A novel series of benzimidazole NR2B-selective NMDA receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2012.01.108

A series of novel benzimidazoles are discussed as NR2B-selective N-methyl-d-aspartate (NMDA) receptor antagonists. High throughput screening (HTS) efforts identified a number of potent and selective NR2B antagonists such as 1. Exploration of the substituents around the core of this template identified a number of compounds with high potency for NR2B (pIC₅₀ >7) and good selectivity against the NR2A subunit (pIC₅₀ <4.3) as defined by FLIPR-Ca²⁺ and radioligand binding studies. These agents offer potential for the development of therapeutics for a range of nervous system disorders including chronic pain, neurodegeneration, migraine and major depression.

Full text of DOI:10.1016/j.bmcl.2012.01.108

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2620–2623
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A novel series of benzimidazole NR2B-selective NMDA receptor antagonists
David J. Davies a, Matthew Crowe a, Nolwenn Lucas a, Joanna Quinn a, David D. Miller a, Sara Pritchard b,
David Grose a, Ezio Bettini c, Novella Calcinaghi c, Caterina Virginio c, Lee Abberley b, Paul Goldsmith b,
Anton D. Michel b, Iain P. Chessell b, James N.C. Kew b, Neil D. Miller b, Martin J. Gunthorpe b,
⇑
a GlaxoSmithKline R&D, Medicines Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
b GlaxoSmithKline R&D, New Frontiers Science Park, Third Avenue, Harlow, Essex CM19 5AW, UK
c GlaxoSmithKline, Via Fleming, Verona, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel benzimidazoles are discussed as NR2B-selective N-methyl-
D
-aspartate (NMDA) receptor
Received 30 September 2011
Revised 25 January 2012
Accepted 27 January 2012
Available online 6 February 2012
antagonists. High throughput screening (HTS) efforts identified a number of potent and selective NR2B
antagonists such as 1. Exploration of the substituents around the core of this template identified a num-
ber of compounds with high potency for NR2B (pIC50 >7) and good selectivity against the NR2A subunit
(pIC50 <4.3) as defined by FLIPR-Ca2+ and radioligand binding studies. These agents offer potential for the
development of therapeutics for a range of nervous system disorders including chronic pain, neurodegen-
eration, migraine and major depression.
Keywords:
NMDA
NR2B
Ó 2012 Elsevier Ltd. All rights reserved.
Glutamate
Ion channel
Depression
Pain
Drug discovery
N-methyl-
D
-aspartate (NMDA) receptors are a molecularly, and
excitotoxicity which has been linked with several CNS disorders
and pathologies.10–12
pharmacologically, distinct sub-family of glutamate-gated ion
channels which are widely distributed throughout the mammalian
central nervous system (CNS) where they play key roles in excit-
atory neurotransmission.1–3 They are assembled from heteromeric
combinations of subunits encoded by distinct NR1, NR2 and NR3
genes. The most common receptor forms in the adult CNS are
thought to be heterotetramers of two NR1 and two NR2A or
NR2B subunits with alternative (and mixed) combinations with
NR2C and NR2D contributing to a diversity of receptors with differ-
ing biophysical and pharmacological properties and specific regio-
nal distributions.3,4 The more recently identified NR3A and 3B
subunits also contribute to NMDA receptor heterogeneity via for-
mation of ternary complexes, however, the functional roles of
these channels are less clear at present.5,6
NMDA receptor antagonists have, for some time, therefore, been
considered as potential therapeutic agents. Non subtype-selective
(or pan-) NMDA receptor antagonists such as memantine and
MK-801 which act at the pore of the open channel were initially
developed for the treatment of stroke, head trauma and neurode-
generation (Fig. 1).13,14
Others such as the anaesthetic agent ketamine have been used
off-label to treat neuropathic pain and more recently in promising
exploratory studies in major depression.7,15 The therapeutic utility
of such agents has however, been severely limited by their CNS
side-effects that include cognitive impairment, psychotomimetic
effects, hallucinations and even catatonia at higher doses.8,16
Due to their synaptic and extrasynaptic locations, voltage-
dependent block by Mg2+ and high Ca2+ permeability, NMDA
receptors play key roles in a range of mechanisms that underpin
physiological processes associated with synaptic plasticity, learn-
ing, and memory.7–9 The ability to elicit high levels of intracellular
Ca2+ is also an established cause of the NMDA receptor-induced
Cl
NH2
O
HN
HN
Memantine
Ketamine
MK-801
⇑
Corresponding author. Tel.: +44 2032 877763.
