-
sity of Bath, Claverton Down, Bath, UK BA2 7AY. E-mail: m.d.threadgill@
bath.ac.uk; Fax: +44 1225 386114; Tel: +44 1225 386840
bCrystallography Unit, Department of Chemistry, University of Bath,
Claverton Down, Bath, UK BA2 7AY
cKuDOS Pharmaceuticals Ltd., 410 Cambridge Science Park, Milton Road,
Cambridge, UK CB4 0PE
† CCDC reference numbers 781105 (26a), 781106 (33) and 781107 (35).
For crystallographic data in CIF or other electronic format see DOI:
10.1039/c0ob00665c
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 881
©

View Article Online
Scheme 1 Previous syntheses of 4,5-disubstituted isoquinolin-1-ones. Reagents: i, (Ph3P)4Pd, Et3N, Bu4NCl, DMF; ii, H2, Pd/C, EtOH, aq. HCl; iii,
Bu3N, Ph2O, 240 ◦C; iv, pyridinium+ Br3-, CH2Cl2; v, (COCl)2, DMF, (CH2Cl)2, vi, NaOMe, MeOH; vii, BuLi, THF; viii, DMF or MeI or MeSSMe; ix,
aq, HBr.
electrophile, followed by demethylation with hydrogen bromide.13
There is thus a strong need to develop efficient syntheses of 4,5-
disubstituted isoquinolin-1-ones with nitrogen substituents in the
5-position and opportunities for diversity at the 4-position to
be able, inter alia, to explore the structure–activity relationships
for inhibition of the numerous isoforms of PARP and other
enzymes.
a small additional amount of 18 but mainly the unbrominated
starting material 16. As it is known that isocoumarins are easily
converted into the corresponding isoquinolin-1-ones, bromination
of 5-nitroisocoumarin 20 was also investigated; in this case,
reaction with bromine in acetic acid furnished only a good yield
of the trans-3,4-dibromo compound 21. Interestingly, the small
coupling constant (3J = 1.7 Hz) between the 3-H and the 4-H in
the 1H NMR spectrum of 21 is only consistent with a trans-diaxial
arrangement of the bromine atoms, presumably to avoid steric
clash of the 3-Br with the adjacent bulky nitro group.
Chemistry
Unfortunately, 4-bromo-5-nitroisoquinolin-1-one 18 failed to
couple in Sonogashira, Stille and Suzuki reactions under a wide
variety of conditions; in most cases, 18 was recovered but with
some of the debrominated analogue 16 also being obtained. The
formation of 16 indicated that some palladation had taken place.
To test whether or not the failure to couple was due to steric
hindrance from the peri nitro group, 18 was reduced selectively
with tin(II) chloride to the 5-amino-4-bromo compound 22, under
conditions designed to avoid hydrogenolysis of the C–Br bond.16
Frustratingly, Pd-catalysed couplings to 22 also failed (Scheme 2).
Many isoquinolin-1-ones have very limited solubility in solvents
which are appropriate for Pd-catalysed couplings and 18 and 22
are no exception. To attempt to alleviate this potential problem,
the lactam moiety was masked. As shown in Scheme 2, reaction
of 18 with the Vilsmeier reagent generated in situ from DMF
and oxalyl chloride gave the 1-chloroisoquinoline 23 in excellent
yield, from which the chloride was displaced by methoxide
to furnish 4-bromo-1-methoxy-5-nitroisoquinoline 24. Following
Sercel’s approach to 15b, exchange of bromine for lithium with
butyl lithium, followed by quench of the anion with iodomethane,
furnished the 4-methyl compound 25 but in very poor yield.
It could be speculated that the peri nitro group had interfered
either with the lithiation or with the subsequent reaction with
the electrophile. This lithiation/quench sequence could also not
be applied to the introduction of aryl groups at the 4-positions,
owing to lack of an appropriate electrophile, so other couplings
were sought.
Palladium-catalysed couplings usually offer great opportunities to
introduce a wide range of substituents onto a heterocyclic core un-
der relatively mild conditions, providing chemical diversity rapidly
and efficiently. Thus these methods were explored for attachment
of the 4-substituents. 5-Nitroisoquinolin-1-one 16 (Scheme 2) is
readily accessible as a starting material carrying the required 5-
nitrogen substituent6,9 and it was expected that a halogen could
be introduced electrophilically to the 4-position, as this is the
most nucleophilic in this heterocycle.14 It proved impossible to
iodinate at this position using a variety of reagents and conditions
(iodine in acetic acid, N-iodosuccinimide in acetic acid, etc.); even
activation of molecular iodine by Ag+ ions (I2/AgOTf) failed to
effect any conversion of the starting isoquinolin-1-one 16. Horning
et al.15 have reported that reaction of 2-methyl-5-nitroisoquinolin-
1-one with bromine in acetic acid afforded an equimolar mix-
ture of 4-bromo-2-methyl-5-nitroisoquinolin-1-one and 4-bromo-
3-hydroxy-2-methyl-5-nitro-3,4-dihydroisoquinolin-1-one and we
sought to investigate whether or not this process could be adapted
to the N-unsubstituted analogue. Pleasingly, treatment of a
concentratedsolution of 16 with bromine inhotacetic acid gave the
required 4-bromo compound 18 in moderate yield, accompanied
by a significant amount of the 3-hydroxy-3,4-dihydro analogue 17.
◦
Heating 17 to 175 C in the absence of solvent for several hours
eliminated water to give a small additional yield of 16. Bromination
of 16 with N-bromosuccinimide (NBS) in hot acetic acid gave 18
in lower yield, together with a moderate yield of the 3-◦acetoxy-
3,4-dihydro analogue 19. Again, thermolysis of 19 (137 C) gave
882 | Org. Biomol. Chem., 2011, 9, 881–891
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Scheme 2 Pd-catalysed couplings to 4-bromo-1-methoxyisoquinolin-1-one 24. Reagents and conditions: i, Br2, AcOH, 60 ◦C, 52% (18), 26% (17); ii,
NBS, AcOH, 33% (19), 21% (18); iii, Br2, AcOH, 52%; iv, SnCl2, EtOH, 80 ◦C, 54%; v, (COCl)2, DMF, (CH2Cl)2, 80 ◦C, 89%; vi, Na, MeOH, reflux,
82%; vii, BuLi, THF, MeI, -78 ◦C → 20 ◦C, 9%; viii, SnMe4, Pd2(dba)3, SPhos, PhMe, 100 ◦C, 72%; ix, PhB(OH)2 or 4-F3CPhB(OH2), Pd2(dba)3, SPhos,
PhMe, 100 ◦C, 86% (28a), 81% (28b); x, PhNH2, Pd2(dba)3, SPhos, KOBut, 1,4-dioxane, 100 ◦C, 45%. Compounds 17, 19 and 21 are racemic.
Palladium-catalysed couplings are relatively insensitive to
the presence of other substituents in the substrates. Stille
Exploring the generality of this reaction, Buchwald–Hartwig
couplings were attempted with aniline, with thiophenol and with
phenol. Only the former coupled effectively, to give the deep
red 4-phenylamino analogue 27. Optimisation of this Buchwald–
Hartwig coupling of 24 with aniline revealed that the optimum
ligand was SPhos (compared with XPhos, BuXPhos and John-
Phos), the optimum base was potassium t-butoxide (compared
with tripotassium phosphate) and the optimum solvent was 1,4-
dioxane (compared with toluene and DMF) (Table 1).
