Phthalazinones. Part 1: The design and synthesis of a novel series of potent inhibitors of poly(ADP-ribose)polymerase

Article abstract of DOI:10.1016/j.bmcl.2005.03.026

Screening of the Maybridge compound collection identified 4-arylphthalazinones as micromolar inhibitors of PARP-1 catalytic activity. Subsequent optimisation of both inhibitory activity and metabolic stability led to a novel series of meta-substituted 4-benzyl-2H-phthalazin-1-ones with low nanomolar, cellular activity as PARP-1 inhibitors and promising metabolic stability in vitro.

Full text of DOI:10.1016/j.bmcl.2005.03.026

Bioorganic & Medicinal Chemistry Letters 15 (2005) 2235–2238
Phthalazinones. Part 1: The design and synthesis of a novel
series of potent inhibitors of poly(ADP-ribose)polymerase
Vincent M. Loh, Jr.,a,* Xiao-ling Cockcroft,a Krystyna J. Dillon,b Lesley Dixon,c
Jan Drzewiecki,a Penny J. Eversley,c Sylvie Gomez,a Janet Hoare,c Frank Kerrigan,c
Ian T. W. Matthews,a Keith A. Menear,a Niall M. B. Martin,b Roger F. Newton,c
Jane Paul,c Graeme C. M. Smith,b Julia Vilec and Alan J. Whittlec
aKuDOS Horsham Ltd, 26 Foundry Lane, Horsham, West Sussex RH13 5PX, UK
bKuDOS Pharmaceuticals Ltd, 327 Cambridge Science Park, Milton Road, Cambridge CB4 0WG, UK
cMaybridge, Trevillett, Tintagel, Cornwall PL34 0HW, UK
Received 20 December 2004; revised 4 March 2005; accepted 7 March 2005
Abstract—Screening of the Maybridge compound collection identified 4-arylphthalazinones as micromolar inhibitors of PARP-1
catalytic activity. Subsequent optimisation of both inhibitory activity and metabolic stability led to a novel series of meta-substituted
4-benzyl-2H-phthalazin-1-ones with low nanomolar, cellular activity as PARP-1 inhibitors and promising metabolic stability in
vitro.
Ó 2005 Elsevier Ltd. All rights reserved.
The mammalian enzyme poly(ADP-ribose) polymerase-
1 (PARP-1) has been implicated in the repair and signal-
ling of DNA damage through its ability to recognise and
rapidly bind to DNA single or double strand breaks.1
Once associated with DNA, activated PARP-1 utilises
nicotinamide adenine dinucleotide (NAD+) to synthesise
poly(ADP-ribose) homopolymers onto a variety of
nuclear target proteins, most notably itself.2 It has been
postulated that the formation of these negatively
charged polymers causes electrostatic repulsion of modi-
fied PARP-1, preventing DNA recombination. It is fur-
ther hypothesised that modified PARP-1 facilitates
recruitment of the base excision repair complex to the
site of the DNA damage.3
enhanced DNA damage and cell killing.6 PARP-1 activa-
tion has also been shown to play a pivotal role in the
development of septic shock, ischaemic injury and in
neurotoxicity.7,8 Inhibitors of PARP-1 catalytic activity
may therefore have wide-ranging therapeutic potential.
Previous investigators have designed inhibitors of
PARP-1 to mimic the substrate–protein interactions of
NAD+ with the enzyme. Mechanistically, these com-
pounds inhibit PARP-1 by blocking the binding of the
substrate, particularly the nicotinamide moiety, to the
active site of the enzyme. Early weak inhibitors such as
3-aminobenzamide 19 have been developed into more
potent PARP-1 inhibitors derived from a range of related
pharmacophoric templates, for example the 5-substi-
tuted dihydroisoquinolin-1-ones 2,10 2,8-disubstituted
quinazolin-4-ones 311 and benzoxazole-4-carboxamides
PARP-1 activation and subsequent poly(ADP ribosyl)-
ation are immediate cellular responses to chemical or
radiation-induced DNA damage.4 Studies have shown
that inhibition of PARP-1 activity enhances the effects
of radiation by suppressing the repair of potentially
lethal damage in cancerous cells.5 Inhibition of PARP-
1 in cells treated with alkylating agents similarly causes
O
O
O
O
H
H
H
H
N
H
N
N
N
H
N
R2
N
R
O
NH2
R
R1
1
2
3
4
Keywords: Phthalazinone; PARP; Poly(ADP)polymerase.
*
Corresponding author. Tel.: +44 (0)1403 248844; fax: +44 (0)1403
248855; e-mail: vloh@kudospharma.co.uk
Figure 1. Previously reported PARP-1 Inhibitors.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.03.026

