Discovery and SAR of novel, potent and selective hexahydrobenzonaphthyridinone inhibitors of poly(ADP-ribose)polymerase-1 (PARP-1)

Article abstract of DOI:10.1016/j.bmcl.2009.12.002

A novel hexahydrobenzonaphthyridinone PARP-1 pharmacophore is reported, subsequent SAR exploration around this scaffold led to selective PARP-1 inhibitors with low nanomolar enzyme potency, displaying good cellular activity and promising rat PK properties.

Full text of DOI:10.1016/j.bmcl.2009.12.002

Bioorganic & Medicinal Chemistry Letters 20 (2010) 448–452
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and SAR of novel, potent and selective hexahydrobenzonaphthyridi-
none inhibitors of poly(ADP-ribose)polymerase-1 (PARP-1)
*
Caterina Torrisi , Monica Bisbocci, Raffaele Ingenito, Jesus M. Ontoria, Michael Rowley,
Carsten Schultz-Fademrecht, Carlo Toniatti, Philip Jones
IRBM—Merck Research Laboratories Rome, Via Pontina km 30,600, Pomezia, 00040 Rome, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel hexahydrobenzonaphthyridinone PARP-1 pharmacophore is reported, subsequent SAR exploration
around this scaffold led to selective PARP-1 inhibitors with low nanomolar enzyme potency, displaying
good cellular activity and promising rat PK properties.
Received 9 November 2009
Revised 1 December 2009
Accepted 1 December 2009
Available online 4 December 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
PARP
Cancer
Hexahydrobenzonaphthyridinone
Poly(ADP-ribose) polymerases (PARPs) constitute a super family
of 18 nuclear and cytoplasmic enzymes which use NAD+ as the
substrate to catalyze the formation of long and branched ADP-ri-
bose polymers on a number of proteins, including: itself, histones,
topoisomerases and p53. This structural modification alters the
function of the targeted proteins, and poly(ADP-ribosyl)ation
(PARylation) is implicated in the regulation of several fundamental
biological processes, including DNA repair, cell death and genomic
stability.1
PARP-1, the founding member of PARP family, accounts for
about 95% of poly(ADP-ribosyl)ation activity in cells and acts as a
DNA damage sensor and signaling protein.2 Through the DNA-
binding domain, PARP-1 can detect and bind to DNA single strand
breaks. Upon binding to DNA nicks, PARP-1 catalytic activity is
promptly stimulated, leading to the formation of poly(ADP-ribose)
(PAR) polymers. These PAR chains act as a beacon and recruit the
enzymes required for DNA repair to the site of damage, and the le-
sion is thus repaired.3 Similarly, PARP-2 elicits DNA-damage-
dependent catalytic activity, being thus responsible for the residual
PARylation activity observed in PARP-1 deficient mouse fibro-
blasts.4 Since DNA repair represents a major mechanism of tumor
cell resistance towards the common DNA-damaging cancer thera-
pies, PARP inhibitors (PARPi) are useful as chemosensitizers for use
in combination with radiotherapy and/or cytotoxic agents (e.g.,
temozolomide, topotecan and cisplatin).5 In addition, PARPi can
also be used as context–specific monotherapy of tumors with spe-
cific defects in the DNA repair. BRCA-1 and BRCA-2 are key proteins
involved in DNA double strand break repair by the homologous
recombination pathway, and mutations in these genes have been
associated with hereditary cancers, notably breast and ovarian. It
has been described that inhibition of PARP activity in BRCA-1 or
BRCA-2 deficient tumor cells becomes lethal, whereas PARP inhibi-
tion is generally non-toxic to normal cells.6 A number of inhibitors
are currently being explored in the clinic as both mono- and com-
bination therapies.7
The vast majority of PARPi known to date are substrate compet-
itive inhibitors and interact with the NAD+ binding domain of the
enzyme through three hydrogen bonds, with Ser904 and
Gly863.8 The byproduct of the PARylation reaction, nicotinamide,
was discovered to be a weak PARP inhibitor and structural ana-
logues of nicotinamide such as benzamide 1 and derivatives
O
O
O
O
NH2
NH
NH
NH
NH2
NH
NH2
PARP-1
IC50 (nM)
3300
410
540
140
1
2
3
4
* Corresponding author at present address: Addex Pharma SA, Chemin des Aulx
12, 1228 Plan-les-Ouates, Switzerland. Tel.: +41 228841555; fax: +41 228841556.