Scheme 3 shows how the coupled products were converted to
the corresponding isoquinolin-1-ones. The lactim unit of 25 was
deprotected to the lactam by demethylation with hydrobromic
acid to give 4-methyl-5-nitroisoquinolin-1-one 28. From here,
simple catalytic hydrogenation of the nitro group provided the
5-amino analogue 29. The 5-aryl-1-methoxyisoquinolines 26a,b
were similarly demethylated to the 4-aryl-5-nitroisoquinolin-1-
ones 30a,b, which were reduced to the 5-amino-4-arylisoquinolin-
1-ones 31a,b.
coupling of 24 with tetramethyltin using
a conventional
tetrakis(triphenylphosphine)palladium(0) catalyst failed but
the Pd2(dba)3/SPhos catalyst/ligand system introduced by
Buchwald17 for enhancing Suzuki couplings was successful in
providing 25 in 72% yield. Similarly, Suzuki coupling of 24
with phenylboronic acid and with 4-trifluoromethylphenylboronic
acid (an electron-poor boronic acid often associated with poor
coupling yields18) proceeded very efficiently with Pd2(dba)3/SPhos
in toluene in providing 26a and 26b, respectively, in high yields.
The structure of coupled product 26a was confirmed by X-ray
crystallography (Fig. 2). Interestingly, rhomboids of dimensions
1 cm ¥ 1 cm ¥ 0.3 cm could be grown readily from ethyl
acetate–hexane. The crystal structure shows clearly that the 5-
nitro is rotated significantly out of the plane of the bicycle
but the 4-phenyl remains coplanar with it, retaining maximum
conjugation.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 883
©

View Article Online
Table 1 Yields of 27 obtained during optimisation of the Buchwald–Hartwig coupling of 24 with aniline
Ligand
PhMe/K3PO4
DMF/K3PO4
Dioxane/K3PO4
PhMe/KOBut
DMF/KOBut
Dioxane/KOBut
XPhosa
30%
32%
11%
0%
36%
38%
5%
38%
36%
10%
0%
32%
34%
9%
38%
40%
9%
34%
45%
14%
0%
SPhosb
t-BuXPhosc
JohnPhosd
0%
0%
0%
a 2-Dicyclohexylphosphino-2¢,4¢,6¢-triisopropylbiphenyl. b 2-Dicyclohexylphosphino-2¢,6¢-dimethoxybiphenyl. c 2-Di-t-butylphosphino-2¢,4¢,6¢-triisopro-
pylbiphenyl. d 2-Di-t-butylphosphinobiphenyl.
Fig. 2 X-Ray crystal structures of 4,5-disubstituted isoquinolines 26a, 33 and 35. Ellipsoids are represented at 30% probability. Solvent in the structure
of 35 is omitted for clarity.
This demethylation is inappropriate to the synthesis of 34 and
31c carrying a 4-(4-methoxyphenyl) group, as this substituent
would also be demethylated by the hydrogen bromide. A modified
sequence is shown in Scheme 4, in which the protecting 1-methoxy
group is replaced by a benzyloxy group which can be removed
under conditions which retain the 4-(4-methoxyphenyl) unit.
Reaction of the 1-chloro compound 23 with sodium benzyloxide
in boiling DMF furnished the required 1-OBn protected com-
pound 33 in moderate yield but a significant amount of the 1-
dimethylaminoisoquinoline 32 was also isolated. This material
arose from thermal degradation of the solvent, liberating highly
nucleophilic dimethylamine which reacted with the electrophilic
23. The structure of 33 was confirmed by X-ray crystallography.
In this structure (Fig. 2), the molecule is essentially planar, with
the benzyloxy function pointing away from the core. Again, the
nitro group is peri to a large group in the 4-position (bromine),
resulting in the nitro group being twisted out of plane and the C–Br
bond being bent away from the nitro group. Pd-catalysed Suzuki
coupling of 33 with 4-methoxyphenylboronic acid smoothly gave a
high yield of 34. From here, removal of the benzyl protecting group
and reduction of the 5-nitro group was achieved in one step, giving
31c. To provide a severe test of any steric constraints on this new
Suzuki coupling to the 4-bromo-5-nitroisoquinolines, coupling of
33 with phenanthrene-9-boronic acid was attempted. Surprisingly
for such a large aromatic group approaching the 4-position of
the isoquinoline with the bulky peri 5-nitro group, coupling was
effective under the standard conditions, giving a 42% yield of
35. The X-ray crystal structure of this highly crowded extended
binaphthyl was determined (Fig. 2). As expected, both the 5-nitro
and the 4-phenanthrene substituents are twisted severely out of
the plane of the isoquinoline to relieve the steric compression.
This is shown by the asymmetry in the angles at C8 involving the
phenanthrene (C7–C8–C17 126.6, C9–C8–C17 116.3◦) and also by
the C8–C17–C30–C29 torsion angle of 171◦. Nearest neighbour
4-phenanthrene groups in the gross structure are approximately
˚
coplanar, with a separation distance of 3.68 A. The O-benzyl group
was removed by hydrogenolysis, with simultaneous reduction of
the nitro group, to give 36. This compound was insoluble in all
common solvents, precluding both characterisation by NMR and
any biological evaluation.
As 5-benzamidoisoquinolin-1-one 4 was highly selective for
inhibition of the PARP-2 isoform,9 31a was benzoylated at the
exocyclic amine to furnish 37 (Scheme 5) as one example of a
4-substituted 5-benzamidoisoquinolin-1-one.
884 | Org. Biomol. Chem., 2011, 9, 881–891
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
more inhibitory towards PARP-2. Thus 31b is almost as selective
(7.6-fold) for PARP-2 as is most selective compound 4 (Ar = Ph;
9.3-fold) reported to date9 and presents a new lead core to explore
further the structural requirements for selectivity. Curiously,
changing the electron-withdrawing trifluoromethyl group in 31b
for an electron-donating methoxy group in 31c decreased binding
to PARP-2 and hence selectivity. The most selective lead inhibitor
4 (Ar = Ph) contains a 5-benzamido group but including this
into the 4-aryl series in 37 completely abolished activity against
both isoforms. This is probably owing to the steric crowding
between the 4- and 5-peri-substituents evident in the MM2-
minimised structure of 37 (Fig. 3), distorting the bicycle and
related to that observed in the crystal structure of 35 (Fig. 2).
Fig. 4 shows the results of post facto modelling of the structure
of 31b complexed to the NAD+-binding site of human PARP-
2. The starting structure was of human PARP-2 complexed with
the non-selective inhibitor ABT888.19 PARP-2-selective inhibitor
31b was then docked into the models using the existing bound
inhibitor as template. Once docked, the inhibitor was subjected
to molecular mechanics and dynamics calculations to establish
optimal docking conformations; during these calculations, the
receptor was restrained to its original conformation. Lastly, both
˚
the inhibitors and binding pockets (radius 10 A) were subjected to
molecular dynamics and finally molecular mechanics calculations
to give the final structure (Fig. 3). The larger binding pocket
of PARP-29 accommodates the 4-(4-trifluoromethyphenyl) group,
whereas the smaller pocket of PARP-19 does not.
Scheme 3 Formation of the 5-aminoisoquinolin-1-ones 29 and 31a,b.
Reagents and conditions: i, aq. HBr, 80 ◦C, 70%; ii, aq. HBr, 50 ◦C, 65%
(30a), 65% (30b); iii, H2, Pd/C, EtOH, aq. HCl, 65%; iv, H2, Pd/C, EtOH,
51% (31a), 53% (31b).
Biochemical evaluation
Fig. 3 MM2-Minimised structure of 37, showing steric crowding between
the peri 4,5-substituents.