2236
V. M. Loh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2235–2238
4.11 In this paper, we describe the synthesis and prelimin-
ary biological evaluation of novel 4-benzylphthalazi-
nones as potent inhibitors of PARP-1 (Fig. 1).
Table 1. Initial structure–activity relationships of the meta-substituted
4-benzyl-2H-phthalazin-1-ones
O
NH
N
High throughput screening of the Maybridge Screening
Collection in a FlashPlate scintillation proximity assay12
identified several 4-aryl-2H-phthalazinones as low
micromolar PARP-1 inhibitors. These appeared a promi-
sing start point, since phthalazinone itself is a known
PARP-1 inhibitor.13 Initial hit exploratory chemistry14
led to 4-benzyl-2H-phthalazin-1-one 5 with a potency
against human PARP-1 of IC50 0.77 lM (Table 1).
Early SAR studies, backed by structure-based design
utilising a homology model derived from the known
crystal structure of chicken PARP-1,15 suggested that
further structural elaboration around the meta position
of the benzyl moiety in 5 could be potency enhancing.
In particular, carbonyl-containing substituents at the
3-position led to potent PARP-1 inhibitors exemplified
by 6 and 7. These observations encouraged the synthesis
of focused analogue libraries using parallel synthesis
methodologies. Preliminary SAR development showed
that chain extension to the propionanilide 8 significantly
increased PARP-1 inhibitory potency (IC50 20 nM).
While further chain extension (9) or branching (10)
tended to reduce potency, constraining the branch into
a cyclopropane ring (11) reversed this trend. A number
of phthalazinones bearing different biaryl substituents
at the meta benzylic position (12–15) were also shown
to be potent PARP-1 inhibitors. Propionanilide 8 exhibi-
ted good PARP-1 inhibitory activity at 1 lM in a cell-
based assay (data not shown) making it a suitable candi-
date for continued optimisation chemistry. However,
preliminary pharmacokinetic studies raised concerns
that 8 would be rapidly cleared since in vitro it exhibited
poor metabolic stability (mouse hepatic microsomes: Cli
12 mL minÀ1 gÀ1 liver). It was postulated that the pen-
dant anilide was the most likely point of metabolic insta-
bility. Therefore, several structural modifications were
undertaken in an attempt to stabilise this group. Triflu-
oro-substitution a to the carbonyl to give 16 retained
potency against isolated PARP-1 enzyme (IC50
13 nM), but reduced in vitro metabolic stability (mouse
hepatic microsomes: Cli > 50 mL minÀ1 gÀ1 liver). Sev-
eral bioisosteric amide replacements, such as carbamate
17, urea 18, sulfonamide 19 and 1,2,4-oxadiazole 20,
were detrimental to PARP-1 inhibitory potency. How-
ever, reversal of the anilide topology in 8 to give amide
21 retained PARP-1 inhibitory potency and consider-
ably enhanced metabolic stability in vitro (mouse hepa-
tic microsomes: Cli < 1 mL minÀ1 gÀ1 liver). This
compound was selected as a primary candidate for the
development of a further novel series of potent PARP-
1 inhibitors, which will be reported in detail in due
course.
R
R
H
IC50 (nM)
770
5
6
O
*
*
*
*
36
O
O
7
8
9
90
20
90
N
H
O
O
N
H
N
H
O
O
*
*
10
11
370
55
N
H
N
H
*
N
N
12
13
14
15
16
10
47
27
56
13
HN
O
O
O
N
N
*
N
HN
*
HN
O
S
N
S
*
HN
*
An alternative approach to stabilising the anilide moiety
in 8 was to constrain the amide into a ring, as in lactam
22. This resulted in a moderate increase in metabolic sta-
bility (mouse hepatic microsomes: Cli 4.9 mL minÀ1 gÀ1
liver), but a reduction in potency. The introduction of a
second carbonyl group into the lactam ring resulted in
imide 23 (Table 2), which exhibited a 10-fold increase
S
O
O
F
F
N
H
F