E-mail address: caterina.torrisi@addexpharma.com (C. Torrisi).
Figure 1. Known PARP-1 inhibitors (1–3) with published IC50 values,8 and novel
hexahydrobenzophthyridinone inhibitor 4.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.12.002

C. Torrisi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 448–452
449
Table 1
(Fig. 1) were among the first compounds to be investigated as PARP
Enzyme and cellular activities for compounds 4 and 7–15
inhibitors. However, these acyclic molecules like 1 had only weak
inhibitory activity. Restriction of the carboxamide moiety in the
bioactive anti conformation by intramolecular hydrogen bonding
or via a ring connection (e.g., incorporation into a lactam ring)
has become a common strategy for PARPi development leading
to much more potent compounds such as substituted dihydroiso-
quinoline 2 and fused tricyclic dihydroisoquinoline 3.
With the aim of finding a new class of potent PARP inhibitors
suitable for clinical development, we designed 1,2,3,8,9,9a-hexa-
hydro-7H-benzo[de]-1,7-naphthyridin-7-one 4 as a new pharma-
cophore capable of interacting with the NAD+ binding site. This
structure was designed bearing a functional handle to allow the
introduction of substituents pointing at the adenine–ribose bind-
ing pocket.
O
NH
N
R
a
a
Compds
R
hPARP-1 IC50 (nM)
140
PARylation EC50 (nM)
NDb
4
H
7
190
5200
840
230
350
O
8
4
Compound 4 was prepared using a known two-step procedure
(Scheme 1).9 Commercially available 1,2,3,4-tetrahydroisoquino-
line-1-acetic acid was subjected to acid-catalyzed cyclization to
give 5, which afforded the desired racemic tricycle 4 as the minor
product in 18% yield, together with the corresponding isomer 6 as
dominant product (61%), via a Schmidt rearrangement.
We were pleased to find that compound 4 exhibited good levels
of inhibition of the activity of both PARP-1 (IC50 = 140 nM) and
PARP-2 (IC50 = 60 nM) with good selectivity over Tankyrase (11%
9
8
O
O
10
4
H
N
11
11
1100
O
O
inhibition at 1
lM). The isomer 6 was not active on PARPs 1 and
2 at 1 M, confirming the importance of the amide bond orienta-
l
12
13
14
12
9
690
1400
130
N
H
tion for PARP activity. Based upon this promising finding a SAR
exploration was started to improve the potency of these com-
pounds. Alkyl, acyl, carbamate, sulfonamide and urea derivatives
7–49 were prepared in parallel from 4, applying standard or poly-
mer-supported solution phase synthesis conditions.10
Initial SAR focused ondetermining the optimalnitrogen substitu-
ent, both in terms of functional group and chain length. The com-
pounds were screened as racemates for both inhibition of human
PARP-1, and inhibition of PARylation in HeLa cells upon stimulation
of DNA damage by treatment with hydrogen peroxide (Table 1).11
While introduction of a benzyl group to give the tertiary amine 7
gave no improvement in activity with respect to 4, elimination of
the basic center using an amide linker (8) led to a boost in activity.
In fact, amide analogue 8 displayed inhibitory activity of PARP-1
enzyme in the low nanomolar range with more than 30-fold
improvement in potency over the lead 4. In terms of activity on
O
O
O
14
O
O
S
O
15
83
2700
a
Values are means of two or more experiments.