Selected isoquinoline-1-ones 22, 30a, 31b,c and 37 were evaluated
for their inhibition of the catalytic activities of PARP-1 and PARP-
2; the data are presented in Table 2. Comparative data are also
given for the non-isoform-selective inhibitor 5-AIQ 3,9 for 4 (Ar =
Ph), which is 9.3-fold selective for inhibition of PARP-2,9 and
for 6, which is 2.75-fold selective for PARP-29 (60-fold claimed
by Pellicciari et al.10). Introduction of the 4-bromo substituent
in 22 led to increase in potency against both PARP isoforms,
relative to the lead non-selective inhibitor 5-AIQ 3; notably, the
activity against PARP-2 was increased over 4-fold, leading to a
2.6-fold selectivity for inhibition of PARP-2 by 22. The 5-nitro-4-
phenyl analogue 30a was non-selective and less potent than was
3, as expected for 5-nitroisoquinolin-1-ones which are generally
weaker inhibitors of PARP enzymes.1 By contrast, 31b, which
corresponds to 5-AIQ 3 but carrying a 4-trifluorophenyl group at
the 4-position, is four-times less potent than 3 against PARP-1 but
Conclusions
In this paper, we report that palladium-catalysed couplings (Stille,
Suzuki, Buchwald–Hartwig) to the sterically very crowded 4-
position of 1-alkoxy-4-bromo-5-nitroisoquinolines 24 and 33
are very efficient in providing 4-alkyl- and 4-aryl-1-alkoxy-5-
nitroisoquinolines. The Suzuki coupling with arylboronic acids is
insensitive to electron-withdrawing (-CF3) and electron-donating
(-OMe) groups on the phenylboronic acid. Surprisingly, major
steric bulk is also tolerated in the coupling reaction, in that
phenanthrene-9-boronic acid is also a satisfactory coupling part-
ner in the formation of 35. Analogous couplings were not possible
using the corresponding isoquinolin-1-ones 18 and 22, probably
owing to poor solubility of these lactams. 1-Alkoxyisoquinolines
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 885
©

View Article Online
Table 2 Inhibition of the activities of PARP-1 and PARP-2 by 4,5-disubstituted isoquinolin-1-ones 22, 30a, 31b,c and 37; data for 5-AIQ 3, 5-
benzamidoisoquinolin-1-one 4 (Ar = Ph) and 5-benzoyloxyisoquinolin-1-one 6 are shown for comparison9
Cpd. No.
4-Substituent
5-Substituent
PARP-1 IC50/mM
PARP-2 IC50/mM
Observed selectivitya
22
Br
Ph
4-F3CPh
4-MeOPh
Ph
H
H
H2N
O2N
H2N
H2N
PhCONH
H2N
PhCONH
PhCO2
0.56
3.8
4.1
0.22
2.5
0.54
2.2
>100
1.05
1.5
2.6
1.5
7.6
2.2
—
0.9
9.3
2.8
30a
31b
31c
37
3
4 (Ar = Ph)
6
4.7
>100
0.94
13.9
4.10
H
1.49
a IC50 (PARP-1)/IC50 (PARP-2)
.
Scheme 4 Pd-catalysed couplings to 1-benzyloxy-4-bromo-5-nitroisoquinolin-1-one 33. Reagents and conditions: i, BnOH, NaH, DMF, 100 ◦C, 71%
(33), 12% (32); ii, 4-MeOPhB(OH)2, Pd2(dba)3, SPhos, K3PO4, PhMe, 100 ◦C, 61%; iii, phenanthrene-9-boronic acid, Pd2(dba)3, SPhos, K3PO4, PhMe,
100 ◦C, 42%; iv, Pd/C, H2, EtOH, 47% (31c), 57% (36).
compounds 25 and 26a,b led to the 5-nitroisoquinolin-1-ones 28
and 30a,b, respectively, for later reduction to the 4-substituted
5-AIQs 29 and 31a,b. Simplifying the syntheses of 4-substituted
5-AIQs further, catalytic hydrogenolysis simultaneously removed
the protecting O-benzyl group and reduced the nitro function of
34 and 35 to access the 4-aryl 5-AIQs 31c and 36, respectively.
Benzoylation of 31a gave 5-benzamido-4-phenylisoquinolin-1-
one 37, despite the severe crowding in the product. Selected
4,5-disubstituted isoquinoline-1-ones were evaluated as isoform-
selective inhibitors of PARP-2; 4-(4-trifluoromethylphenyl)-5-
AIQ 31b was particularly potent and selective and is a new lead
in the search for isoform-selectivity for this important family of
enzymes.
Experimental
General
Fig. 4 Molecular modelling of 31b bound to the NAD+-binding site of
NMR spectra were recorded on JEOL Delta 270 and Var-
ian Mercury 400 spectrometers of solutions in deuteriochloro-
form, unless otherwise stated; coupling constants (J) are given
in Hz. Mass spectra were obtained using VG7070E and Bruker
human PARP-2.
can be considered as masked isoquinoline-1-ones22 and 25, 26a,b,
34 and 35 are no exception. Demethylation of the 1-methoxy
886 | Org. Biomol. Chem., 2011, 9, 881–891
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
8.0, 1.2, 8-H). Further elution gave 18 (300 mg, 21%), with data
as above.
( )-trans-3,4-Dibromo-5-nitroisocoumarin (21)
Bromine (210 mg, 1.3 mmol) in acetic acid (1.0 mL) was added
to 5-nitroisocoumarin 206 (250 mg, 1.3 mmol) in acetic acid
(2.5 mL) and the mixture was stirred for 2 h before being
poured into ice–H2O (100 mL). The suspension was stirred
for 10 min, then extracted (CH2Cl2). Drying, evaporation and
chromatography (hexane–EtOAc 4 : 1) yielded 21 (240 mg, 52%)
as white crystals: mp 116–118 ◦C; nmax 1761, 1528, 1346 cm-1;
dH 6.32 (1 H, d, J 1.7, 4-H), 6.90 (1 H, d, J 1.7, 3-H), 7.80
(1 H, t, J 8.2, 7-H), 8.52 (1 H, s, 6-H), 8.64 (1 H, s, 8-H); dC
(HMQC/HMBC) 38.4 (4-C), 78.2 (3-C), 125.1 (9-C), 130.9 (7-C),
131.3 (8-C), 133.0 (10-C), 135.7 (6-C), 145.3 (5-C), 158.3 (1-C);
m/z 373.8455 (M + Na) (C9H679Br81BrNO4Na requires 373.8463),
353.8621 (M + H) (C9H681Br2NO4 requires 353.8623), 351.8647
(M + H) (C9H679Br81BrNO4 requires 351.8642), 349.8658 (M + H)
(C9H679Br2NO4 requires 349.8664).
Scheme 5 Benzoylation of 31a. Reagents and conditions: i, PhCOCl,
pyridine, 90 ◦C, 36%.
microTOFTM spectrometers in the ES+ mode. IR spectra were
measured on a Perkin–Elmer RXI FTIR spectrometer as KBr
discs. The stationary phase for chromatography was silica gel.
All reactions were carried out at ambient temperature, unless
otherwise stated. Solvents were evaporated under reduced pres-
sure. Melting points were determined using a Reichert–Jung
Thermo Galen instrument and are uncorrected. SPhos refers to
2-dicyclohexylphosphino-2¢,6¢-dimethoxybiphenyl and Pd2dba3
refers to tris(dibenzylideneacetone)dipalladium.
5-Amino-4-bromoisoquinolin-1(2H)-one (22)
Compound 18 (540 mg, 2.0 mmol) was heated with tin(II) chloride
(1.21 g, 6.4 mmol) in ethanol (20 mL) at 80 ◦C for 4 h, then carefully
poured into ice–H2O (75 mL). The suspension was made alkaline
with aq. NaOH and the precipitate was filtered. Extraction of the
filtrate (EtOAc), evaporation and chromatography (ethyl acetate–
hexane 4 : 1) gave 22 (250 mg, 54%) as a pale buff powder: mp
210–212 ◦C; nmax 3443, 3321, 1661, 1624 cm-1; dH ((CD3)2SO) 5.92
(2 H, br s, NH2), 7.73 (1 H, s, 3-H), 7.02 (1 H, dd, J 8.2, 1.6, 6-H),
7.22 (1 H, s, 3-H), 7.25 (1 H, t, J 8.2, 7-H), 7.54 (1 H, dd, J 7.8,
1.2, 8-H), 11.34 (1 H, br s, NH); dC ((CD3)2SO) (HMQC/HMBC)
93.10 (4-C), 115.83 (8-C), 119.17 (6-C), 119.42 (10-C), 128.04 (3-
C), 128.44 (9-C), 128.56 (7-C), 144.72 (5-C), 160.83 (1-C); m/z
238.9815 (M + H) (C9H879BrN2O requires 238.9820).
