Synthesis of 3-substituted benzamides and 5-substituted isoquinolin-1(2H)-ones and preliminary evaluation as inhibitors of poly(ADP-ribose)polymerase (PARP)

Article abstract of DOI:10.1016/S0968-0896(98)00029-7

Inhibitors of poly(ADP-ribose)polymerase (PARP) inhibit repair of damaged DNA and thus potentiate radiotherapy and chemotherapy of cancer. 3-Substituted benzamides and 5-substituted isoquinolin-1-ones have been synthesised and evaluated for inhibition of PARP. Reduction of 3-(bromoacetyl)benzamide, followed by treatment with base, gave RS-3-oxiranylbenzamide. Reduction of 3-(hydroxyacetyl)benzonitrile with bakers' yeast gave the R-diol which was converted to R-3-(1,2-dihydroxyethyl)benzamide. Similar reduction of 3-(acetoxyacetyl)benzonitrile led towards the S-diol which was converted to its cyclic acetonide. E-2-(2,6-Dicyanophenyl)-N,N-dimethylethenamine was formed by condensation of 2,6-dicyanotoluene with dimethylformamide dimethyl acetal (DMFDMA); cyclisation under acidic conditions afforded 5-cyanoisoquinolin-1-one. Heck coupling of 5-iodoisoquinolin-1-one with propenoic acid formed E-3-(1-oxoisoquinolin-5-yl)propenoic acid. 3-Oxiranylbenzamide, 5-bromoisoquinolin-1-one and 5-iodoisoquinolin-1-one were among the most potent inhibitors of PARP activity in a preliminary screen in vitro. Copyright (C) 1998 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(98)00029-7

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 721±734
Synthesis of 3-Substituted Benzamides and 5-Substituted
Isoquinolin-1(2H)-ones and Preliminary Evaluation as
Inhibitors of Poly(ADP-ribose)polymerase (PARP)
Corrine Y. Watson,a William J. D. Whishb and Michael D. Threadgilla,*
aDepartment of Pharmacy & Pharmacology, University of Bath, Claverton Down, Bath, BA2 7AY, UK
bDepartment of Biology & Biochemistry, University of Bath, Claverton Down, Bath, BA2 7AY, UK
Received 18 September 1997; accepted 1 December 1997
AbstractÐInhibitors of poly(ADP-ribose)polymerase (PARP) inhibit repair of damaged DNA and thus potentiate
radiotherapy and chemotherapy of cancer. 3-Substituted benzamides and 5-substituted isoquinolin-1-ones have been
synthesised and evaluated for inhibition of PARP. Reduction of 3-(bromoacetyl)benzamide, followed by treatment
with base, gave RS-3-oxiranylbenzamide. Reduction of 3-(hydroxyacetyl)benzonitrile with bakers' yeast gave the R-
diol which was converted to R-3-(1,2-dihydroxyethyl)benzamide. Similar reduction of 3-(acetoxyacetyl)benzonitrile led
towards the S-diol which was converted to its cyclic acetonide. E-2-(2,6-Dicyanophenyl)-N,N-dimethylethenamine was
formed by condensation of 2,6-dicyanotoluene with dimethylformamide dimethyl acetal (DMFDMA); cyclisation
under acidic conditions a€orded 5-cyanoisoquinolin-1-one. Heck coupling of 5-iodoisoquinolin-1-one with propenoic
acid formed E-3-(1-oxoisoquinolin-5-yl)propenoic acid. 3-Oxiranylbenzamide, 5-bromoisoquinolin-1-one and 5-iodo-
isoquinolin-1-one were among the most potent inhibitors of PARP activity in a preliminary screen in vitro. # 1998
Elsevier Science Ltd. All rights reserved.
Introduction
to DNA to have enzymatic activity.12,13 PARP catalyses
the transfer of ADP-ribose units from its substrate
NAD+ 1 (Figure 1) to several protein acceptors. Most
of these are nuclear proteins involved in chromatin
architecture and DNA metabolism (heteromodi®cation)
but the main protein to be poly(ADP-ribosyl)ated is
PARP itself (automodi®cation).14,15
Radiotherapy and many forms of chemotherapy of
cancer kill proliferating malignant cells by damaging
DNA. Ecient repair of this damage can lead to resis-
tance to these therapeutic strategies.1 Poly(ADP-ribose)
polymerase (PARP) is the enzyme that is responsible for
controlling excision repair processes and inhibitors of
this enzyme have been shown to act as potentiators to
the lethal e€ects of radiation2±7 and DNA-damaging
chemotherapeutic drugs.8±11 PARP is abundant in most
cell nuclei and ca. 90% of the PARP in cells is bound to
chromatin. It comprises a 113 KDa protein with three
domains: a 46 KDa DNA-binding domain, a 22 KDa
Most known inhibitors of PARP mimic the nicotin-
amide moiety of the substrate 1. The ®rst enzyme-selec-
tive inhibitor16 was 3-aminobenzamide 2 (Figure 1)
(IC50 22 mM17). Further structure±activity studies on
substituted benzamides16±20 showed that optimum
activity in this monocyclic series was achieved with an
electron-donating group at position 3 but with no sub-
stituent in positions 2 or 4±6 on the benzene ring.
Recognition that the conformation of the benzamide
pharmacophore was important20 led to development of
5-substituted isoquinolin-1-ones17,20 and 3,4-dihy-
droisoquinolin-1-ones20 3 and of 2,8-disubstituted qui-
nazolin-4-ones17,21,22 4 (Figure 1), which are some 10- to
50-fold more potent as inhibitors (3 R1=OH, R2=H,
automodi®cation domain and
a 54 KDa catalytic
domain but acts catalytically as a homodimer. The pro-
tein binds to DNA strand-breaks through the two zinc
®ngers of the DNA-binding domain and must be bound
Key words: Poly(ADP-ribose)polymerase; benzamide; iso-
quinolinone; asymmetric reduction; Heck coupling.
*Corresponding author. E-mail: prsmdt@bath.ac.uk
0968-0896/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00029-7

722
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
variety of cyclic amides and related compounds con-
®rmed the need for the trans-arylamide structural ele-
ment; large planar lactams such as 4-amino-1,8-
naphthalimide (IC50 0.18 mM) and phenanthridin-6(5H)-
one (IC50 0.30 mM) were particularly potent as inhibitors
of PARP. In structures 2±5 (Figure 1), the consensus
pharmacophore is shown in bold.
As part of our search for more active radiosensitising
drugs,23,24 we rationalised that design of a mechanism-
based irreversible inhibitor of PARP may be more
e€ective in inhibiting DNA repair in a therapeutic con-
text. Scheme 1 shows a proposed PARP-catalysed
mechanism in which the nicotinamide leaves the sub-
strate NAD+ 1, giving an intermediate oxonium ion 6.
The incoming nucleophile (Nu), which is a carboxylate
side-chain in the acceptor protein or the 20-OH of the
ribose of the growing poly(ADP-ribose) chain, then
approaches from the b-face. A benzamide or iso-
quinolin-1-one 7 with an electrophilic substituent E
(Scheme 1) may trap the nucleophile of the Glu or the
poly(ADP-ribose) chain, capping the polymer, or may
react with an appropriately placed nucleophile in the
active site of PARP. We present here the synthesis and
evaluation of benzamides and isoquinolin-1-ones with
substituents in the 3- and 5-positions, respectively,
which are electrophilic or which may make other inter-
actions with the NAD+-binding site of PARP.
Figure 1. Structures of NAD+ 1 and known inhibitors of
PARP 2±5.
X=CH: IC50 0.14 mM; 4 R1=OH, R2=Me, X=N:
IC50 0.44 mM). In 3 and 4, the heterocyclic ring con-
strains the conformation of the benzamide with the N±
H bond trans to the carbonyl-arene bond. An interesting
new approach to maintaining this required conforma-
tion was developed by Grin et al.22 who used a
hydrogen-bond in the benzoxazolecarboxamides
5
(Figure 1) (5 R=Ph: IC50 2.1 mM). A large study17 of a
Scheme 1. Chemical mechanism for the PARP-catalysed formation of poly(ADP-ribose) from NAD+ 1 via the intermediate oxonium
ion 6.

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
723
Chemical synthesis: Benzamides
corresponding to the amine in the lead compound 3-
aminobenzamide 2 and to the 5-substituents in the iso-
quinolinones 3. Target compounds 11, 15R, 15S, 21, 22
and 24 all have an oxygen atom located at a site
designed to be able to make hydrogen-bonding inter-
actions with the active site of PARP corresponding to
those of the ribose oxygen of the substrate NAD+ 1.
Since all these target compounds have a 2-carbon unit
attached at the 3-position of the benzamide, 3-acetyl-
benzamide 11 appeared to be a suitable starting mate-
rial, being also one of the targets. This is not
commercially available but was readily synthesised by
The simplest target benzamide with a benzylic electro-
phile in the 3-position for potential irreversible inhibi-
tion of PARP is 3-(chloromethyl)benzamide 9, which
was prepared in high yield by treatment of the corre-
sponding acyl chloride 8 with ethereal ammonia, using a
short contact time to avoid substitution at the benzylic
position (Scheme 2).
Scheme 2 shows the synthetic routes to benzamides
with more complex functional groups in the 3-position,
Scheme 2. Synthesis of 3-substituted benzamides which are inhibitors of PARP. Reagents: (i) NH3, Et2O; (ii) H2O2, NaOH, EtOH,
H2O; (iii) NaBH4, EtOH; (iv) PhI(OAc)2, CF3CO2H, MeCN, H2O; (v) bakers' yeast, sucrose, H2O; (vi) Br2, CHCl3; (vii) KOAc, NaI,
EtOH; (viii) NH3, MeOH, H2O; (ix) Me2C(OMe)2, TsOH; (x) Br2, HOAc; (xi) NaOH, H2O.