E-mail address: Martin.Gunthorpe@gmail.com (M.J. Gunthorpe).
Figure 1. Non subtype-selective NMDA antagonists.
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.01.108

D. J. Davies et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2620–2623
2621
HO
OH
OH
OH
CH3
N
N
N
CH3
CH3
OH
OH
OH
Pfizer CP-101,606
Roche Ro 25,6981
Ifenprodil
O
H
F
N
O
O
N
H
N
O
N
F
N
N
H
O
O
N
Merck MK-0657
Gedeon Richter RGH-896
Figure 2. NR2B-selective antagonists.
characterized, such NR2B-selective antagonists have shown good
activity and side effect profiles in animal models and, encourag-
ingly, in the case of the CP-101,606, have also demonstrated efficacy
in clinical studies in patients with chronic pain26 and treatment-
resistant major depressive disorder.27
N
HO
N
In this Letter we disclose the discovery of a novel series of benz-
imidazoles as NR2B-selective NMDA antagonists.
High throughput screening identified a benzimidazole (1) as a
potent inhibitor of NR2B with a sub-micromolar IC50 in a plate-
based Ca2+ flux assay (Fig. 3).
N
H
Figure 3. Compound 1.
Biological activity was measured in two assay formats. Firstly, a
FLIPR (FLuorescent Imaging Plate Reader) based-Ca2+ assay, based
on a transient transfection protocol with human NR1 and NR2B sub-
units, was used for primary hit identification. Secondly, the SAR re-
ported herein was generated using a recombinant binding assay
employing [3H]-Ro 25,6981 as a specific ifenprodil-site radioligand.
More detailed mechanistic studies were performed using whole-
cell voltage-clamp electrophysiology using transiently transfected
NR1-NR2B HEK-293T cells where NMDA receptor activity was
MeO
MeO
NH2
NH2
N
RCHO
R
a
N
H
R1R2N
HO
N
HO
HNR1R2
N
HBr
R
R
b
c
N
H
evoked in response to 100 lM glutamate and 10 lM glycine.
N
H
The discovery of compound 1 prompted the synthesis of a series
of benzimidazoles which were prepared according to the synthetic
routes outlined in reaction Schemes 1 and 2. The starting materials
were generally commercially available diamines, but where neces-
sary the diamine was prepared by palladium catalysed reduction of
the corresponding nitro aniline. In Scheme 1, the diamines were
condensed with aldehydes in aqueous DMF in the presence of ‘Ox-
one’. The reaction was often accompanied by a significant degree of
di-imine formation but the desired product could be isolated by an
acid/base work up procedure. The methoxy benzimidazoles were
subsequently demethylated under acidic conditions at elevated
temperature to give the corresponding phenols. The phenols were
Scheme 1. (a) Oxone, rt, DMF:H2O (97:3) 1–18 h (b) 48% HBr 80 °C, 1–18 h (c)
CH2O, AcOH, R1R2NH, THF, 70 °C, 5 h.