4-Bromo-5-nitroisoquinolin-1-one (18) and 4-bromo-3-hydroxy-5-
nitro-3,4-dihydroisoquinolin-1-one (17)
Bromine (5.0 g, 32 mmol) in acetic acid (5.0 mL) was added
slowly to a suspension of 16 (6.0 g, 32 mmol) in acetic acid
(15 mL). The mixture was stirred at 60 ◦C for 16 h, then
cooled and poured onto ice–water (60 mL). The precipitate
was collected, washed (methanol) and dried. Chromatography
(hexane–ethyl acetate 6 : 1) gave 18 (4.5 g, 52%) as a pale orange
solid: mp 229–232 ◦C; nmax 3467, 1674, 1534, 1368 cm-1; dH
((CD3)2CO) 7.73 (1 H, s, 3-H), 7.77 (1 H, t, J 7.8, 7-H), 8.10
(1 H, dd, J 7.8, 1.6, 6-H), 8.61 (1 H, dd, J 8.6, 1.9, 8-H); dC
((CD3)2CO) (HMQC/HMBC) 90.4 (4-C), 128.0 (10-C), 128.3
(7-C), 129.5 (6-C), 129.9 (9-C), 132.1 (8-C), 135.6 (3-C), 147.3
(5-C), 159.0 (1-C); m/z 292.9354 (M + Na) (C9H581BrN2O3Na
requires 292.9312), 290.9376 (M + Na) (C9H579BrN2O3Na re-
quires 290.9332), 267.9478 (M) (C9H579BrN2O3 requires 267.9484);
Found: C, 40.60; H, 1.61; N, 10.19. Calc. for C9H5BrN2O3: C,
40.18; H, 1.87; N, 10.41%. Further elution gave 17 (2.4 g, 26%) as
a pale yellow solid: mp 170–172 ◦C; dH ((CD3)2CO) 5.52 (1 H, dd,
J 5.7, 1.2, 3-H), 6.73 (1 H, d, J 5.9, 4-H), 7.82 (1 H, t, J 7.9, 7-H),
8.00 (1 H, dd, J 7.9, 1.5, 6-H), 8.30 (1 H, dd, J 7.7, 1.5, 8-H), 8.55
(1 H, br, NH).
4-Bromo-1-chloro-5-nitroisoquinoline (23)
Oxalyl chloride (5.3 mL, 7.67 g, 60.4 mmol) was added drop-
wise during 30 min to dry dimethylformamide (4.7 mL, 4.4 g,
60.4 mmol) in 1,2-dichloroethane (35 mL) at 0 ◦C. The suspension
was stirred at room temperature for 10 min, then 18 (7.3 g,
◦
27 mmol) was added. The mixture was then heated at 80 C for
6 h, allowed to cool and diluted with CH2Cl2. Washing (water),
drying and evaporation gave 23 (7.0 g, 89%) as a yellow solid: mp
164–166 ◦C; dH 7.82 (1 H, t, J 7.6, 7-H), 8.01 (1 H, dd, J 7.6, 1.2,
6-H), 8.62 (1 H, s, 3-H), 8.65 (1 H, dd, J 7.6, 1.2, 8-H); dC 112.4,
127.1, 127.8, 128.2, 128.6, 131.0, 147.4, 147.6, 152.0.
4-Bromo-5-nitroisoquinolin-1-one (18) and 3-acetoxy-4-bromo-5-
nitro-3,4-dihydroisoquinolin-1-one (19)
4-Bromo-1-methoxy-5-nitroisoquinoline (24)
N-Bromosuccinimide (90 mg, 0.5 mmol) was stirred with 16
(100 mg, 0.5 mmol) in acetic acid (5 mL) for 30 min. The mixture
was poured into ice–H2O (100 mL) and stirred for 10 min. Ex-
traction (ethyl acetate), washing (aq. sodium hydrogen carbonate,
water), drying, evaporation and chromatography (hexane–ethyl
acetate 4 : 1) yielded 19 (70 mg, 33%) as a pale buff solid: mp
137 ◦C; nmax 3462, 1675, 1534, 1335 cm-1; dH ((CD3)2CO) 2.08 (3
H, s, Me), 5.79 (1 H, d, J 4.7, 4-H), 5.95 (1 H, m, 3-H), 7.82 (1
H, t, J 7.9, 7-H), 8.33 (1 H, dd, J 8.2, 1.5, 6-H), 8.45 (1 H, dd, J
Finely divided sodium (700 mg, 31 mmol) was added to 23 (5.0 g,
17 mmol) in dry methanol (90 mL) and the mixture was boiled
under reflux for 16 h. The solvent was then evaporated until
20 mL remained; the residue was diluted with H2O and extracted
(chloroform). Drying and evaporation gave 24 (4.0 g, 82%) as a
yellow solid: mp 154–157 ◦C; dH 4.18 (3 H, s, Me), 7.63 (1 H, t, J
7.8, 7-H), 7.88 (1 H, dd, J 7.8, 1.1, 6-H), 8.29 (1 H, s, 3-H), 8.48 (1
H, dd, J 7.8, 1.1, 8-H); dC 54.6, 104.2, 110.0, 121.9, 126.3, 126.9,
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 887
©

View Article Online
128.5, 146.2, 147.0, 160.3; Found: C, 42.43; H, 2.63; N, 9.69. Calc.
For C10H7BrN2O3: C, 42.43; H, 2.49; N, 9.90%.
(M + H) (C17H12F3N2O3 requires 349.0800); Found: C, 58.47; H,
3.23; N, 7.96. Calc. for C17H11F3N2O3: C, 58.63; H, 3.18; N, 8.05%.
1-Methoxy-5-nitro-4-phenylaminoisoquinoline (27)
1-Methoxy-4-methyl-5-nitroisoquinoline (25) Method A
Compound 24 (1.00 g, 3.5 mmol), Pd2(dba)3 (180 mg, 0.35 mmol),
SPhos (140 mg, 0.70 mmol), potassium t-butoxide (790 mg,
7.06 mmol) and aniline (0.49 g, 5.3 mmol) were placed in a dry
flask. Degassed 1,4-dioxane (40 mL) was added and the mixture
was stirred at 100 ◦C for 16 h. Evaporation and chromatography
(hexane–ethyl acetate 10 : 1) gave 27 (470 mg, 45%) as a deep
red solid: mp 124–126 ◦C; dH 4.17 (3 H, s, OMe), 5.56 (1 H, s,
NH), 6.61 (2 H, dd, J 7.4, 1.1, Ph 2,6-H2), 6.79 (1 H, t, J 7.4,
Ph 4-H), 7.14 (2 H, t, J 7.4, Ph 3,5-H2), 7.59 (1 H, d, J 8.2,
7-H), 7.80 (1 H, dt, J 8.2, 1.2, 8-H), 8.14 (1 H, d, J 1.1, 3-H),
8.51 (1 H, dd, J 8.2, 1.2, 6-H); dC 54.3, 114.2, 119.5, 121.2, 124.7,
125.8, 126.3, 127.8, 128.7, 129.3, 142.3, 146.8, 158.7; m/z 318.0850
(M + Na) (C16H13N3NaO3 requires 318.0855), 296.1027 (M + H)
(C16H14N3O3 requires 296.1035.