724
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
hydrogen peroxide-assisted hydrolysis of the corre-
sponding benzonitrile 10 under basic conditions.
Straight forward reduction with sodium borohydride
then a€orded 3-(1-hydroxyethyl)benzamide 12, which
was required for later development of NMR methods
for determination of optical purities of the enantiomers
of the corresponding ethane-1,2-diols.
bearing electron-withdrawing groups on the benzene
ring. Following the procedure of Manzocchi et al.,36 13
was reduced during 7 days in a fermenting preparation
of bakers' yeast to the dihydroxyethylbenzonitrile 14R
with [a]D20= 46.6ꢀ. In parallel, the ketone 13 was
reduced with sodium borohydride in the usual manner
to give the racemic diol 14RS for use in developing the
methods for determination of ee by NMR (see below).
The enantiomeric diols 15R and 15S were important
synthetic intermediates and targets in their own right. In
15S, the secondary alcohol may occupy the binding site
of the ribose oxygen of NAD+ 1, while the primary
alcohol has potential for mimicking the 20-hydroxyl of
that ribose. The crystal structure of the catalytic frag-
ment of chicken PARP has been reported25 and it is
claimed that the binding site and interactions for
NAD+ 1 are very similar in PARP and in another
ADP-ribosylating protein, diphtheria toxin.26±30 In the
latter, the 20-OH of the nicotinamide ribose appears to
make a weak hydrogen bonding interaction with an
amino-acid residue. It was predicted that the enantiomer
15R should bind and inhibit PARP less eciently and
that comparison of the inhibitory data would enable a
preliminary test of this hypothesis. Synthesis of the
required enantiomeric diols 15R and 15S could, in
principle, be achieved though several routes, including
asymmetric dihydroxylation of a corresponding styrene
and asymmetric reduction of a corresponding sub-
stituted acetylbenzamide. Although several methods of
asymmetric reduction of alkyl±aryl ketones with `che-
mical' reagents are known,31±33 biotransformations fre-
quently o€er more reliably higher enantiomeric excesses.
The reduction of ketones by bakers' yeast (Saccharo-
myces cerevisiae) is one of the most widely used bio-
transformations for the preparation of enantiomerically
pure alcohols.34 Aromatic a-hydroxyketones have been
reduced asymmetrically35±37 to diols by this organism.
For example, (hydroxyacetyl)benzene is reduced to R-
phenylethane-1,2-diol in high yield and high optical
purity.35,36 The opposite stereochemistry of reduction is
observed when the hydroxy group is esteri®ed.36 How-
ever, the bakers' yeast reduction is intolerant37,38 of
some substituents on the aromatic ring, thus 3-(sub-
stituted acetyl)benzamides were discounted as suitable
substrates. Relying on the ecient hydrogen peroxide-
mediated hydrolysis of nitriles to amides,39 bioreduc-
tions of 3-(hydroxyacetyl)benzonitrile 13 and 3-(acetoxy-
acetyl)benzonitrile 18 were mooted as potentially
enantiomerically ecient routes to the target enantio-
meric diols 15R and 15S, respectively. 3-Acetylbenzo-
nitrile 10 was hydroxylated directly at the methyl group
using the hypervalent iodine reagent I,I-bis(tri¯uoro-
acetoxy)iodobenzene (iodosobenzene bis(tri¯uoroace-
tate)) developed by Moriarty et al.40 The modest yield of
13 is consistent with the observation40 of generally
inferior yields during hydroxylations of acetophenones
To form the S-diol 14S, the acetoxyacetylbenzonitrile 18
was required as substrate for the bioreduction. Bromi-
nation of 10 with bromine in chloroform41 gave a 78%
yield of the bromomethyl ketone 16 with 6% of the
overoxidation product, the dibromomethylketone 17.
Nucleophilic substitution of acetoxy for bromine was
e€ected with potassium acetate in the presence of cata-
lytic sodium iodide, according to the general method of
Kaufmann and Muller.42 In addition to the 64% yield
of 18, two other products were recovered and char-
acterised, the methyl ketone 10 and 3-(diethoxy-
acetyl)benzonitrile 19. The former arises from
a
reduction and the latter from an oxidation under the
reaction conditions and proposed mechanisms for these
processes are shown in Scheme 3. Clearly, the iodide
acts as a nucleophilic catalyst, forming the iodomethyl
ketone 25 as an intermediate. This can su€er three fates.
Nucleophilic substitution with acetate to form 18 will be
the main reaction early in the reaction, when large
amounts of AcO are present. When most of the acetate
has been consumed, slower reactions become important,
such as reduction of 25 by iodide to give 10 and elemental
iodine and substitution by the weaker nucleophile
Scheme 3. Proposed mechanism of formation of the reduced
product 10 and the oxidised product 19 during treatment of the
16 with KOAc and NaI.

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
725
ethoxide, giving the intermediate ethoxymethyl ketone
26. Formation of the enolate 27 is favoured by the pres-
ence of the ethoxy group; 27 can then react with the
molecular iodine and substitution of ethoxide for iodine
in 28 forms the acetal 19. Treatment of the acetoxy-
acetylbenzonitrile 18 with fermenting bakers' yeast for a
much shorter contact time that that used for 13 a€orded
a high yield of the S-enantiomer of the monoester 20.
This was separated from the trace of 3-(1,2-dihydroxy-
ethyl)benzonitrile which was also formed; this was of
unpredictable stereochemistry since it could have arisen
from hydrolysis of the acetate ester followed by reduction
to the R-diol 14R or hydrolysis of the acetate of 20 to
give the S-diol 14S. The enantiomeric purity of 20 was
assessed after conversion to its MTPA ester by treatment
with the freshly prepared acyl chloride of R-Mosher's
acid.43 As a control, the derivatisation was repeated
fused ring. By conformationally restricting the amide
group in this way, the activity of these compounds as
inhibitors of PARP is greatly increased.17,20 An iso-
quinolin-1-one with a carbon substituent at the 5-posi-
tion would be a good precursor for a variety of
candidate PARP inhibitors of this general type 7. How-
ever, there are few examples of such substituted iso-
quinolinones in the literature.
The synthesis of 5-cyanoisoquinolinone 30 has been
reported by Wenkert et al.44 This compound gives
potential for elaboration of the 5-substituent by hydro-
lysis to the carboxylic acid 36 or by Pinner reaction a
the corresponding ester; reduction and further elabora-
tion could give a variety of benzylic electrophiles. Fol-
lowing the published procedure,44 2,6-dicyanotoluene 29
was condensed with ethyl formate under basic condi-
tions, followed by acid hydrolytic workup, to give the
isoquinolinone 30 in low yield, together with a sig-
ni®cant amount of 5-cyanoisocoumarin 31. Higher
yielding approaches to 30 were therefore sought
(Scheme 4). The dinitrile 29 was una€ected by moder-
ately forcing Pinner conditions but hydrogen peroxide-
mediated hydrolysis under controlled conditions gave
selectively the monocyanobenzamide 32. Cyclo-
condensation with a formate synthon was envisaged but
32 failed to react with triethyl orthoformate and, on
treatment with the more reactive dimethylformamide
dimethyl acetal, a€orded a high yield of the benzoylfor-
1
using racemic Mosher's acid chloride; the H NMR sig-
nals from the R,R diastereoisomeric monoester were
observed with identical intensity to those from the S,R
diastereoisomer. Similar derivatisations of racemic 14RS
with enantiomerically pure Mosher's acid chlorides gave
identical spectra. Thus 20 was demonstrated to have
been formed with >96% ee. The ester was cleaved with
20
ammonia to reveal the S-diol 14S, which had [a]D
=
+46.6ꢀ. Thus 14R also had ee > 96% (vide supra). As
expected, the enantiomeric dihydroxyethylbenzonitriles
14R and 14S were eciently converted to the corre-
sponding 3-(1,2-dihydroxyethyl)benzamides 15R and 15S,
respectively, by hydrogen peroxide-mediated hydrolysis.
The S-dihydroxyethylbenzamide 15S was cyclised to the
ketal 21 with acetone dimethyl acetal under acidic con-
ditions. In this compound, the dioxolane may mimic the
ribose ring in the PARP substrate NAD+ 1, with ether
oxygens occupying the same relative positions in the
®ve-membered rings.
In the synthetic sequence to the 3-oxiranylbenzamide
24, it was essential that the carboxamide be carried
through the assembly of the heterocycle, as the condi-
tions required for hydrolysis of the nitrile would be
deleterious to the integrity of the oxirane. Bromination
of 11 with bromine in acetic acid gave the bromomethyl
ketone 22 that, surprisingly, is previously unreported in
the literature. Attempts at asymmetric reduction of the
ketone using bakers' yeast were unsuccessful but treat-
ment with sodium borohydride, followed by ring-clo-
sure of the oxirane under basic conditions gave the
racemic oxiranylbenzamide 24 in excellent yield in a
one-pot process.
Scheme 4. Routes to 5-cyanoisoquinolin-1-one 30. Reagents:
(i) EtOCHO, KOBut; (ii) H2O2, NaOH, EtOH, H2O; (iii)
(MeOH)2CHNMe2, DMF; (iv) TsOH, PhMe; (v) HCl, MeOH;
(vi) KOH, EtOH.
Chemical synthesis: Isoquinolin-1-ones
The isoquinolin-1-ones have the benzamide core in their
structure but have the amide as part of a heterocyclic