Following on from the discovery of ifenprodil,17 the first NR2B-
selective NMDA antagonist, the potential for a marked improve-
ment in the therapeutic index achievable with respect to the
dose-range required for efficacy versus doses resulting in typical
NMDA related side effects has been demonstrated.18,19 Thus, a range
of chemotypes which potently and selectively antagonise NR2B-
containing NMDA receptors is now available (Fig. 2).20–25 Where
O
R
R
O
MeO
NH2
NO2
MeO
NH
MeO
NH
RCOCl
H2
b
a
NO2
NH2
R1R2N
HO
MeO
HO
N
N
H+
N
HBr
HNR1R2
R
R
R
N
d
c
e
N
H
N
H
H
Scheme 2. (a) TEA, CH2Cl2, 2 h (b) Pd(0), EtOH, 1–18 h (c) H2SO4, isoamyl alcohol, 18 h (d) 48% HBr 80 °C, 1–18 h (e) CH2O, AcOH, THF, 70 °C, 5 h.

2622
D. J. Davies et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2620–2623
Table 1
Table 2
[3H]-Ro 25,6981 binding assay results for compounds 1–15
[3H]-Ro 25,6981 binding assay results for compounds 16–24
R1
R1
R2
HO
N
N
R
N
H
N
H
R1
NR2B pIC50
a
a
Compd
R2
R1
NR2B pIC50
Compd
R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
Piperidinylmethyl
Hexahydro-azepinylmethyl
Pyrrolidinylmethyl
Cyclopentylaminomethyl
(Propylamino) methyl
4-Morpholinomethyl
(4-Phenyl-1-piperidinyl)methyl 7.1
Dimethylaminomethyl 6.6
(4-Methyl-1-piperazinyl)methyl 6.4
7.9
8.1
8.1
8.1
7.5
7.4
7.4
16
17
18
19
20
21
22
23
24
H
H
H
F
F
F
F
F
F
Piperidinylmethyl
Methyl
1-Piperidinylcarbonyl
Pyrrolidinylmethyl
Piperidinylmethyl
Hexahydro-azepinylmethyl
Methyl(1-methylethyl)amine
(2-Ethyl-piperidinyl)methyl
4-Morpholinomethyl
6.4
<5
<5
7.1
7.1
6.7
6.1
6.0
5.6
2,3 Difluorophenyl Pyrrolidinylmethyl
2,3-Difluorophenyl (2-Thienylmethyl)aminomethyl 6.2
2,3-Difluorophenyl (2-Furanylmethyl)aminomethyl 6.1
4-Methyl phenyl
Benzyl
H
a
Values are means of 2–4 experiments, standard deviation is <0.2.
Cyclopropylaminomethyl
Pyrrolidinylmethyl
Piperidinylmethyl
7.4
5.8
<5
a
Values are means of 2–4 experiments, standard deviation <0.2.
NH2
NH2
N
PhCHO
N
H
a
NR1R2
CHO
Figure 4. Whole-cell patch clamp recordings define use-dependent inhibition of
human NR1-NR2B receptors by 1. The dotted lines highlight activity-dependent
inhibition achieved by the continuous application of 1, above the otherwise stable
CAN
R1R2NH
N
N
baseline response to the application of agonists (100
applied for 3 s at 1 min intervals).
lM glutamate, 10 lM glycine;
b
N
H
c
N
H
Scheme 3. (a) Oxone, rt, DMF/H2O (97:3) 2 h (b) CAN 2–10 days (c) R1R2NH, NaBH4,
rt, 18 h.
Table 1 shows data for a selection of compounds from this
series.
The R1 position proved tolerant to a range of substituents
although small alkyl groups or rings such as piperidine or morpho-
line were preferred. The R position proved less tolerant with phe-
nyl derivatives being strongly preferred. Aliphatic groups in this
position were significantly less active.
Further chemistry effort was initiated, with the aim of expand-
ing the SAR and addressing the potential lack of CNS penetration
predicted by our in-house in-silico models. We were prompted
to focus on the synthesis of the des-phenol (Scheme 3) and fluoro
analogues (Scheme 4) since replacing the phenolic OH would lower
reacted with primary and secondary amines under typical Mannich
reaction conditions to give the required products after HPLC
purification.