Butyllithium in tetrahydrofuran (1.6 M, 0.24 mL, 0.38 mmol) was
added to 24 (100 mg, 0.35 mmol) in dry tetrahydrofuran (9 mL)
◦
at -78 C. The suspension was stirred for 20 min. Iodomethane
(55.4 mg, 0.39 mmol) in tetrahydrofuran (1 mL) was added and
the mixture allowed to warm to 20 ◦C during 1 h. The reaction was
quenched with water. Extraction (dichloromethane), evaporation
and chromatography (hexane–ethyl acetate 15 : 1) gave 25 (7 mg,
9%) as a yellow–orange solid: mp 90–93 ◦C; dH 2.91 (3 H, s,
4-Me), 4.09 (3 H, s, OMe), 7.50 (1 H, t, J 8.6, 7-H), 7.75 (1
H, d, J 7.4, 8-H), 7.87 (1 H, s, 3-H) 8.42 (1 H, d, J 7.4, 6-H);
dC (HMBC/HMQC) 16.00 (4-Me), 53.94 (OMe), 120.00 (4-C),
125.02 (7-C), 125.34 (4a-C), 125.66 (8-C), 128.34 (6-C), 128.91
(8a-C), 143.24 (5-C), 143.58 (3-C), 159.99 (1-C); m/z 241.0582
(M + Na) (C10H10N2NaO3 requires 241.0589), 219.0772 (M + H)
(C10H11N2O3 requires 219.0770).
4-Methyl-5-nitroisoquinolin-1-one (28)
Compound 25 (200 mg, 0.92 mmol) was stirred in aq. hydrobromic
acid (48%, 30 mL) at 80 ◦C for 4 h. Evaporation and recrystalli-
sation (hexane–ethyl ac◦etate) gave 28 (131 mg, 70%) as a pale buff
solid mp: mp 211–214 C (lit.11 mp 209–211 ◦C); dH ((CD3)2CO)
2.02 (3 H, s, Me), 7.22 (1 H, d, J 5.1, 3-H), 7.66 (1 H, t, J 7.8,
7-H), 8.13 (1 H, dd, J 7.8, 1.3, 6-H), 8.50 (1 H, dd, J 7.8, 1.3, 8-H),
11.64 (1 H, br, NH).
1-Methoxy-4-methyl-5-nitroisoquinoline (25) Method B
Compound 24 (1.00 g, 3.52 mmol), Pd2(dba)3 (180 mg, 0.35 mmol),
SPhos (140 mg, 0.70 mmol) and tetramethyltin (0.95 g, 5.28 mol)
were placed in a dry flask. Degassed toluene (20 mL) was added
and the mixture was stirred at 70 ◦C for 7 d. Evaporation and
chromatography (hexane–ethyl acetate 15 : 1) gave 25 (0.77 g, 72%)
as a yellow–orange solid with data as above.
5-Amino-4-methylisoquinolin-1-one hydrochloride (29)
1-Methoxy-5-nitro-4-phenylisoquinoline (26a)
Compound 28 (116 mg, 0.56 mmol) was stirred with palladium on
charcoal (10%, 100 mg) in ethanol (14 mL) and aq. hydrochloric
acid (34%, 0.4 mL) under hydrogen for 2 h. The suspension
Compound 24 (1.00 g, 3.53 mmol), Pd2(dba)3 (180 mg, 0.35 mmol),
SPhos (140 mg, 0.70 mmol), tripotassium phosphate (1.50 g,
7.06 mmol) and phenylboronic acid (640 mg, 5.30 mmol) were
placed in a dry flask. Degassed toluene (40 mL) was added and
the mixture was stirred at 100 ◦C for 16 h. Evaporation and
chromatography (hexane–ethyl acetat◦e 10 : 1) gave 26a (850 mg,
86%) as yellow crystals: mp 118–120 C; dH 4.20 (3 H, s, OMe),
7.27–7.31 (2 H, m, Ph 2,6-H2), 7.38–7.43 (3 H, m, Ph 3,4,5-H3),
7.62 (1 H, t, J 8.0, 7-H), 7.97 (1 H, dd, J 8.0, 1.2, 6-H or 8-H), 8.06
(1 H, s, 3-H), 8.60 (1 H, dd, J 8.0, 1.2, 8-H or 6-H); dC 54.4, 120.7,
124.5, 125.5, 127.4, 127.7, 127.8, 128.1, 128.4, 129.1, 137.5, 144.7,
147.6, 160.3; m/z 303.0740 (M + Na) (C16H11NaN2O3 requires
303.0746); 281.0915 (M + H) (C16H12N2O3 requires 281.0926);
Found: C, 68.50; H, 4.26; N, 10.19. Calc. for C16H12N2O3: C, 68.57;
H, 4.32; N, 10.00%.
R
R
was filtered through Celiteꢀ. The Celiteꢀ pad and residue were
suspended in water (100 mL) and heated. The hot suspension was
R
filtered through a second Celiteꢀ pad. Evaporation of the solvent
and◦drying gave 29 (78 mg, 65%) as a pale buff solid: mp 225–
228 C (lit.11 mp 227–229 ◦C); dH (D2O) 2.37 (3 H, s, Me), 6.94 (1
H, s, 3-H), 7.42 (1 H, t, J = 8.2, 7-H), 7.63 (1 H, d, J = 7.8, 6-H),
8.14 (1 H, d, J = 8.2, 8-H).
5-Nitro-4-phenylisoquinolin-1-one (30a)
Compound 26a was treated with aq. hydrobromic acid, as for the
synthesis of 28, to give 30a (65%) as yellow crystals: mp 211–
214 ◦C; dH ((CD3)2SO) 7.20 (3 H, m, 3-H and Ph 2,6-H2), 7.32 (3
H, m, Ph 3,4,5-H3), 7.70 (1 H, t, J 7.6, 7-H), 8.14 (1 H, dd, J 7.8,
1.2, 6-H), 8.58 (1 H, dd, J 7.8, 1.2, 8-H); dC ((CD3)2SO) 113.5,
126.4, 127.2, 127.6, 128.0, 128.2, 128.5, 129.0, 131.6, 133.1, 136.7,
147.0, 159.7; m/z 289.0598 (M + Na) (C15H10NaN2O3 requires
289.0589); 267.0761 (M + H) (C15H11N2O3 requires 267.0770);
Found: C, 68.60; H, 3.48; N, 10.49. Calc. for C15H10N2O3: C, 67.67;
H, 3.59; N, 10.52%.
1-Methoxy-5-nitro-4-(4-trifluoromethylphenyl)isoquinoline (26b)
Compound 24 was treated with 4-trifluoromethylphenylboronic
acid, Pd2(dba)3, SPhos and K3PO4 in toluene, as for the synthesis
of 26a, to give 26b (81%) as yellow crystals: mp 95–97 ◦C; dH
4.25 (3 H, s, OMe), 7.43 (2 H, d, J 8.8, Ar 2,6-H2), 7.68–7.72
(3 H, m, 7-H and Ar 3,5-H3), 8.05 (1 H, dd, J 8.4, 1.2, 6-H),
8.07 (1 H, s, 3-H), 8.66 (1 H, dd, J 8.4, 1.2, 8-H); dC 54.5, 120.7,
123.1, 125.4 (q, J 3.7, Ar 3.5-C2), 125.8, 127.3 (m, CF3), 127.6,
128.3, 129.4, 130.0 (m, Ar 4-C), 141.2, 145.0, 160.8; m/z 371.0631
(M + Na) (C17H11F3NaN2O3 requires 371.0619), 349.0805
5-Nitro-4-(4-trifluoromethylphenyl)isoquinolin-1-one (30b)
Compound 23b was treated with aq. hydrobromic acid, as for the
synthesis of 28, to give 30b (65%) as yellow crystals: mp 283–
285 ◦C; dH ((CD3)2SO) 7.34 (1 H, d, J 6.5, 3-H), 7.46 (2 H, d,
888 | Org. Biomol. Chem., 2011, 9, 881–891
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
J 7.8, Ar 3,5-H2), 7.70 (2 H, d, J 7.8, Ar 2,6-H2), 7.73 (1 H, t, J
8.2, 7-H), 8.22 (1 H, dd, J 8.2, 1.2, 6-H), 8.61 (1 H, dd, J 8.2, 1.2,
8-H), 12.09 (1 H, d, J ca. 5.5, NH); dC ((CD3)2SO) 112.0, 125.1 (q,
J 3.8, Ar 3,5-C2), 126.7, 127.7, 128.0, 128.1, 128.2, 129.3, 131.9 (m,
Ph C-4), 134.0 (m, CF3), 141.1, 146.7, 159.8; Found: C, 57.14; H,
2.67; N, 8.04. Calc. for C16H9F3N2O3: C, 57.50; H, 2.71; N, 8.38%.