Not determined.
b
PARP-1, essentially little or no influence of the length and the nature
of the linker was observed (9–14). Urea and carbamate groups were
O
H
O
N
O
NH
i
ii
OH
+
NH
O
NH
NH
NH
6
4
5
iii
iv
v
vi
vii
O
N
O
N
O
O
N
O
N
NH
R
NH
NH
NH
NH
H
R
N
N
O
R
R
R
S
O
O
O
O
O
7
8-10
16-49
13-14
15
11-12
Scheme 1. General synthesis of hexahydrobenzophthyridinone compounds. Reagents and conditions: (i) PPA, 150 °C (85%); (ii) NaN3, H2SO4, 0 °C to rt (4, 18%; 6, 61%); (iii)
PhCHO, NaBH3CN, MeOH; (iv) RCOOH, HBTU, DIPEA, DMF or RCOOH, PS-DCC, DIPEA, DMF; (v) RNCO, PS-DIPEA, DCM; (vi) ROCOCl, PS-DIPEA, DCM; (vii) RSO2Cl, PS-DIPEA,
DCM.

450
C. Torrisi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 448–452
Table 2
Table 2 (continued)
PARP-1 binding assay and PARylation assay results for compounds 16–38
a
a
Compds
R
hPARP-1 IC50
(nM)
PARylation EC50
(nM)
O
NH
35
13
6
1300
330
S
N
R
N
N
O
36
a
a
Compds
R
hPARP-1 IC50
(nM)
PARylation EC50
(nM)
N
37
38
1
1600
420
N
16
17
7
260
NDb
Cl
Cl
N
N
14
37
a
Values are means of two or more experiments.
Not determined.
Cl
b
18
35
NDb
also well tolerated, whereas the sulfonamide 15 was detrimental.
However, despite the similarities in enzymatic activity, the amides
were generally more active than the corresponding carbamates
and ureas in cells, the former typically displaying sub-micromolar
inhibition of PARylation. Based on this result, the SAR on this new
series of PARPi was extensively explored further focusing on the syn-
thesis of diverse libraries of aromatic, heteroaromatic and aliphatic
amides.
19
20
13
310
NDb
CH2NEt2
O
130
N
N
NH
NH
21
50
NDb
Initially, substituted benzamides were explored (Table 2), and
the effect of a simple chlorine substituent was investigated. Meta
and ortho substitution (17–18) were detrimental for activity, while
introduction of a chlorine in the para position (16) gave a single di-
git nanomolar PARP inhibitor similar to 8, but with a three-fold
boost in the cellular activity (EC50 = 260 nM). Also a basic amine
in the para position (19) gave improved cellular activity despite
slightly weaker enzyme activity. Larger groups such as homopiper-
azine amides 20–21 resulted in a loss of potency. Homologation of
8 into the corresponding phenylacetamide 9 maintained enzyme
activity but further improved cellular activity, with 9 displaying
EC50 = 230 nM. However, preliminary DMPK studies indicated that
O
22
23
5
5
650
Cl
Cl
1100
Cl
24
16
900
25
26
6
5
160
OMe
OMe
9 was highly turned over in rat liver microsomes (Clint = 240 ll/
min/mg P), thus further modifications to these phenylacetamides
were aimed at stabilizing the compound toward hepatic metabo-
lism, as well as at improving activity. Again, the para position
was confirmed to be overall the preferred site for the introduction
of substituents (22–27), especially for cellular activity, although
the regioselectivity was more modest. A variety of substituents
were tolerated in position para, including methoxy (25), bromine
(28) and cyano groups (30). All three showed modestly improved
activity in cells compared to 9. However, a streric limitation was
determined as the para-benzyloxy group (29) resulted in a three-
fold loss in enzyme activity compared to 25.
Related heterocyclic replacements were also investigated. Intro-
duction of a heteroatom was better tolerated in the 3-position
(31–33), with 33 being a very potent enzyme inhibitor with good
cellular activity. The 3-pyridyl group also gave excellent metabolic
stability in vitro (Rat Clint = 3 ll/min/mg P). Other heterocycles
(34–35) displayed inferior cellular activity.