V. M. Loh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2235–2238
Table 2. Structure–activity relationships of the imide series
2237
Table 1 (continued)
R
IC50 (nM)
O
O
NH
N
*
17
18
19
116
N
H
O
R
X
O
540
290
N
H
N
H
R
X
H
IC50 (nM)
12
PF50
O
*
*
S
O
N
23
1.74
H O
N
N
O
O
O
*
O
N
20
21
189
50
N
24
25
H
H
13
0.94
O
*
N
H
*
*
O
180
—
N
N
O
N
*
O
22
120
O
26
27
28
Cl
19
33
5.0
1.5
O
O
O
*
O
in PARP-1 inhibitory potency. In addition, in vitro
stability studies showed that 23 (human hepatic micro-
somes: Cli < 0.05 mL minÀ1 gÀ1 liver) has significantly
increased metabolic stability over 8. However, the initial
imide lead 23 exhibited only moderate activity in a
cell-based assay (PF50 1.74, Table 2).16 Therefore,
further optimisation of 23 was required to improve
cellular efficacy whilst maintaining in vitro potency
and stability.
N
N
OMe
1.5
*
O
F
5.62
*
O
N
N
29
30
31
F
F
F
3.8
9.8
6.8
18.2
O
O
Introduction of a double bond into the imide ring re-
sulted in compound 24, which retained potency against
the isolated enzyme but abolished activity in the cell-
based assay, as indicated by the low PF50 value. Fusion
of 23 with a benzene ring (25) reduced PARP-1 inhibi-
tory activity overall. Substitution ortho to the imide
moiety with either a chloro or a methoxy group (26
and 27, respectively), did not increase PARP-1 inhibi-
tory potency. However, a significant enhancement of
potency in both enzyme and cell-based assays was ob-
served when a fluoro-substituent was introduced ortho
to the imide ring to give 28. Indeed, compound 28 re-
tains activity in the cellular assay at concentrations be-
low 10 nM (data not shown). This compound also
retained good metabolic stability in vitro (mouse hepatic
microsomes: Cli < 1 mL minÀ1 gÀ1 liver). Testing of
compound 28 against other PARP family members indi-
cated a similar level of potency for PARP-2, while for
Tankyrase and VaultPARP potency was >100-fold less.
Solubility for this compound was limited (0.25
mg mLÀ1) but in a mouse pharmacokinetic study good
bioavailability (50%) was obtained (data not shown).
The enhancement of potency via the fluoro substituent
appeared quite general for this series as exemplified by
compounds 29–34. Furthermore, addition of a methyl
*
O
2.9
5.4
*
O
N
N
O
O
*
O
32
33
F
F
5.0
4.1
14.4
3.6
*
O
N
O
O
*
N
34
F
9.5
3.1
N
*
O