in dry dimethylformamide (30 mL) was added and the suspension
was heated at 100 C for 48 h. The solvent was evaporated until
◦
5 mL remained. The residue was diluted with water and extracted
(chloroform). Evaporation and chromatography (hexane–ethyl
acetate 15 : 1) gave 33 (1.3 g, 71%) as a yellow solid: mp 106–
108 ◦C; dH 5.58 (2 H, s, CH2), 7.40 (3 H, m, Ph 3,4,5-H3),
7.52 (2 H, d, J 7.1, Ph 2,6-H2), 7.62 (1 H, t, J 7.8, 7-H), 7.90
(1 H, dd, J 7.1, 1.2, 8-H), 8.33 (1 H, s, 3-H), 8.55 (1 H, dd, J
7.1, 1.2, 6-H); dC 68.96 (CH2), 104.5, 122.0, 126.3, 126.9, 127.0,
128.2, 128.3, 128.5, 128.6, 136.1, 146.2, 147.1, 159.7; m/z 382.9826
(M + Na) (C16H1181BrNaN2O3 requires 382.9830); 380.9860 (M +
Na) (C16H1179BrNaN2O3 requires 380.9851), 359.0039 (M + H)
(C16H1279BrN2O3 requires 359.0031). Further elution gave 32
(184 mg, 12%) as a red–orange solid mp 127–130 ◦C; dH 3.14
(6 H, s, NMe)2, 7.52 (1 H, t, J = 7.6, 7-H), 7.88 (1 H, dd, J =
7.6, 1.2, 6-H), 8.26 (1 H, dd, J = 7.6, 1.2, 8-H), 8.30 (1 H, s,
3-H); dC (HMQC/HMBC) 43.06 (NMe2), 103.28 (4-C), 122.78
(8a-C), 124.21 (4-C), 126.38 (6-C), 127.95 (4a-C), 130.65 (8-C),
147.04 (3-C), 147.42 (5-C), 160.87 (1-C); m/z 319.9840 (M +
Na) (C11H1081BrNaN3O2 requires 319.9834), 317.9848 (M + Na)
(C11H1079BrNaN3O2 requires 317.9854).
5-Amino-4-phenylisoquinolin-1-one (31a)
Compound 30a (46 mg, 0.17 mmol) was stirred with palladium on
charcoal (10%, 50 mg) in ethanol (15 mL) under hydrogen for 6 h.
R
The suspension was then filtered through Celiteꢀ. Evaporation of
the solvent and drying gave 31a (21 mg, 51%) as a pale yellow
solid: mp 236–240 ◦C; dH ((CD3)2SO) 4.45 (2 H, s, NH2), 6.70 (1
H, brs, 3-H), 6.86 (1 H, dd, J 7.8, 1.2, 6-H), 7.23 (1 H, t, J =
7.8, 7-H), 7.36 (2 H, dd, J 7.3, 1.2, Ph 2,6-H2), 7.41–7.47 (3 H,
m, Ph 3,4,5-H3), 7.60 (1 H, d, J 7.8, 1.2, 8-H), 11.20 (1 H, br,
NH); dC ((CD3)2SO) (HMBC/HMQC) 116.2 (8-C), 118.40 (6-
C), 122.6 (4a-C) 125.4 (4-C), 127.4 (3-C), 128.0 (7-C), 128.3 (Ph
4-C), 129.1 (Ph 3,5-C2), 130.2 (8a-C), 130.3 (Ph 2,6-C2), 141.1
(Ph 1-C), 142.3 (5a-C), 160.1 (1-C); m/z 259.0841 (M + Na)
(C15H11NaN2O requires 259.0847), 237.1017 (M + H) (C15H12N2O
requires 237.1022); Found: C, 76.68; H, 5.46; N, 11.43. Calc. for
C15H12N2O: C, 76.26; H, 5.12; N, 11.37%.
1-Benzyloxy-4-(4-methoxyphenyl)-5-nitroisoquinoline (34)
Compound 33 (500 mg, 1.4 mmol), Pd2(dba)3 (72 mg, 0.14 mmol),
SPhos (168 mg, 0.70 mmol), tripotassium phosphate (594 mg,
2.8 mmol) and 4-methoxyphenylboronic acid (317 mg, 2.1 mmol)
were placed in a dry flask. Degassed toluene (20 mL) was added
and the mixture was stirred at 100 ◦C for 16 h. Chromatography
(hexane–ethyl acetate, 20 : 1) gave 34 (330 mg, 61%) as a yellow
solid: mp 162–164 ◦C; dH 3.85 (3 H, s, OMe), 5.64 (2 H, s, CH2),
6.93 (2 H, d, J 8.6, Ar 3,5-H2), 7.21 (2 H, d, J 8.6, Ar 2,6-H2), 7.36–
7.45 (3 H, m, Ph 3,4,5-H3), 7.55 (2 H, d, J 7.5, Ph 2,6-H2), 7.60 (1
H, t, J 8.2, 7-H), 7.95 (1 H, d, J 7.4, 6-H or 8-H), 8.04 (1 H, s, 3-H),
8.63 (1 H, d, J 7.4, 8-H or 6-H); dC (HMBC/HMQC) d 55.2 (Me),
68.6 (CH2), 108.7, 113.8 (Ar 3,5-C2), 120.7 (8a-C), 124.4 (Ar 1-C),
125.4 (7-C), 127.2 (6-C), 128.0 (Ph 4-C), 128.1 (4a-C), 128.2 (Ph
3.5-C2), 128.6 (Ph 2,6-C2), 129.0 (8-C), 129.3 (Ar 2,6-C2), 129.8 (4-
C), 131.5 (Ar 4-C), 136.7 (Ph 1-C), 147.7 (3-C), 159.2 (5-C), 159.5
(1-C); m/z 409.1164 (M + Na) (C23H17N2NaO4 requires 409.1164),
387.1366 (M + H) (C23H19N2O4 requires 387.1345); Anal. Found:
C, 71.56; H, 4.82; N, 7.31. Calc. for C23H18N2O4: C, 71.49; H, 4.70;
N, 7.25%.