1200
OMe
27
27
1700
28
29
7
200
580
Br
16
OCH2Ph
CN
30
31
9
200
230
N
13
In the phenylpropamide series, replacement of the phenyl ring
of 10 with pyrazine (36) did not affect significantly activity com-
pared to 9. However, 36 showed much better rat liver microsome
32
33
10
1
4600
360
N
N
stability (Clint = 24 ll/min/mg P). Benzimidazole 37 was a very po-
tent PARP inhibitor (IC50 = 1 nM) but in contrast had only micro-
molar cellular activity. On the contrary pyrazole 38, despite a
decreased potency on the enzyme, showed an acceptable level of
cellular potency.
N
34
6
2900
S

C. Torrisi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 448–452
451
In parallel to the (hetero)aromatic derivatives, a library of ali-
Table 4
Activity against various PARP isoforms for compounds 33 and 36a
phatic amides was evaluated (Table 3). The simple cyclopentyl
derivative 39 displayed PARP-1 IC50 = 11 nM and EC50 = 550 nM
in cells. To try to pick up additional hydrogen-bond interactions,
Compds PARP-1
IC50 (nM)
PARP-2
PARP-3
v-PARP
TNKS1 IC50
IC50 (nM)
IC50 (nM)
IC50 (nM)
(nM)
an amine moiety was incorporated resulting in the L-prolyl deriva-
33
36
1
6
1
1
50
70
440
660
3500
2000
tive 40 which improved both enzyme and cell activity about three-
fold. The nitrogen needs to be basic to maintain activity as demon-
strated by the acetyl analogue 41 which displayed only weak cel-
a
Values are means of two or more experiments.
lular activity (EC50 = 2.8 lM). Moving the nitrogen to give the
pyrrolidin-3-yl derivative 42 and reducing the size of the ring to
give azetidine 43 were both changes detrimental to cellular activ-
ity. Other cyclic amines such as (piperidin-1-yl) acetyl 44 were also
active cell permeable PARPi, with 44 displaying IC50 = 6 nM and
EC50 = 280 nM, and a remarkable improvement in microsome sta-
Table 5
Rat PK summary for compounds 33 and 36a
Compds
t1/2 (h)
Vd (L/kg)
CL (ml/min/kg)
%F
po AUC (lM h)
33
36
1.7
1.2
3
2
38
31
51
2.3
NDb
NDb
bility (Clint = 18 ll/min/mg P). Phenyl piperidines have been
claimed as potent PARPi,12 and while N-methyl piperazine 45
was only weakly active, possibly due to the presence of the second
amine group, the introduction of a phenyl substituent in the 4-po-
sition was confirmed to give a 10-fold boost in activity on the en-
zyme in both the piperazine and corresponding piperidine
compounds (46–47), with 47 being a sub-nanomolar PARP inhibi-
tor. Other functionalities like an amide (48) or a carboxylic acid
(49) were evaluated but despite giving potent enzyme inhibitors,
these modifications were not tolerated in cells.
a
Compounds were administered iv and po at 3 mg/kg.
Not determined.
b
Representative compounds such as 33 and 36, which were iden-
tified as potent PARP-1 inhibitors endowed with sub-micromolar
cellular activity and good metabolic stability, were screened
against a panel of other PARP isoforms (Table 4). Both compounds
resulted in very potent inhibitors of PARPs 1 and 2, with good
selectivity towards PARP-3, and excellent selectivity over vPARP
and Tankyrase-1.
Given the interest in 33 and 36, these compounds were evalu-
ated in rat PK studies and results are summarized in Table 5. Both
compounds displayed acceptable clearance in vivo, with Cl = 38
and 31 mL/min/kg respectively, which reflects the good microsome
Table 3
Enzyme and cellular activities for compounds 39–49
O
NH
stability (Clint = 3, and 24 ll/min/mg P, respectively). The pyridyl
derivative 33 also showed good oral bioavailability (F = 51%). Fur-
N
R
thermore, 33 and 36 did not show any CYP inhibition or binding
O
to IKr at 10 lM.
a
a
Compds
R
hPARP-1 IC50 (nM) PARylation EC50 (nM)
In conclusion, we have described the discovery, SAR and preli-
minary biological evaluation of a novel series of hexahydrobenzo-
naphthyridinone carboxamides. These compounds represent a
novel class of potent PARP inhibitors, capable of inhibiting the
PARylation reaction in whole cells at sub-micromolar concentra-
tions. Furthermore, selective PARP-1 and -2 inhibitors like 33 and
36 having good metabolic stability and promising rat PK profiles
were identified.