2238
V. M. Loh et al. / Bioorg. Med. Chem. Lett. 15 (2005) 2235–2238
O
O
O
NO2
NO2
X
i
H
O
X
+
+
iii
O
2a: X = H
O
1
5
3a,b
NH
N
2b: X = OMe
O
O
O
iii
NO2
ii
H
NH2
X
O
O
Y
X
OMe
OMe
P
4a-d
O
X
2c: X = Cl
2d: X = F
O
6c,d: Y = NO2
7c,d: Y = NH2
iv
O
NH
N
NH
N
v
R1
vi
O
R1
O
H
N
R2
N
X
OH
O
R2
O
X
8a-d
Compounds 23-34 (Table 2)
Scheme 1. Synthesis route for phthalazinone imides. Reagents: (i) NaOMe, MeOH, reflux; (ii) NEt3, THF; (iii) NH2NH2ÆH2O, reflux; (iv) Fe powder,
NH4Cl; (v) Cyclic anhydride, MeCN, reflux; (vi) HBTU, DMF.
group a to the carbonyl of the imide 28 to give 29 led to
a further increase in cellular activity (PF50 18.2). This
could be due to enhanced cell penetration as a result
of increased lipophilicity. However, geminal-dialkyla-
tion a to the carbonyl to give 30 resulted in a slight de-
crease in potency in both isolated enzyme and cell-based
assays. Larger aromatic substituents a to the carbonyl of
imide 28 were tolerated, as exemplified by 31, which re-
tained good PARP-1 inhibitory potency. A high potency
PARP-1 inhibitor, 32, was also obtained by fusion of a
cyclopropane ring. Ring expansion of the imide 28 to a
piperidinedione such as 33, or inclusion of a further het-
eroatom to give a piperazinedione 34 also retained good
activity on the free enzyme and in cells.
References and notes
1. DꢀAmours, D.; Desnoyers, S.; DꢀSilva, I.; Poirier, G. G.
Biochem. J. 1999, 342, 249–268.
2. Ogata, N.; Ueda, K.; Kawaichi, M.; Hayaishi, O. J. Biol.
Chem. 1981, 256, 4135–4137.
3. Lindahl, T.; Satoh, M. S.; Poirier, G. G.; Klungland, A.
TIBS 1995, 20, 405–411.
4. Satoh, M.; Lindahl, T. Nature 1992, 356, 356–358.
5. Schlicker, A.; Peschke, P.; Burkle, A.; Hahn, E. W.; Kim,
J. H. Int. J. Radiat. Biol. 1999, 75, 91–100.
6. Durkacz, B. W.; Omidiji, O.; Gray, D. A.; Shall, S. Nature
1980, 283, 593–596.
7. Szabo, S.; Zingarelli, B.; OꢀConnor, M.; Salzman, A. L. J.
Clin. Invest. 1997, 100, 723–735.
8. Cantoni, O.; Cattbeni, F.; Stocchi, V.; Meyn, R. E.;
Cerrutti, P.; Murray, D. Biochem. Biophys. Acta 1989,
1014, 1–7.
9. Purnell, M. R.; Whish, W. J. D. Biochem. J. 1980, 185, 775.
10. Suto, M. J.; Turner, W. R.; Arundel-Suto, C. M.; Werbel,
L. M.; Sebolt-Leopold, J. S. Anti-Cancer Drug Des. 1991,
6, 107–117.
11. Griffin, 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 Des. 1995, 10, 507–514.
12. Dillon, K. J.; Smith, G. C. M.; Martin, N. M. B. J.
Biomol. Screen. 2003, 8, 347–352.
13. Banasik, M.; Komura, H.; Shimoyama, M.; Ueda, K. J.
Biol. Chem. 1992, 267, 1569–1575.
14. Douglas, D. G.; Eversley, P. J.; Martin, N. M. B.;
Newton, R. F.; Smith, G. C. M.; Vile, J.; White, C. R. Int.
Patent WO 02/36576, 2002.
The synthetic route for the phthalazinone imide ana-
logues is outlined in Scheme 1. For the non-haloge-
nated analogues (X = H and OMe) simple
nucleophilic addition of phthalide 1 to the aldehydes
2a,b in the presence of sodium methoxide, followed
by simultaneous cyclisation/nitro reduction of the
indanediones 3a,b with hydrazine hydrate gave the
key amine intermediates 4a,b. For the halogenated ana-
logues (X = Cl and F), in which the halogen ortho to
the nitro group is labile to nucleophilic displacement,
the alternative, milder methodology starting from the
phthalide phosphonate 5 was required. Formation of
the cyclic imides (23–34) was problematic under stan-
dard conditions. However, a two-step process involving
reaction with a cyclic anhydride to give 8 and subse-
quent cyclisation with a peptide-coupling reagent
proved generally applicable.
15. Ruf, A.; de Murcia, G. M.; Schulz, G. E. Biochemistry
1998, 37, 3893–3900.
16. The PF50 value is the potentiation factor, calculated as the
ratio of the IC50 growth curve for the alkylating agent,
methyl methane-sulfonate (MMS) divided by the IC50 of
the curve of MMS + PARP inhibitor. The cells used were
HeLa B and the test compounds were used at a fixed
concentration of 200 nM.
In conclusion, the synthesis and preliminary biological
evaluation of a novel class of phthalazinones as potent
PARP-1 inhibitors have been described. Further evalua-
tion of these compounds is ongoing and results will be
reported in due course.