5-Amino-4-(4-trifluoromethylphenyl)isoquinolin-1-one (31b)
Compound 30b was treated with hydrogen and palladium on
charcoal, as for the synthesis of 31a, to give 31b (53%) as a
pale yellow solid: mp 265–267 ◦C; dH ((CD3)2SO) 4.37 (2 H, s,
NH2), 6.81 (1 H, brs, 3-H), 6.93 (1 H, dd, J 7.8, 1.2, 6-H), 7.27
(1 H, t, J 7.8, 7-H), 7.57 (2 H, d, J 7.6, Ar 2,6-H2), 7.63 (1 H,
dd, J 7.8, 1.2, 6-H), 7.77 (1 H, d, J 7.6, 8-H), 11.31 (1 H, brs,
NH); dC ((CD3)2SO) 112.9, 114.7, 115.8, 118.2, 120.9, 124.9 (m,
Ph 3,5-C2), 127.5, 127.8, 127.9 (m, CF3 or Ar C-4), 128.8 (m,
Ar C-4 or CF3), 130.2, 130.5, 143.0, 144.2, 161.5; m/z 327.0729
(M + Na) (C17H11F3NaN2O requires 327.0712), 305.0905 (M + H)
(C17H12F3N2O requires 305.0902); Found: C, 63.21; H, 3.69; N,
9.30. Calc. for C17H11F3N2O: C, 63.16; H, 3.64; N, 9.21%.
5-Amino-4-(4-methoxyphenyl)isoquinolin-1-one (31c)
Compound 34 was treated with hydrogen and palladium on
charcoal, as for the synthesis of 31a, to give 31c (32 mg, 47%)
as a pale buff solid: mp 240–243 ◦C; dH ((CD3)2SO) 4.03 (3 H, s,
Me), 4.62 (2 H, br s, NH2), 6.85 (1 H, dd, J 7.8, 1.2, 6-H), 6.93 (2
H, d, J 8.6, Ar 3,5-H2), 7.34 (2 H, d, J = 8.6, Ar 2,6-H2), 7.39 (1 H,
d, J 8.6, Ar 3.5-H2), 7.46 (1 H, t, J 7.8, 7-H), 7.60 (1 H, d, J 7.8, 1.2,
8-H), 10.86 (1 H, br, NH); dC ((CD3)2SO) 55.4, 112.1, 112.7, 113.9,
114.3, 115.0, 122.4, 126.8, 127.4, 127.5, 130.5, 143.4, 159.0, 161.4;
m/z 267.1105 (M + H) (C16H15N2O2 requires 267.1134); Found:
C, 76.68; H, 5.46; N, 11.43. Calc. for C16H14N2O2: C, 76.26; H,
5.12; N, 11.37%.
1-Benzyloxy-5-nitro-4-(phenanthren-9-yl)isoquinoline (35)
Compound 33 was treated with Pd2(dba)3, SPhos, tripotassium
phosphate and phenanthrene-9-boronic acid, as for the synthesis
of 34, to give 35 (42%) as yellow crystals: mp 102–105 ◦C; dH 5.72
(2 H, s, CH2), 7.41 (1 H, t, J 7.8, Ph 4-H), 7.47 (2 H, t, J 7.8,
Ph 3,5-H2), 7.52 (1 H, td, J 7.8, 1.2, Phen 3 or 6-H), 7.56 (1 H, s,
Phen 9-H), 7.60–7.65 (3 H, m, Ph 2,6-H2 and 7-H), 7.66–7.71 (2 H,
m, Phen 1,8-H2), 7.82 (2 H, t, J 7.4, Phen 2,7-H2), 7.85 (1 H, dd,
J 7.6, 1.2, 6-H), 8.71 (1 H, dd, J 7.6, 1.2, 8-H), 8.74 (1 H, d, J
7.8, Phen 4-H or 5-H), 8.79 (1 H, d, J 7.8, Phen 5-H or 4-H);
dC (HMQC/HMBC) 68.8 (CH2), 120.8, 122.1 (Phen 1-C), 122.7,
123.1, 125.6, 126.4 (Phen 2-C or 7-C), 126.6, 126.8, 127.0 (8-C),
127.0, 128.2 (Ph 2,6-C2), 128.3 (Ph 4-C), 128.4, 128.7, 128.8 (Ph
3,5-C2), 129.0 (Phen 7-C or 2-C), 129.2, 130.3, 130.6, 131.0, 131.4,
1-Benzyloxy-4-bromo-5-nitroisoquinoline (33) and
4-bromo-1-dimethylamino-5-nitroisoquinoline (32)
Benzyl alcohol (680 mg, 6.3 mmol) was added to sodium hydride
(300 mg, 12.5 mmol) in dry dimethylformamide (10 mL) and the
mixture was stirred for 30 min. Compound 20 (1.5 g, 5.2 mmol)
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 889
©

View Article Online
132.7, 136.7 (Ph 1-C), 145.8 (3-C), 147.7 (5-C), 160.0 (1-C); m/z
479.1360 (M + Na) (C30H20N2NaO3 requires 479.1372), 457.1571
(M + H) (C30H21N2O3 requires 457.1552).
2s > (I)]; R1 = 0. 0649, wR2 = 0. 1729. R indices (all data); R1 = 0.
0836, wR2 = 0. 1860. Max/min peak and hole = 0.442, -1.102 e A .
-3
˚
PARP-1 inhibition assay
5-Amino-4-(phenanthren-9-yl)isoquinolin-1-one (36)
Compounds were assayed for inhibition of the catalytic activity
of PARP-1 using the FlashPlate scintillation proximity assay
previously developed at KuDOS.20 Compounds were evaluated
at eight different concentrations in triplicate; data were fitted to
the dose–response curve using a log10 concentration scale using
SigmaPlot, IC50 values were measured in two or three independent
experiments and the mean values are reported.
Compound 35 was treated with hydrogen and palladium on
charcoal, as for the synthesis of 31c, to give 36 (57%) as a buff solid:
mp >300 ◦C; m/z 359.1196 (M + Na) (C23H16NaN2O requires
359.1160), 337.1331 (M + H) (C23H17N2O requires 337.1341).
5-Benzamido-4-phenylisoquinolin-1-one (37)
Compound 31a (68 mg, 0.25 mmol) was stirred with benzoyl
chloride (39 mg, 0.28 mmol) in pyridine (2.0 mL) at 90 ◦C for
16 h. Evaporation and chromatography (ethyl acetate → ethyl
acetate–methanol 4 : 1) gave 31 (31 mg, 36%) as a very pale pink
solid: mp 230–232 ◦C (decomp.); dH (CD3OD) 7.10–7.16 (3 H, m,
Ph 3,4,5-H3), 7.24 (1 H, s, 3-H), 7.26–7.36 (4 H, m, 7-H and COPh
3,4,5-H3), 7.62 (2 H, d, J 7.2, Ph 2,6-H2), 7.71 (1 H, d, J 7.8, 6-H),
7.78 (1 H, d, J 7.4, COPh 2,6-H2), 8.42 (1 H, d, J 7.8, 8-H); dC
(CD3OD) 119.5, 127.7, 127.6, 128.4, 128.6, 128.9, 129.2, 129.4,
129.9, 130.3, 130.9, 131.8, 133.0, 134.6, 140.2, 164.1, 168.3; m/z
363.1133 (M + Na) (C22H16NaN2O2 requires 363.1109), 341.1312
(M + H) (C17H15N2O2 requires 341.1290).
PARP-2 inhibition assay
Compounds were assayed for inhibition of the catalytic activity
of PARP-2 using a method in which recombinant PARP-2
protein (recombinant) was bound down by a PARP-2-specific
antibody in a 96-well white-walled plate. PARP-2 activity was
measured following addition of 3H-NAD+ and DNA.21 After
3
washing, scintillant was added to measure the H-incorporated.
Compounds were evaluated at eight different concentrations in
triplicate; data were fitted to the dose–response curve using a log10
concentration scale using SigmaPlot, IC50 values were measured
in two or three independent experiments and the mean values are
reported.
X-Ray crystallography
General. All data were collected at 150 K on a Nonius
kappaCCD diffractometer. The structures were uniformly solved
using SHELXS-9723 and refined using full-matrix least squares in
SHELXL-97.
Acknowledgements
We wish to 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) for the kind gift of the FEK4 cells. We are
very grateful to KuDOS Pharmaceuticals Ltd. and the University
of Bath for financial support. MDT, MDL and AST are members
of Cancer Research at Bath (CR@B).