In Scheme 2, 2-nitro-5-methoxy-aniline was acylated under
standard conditions and subsequent reduction of the nitro group
by catalytic hydrogenation yielded the aniline intermediate. Cycli-
sation was then carried out under acid conditions to provide the
required benzimidazoles. De-alkylation of the methoxy group
and subsequent Mannich reaction was carried out as described in
Scheme 1.
F
O
F
O
F
O
F
F
O
F
OMe
NH2
c
OMe
OMe
b
OMe
a
(1)
NH2
NO2
NH2
NO2
d
OH
N
COOMe
N
F
f
F
e
N
N
F
N
H
N
H
N
H
(3)
(2)
Scheme 4. (a) H2SO4/HNO3 2 h (b) NH4OH, 1 h (c) Pd(0), H2, 18 h (d) PhCHO, Oxone, rt, DMF/H2O (97:3), 1 h (e) DIBAL, THF, 1 h (f) MsCl, rt, 5 h, piperidine, 18 h.

D. J. Davies et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2620–2623
2623
Table 3
[3H]-Ro 25,6981 binding and in vitro developability data for compounds 1, 16 and 19
Compds
NR2B pIC50
8.1
CYP450 inhibition IC50
(
l
M)a
CLi mL/min/g (h = human, r = rat)
1
2D6 = 6, 1A2 = 3
others >30
3.9 (h)
24 (r)
16
19
6.4
7.1
2D6 = 5, 1A2 = 7
others >50
2D6 = 5, 1A2 = 16
others >30
3.6 (h)
9 (r)
<0.5 (h)
<0.5 (r)
a
CYP enzymes profiled were: 1A2, 2C9, 2C19, 2D6 and 3A4.
the polar surface area and reduce the number of hydrogen bond
donors, and therefore likely have a beneficial effect on the degree
of CNS penetration.
molecular hydrogen bonding between the phenol and the basic
side chain, which was not taken into account by the in-silico
model.
4-Methyl-2-phenyl benzimidazole was prepared as described in
Scheme 3 and oxidized using ceric ammonium nitrate-(CAN). We
found this reaction to be capricious and poor yielding but with
an excess of CAN and an extended reaction time of 10 days we
managed to obtain gram quantities of material, sufficient to pro-
ceed with the subsequent reductive amination.
In summary, high throughput screening identified a structurally
novel series of potent NMDA receptor NR2B subtype-selective
antagonists which showed good selectivity over the NR2A subtype.
Preliminary data indicated that exemplars showed promising dev-
elopability profiles and were CNS penetrant, thus supporting fur-
ther progression of the series.
The desired 5-fluoro analogues were prepared according to the
route outlined in Scheme 4. This involved preparation of the re-
quired diamine in two steps from a commercial di-fluoro ester.
Cyclisation with oxone gave the required benzimidazole and the
ester group was readily reduced by Dibal to yield the correspond-
ing alcohol. Reaction with methanesulphonyl chloride followed by
amine displacement installed the desired side chain.
Acknowledgement
We would like to thank the GSK Computational and Analytical
Science group for assistance with purification and for the genera-
tion of physicochemical data.
Table 2 shows data for exemplars from the fluoro and des-
phenol series.
References and notes
1. Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F. Pharmacol. Rev. 1999, 51, 7.
2. Kew, J. N.; Kemp, J. A. Psychopharmacol. (Berl) 2005, 179, 4.
3. Paoletti, P.; Neyton, J. Curr. Opin. Pharmacol. 2007, 7, 39.
4. Loftis, J. M.; Janowsky, A. Pharmacol. Ther. 2003, 97, 55.
5. Matsuda, K.; Kamiya, Y.; Matsuda, S.; Yuzaki, M. Brain Res. Mol. Brain Res. 2002,
100, 43.