726
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
mamidine 33 from condensation with the carboxamide
only. An attempt to complete the cyclisation by con-
densation with the activated methyl group under acidic
conditions gave the N1,N3-diacylformamidine 34 as the
sole isolable product. The mechanistic origin of this
unsymmetrical formamidine is unclear but no iso-
quinolinones were found in the product mixture. Since
the carboxamide was proving to be the sole nucleophile
in 32, attention was turned to condensation of a formate
synthon with the methyl group of 29, in which the
nitriles cannot compete as nucleophiles. Condensation
of 29 with dimethylformamide dimethyl acetal eciently
led to the E-dimethylaminostyrene 35. Cyclisation to the
required 5-cyanoisoquinolinone 30 was then achieved in
high yield by treatment with methanolic hydrogen
chloride, followed by aqueous workup. The remaining
nitrile was, interestingly, completely una€ected by these
Pinner reagents. Treatment of 30 under more forcing
acidic conditions failed to produce the corresponding
ester. In contrast, the nitrile was hydrolysed with base to
give the carboxylic acid 36. The insolubility of 36 in
common solvents precluded ecient further elaboration
from this sequence and alternatives were sought.
Scheme 5. Reaction of 5-lithioisoquinoline 38 with benzaldehyde
and structure of 41. Reagents: (i) BuLi, C6H14, THF; (ii)
PhCHO, THF.
Krohn et al.45 reported the synthesis of benzamide-3-(b-
D-riboside) using addition of 3-(4,5-dihydro-4,4-di-
methyloxazol-2-yl)phenyl lithium to the protected
ribose-derived lactone 41 as the critical step of assembly.
An analogueous addition of an aryl lithium derived
from an isoquinolinone would represent a rapid entry
into isoquinolinones with highly functionalised carbon
chains in the 5-position. Lithiation of isoquinolinones is
unfeasible, owing to the presence of the carbonyl group.
1-Lithioisoquinolines and 4-lithioisoquinolines are
known46±48 to react with carbon electrophiles but 5-
lithioisoquinoline is hitherto unreported. A two-step
sequence (N-oxidation and rearrangement with acetic
anhydride) has been described44 for conversion of iso-
quinolines to the corresponding isoquinolin-1-ones,
although the conditions required are harsh. 5-Bromoiso-
quinoline 37 (prepared49 from 5-aminoisoquinoline by
diazotisation and treatment with copper (I) bromide)
was lithiated by transmetallation with butyl lithium at
116 ꢀC. Treatment of the intermediate 38 (Scheme 5)
with a model carbon electrophile, benzaldehyde, gave a
moderate yield of the alcohol 39, along with a small
amount of isoquinoline 40, derived from quenching
unreacted anion 38 during aqueous workup. However,
treatment of 38 with 41 gave only products of degrada-
tion of the lactone, together with large quantities of
isoquinoline 40, suggesting that the highly basic anion
38 had deprotonated the 2-position of the lactone.
5-position of the isoquinolinone pharmacophore.
5-Iodoisoquinolin-1-one 42 was synthesised by one-pot
Curtius rearrangement and cyclisation from 3-(2-iodo-
phenyl)propenoic acid at 260 ꢀC, as previously descri-
bed.24 The 5-bromo analogue 51 was prepared
similarly24 from 3-(2-bromophenyl)propenoic acid.
Although Plevyak and Heck used acetonitrile51 as the
reaction solvent in a sealed tube at 100 ꢀC, the proce-
dure is facilitated by conducting the coupling of 42 with
propenoic acid in the boiling homologue propanenitrile
to give the isoquinolinone-5-propenoic acid 43 in excel-
lent yield. Attempted extension of this procedure to
coupling of more highly functionalised 5-carbon alkenes
was less successful. Grignard reaction of R-glycer-
aldehyde acetonide with vinyl magnesium bromide52,53
gave the allylic alcohol 44 as a 10:7 mixture of diastereo-
isomers. This contains a ®ve-carbon unit appropriately
functionalised for potential later elaboration to mimics
of the ribose unit of NAD+ 1. Palladium-catalysed
coupling of alkene 44 to 42 gave a low yield of the
expected mixture of diastereoisomers of the alcohol 45
(Scheme 6) which could be separated chromato-
graphically from the ketone 46. The latter arises from
Pd-catalysed migration of the C=C double bond of the
allylic alcohol 45 into the enolic position, followed by
tautomerisation. Similar migration of a C=C double
bond also took place during the coupling of 42 with the
simpler ®ve-carbon unit, pent-4-enol. The observed
product mixture was also complicated in this case by
poor regioselectivity during the initial Heck coupling.
Product alcohols 47 and 48 were obtained as a chromato-
graphically inseparable mixture and arise from reaction
The Heck reaction50,51 was investigated as an alternative
metal-mediated carbon±carbon bond forming process
for attachment of functionalised carbon chains at the

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
727
Scheme 6. Heck couplings of 42 with various alkenes and structure of 51. Reagents: (i) CH2CHCO2H, Pd(OAc)2, Et3N, EtCN; (ii) 44,
Pd(OAc)2, Et3N, EtCN; (iii) CH2CH(CH2)OH, Pd(OAc)2, Et3N, EtCN.
of the alkene at the less hindered and more hindered
ends, respectively, in contrast to the usual sensitivity of
the Heck coupling to steric e€ects.50 Migration of the
C=C double bond along the alkyl chain of 47 and 48
was terminated by enol keto tautomerisation, as for the
ketone 46, giving an inseparable mixture of the isomeric
aldehydes 49 and 50, respectively.
have suggested that electron-withdrawing groups (e.g.
nitro) are deleterious to activity. Also notably potent
was the oxirane 24. All the isoquinolinones inhibited
PARP strongly at ca. 10 mM, with the activity of the
enzyme being virtually abolished by 5-iodoisoquino-
linone 42 and 5-bromoisoquinolinone 51. However,
there was no evidence of increased inhibition of PARP
activity by the potentially electrophilic benzamides 9
and 24 and isoquinolinone 43 when the nuclear pre-
paration was incubated with the test compounds for
di€erent time intervals up to 30 min before the assay
was started by addition of [3H]-NAD+. These observa-
tions suggest that these compounds are not acting as
time-dependent irreversible inhibitors of PARP.
Biological Evaluation and Discussion
As a preliminary screen, the inhibitory e€ects of 3-sub-
stituted benzamides and 5-substituted isoquinolin-1-
ones on PARP were examined in vitro at ca. 10 mM
concentration. The enzyme was extracted from nuclei
isolated from L929 cells (mouse areolar and adipose
tissue cells). PARP activity was estimated by incor-
poration of radioactivity from [3H]adenosine-NAD+
into acid-insoluble macromolecules, according to the
method published previously.16 The acid-precipitated
radioactivity of the test incubations was expressed as a
percentage of the control to which the solvent bu€er
alone was added. The results are given in Table 1. All
the 3-substituted benzamides inhibited PARP activity,
as expected.16 Interestingly, particularly potent inhibi-
tion was shown by 11 and 22, which have electron-
withdrawing groups in the 3-position. Previous reports16
A preliminary study of the e€ect of concentration of 55
on the inhibition of PARP activity showed IC50 ca.
300 nM, which is comparable with the value reported by
Suto et al.20 using a di€erent assay system employing a
di€erent source of the enzyme. Indeed, it is notable that
a variety of enzyme preparations and assay method
have been reported2±11,16±22 to have been used in the
development and evaluation of inhibitors of PARP.
Compound 55 was also evaluated for its inhibition of
the mono-ADP-ribosylating activity of diphtheria
toxin.54,55 The isoquinolinone (0±250 mM) was incu-
bated in bu€er with diphtheria toxin (100 ng), its natural