˚
Crystal data for 26a. C16H12N2O3, M = 280.28, l = 0.71073 A,
orthorhombic, space group = P212121, a = 7.4710(1), b = 8.3050(1),
3
-3
˚
˚
c = 21.1940(4) A, U = 1315.02(3) A , Z = 4, Dc = 1.416 g cm ,
m = 0.100 mm-1, F(000) = 584, crystal size = 0.50 ¥ 0.40 ¥
0.30 mm. Reflections collected = 23576, unique reflections =
3016 [Rint = 0.0757], reflections observed (I > 2s > (I)) = 2288,
data/restraints/parameters = 3016/0/192. Final R indices [I >
2s > (I)]; R1 = 0.0426, wR2 = 0.0893. R indices (all data);
R1 = 0.0725, wR2 = 0.1022. Max/min peak and hole = 0.468,
Notes and references
1 M. J. Suto, W. R. Turner, C. M. Arundel-Suto, L. M. Werbel and J. S.
Sebolt-Leopold, Anti-Cancer Drug Des., 1991, 6, 107–117.
2 C. M. Arundel-Suto, S. C. Scavone, W. R. Turner, M. J. Suto and J. S.
Sebolt-Leopold, Radiat. Res., 1991, 126, 367–371.
3 M. Banasik, H. Komura, M. Shimoyama and K. Ueda, J. Biol. Chem.,
1992, 267, 1569–1575; S. Sa´nchez-Fidalgo, I. Villegas, A. Mart´ın, M.
Sa´nchez-Hidalgo, C. Alarco´n and C. de la Lastra, Eur. J. Pharmacol.,
2007, 563, 216–223.
4 S. Cuzzocrea, E. Mazzon, R. Di Paola, T. Genovese, N. S. A. Patel, C.
Muia`, M. D. Threadgill, A. De Sarro and C. Thiemermann, Naunyn-
Schmiedebergs Arch. Pharmacol., 2004, 370, 464–473.
5 N. Wayman, M. C. McDonald, A. S. Thompson, M. D. Threadgill and
C. Thiemermann, Eur. J. Pharmacol., 2001, 430, 93–100.
6 M. C. McDonald, H. Mota-Filipe, J. A. Wright, M. Abdelrahman,
M. D. Threadgill, A. S. Thompson and C. Thiemermann, Br. J. Phar-
macol., 2000, 130, 843–850.
-3
˚
-0.407 e A .
˚
Crystal data for 33. C16H11BrN2O3, M = 359.18, l = 0.71073 A,
¯
triclinic, space group = P1(No. 2), a = 8.6180(4), b = 9.9120(4), c =
◦
˚
10.2260(5) A, a = 114.347(2), b = 108.553(2), g = 100.738(2) , U =
3
-3
-1
˚
701.65(6) A , Z = 2, Dc = 1.700g cm , m = 2.944mm , F(000) = 360,
crystal size = 0.40 ¥ 0.30 ¥ 0.30 mm. Reflections collected = 11401,
unique reflections = 3168 [Rint = 0.0484], reflections observed (I >
2s > (I)) = 2680, data/restraints/parameters = 3168/0/200. Final
R indices [I > 2s > (I)]; R1 = 0. 0303, wR2 = 0. 0642. R indices
(all data); R1 = 0. 0421, wR2 = 0. 0680. Max/min peak and hole =
-3
˚
0.369, -0.388 e A .
7 T. Genovese, E. Mazzon, C. Muia`, N. S. A. Patel, M. D. Threadgill,
P. Bramanti, A. De Sarro, C. Thiemermann and S. Cuzzocrea,
J. Pharmacol. Exp. Ther., 2005, 312, 449–457.
8 M. Li, M. D. Threadgill, Y. Wang, L. Cai and X. Lin, Pathobiology,
2009, 76, 108–116; Y. Qin, Y. Wang and Y.-Y. Li, J. Third Military Med.
Univ, 2008, 30, 1330–1333.
9 P. T. Sunderland, E. C. Y. Woon, A. Dhami, M. F. Mahon, P. J. Wood,
L. A. Jones, S. R. Tully, M. D. Lloyd, A. S. Thompson, H. Javaid,
N. M. B. Martin and M. D. Threadgill, J. Med. Chem. submitted.
˚
Crystal data for 35. C32H24N2O4, M = 500.53, l = 0.71073 A,
monoclinic, space group = P2/◦c, a = 13.9880(2), b = 10.9970(2),
3
˚
˚
c = 16.1970(2) A, b = 95.645(1) U = 2479.44(7) A , Z = 4, Dc =
1.341 g cm-3, b = 0.089 mm-1, F(000) = 1048, crystal size = 0.40 ¥
0.20 ¥ 0.20 mm. Reflections collected = 44827, unique reflections =
5661 [Rint = 0.0507], reflections observed (I > 2s > (I)) = 4556,
data/restraints/parameters = 5661/0/350. Final R indices [I >
890 | Org. Biomol. Chem., 2011, 9, 881–891
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
10 R. Pellicciari, E. Camaioni, G. Costantino, L. Formentini, P. Sabbatini,
F. Venturoni, G. Eren, D. Bellocchi, A. Chiarugi and F. Moroni,
ChemMedChem, 2008, 3, 914–923.
11 A. Dhami, M. F. Mahon, M. D. Lloyd and M. D. Threadgill,
Tetrahedron, 2009, 65, 4751–4765.
17 T. E. Barder, S. D. Walker, J. R. Martinelli and S. L. Buchwald, J. Am.
Chem. Soc., 2005, 127, 4685–4696.
18 M. S. Wong and X. L. Zhang, Tetrahedron Lett., 2001, 42, 4087–4089.
19 T. Karlberg, M. Hammarstro¨m, P. Schu¨tz, L. Svensson and H. Schu¨ler,
Biochemistry, 2010, 49, 1056–1058.
12 M. Croisy-Delcey, E. Bisagni, C. Huel, D. Zilberfarb and A. Croisy,
J. Heterocycl. Chem., 1991, 28, 65–71.
13 A. D. Sercel, J. P. Sanchez and H. D. H. Showalter, Synth. Commun.,
2007, 37, 4199–4208.
14 J. M. Berry, C. Y. Watson, W. J. D. Whish and M. D. Threadgill,
J. Chem. Soc., Perkin Trans. 1, 1997, 1147–1156; P. T. Sunderland,
A. S. Thompson and M. D. Threadgill, J. Org. Chem., 2007, 72, 7409–
7412.
20 K. J. Dillon, G. C. M. Smith and N. M. B. Martin, J. Biomol. Screening,
2003, 8, 347–352.
21 K. A. Menear, C. Adcock, R. Boulter, X.-L. Cockcroft, L. Copsey, A.
Cranston, K. J. Dillon, J. Drzewiecki, S. Garman, S. Gomez, H. Javaid,
F. Kerrigan, C. Knights, A. Lau, V. M. Loh, I. T. W. Matthews, S.
Moore, M. J. O’Connor, G. C. M. Smith and N. M. B. Martin, J. Med.
Chem., 2008, 51, 6581–6591.
22 S. Ferrer, D. P. Naughton and M. D. Threadgill, Tetrahedron, 2003,
59, 3437–3444; S. Ferrer, D. P. Naughton and M. D. Threadgill,
Tetrahedron, 2003, 59, 3445–3454.
15 D. E. Horning, G. Lacasse and J. M. Muckowski, Can. J. Chem., 1971,
49, 2785–2796.
16 I. Parveen, D. P. Naughton, W. J. D. Whish and M. D. Threadgill, M,
Bioorg. Med. Chem. Lett., 1999, 9, 2031–2036.
23 G. M. Sheldrick, SHELXL-97, a computer program for crystal structure
refinement, University of Go¨ttingen: Go¨ttingen, 1997.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 881–891 | 891
©