39
11
4
550
180
40b
41b
N
H
N
14
2800
O
42
43
1
1
3200
660
Acknowledgments
NH
NH
We are grateful to Massimiliano Fonsi and Maria Vittoria Orsale
for DMPK support and to Fabrizio Fiore for PK studies.
N
N
44
45
6
280
References and notes
97
3500
N
Me
1. (a) Nguewa, P. A.; Fuertes, M. A.; Valladares, B.; Alonso, C.; Perez, J. M. Prog.
Biophys. Mol. Biol 2005, 88, 143; (b) Amé, J. C.; Spenlehauer, C.; de Murcia, G.
BioEssays 2004, 26, 882.
2. Szabó, C.; Jagtap, P. Nat. Rev. Drug Disc. 2005, 4, 421.
3. Schreiber, V.; Dantzer, F.; Ame, J. C.; de Murcia, G. Nat. Rev. Mol. Cell. Biol. 2006,
7, 517.
4. (a) Eliasson, M. J.; Sampei, K.; Mandir, A. S.; Hurn, P. D.; Traystman, R. J.; Bao, J.;
Pieper, A.; Wang, Z. Q.; Dawson, T. M.; Snyder, S. H.; Dawson, V. L. Nat. Med.
1997, 3, 1089; (b) Shieh, W. M.; Ame, J. C.; Wilson, M. V.; Wang, Z. Q.; Koh, D.
W.; Jacobson, E. L. J. Biol. Chem. 1998, 273, 30069.
N
N
N
46
47
11
440
410
0.5
5. Tentori, L.; Graziani, G. Pharmacol. Res. 2005, 52, 25.
O
6. (a) Bryant, H. E.; Schultz, N.; Thomas, H. D.; Parker, K. M.; Flower, D.; Lopez, E.;
Kyle, S.; Meuth, M.; Curtin, N. J.; Helleday, T. Nature 2005, 434, 913; (b) Farmer,
H.; McCabe, N.; Lord, C. J.; Tutt, A. N. J.; Johnson, D. A.; Richardson, T. B.;
Santarosa, M.; Dillon, K. J.; Hickson, I.; Knights, C.; Martin, N. M. B.; Jackson, S.
P.; Smith, G. C. M.; Ashworth, A. Nature 2005, 434, 917.
7. Rodon, J.; Iniesta, M. D.; Papadopoulos, K. Expert Opin. Investig. Drugs 2009, 18,
31.
8. Peukert, S.; Schwahn, U. Exp.Opin. Ther. Patents 2004, 14, 1531.
9. Pelletier, J. C.; Cava, M. P. J. Org. Chem. 1987, 52, 616.
48
49
9
2800
5000
N
O
15
OH
a
Values are means of two or more experiments.
Mixture of two diastereoisomers.
b

452
C. Torrisi et al. / Bioorg. Med. Chem. Lett. 20 (2010) 448–452
10. Ingenito, R.; Jones, P.; Ontoria, J. M., Schultz-Fademrecht, C.; Torrisi, C. PCT
Patent Application WO2008/001134 A1, 2008. This patent describes the
general procedures for the synthesis of alkyl, acyl, carbamate, sulfonamide
and urea derivatives from compound 4. Characterization data for
representative compounds of each class are given.
11. (a) Cheung, A.; Zhang, J. Anal. Biochem. 2000, 282, 24; (b) Alvarez-Gonzalez, R.;
Spring, H.; Muller, M.; Burkle, A. J. Biol. Chem. 1999, 274, 32122.
12. Ishida, J.; Hattori, K.; Yamamoto, H.; Iwashita, A.; Mihara, K.; Matsuoka, N.
Bioorg. Med. Chem. Lett. 2005, 15, 4221.