728
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
Table 1. Results of preliminary studies of inhibition of PARP
activity by 3-substituted benzamides and 5-substituted isoqui-
nolin-1(2H)-ones
may lead to tumour-selective inhibition of DNA repair
and to selective potentiation of radiotherapy and chemo-
therapy of cancer.
Compd
Inhibition of PARP activity at ca. 10 mM
Benzamides
Experimental
9
11
62% @ 10.5 mM
83% @ 11.0 mM
32% @ 9.3 mM
44% @ 9.4 mM
54% @ 10.0 mM
84% @ 10.7 mM
88% @ 9.2 mM
General methods
15R
15S
21
Solutions in organic solvents were dried with anhydrous
MgSO4. The chromatographic stationary phase was
silica gel. IR Spectra were obtained of samples as KBr
discs, unless otherwise noted. NMR spectra were
obtained of solutions in CDCl3, except where noted.
Dilute H2SO4 refers to a 10% aqueous solution, dilute
HCl refers to a 2 M aqueous solution. The brine was
saturated. The bakers' yeast was obtained from Phipps
Bakery, Bath, UK. The sucrose was `Silver Spoon'
brand obtained from J. Sainsbury p.l.c., Bath, UK. 5-
Iodoisoquinolin-1-one 42 and 5-bromoisoquinolin-1-
one 51 were prepared as described previously by us.24 5-
Aminoisoquinolin-1-one was prepared essentially by the
method of Wenkert et al.44
22
24
Isoquinolinones
36
42
43
51
55
79% @ 13.2 mM
96% @ 8.0 mM
81% @ 11.6 mM
95% @ 11.2 mM
92% @ 4.0 mM
[IC50=ca. 300 nM]
substrate elongation factor 2 (eEF-2)56,57 (4.0 mg) and
[14C]-NAD+ in a total reaction volume of 55 mL for
45 min. Trichloroacetic acid was added to precipitate the
macromolecular fraction including the eEF-2; this was
collected and the incorporated radioactivity was
assayed. Compound 55 inhibited this activity with IC50
ca. 80 mM, giving an apparent >200-fold selectivity for
inhibition of PARP. Direct comparisons of structural
requirements for inhibition of PARP and of diphtheria
toxin have not previously been made but Banasik et
al.17 noted selectivities of >1000-fold for PARP inhibi-
tion by 3-hydroxybenzamide and by phenanthridin-
6(5H)-one over inhibition of an arginine-speci®c
mono(ADP-ribose)transferase by these compounds.
These observations tend to indicate that there are sig-
ni®cant di€erences between the binding site for NAD+
in PARP and those in the mono(ADP-ribosyl)ating
enzymes, in contrast to the claim of similarity by Ruf et
al.25 Clearly, further studies are required in this area.
3-(Chloromethyl)benzamide (9). NH3 in dry Et2O (4%
w/v, 200 mL) was added to 3-chloromethylbenzoyl
chloride (8) (2.85 g, 15 mmol) in Et2O (25 mL) during
10 min. The mixture was stirred for 30 min then washed
with water, with dilute H2SO4 (2Â) and with water and
was dried. Evaporation and recrystallisation (EtOAc:
hexane) gave 9 (2.27 g, 89%) as white needles: mp 118±
120ꢀC (lit.58 mp 119±120 ꢀC); IR 3360, 3190, 1655, 1625,
1
705 cm
;
1H NMR ((CD3)2SO) d 4.64 (2 H, s, CH2),
6.34 (1 H, bs, NH), 7.30 (1 H, bs, NH), 7.44 (1 H, t,
J=7.7 Hz, 5-H), 7.54 (1 H, d, J=7.7 Hz, 4-H), 7.84 (1
H, d, J=7.7 Hz, 6-H), 7.94 (1 H, s, 2-H); MS (EI) m/z
171/169 (M).
3-Acetylbenzamide (11). H2O2 (27.5% in water, 15 mL,
121 mmol) was added during 30 min to 3-acetylbenzo-
nitrile (10) (5.0g, 34 mmol) and NaOH (340 mg, 8.5mmol)
in EtOH (50 mL) and water (17 mL) was added, keeping
the temperature <50 ꢀC. The mixture was stirred at
50 ꢀC for 1 h. After neutralisation with dilute H2SO4, the
EtOH was evaporated. The residue, in CH2Cl2, was
washed with water and with brine and was dried. Eva-
poration gave 11 (5.48 g, 97%) as white crystals: mp
125.5±126.5 ꢀC; Found: C, 66.1; H, 5.55; N, 8.20.
C9H9NO2 requires C, 66.25; H, 5.56; N, 8.58%; IR
3380, 3185, 1680, 1655, 1630 cm 1; 1H NMR d 2.67 (3
H, s, Me), 5.93 (1 H, bs, NH), 6.30 (1 H, bs, NH), 7.58
(1 H, t, J=7.8 Hz, 5-H), 8.06 (1 H, dt, J=7.8, 1.5 Hz, 4-
H), 8.12 (1 H, dt, J=7.8, 1.5 Hz, 6-H), 8.40 (1 H, t,
J=1.5 Hz, 2-H); MS (EI) m/z 164.0671 (M)
(13C12C9H9NO2 requires 164.0667), 163.0636 (M)
(12C9H8NO2 requires 163.0633), 148 (100%).
Conclusion
Routes have been developed for syntheses of series of 3-
substituted benzamides and 5-substituted isoquinolin-1-
ones. In a preliminary screen in vitro, all the test com-
pounds inhibited PARP activity. Further studies are
being undertaken to establish the nature of the inhibi-
tion by the more potent members of these series; the
results of these studies will be published elsewhere.
Modi®cations of the potent lead compounds identi®ed
here, by incorporating appropriate targetting moieties
or by development of selectively-activated prodrugs,

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
729
RS-3-(1-Hydroxyethyl)benzamide (12). 3-Acetylbenz-
amide (11) (500 mg, 3.1 mmol) was stirred with NaBH4
(156 mg, 4.1 mmol) in EtOH (12 mL) at 0 ꢀC for 30 min
and at 20 ꢀC for 1 h. The reaction was quenched by
addition of dilute HCl and the solvent was evaporated.
The residue, in EtOAc, was washed with water (2Â) and
was dried. Evaporation gave 12 (311 mg, 62%) as white
crystals: mp 64±66 ꢀC; [a]D20 +46.6ꢀ (c 2.0, CH2Cl2); 1H
NMR d 2.86 (2 H, bs, 2ÂOH), 3.60 (1 H, dd, J=11.4,
7.9 Hz) and 3.78 (1 H, dd, J=11.4, 3.3 Hz)(CH2), 4.84 (1
H, dd, J=7.9, 3.3 Hz, ArCH), 7.46 (1 H, t, J=7.7 Hz, 5-
H), 7.60 (2 H, m, 4,6-H2), 7.71 (1 H, s, 2-H); MS (FAB)
m/z 164.0657 (100%) (M+H) (12C9H10NO2 requires
164.0712).
crystals: mp 127±129 ꢀC; IR 3360, 3190, 1670,
1
1620 cm
;
1H NMR ((CD3)2SO) d 1.33 (3 H, d,
RS-3-(1,2-Dihydroxyethyl)benzonitrile (14RS). Com-
pound 13 (80 mg, 500 mmol) was stirred with NaBH4
(19 mg, 500 mmol) in EtOH (3 mL) for 3 h. The eva-
poration residue, in EtOAc, was washed with water.
J=6.6 Hz, Me), 4.76 (1 H, q, J=6 Hz, CH), 5.24 (1 H,
d, J=4.0 Hz, OH), 7.32 (1 H, bs, NH), 7.38 (1 H, t,
J=7.7 Hz, 5-H), 7.49 (1 H, d, J=7.7 Hz, 4-H), 7.72 (1
H, d, J=7.7 Hz, 6-H), 7.86 (1 H, s, 2-H), 7.95 (1 H, bs,
NH); MS (EI) m/z 166.0867 (100%) (M) (12C9H11NO2
requires 166.0868).
Drying and evaporation gave 14RS (70 mg, 86%) as a
1
colourless viscous oil: IR 3340±3410, 2220 cm
;
1H
NMR d 2.9 (1 H, bs, OH), 3.5 (1 H, bs, OH), 3.60 (1 H,
dd, J=11.3, 3.2 Hz) and 3.78 (1 H, dd, J=11.3,
8.1 Hz)(CH2), 4.85 (1 H, dd, J=3.2, 8.1 Hz, ArCHOH),
7.47 (1 H, t, J=7.7 Hz, 5-H), 7.6 (2 H, m, 4,6-H2), 7.72
(1 H, s, 2-H).
3-(Hydroxyacetyl)benzonitrile (13). 3-Acetylbenzonitrile
10 (1.00 g, 6.9 mmol), PhI(O2CCF3)2 (5.9 g, 13.8 mmol)
and CF3CO2H (1.06 mL, 13.8 mmol) were boiled under
re¯ux in water (8 mL) and MeCN (40 mL) for 8 h. The
evaporation residue, in EtOAc, was washed with water.
The aqueous phase was extracted with EtOAc (3Â). The
organic extracts were washed with saturated aq
NaHCO3 and were dried. Evaporation and chromato-
graphy (EtOAc:hexane, 1:1) gave 13 (310 mg, 28%) as
white crystals: mp 102±105 ꢀC; IR 3400±3600, 2200,
R-3-(1,2-Dihydroxyethyl)benzamide (15R). The R-diol
14R was treated with H2O2 and NaOH, as for the
synthesis of 15S, to give 15R (quant) as a hygroscopic
viscous gum: 1H NMR ((CD3)2SO) 3.45 (2 H, d,
J=6.4 Hz, CH2), 4.0 (2 H, bs, 2ÂOH), 4.57 (1 H, t,
J=6.4 Hz, ArCHOH), 7.32 (1 H, bs, NH), 7.37 (1 H, t,
J=7.7 Hz, 5-H), 7.48 (1 H, d, J=7.7 Hz, 4-H), 7.74 (1
H, d, J=7.7 Hz, 6-H), 7.85 (1 H, s, 2-H), 7.96 (1 H, bs,
NH); MS (FAB) m/z 182.0816 (100%) (M+H)
(12C9H12NO3 requires 182.0817).
1
1650±1700 cm 1; H NMR ((CD3)2SO) d 3.37 (1 H, bs,
OH), 4.91 (2 H, s, CH2), 7.68 (1 H, t, J=7.7 Hz, 5-H),
7.92 (1 H, d, J=7.7 Hz, 4-H), 8.15 (1 H, d, J=7.7 Hz, 6-
H), 8.22 (1 H, s, 2-H); MS (EI) m/z 161.0471 (M)
(12C9H7NO2 requires 161.0432) 130 (100%).
S-3-(1,2-Dihydroxyethyl)benzamide (15S). The S-diol
14S (330 mg, 2.2 mmol), in EtOH (6.5 mL), was stirred
at 50 ꢀC for 1 h with aq NaOH (0.5 M, 1.0 mL,
0.5 mmol) and aq H2O2 (27.5%, 0.85 mL, 6.8 mmol).
The solution was neutralised with dilute HCl. Filtration
and evaporation gave 15S (366 mg, quant.) as a hygro-
R-3-(1,2-Dihydroxyethyl)benzonitrile (14R). To a fer-
menting slurry of bakers' yeast (88 g), water (880 mL)
and sucrose (88 g) was added 13 (600 mg, 3.7 mmol).
The mixture was stirred at 30 ꢀC for 7 days. Et2O
(100 mL) and Celite1 (100 g) were added and stirring
continued for a further 15 min. The slurry was ®ltered
and the solids were washed with Et2O (2Â50 mL). The
®ltrate was extracted with Et2O (5Â100 mL). Drying,
evaporation and chromatography (EtOAc:hexane 1:1
3:2) gave 14R (268 mg, 44%) as a white solid: mp 64±
1
scopic viscous gum: H NMR ((CD3)2SO) d 3.45 (2 H,
d, J=6.4 Hz, CH2), 4.0 (2 H, bs, 2ÂOH), 4.57 (1 H, t,
J=6.4 Hz, ArCHOH), 7.32 (1 H, bs, NH), 7.37 (1 H, t,
J=7.7 Hz, 5-H), 7.48 (1 H, d, J=7.7 Hz, 4-H), 7.74 (1
H, d, J=7.7 Hz, 6-H), 7.85 (1 H, s, 2-H), 7.96 (1 H, bs,
NH); MS (FAB) m/z 182.0822 (100%) (M+H)
(12C9H12NO3 requires 182.0817).
66 ꢀC; [a]D
46.6ꢀ (c 1.0, CH2Cl2); H NMR d 2.33 (2
20
1
H, bs, OH), 3.62 (1 H, dd, J=11.4, 7.7 Hz) and 3.81 (1
H, dd, J=11.4, 3.6 Hz)(CH2OH), 4.87 (1 H, dd, J=7.7,
3.6 Hz, ArCH), 7.48 (1 H, t, J=7.7 Hz, 5-H), 7.6 (2 H,
m, 4,6-H2), 7.71 (1 H, s, 2-H); MS(CI) m/z 164 (M+H)
(100%), 146.
3-(Bromoacetyl)benzonitrile (16) and 3-(dibromoacetyl)-
benzonitrile (17). Br2 (1.1 mL, 21.3 mmol) was added
dropwise to 10 (3.00 g, 20.7 mmol) in CHCl3 (15 mL) at
0 ꢀC during 30 min. The mixture was stirred at 20 ꢀC for
2 h. Evaporation and chromatography (EtOAc:hexane,
1:9) gave 17 (500 g, 8%) as white crystals: mp 84.5±
86 ꢀC; Found: C, 35.65; H, 1.62; N, 4.58. C9H5Br2NO
requires C, 35.68; H, 1.66; N, 4.62%; IR 2240,
S-3-(1,2-Dihydroxyethyl)benzonitrile (14S). The S-ester
20 (500 mg, 2.4 mmol) was stirred with aq NH3 (35%,
5.0 mL) in MeOH (15 mL) for 45 min. The volatile
materials were evaporated. THF (50 mL) was added and
the volatile materials were evaporated; this procedure
was repeated twice. CH2Cl2 (100 mL) was added. Dry-
ing and evaporation gave 14S (360 mg, 92%) as white
1
1705 cm 1; H NMR d 6.56 (1 H, s, CHBr2), 7.68 (1 H,
t, J=7.9 Hz, 5-H), 7.92 (1 H, dt, J=7.9, 1.5 Hz, 6-H),
8.37 (1 H, dt, J=7.9, 1.5 Hz, 4-H), 8.42 (1 H, t,

730
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
J=1.5 Hz, 2-H); MS (CI) m/z 306/304/302 (M). Further
elution gave 16 (3.60 g, 78%) as white needles: mp 70±
washed with 10% aq Na2CO3 and with brine. Drying
and evaporation gave 21 (55 mg, 82%) as a white solid:
mp 128±130 ꢀC; [a]D20 +39.2ꢀ (c 2.5, CH2Cl2); 1H NMR
d 1.50 (3 H, s, Me), 1.57 (3 H, s, Me), 3.70 (1 H, t,
J=8.0 Hz, dioxole 5-H), 4.35 (1 H, dd, J=8.0, 6.2 Hz,
dioxole 5-H), 5.13 (1 H, dd, J=8.0, 6.2 Hz, dioxole 4-
H), 5.71 (1 H, bs, NH), 6.11 (1 H, bs, NH), 7.57 (1 H, t,
J=7.6 Hz, Ar 5-H), 7.55 (1 H, dt, J=7.6, 1.5 Hz, Ar 4-
H), 7.73 (1 H, dt, J=7.6, 1.5 Hz, Ar 6-H), 7.82 (1 H, t,
J=1.5 Hz, 2-H); MS (FAB) m/z 222.1153 (M+H)
(100%) (12C12H15NO3 requires 222.1130), 164.
71 ꢀC (lit.59 mp 65.5±66.5 ꢀC); IR 2240, 1710 cm
;
1H
1
NMR d 4.43 (2 H, s, CH2Br), 7.67 (1 H, t, J=7.9 Hz, 5-
H), 7.90 (1 H, dt, J=7.9, 1.5 Hz, 6-H), 8.22 (1 H, dt,
J=7.9, 1.5 Hz, 4-H), 8.28 (1 H, t, J=1.5 Hz, 2-H);
MS(CI) m/z 224/222 (M).
3-(Acetoxyacetyl)benzonitrile (18), 3-acetylbenzonitrile
(10) and 3-(diethoxyacetyl)benzonitrile (19). Compound
16 (5.50 g, 24.5 mmol) was boiled under re¯ux with
KOAc (2.40 g, 24.5 mmol) and NaI (360 mg, 2.4 mmol)
in EtOH (100 mL) for 6 h. The evaporation residue, in
EtOAc, was washed with water (3Â). Drying, evapora-
tion and chromatography (hexane:EtOAc, 4:1) gave 19
(160 mg, 3%) as a pale-yellow oil: IR (®lm) 2240,
1705 cm 1; 1H NMR d 1.26 (6 H, t, J=7.2 Hz, 2 ÂMe),
3.65 (2 H, dq, J=9.5, 7.2 Hz) and 3.81 (2 H, dq, J=9.5,
7.2 Hz) (2ÂCH2), 7.59 (1 H, t, J=7.7 Hz, 5-H), 7.83 (1
H, dt, J=7.7, 1.5 Hz, 6-H), 8.39 (1 H, dt, J=7.7, 1.5 Hz,
4-H), 8.51 (1 H, t, J=1.1 Hz, 2-H); 13C NMR d 15.00,
64.01, 103.68, 112.58, 117.92, 129.14, 133.71), 133.78,
134.07, 135.97, 192.03; MS (FAB) m/z 234.1137 (M+H)
(100%) (12C13H16NO3 requires 234.1130). Further elu-
tion gave 10 (1.15 g, 32%). Further elution gave 18
(3.20 g, 64%) as white crystals: mp 70±72 ꢀC; Found: C,
64.9; H, 4.42; N, 6.76. C11H9NO3 requires C, 65.02; H,
3-(Bromoacetyl)benzamide (22). Br2 (2.0 g, 12 mmol) was
added to 11 (2.0 g, 12 mmol) in AcOH (180 mL) at
60 ꢀC. The mixture was stirred for 2.5 h. Evaporation
and recrystallisation (MeCN) gave 22 (1.74 g, 48%) as
white crystals: mp 133±135 ꢀC; Found: C, 44.8; H, 3.33;
N, 5.78. C9H8BrNO2 requires C, 44.66; H, 3.33; N,
5.79%; 1H NMR d 4.67 (2 H, s, CH2), 6.82 (1 H, bs,
NH), 7.59 (1 H, t, J=7.8 Hz, 5-H), 8.01 (1 H, bs, NH),
8.12 (1 H, dt, J=7.8, 1.5 Hz) and 8.21 (1 H, dt, J=7.8,
1.5 Hz) (4,6-H2), 8.57 (1 H, t, J=1.5 Hz, 2-H); MS
(FAB) m/z 243.9796 (100%) (M+H) (12C9H981BrNO2
requires 243.9796), 241.9810 (100%) (M+H)
(12C9H979BrNO2 requires 241.9817).
RS-3-(Oxiran-2-yl)benzamide (24). Compound 22 (500 mg,
2.1 mmol) was stirred with NaBH4 (47 mg, 1.2 mmol) in
EtOH (6 mL) at 0 ꢀC for 30 min and at 20 ꢀC for 2 h. Aq
NaOH (1.0 M, 1.5 mL, 1.5 mmol) was added and the
mixture was stirred at 20 ꢀC for 20 min and at 50 ꢀC for
10 min. Water (30 mL) was added and the mixture was
extracted with CH2Cl2 (2Â). The extracts were washed
with brine. Drying and evaporation gave 24 (331 mg,
98%) as white crystals: mp 95.5±97 ꢀC; Found: C, 66.0;
H, 5.67; N, 8.38. C9H9NO2 requires C, 66.25; H, 5.56;
1
4.46; N, 6.89%; IR 2240, 1755, 1710 cm 1; H NMR d
2.24 (3 H, s, Me), 5.31 (2 H, s, CH2), 7.66 (1 H, t,
J=7.8 Hz, 5-H), 7.90 (1 H, dt, J=7.8, 1.3 Hz, 6-H), 8.14
(1H, dt, J=7.8, 1.3 Hz, 4-H), 8.20 (1 H, t, J=1.3 Hz, 2-
H); MS(CI) m/z 204 (M+H) (100%).
S-3-(2-Acetoxy-1-hydroxyethyl)benzonitrile (20). Bakers'
yeast (7.5 g) was suspended in water (56 mL) at 30 ꢀC
and sucrose (11.2 g) was added. To this fermenting
slurry, 18 (760 mg, 3.7 mmol) was added and the mix-
ture was stirred for 8 h before being ®ltered through
Celite1. The ®ltrate was extracted with Et2O
(4Â30 mL). Drying, evaporation and chromatography
(EtOAc:hexane 2:3) gave 20 (650 mg, 85%) as a colour-
1
N, 8.58%; H NMR d 2.89 (1 H, m, oxirane 3-H), 3.15
(1 H, t, J=4.8 Hz, oxirane 3-H), 3.99 (1 H, m, oxirane
2-H), 7.40 (1 H, bs, NH), 7.44 (2 H, m, 4,5-H2), 7.79 (2
H, m, 2,6-H2), 8.01 (1 H, bs, NH); MS (FAB) m/z
164.0736 (100%) (M+H) (12C9H10NO2 requires
164.0712).
less oil: [a]D +34.9ꢀ (c 2.9, CH2Cl2); IR (®lm) 3400±
20
1
3500, 2240, 1740 cm
;
1H NMR d 2.11 (3 H, s, Me),
3.04 (1 H, bs, OH), 4.12 (1 H, dd, J=11.6, 7.9 Hz) and
4.30 (1 H, dd, J=11.6, 3.5 Hz)(CH2O), 5.01 (1 H, dd,
J=7.9, 3.5 Hz, ArCH), 7.49 (1 H, t, J=7.7 Hz, 5-H),
7.62 (2 H, m, 4,6-H2), 7.73 (1 H, t, J=1.6 Hz, 2-H); MS
(FAB) m/z 206.0869 (M+H) (12C11H12NO3 requires
206.0817), 188 (100%). Further elution gave 3-(1,2-
dihydroxyethyl)benzonitrile of unknown stereo-
chemistry (14 mg, 2%) as a colourless oil.
5-Cyanoisoquinolin-1-one (30). Method A. 2,6-Dicyano-
toluene 29 (2.0 g, 14 mmol) was stirred at 0±5 ꢀC with
KOBut (8.7 g, 77 mmol) in freshly distilled EtOCHO
(50 mL) for 25 min. Et2O (100 mL) was added and the
suspension was ®ltered. The solid was dissolved in water
(20 mL). The solution was acidi®ed with AcOH, satu-
rated with NaCl and extracted with CHCl3. Drying,
evaporation, chromatography (CHCl3:EtOAc, 2:1) and
chromatography (hexane:EtOAc, 4:1) yielded 31
(330 mg, 14%) as white crystals: mp 212±214 ꢀC; IR
2240, 1735 cm 1; 1H NMR d 6.88 (1 H, d, J=5.7 Hz, 4-
H), 7.46 (1 H, d, J=5.7 Hz, 3-H), 7.64 (1 H, t,
J=7.7 Hz, 7-H), 8.05 (1 H, d, J=7.7 Hz, 6-H), 8.52 (1
S-3-(4,5-Dihydro-2,2-dimethyl-1,3-dioxol-4-yl)benzamide
(21). Freeze-dried 15S (62 mg, 34 mmol) was stirred with
2,2-dimethoxypropane (1.0 mL) and TsOH (1.0 mg) for
3 days. EtOAc (50 mL) was added and the solution was

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
731
H, d, J=7.7 Hz, 8-H); MS (EI) m/z 172.0353 (M)
(13C12C9H5NO2 requires 172.0353), 171.0319 (M)
(12C10H5NO2 requires 171.0320), 143 (100%). Further
elution gave 30 (300 mg, 13%) as a pale-yellow solid: mp
10% aq NNa2CO3. Drying, evaporation and recrys-
tallisation CH2Cl2) gave 34 (132 mg, 27%) as white
crystals: mp 160±162 ꢀC; Found: C, 64.25; H, 4.55; N,
15.6. C19H16N4O3.0.5H2O requires C, 63.95; H, 4.8; N,
296±300 ꢀC (lit.44 mp 275±276 ꢀC); IR 3500, 3300, 2230,
15.7%; IR 3370, 3190, 2210, 1735, 1690, 1655 cm 1; H
1
1
1690, 1660 cm
;
1H NMR ((CD3)2SO) d 6.60 (1 H,
NMR ((CD3)2SO) d 2.59 (3 H, s, Me), 2.63 (3 H, s, Me),
7.50 (1 H, t, J=7.7 Hz, 5-H), 7.60 (1 H, t, J=7.7 Hz, 50-
H), 7.71 (1 H, d, J=7.7 Hz, 40-H), 7.72 (1 H, bs,
CONH2), 7.90 (2 H, d, J=7.7 Hz, 4,6-H2), 8.00 (1 H, bs,
CONH2), 8.04 (1 H, d, J=7.7 Hz, 60-H), 9.22 (1 H, s,
formamidine CH), 11.78 (1 H, bs, formamidine NH);
MS (FAB) m/z 349 (M+H), 172 (M- H2NOC(-
Me)C6H5CON), 161 (NC(Me)C6H5CONH2+H).
J=7.0 Hz, 4-H), 7.47 (1 H, d, J=7.0 Hz, 3-H), 7.63 (1
H, t, J=7.7 Hz, 7-H), 8.26 (1 H, d, J=7.7 Hz, 6-H), 8.47
(1 H, d, J=7.7 Hz, 8-H), 11.72 (1 H, bs, NH); m/z (EI)
171.0512 (M) (13C12C9H6N2O requires 171.0514),
170.0480 (M) (12C10H6N2O requires 170.0480).
5-Cyanoisoquinolin-1-one (30). Method B. Compound
35 (300 mg, 1.5 mmol) in dry MeOH (50 mL) was satu-
rated with hydrogen chloride. The mixture was boiled
under re¯ux for 3 days. The evaporation residue was
stirred with 10% aq NNa2CO3 for 30 min. The suspen-
sion was extracted with EtOAc (4Â). Drying, evapora-
tion, and recrystallisation, followed by chromatography
of the mother liquor (hexane:EtOAc, 4:1), gave 30
(141 mg, 55%) as an o€-white solid, mp 296±
300 ꢀC. Also isolated by chromatography was 31 (13 mg,
5%).
E-2-(2,6-Dicyanophenyl)-N,N-dimethylethenamine (35).
2,6-Dicyanotoluene 29 (1.00 g, 7.0 mmol) was boiled
under re¯ux in Me2NCH(OMe)2 (10 mL) for 4 days.
Evaporation and recrystallisation (EtOH) gave 35
(973 mg, 71%) as a yellow solid: mp 117±119 ꢀC; 1H
NMR d 3.00 (6 H, s, NMe2), 5.36 (1 H, d, J=13.6 Hz,
ArCH=C), 6.94 (1 H, t, J=7.7 Hz, 5-H), 7.63 (2 H, d,
J=7.7 Hz, 4,6-H2), 7.77 (1 H, d, J=13.6 Hz, C=CHN);
MS (EI) m/z 197.0960 (100%) (M) (12C12H11N3 requires
197.0953), 182 (M-Me), 155.
3-Cyano-2-methylbenzamide (32). 2,6-Dicyanotoluene 29
(13.0 g, 21 mmol) was stirred with aq H2O2 (27.5% w/v,
7.0 mL, 57 mmol) and NaOH (210 mg, 5.25 mmol) in
EtOH (75 mL) and water (10 mL) for 1 h before being
extracted with EtOAc (4Â100 mL). Drying, evapora-
tion, and recrystallisation (EtOAc) gave 32 (1.62 g,
48%) as white crystals: mp 186±188 ꢀC; IR 3420, 3350,
Isoquinoline-5-carboxylic acid (36). Compound 30
(427 mg, 2.5 mmol) was boiled under re¯ux with KOH
(2.4 g, 43 mmol) in EtOH (10 mL) under N2 for 3 days.
The mixture was acidi®ed with aq HCl (9 M). The eva-
poration residue, in MeOH, was ®ltered. Evaporation of
the solvent from the ®ltrate gave 36 (394 mg, 83%) as a
2218, 1680, 1625 cm 1; H NMR ((CD3)2SO) d 2.65 (3
white solid: mp >300 ꢀC; IR 3400, 3400±2700, 1705,
1
1
H, s, Me), 6.94 (1 H, bs, NH), 7.36 (1 H, t, J=7.7 Hz, 5-
H), 7.52 (1 H, bs, NH), 7.65 (2 H, m, 4,6-H2); MS (EI)
m/z 160.0634 (M) (100%) (12C9H8N2O requires
160.0637).
1655, 1625 cm
;
1H NMR d (CD3OD) 7.26 (1 H, d,
J=7.7 Hz, 4-H), 7.58 (1 H, t, J=7.7 Hz, 7-H), 7.75 (1 H,
d, J=7.7 Hz, 3-H), 8.41 (1 H, d, J=7.7 Hz, 8-H), 8.55 (1
H, d, J=7.7 Hz, 6-H); MS (EI) m/z 190.0458 (M)
(13C12C9H7NO3 requires 190.0459), 189.0425 (100%)
(M) (12C10H7NO3 requires 189.0426).
3-Cyano-N-(dimethylaminomethylene)-2-methylbenzamide
(33). 3-Cyano-2-methylbenzamide 32 (102 mg, 640 mmol)
was boiled under re¯ux with Me2NCH (OMe)2 (164 mg,
1.25 mmol) in DMF (1 mL) for 3 h. Water (20 mL) was
added and the mixture was extracted with CH2Cl2 (4Â).
The combined extracts were washed with water and with
brine. Drying and evaporation gave 33 (114 mg, 83%) as
RS-Isoquinolin-5-ylphenylmethanol (39). BuLi in hex-
anes (2.5 M, 0.40 mL, 1.0 mmol) was added to 37
(208 mg, 1.0 mmol) in dry THF (6.2 mL) at 116 ꢀC
under N2 and stirring continued for 10 min. PhCHO
(104 mg, 980 mmol) in dry THF (1.0 mL) was added and
the mixture was allowed to warm to 20 ꢀC during
30 min. Aq NH4Cl (20%, 1.0 mL) was added. After
5 min, the solvents were evaporated. The residue, in
CH2Cl2, was washed with saturated aq NaHCO3 and
was dried. Evaporation and chromatography (hex-
ane:EtOAc, 3:2) gave 40 (4 mg, 3%) as a pale-yellow oil:
1H NMR d 7.66 (3 H, m, 4,6,7-H3), 7.82 (1 H, d,
J=8.1 Hz, 5-H), 7.98 (1 H, d, J=8.1 Hz, 8-H), 8.53 (1
H, d, J=5.9 Hz, 3-H), 9.26 (1 H, s, 1-H). Further elu-
tion gave recovered 37 (26 mg, 13%). Further elution
gave 39 (79 mg, 34%) as white crystals: mp 124±126 ꢀC;
1H NMR d 1.7 (1 H, bs, OH), 6.48 (1 H, s, CHOH), 7.36
a yellow solid: mp 98±100 ꢀC; H NMR d 2.81 (3 H, s,
1
Ar-Me), 3.19 (3 H, s, NMe), 3.23 (3 H, s, NMe), 7.33 (1
H, t, J=7.7 Hz, 5-H), 7.65 (1 H, d, J=7.7 Hz, 4-H), 8.18
(1 H, d, J=7.7 Hz, 6-H), 8.62 (1 H, s, NCH); MS (FAB)
m/z 216.1139 (M+H) (100%) (12C12H14N3O requires
216.1137).
N-(3-Carbamoyl-2-methylbenzoyl)-N0-(3-cyano-2-methyl-
benzoyl)formamidine (34). Compound 33 (300 mg,
1.4 mmol) was boiled under re¯ux with TsOH (361 mg,
2.1 mmol) in PhMe (10 mL) for 13 h. The evaporation
residue, in CH2Cl2, was washed with water and with

732
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
(5 H, m, Ph-H5), 7.65 (1 H, t, J=7.7 Hz, isoquinoline 7-
H), 7.78 (1 H, d, J=6.1 Hz, isoquinoline 4-H), 7.9 (2 H,
m, isoquinoline 6,8-H2), 8.41 (1 H, d, J=6.1 Hz, iso-
quinoline 3-H), 9.20 (1 H, s, isoquinoline 1-H); MS (EI)
m/z 236.1030 (M) (13C12C15H13NO requires 236.1030),
235.0993 (M) (12C16H13NO requires 235.0997).
7.8 Hz, dioxole 5-H), 4.45 (1 H, dd, J=7.8, 5.5 Hz,
dioxole 4-H), 6.87 (1 H, d, J=7.3 Hz, isoquinoline 4-H),
7.29 (1 H, d, J=7.3 Hz, isoquinoline 3-H), 7.50 (1 H, t,
J=7.9 Hz, isoquinoline 7-H), 7.61 (1 H, d, J=7.9 Hz,
isoquinoline 6-H), 8.34 (1 H, d, J=7.9 Hz, isoquinoline
8-H), 11.15 (1 H, br, NH); MS (FAB) m/z 302.1384
(M+H) (12C17H20NO4 requires 302.1392).
E-3-(1-Oxoisoquinolin-5-yl)propenoic acid (43). 5-Iodoiso-
quinolin-1-one 4224 (100mg, 370 mmol) was boiled
under re¯ux with propenoic acid (35 mg, 490 mmol),
Pd(OAc)2 (8.0 mg, 37 mmol) and Et3N (93 mg, 920 mmol)
in EtCN (0.3 mL) for 1 h. Dilute HCl (10 mL) was
added. The precipitate was washed with water and was
5-(E-5-Hydroxypent-1-enyl)isoquinolin-1-one (47), 5-(4-
hydroxy-1-methylenebutyl)isoquinolin-1-one (48), 5-(5-
oxopentyl)isoquinolin-1-one (49) and 5-(RS-4-oxo-1-
methylbutyl)isoquinolin-1-one (50). 5-Iodoisoquinolin-1-
one 4224 (400 mg, 1.5 mmol) was boiled under re¯ux
with pent-4-enol (170 mg, 1.9 mmol), Pd(OAc)2 (32 mg,
140 mmol) and Et3N (370 mg, 3.7 mmol) in EtCN
(1.5 mL) for 2 h. The evaporation residue, in EtOAc,
was washed with dilute HCl, with water and with brine,
and was dried. Evaporation and chromatography
(EtOAc:MeOH, 99:1) gave unreacted 42 (221 mg, 55%).
Further elution gave a mixture of 49 (5% by NMR) and
dried to give 43 (76 mg, 97%) as an o€-white solid: mp
315±318 ꢀC; IR 3550, 1695, 1660 cm
1
;
1H NMR
((CD3)2SO) d 6.57 (1 H, d, J=15.8 Hz, 2-H), 6.76 (1 H,
d, J=7.3 Hz, isoquinoline 4-H), 7.28 (1 H, d, J=7.3 Hz,
isoquinoline 3-H), 7.52 (1 H, t, J=7.7 Hz, isoquinoline
7-H), 8.10 (1 H, d, J=15.8 Hz, 3-H), 8.12 (1 H, d,
J=7.7 Hz) and 8.27 (1 H, d, J=7.7 Hz) (isoquinoline 6,
8-H2); MS (FAB) m/z 216.0616 (M+H) (12C12H10NO3
requires 216.0661), 215.0584 (M) (12C12H9NO3 requires
215.0582).
1
50 (2.5% by NMR). 49: H NMR ((CD3)2SO) d 1.70 (4
H, m, pentyl 2, 3-H4), 2.50 (2 H, m, pentyl 4-H2), 2.90 (2
H, m, pentyl 1-H2), 6.73 (1 H, d, J=7.7 Hz, isoquinoline
4-H), 7.23 (1 H, d, J=7.7 Hz, isoquinoline 3-H), 7.50 (2
H, m, isoquinoline 6, 7-H2), 8.30 (1 H, m, isoquinoline
5-(1E-3RS-3-(4R-2,2-Dimethyldihydro-1,3-dioxol-4-yl)-
3-hydroxyprop-1-enyl)isoquinolin-1-one (45) and 5-(3-
(4R-2,2-dimethyldihydro-1,3-dioxol-4-yl)-3-oxopropyl)iso-
quinolin-1-one (46). 5-Iodoisoquinolin-1-one 4224
(400 mg, 1.5 mmol) was boiled under re¯ux with 4R-2,2-
dimethyl-4-(1RS-1-hydroxyprop-2-enyl)dihydro-1,3-
dioxole 4452,53 (315 mg, 1.9 mmol), Pd(OAc)2 (32 mg,
140 mmol) and Et3N (370 mg, 3.7 mmol) in EtCN
(1.5 mL) for 2 h. The evaporation residue, in EtOAc,
was washed with dilute HCl (2 M) and with brine. Dry-
ing, evaporation and chromatography (EtOAcMe2CO:-
hexane, 2:1) gave 45 (76 mg, 18%) as an o€-white solid:
mp 164±166 ꢀC; 1H NMR d 1.40 (3 H, s, Me), 1.50 (3 H,
s, Me), 2.17 (1 H, br, OH), 3.93 (0.5 H, dd, J=8.5,
5.5 Hz, dioxole 5-H, diastereoisomer A), 4.04 (1 H, m,
dioxole 5-H2, diastereoisomer B), 4.1 (0.5 H, m, dioxole
5-H, A), 4.20 (0.5 H, m, dioxole 4-H, A), 4.28 (0.5 H, m,
dioxole 4-H, B), 4.34 (0.5 H, m, propenyl 3-H, A), 4.56
(0.5 H, m, propenyl 3-H, B), 6.18 (1 H, m, propenyl 4-
H), 6.81 (0.5 H, d, J=7.3 Hz, isoquinoline 4-H, B), 6.82
(0.5 H, d, J=7.3 Hz, isoquinoline 4-H, A), 7.19 (1 H, d,
J=7.3 Hz, isoquinoline 3-H), 7.24 (1 H, m, propenyl 1-
H), 7.48 (1 H, t, J=7.8 Hz, isoquinoline 7-H), 7.78 (1 H,
d, J=7.8 Hz, isoquinoline 6-H), 8.37 (1 H, d, J=7.8 Hz,
isoquinoline 8-H), 10.6 (1 H, br, NH); MS (FAB) m/z
302.1373 (M+H) (12C17H20NO4 requires 302.1392).
Further elution gave 46 (176 mg, 42%) as an o€-white
solid: mp 138±141 ꢀC; IR 3300, 3160, 1715, 1660,
1
8-H), 9.78 (1 H, br, CHO), 11.37 (1 H, br, NH). 50: H
NMR ((CD3)2SO) d 1.30 (5 H, m, Me + butyl 2-H2),
2.40 (2 H, t, J=7.7 Hz, butyl 2-H2), 3.40 (1 H, m, butyl
1-H), 6.83 (1 H, d, J=7.7 Hz, isoquinoline 4-H), 7.50 (1
H, m, isoquinoline 3,6,7-H3), 8.30 (1 H, m, isoquinoline
8-H), 9.74 (1 H, br, CHO), 11.37 (1 H, br, NH); MS
(FAB) m/z 230.1184 (M+H) (12C14H16NO2 requires
230.1181). Further elution gave a mixture of 47 (13% by
NMR) and 48 (11% by NMR). 47: 1H NMR
((CD3)2SO) d 1.60 (2 H, quintet, J=6.8 Hz, pentenyl 2-
H2), 2.30 (2 H, m, pentenyl 3-H2), 3.50 (2 H, m, pente-
nyl 1-H2), 6.3 (1 H, m, pentenyl 4-H), 6.75 (1 H, d,
J=7.5 Hz, isoquinoline 4-H), 6.93 (1 H, d, J=15.6 Hz,
pentenyl 5-H), 7.18 (1 H, d, J=7.5 Hz, isoquinoline 3-
H), 7.48 (1 H, m, isoquinoline 7-H), 7.80 (1 H, d,
J=7.5 Hz, isoquinoline 6-H), 8.10 (1 H, m, isoquinoline
1
8-H). 48: H NMR ((CD3)2SO) d 2.10 (2 H, m, butyl 3-
H2), 2.50 (2 H, m, butyl 2-H2), 3.50 (2 H, m, butyl 1-
H2), 5.5 (2 H, m, H2C=C), 6.62 (1 H, d, J=7.7 Hz,
isoquinoline 4-H), 7.18 (1 H, d, J=7.7 Hz, isoquinoline
3-H), 7.48 (2 H, m, isoquinoline 6,7-H2), 8.10 (1 H, m,
isoquinoline 8-H); MS (FAB) m/z 230.1178 (M+H)
(12C14H16NO2 requires 230.1181).
Acknowledgements
The authors thank Mr R. R. Hartell and Mr D. J.
Wood for the NMR spectra and Mr C. Cryer for the
mass spectra. The support of the EPSRC Mass Spec-
trometry Service Centre in providing some of the
high-resolution mass spectra is acknowledged. We are
1
1630 cm 1; H NMR d 1.38 (3 H, s, Me), 1.43 (3 H, s,
Me), 2.99 (1 H, t, J=7.6 Hz) and 3.00 (1 H, t,
J=7.6 Hz) and 3.20 (2 H, m) (propyl 1,2-H4), 3.96 (1 H,
dd, J=8.5, 5.5 Hz, dioxole 5-H), 4.19 (1 H, dd, J=8.5,

C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
733
particularly grateful to Mrs J. Whish and Dr S. Kappus
for help with the PARP assays and for useful discus-
sions and to Dr S. Old®eld (University of Bristol) for
the diphtheria toxin assay and for help with the pre-
paration of the manuscript. C.Y.W. thanks the Royal
Pharmaceutical Society of Great Britain and Glaxo-
Wellcome for a studentship.
Stratford, I. J.; Stephens, M. A.; Fielden, E. M.; Adams, G. E.
Drug Design and Discovery 1993, 10, 249.
24. Berry, J. M.; Watson, C. Y.; Whish, W. J. D.; Threadgill,
M. D. J. Chem. Soc., Perkin Trans. 1 1997, 1147.
25. Ruf, A.; Menissier de Murcia, J.; de Murcia, G. M.;
Schultz, G. E. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7481.
26. Choe, S.; Bennett, M. J.; Fugii, G.; Curmi, P. M. G.;
Kantardjie€, K. A.; Collier, R. J.; Eisenberg, D. Nature
(London) 1992, 357, 216.
References
27. Bennett, M. J.; Eisenberg, D. Prot. Sci. 1994, 3, 1464.
1. Harris, A. L. Int. J. Radiat. Biol. 1985, 48, 675.
2. Shall, S. Adv. Radiat. Biol. 1984, 11, 1.
28. Bennett, M. J.; Choe, S.; Eisenberg, D. Prot. Sci. 1994, 3,
1444.
3. Ben-Hur, E. Int. J. Radiat. Biol. 1984, 46, 659.
29. Bell, C. E; Eisenberg, D. Biochemistry 1996, 35, 1137.
30. Bell, C. E; Eisenberg, D. Biochemistry 1997, 36, 481.
4. Ben-Hur, E.; Utsumi, H.; Elkind, M. M. Radiat. Res. 1984,
97, 546.
31. Brown, H. C.; Park, W. S.; Cho, B. T.; Ramachandran,
P. V. J. Org. Chem. 1987, 52, 5406.
5. Ueno, A. M.; Tanaka O.; Matsudaira, H. Radiat. Res. 1984,
98, 574.
32. Itsuno, S.; Nakano, M.; Miyazaki, K.; Masuda, H.; Ito,
K.; Hirao, A.; Nakahama, S. J. Chem. Soc., Perkin Trans. 1
1985, 2039.
6. Thraves, P.; Mossman, K. L.; Brennan, T.; Dritschilo, A.
Radiat. Res. 1985, 104, 119.
7. Grin, R. J.; Curtin, N. J.; Newell, D. R.; Golding, B. T.;
Durkacz, B. W.; Calvert, A. H. Biochimie 1995, 77, 408.
33. Kawasaki, M.; Suzuki, Y.; Terashima, S. Chem. Lett. 1984,
239.
8. Sebolt-Leopold, J. S.; Scavone, S. V. Int. J. Radiat. Oncol.
Biol. Phys. 1992, 22, 619.
34. Servi, S. Synthesis 1990, 1.
35. Ridley, D. D.; Stralow, M. J. Chem. Soc., Chem. Commun.
1975, 400.
9. Chen, G.; Pan, Q.-C. Cancer Chemother. Pharmacol. 1988,
22, 303.
36. Manzocchi, A.; Fiecchi, A.; Santaniello, E. J. Org. Chem.
1988, 53, 4405.
10. Wedge, S. R.; Porteous, J. K.; Newlands, E. S. Br. J. Can-
cer 1996, 74, 1030.
37. Ferraboschi, P.; Grisenti, P.; Manzocchi, A.; Santaniello,
E. Tetrahedron 1994, 50, 10539.
11. Boulton, S.; Pemberton, L. C.; Porteous, J. K.; Curtin, N.
J.; Grin, R. J.; Golding, B. T.; Durkacz, B. W. Br. J. Cancer
1995, 72, 849.
38. Ferraboschi, P.; Grisenti, P.; Manzocchi, A.; Santaniello,
E. J. Chem. Soc., Perkin Trans. 1 1990, 2469.
12. Benjamin, R. C.; Gill, D. M. J. Biol. Chem. 1980, 255,
10493.
39. Noller, C. R.; Hartman, W. W.; Smith, L. A. Org. Synth.
Wiley, New York, 1943; Coll. Vol. II, p 586.
13. Benjamin, R. C.; Gill, D. M. J. Biol. Chem. 1980, 255,
10502.
40. Moriarty, R. M.; Berglund, B. A.; Penmasta, R. Tetra-
hedron Lett. 1992, 33, 6065.
14. Mendoza-Alvarez, H.; Alvarez-Gonzalez, R. J. Biol. Chem.
1993, 268, 22575.
41. Holshouser, M. H.; Kolb, M. J. Pharm. Sci. 1986, 75,
619.
15. Panzeter, P. L.; Althaus, F. R. Biochemistry 1994, 33, 9600.
42. Kaufmann. A.; Muller, H. Ber. Deutsch. Chem. Ges. 1918,
51, 123.
16. Purnell, M. R.; Whish, W. J. D. Biochem. J. 1980, 185,
775.
43. Dale, J. A.; Dull D. L.; Mosher, H. S. J. Org. Chem. 1969,
34, 2543.
17. Banasik, M.; Komura, H.; Shimoyama, M.; Ueda, K. J.
Biol. Chem. 1992, 267, 1569.
44. Wenkert, E.; Johnston, D. B. R.; Dave, K. G. J. Org.
Chem. 1964, 29, 2534.
18. Sestili, P.; Balsamini, C.; Spadoni, G.; Cattabeni, F.; Can-
toni, O. Pharmacol. Res. Commun. 1988, 20, 613.
45. Krohn, K.; Heins, H.; Wielckens, K. J. Med. Chem. 1992,
35, 511.
19. Cantoni, O.; Sestili, P.; Spadoni, G.; Balsamini, C.;
Cucchiarini, L.; Cattabeni, F. Biochem. Internat. 1987, 15, 329.
46. Gilman H.; Soddy, T. S. J. Org. Chem. 1957, 22, 565.
20. Suto, M. J.; Turner, W. R.; Arundel-Suto, C. M.; Werbel,
L. M.; Sebolt-Leopold, J. S. Anti-Cancer Drug Design 1991, 7,
107.
47. Wommack, J. B.; Barbee, T. G.; Thoennes, D. J.; McDo-
nald, M. A.; Pearson, D. E. J. Heterocyclic Chem. 1969, 6,
243.
21. Yoshida, S.; Aoyagi, T.; Harada, S.; Matsude, N.; Ikeda,
T.; Naganawa, H.; Hamada, M.; Takeuchi, T. J. Antibiot.
1991, 44, 111.
48. Clipperton, R. D. J. Ph.D. Thesis, University of Bath, 1969.
49. Osborn, A. R.; Scho®eld, K.; Short, L. N. J. Chem. Soc.
1956, 4191.
22. Grin, R. J.; Pemberton, L. C.; Rhodes, D.; Bleasdale, C.;
Bowman, K.; Calvert, A. H.; Curtin, N. J.; Durkacz, B. W.;
Newell, D. R.; Porteous, J. K.; Golding, B. T. Anti-Cancer
Drug Design 1995, 10, 507.
50. Heck, R. F. Org. React. 1982, 27, 345.
51. Plevyak, J. E.; Dickerson, J. E.; Heck, R. F. J. Org. Chem.
1979, 44, 4078.
23. Naylor, M. A.; Threadgill, M. D.; Showalter, H. D. H.;
52. Dumont, R.; Pfander, H. Helv. Chim. Acta 1983, 66, 814.

734
C. Y. Watson et al./Bioorg. Med. Chem. 6 (1998) 721±734
53. Jager, V.; Schroter, D.; Koppenhoe€er, B. Tetrahedron
1991, 47, 2195.
56. Redpath, N. T.; Proud, C. G. Biochim. Biophys. Acta 1994,
1220, 147.
57. Iglewski, W. J.; Dewhurst, S. FEBS Lett. 1991, 283, 235.
54. Van Ness, B. G.; Howard, J. B.; Bodley, J. W. J. Biol.
Chem. 1978, 253, 8687.
58. Reddy, M. V. R.; Vijayalakshmi, S.; Reddy, D. B; Reddy,
P. V. R. Phosphorus, Sulfur Silicon 1991, 60, 209.
55. Oppenheimer, N. J.; Bodley, J. W. J. Biol. Chem. 1981,
256, 8579.
59. Morozov, A. A. Zh. Obshch. Khim., 1989, 59, 215.