The requirement for a basic centre in the R1 side-chain, is
illustrated with amide 18 which was devoid of activity. In addition,
removal of the benzimidazole NH in compound 24 through
N-methylation was not tolerated (data not shown). All attempts
to replace the benzimidazole core with alternative 5,6-bicyclic
heterocycles e.g. indole were unsuccessful.
Electrophysiological examination of exemplar compounds
showed them to be potent, reversible inhibitors. The compounds
also exhibited use-dependent activity consistent with previous re-
ports for ‘ifenprodil-site’ ligands,28 a feature which may underlie
their greater tolerability at efficacious doses (Fig. 4). The pIC50
determined at steady state for compound 1 was 7.0 in good agree-
ment with the binding data. Consistent with activity resulting from
binding at the ifenprodil site on NR2B containing NMDA receptors,
all compounds included here had a pIC50 <4.3 in an NR1-NR2A-
based FLIPR-Ca2+ assay. They also had high selectivity over the
hERG potassium channel (>100-fold) as determined in a difetolide
binding assay and, with the exception of compound 7, showed
good selectivity over a1 adrengeric receptors (<50% inhibition at
6. Yao, Y.; Mayer, M. L. J. Neurosci. 2006, 26, 4559.
7. Chizh, B. A. J. Psychopharmacol. 2007, 21, 259.
8. Kemp, J. A.; McKernan, R. M. Nat. Neurosci. 2002, 5, 1039.
9. Liu, L.; Wong, T. P.; Pozza, M. F.; Lingenhoehl, K.; Wang, Y.; Sheng, M.;
Auberson, Y. P.; Wang, Y. T. Science 2004, 304, 1021.
10. Obrenovitch, T. P.; Urenjak, J.; Zilkha, E.; Jay, T. M. Int. J. Dev. Neurosci. 2000, 18,
281.
11. Sinor, J. D.; Du, S.; Venneti, S.; Blitzblau, R. C.; Leszkiewicz, D. N.; Rosenberg, P.
A.; Aizenman, E. J. Neurosci. 2000, 20, 8831.
12. Wang, C. X.; Shuaib, A. Curr. Drug Targets CNS Neurol. Disord. 2005, 4, 143.
13. Ikonomidou, C.; Turski, L. Lancet Neurol. 2002, 1, 383.
14. Palmer, G. C. Curr. Drug Targets 2001, 2, 241.
15. Machado-Vieira, R.; Salvadore, G.; Diazgranados, N.; Zarate, C. A., Jr. Pharmacol.
Ther. 2009, 123, 143.
16. Chizh, B. A.; Headley, P. M. Curr. Pharm. Des. 2005, 11, 2977.
17. Williams, K. Curr. Drug Targets 2001, 2, 285.
18. Chazot, P. L. Curr. Med. Chem. 2004, 11, 389.
19. Chizh, B. A.; Headley, P. M.; Tzschentke, T. M. Trends Pharmacol. Sci. 2001, 22,
636.
20. Borza, I.; Kolok, S.; Gere, A.; Nagy, J.; Fodor, L.; Galgoczy, K.; Fetter, J.; Bertha, F.;
Agai, B.; Horvath, C.; Farkas, S.; Domany, G. Bioorg. Med. Chem. Lett. 2006, 16,
4638.
10
lM).
The in vitro developability profile of the series was encouraging
as illustrated by the fact that analogues in which the OH was re-
placed or deleted retained activity, albeit with a slight reduction
in potency. The drop in potency observed in compounds 16 and
19 was somewhat mitigated by improvements in their metabolic
stability (Intrinsic clearance, CLi; determined using liver micro-
somes) and CYP450 inhibition profiles (Table 3).
Preliminary in vivo pharmacokinetic evaluation of key
exemplars from the phenolic series in rat was also encouraging
e.g. compound 1 achieved peak blood concentrations of 0.6 lM fol-
21. Borza, I.; Domany, G. Curr. Top. Med. Chem. 2006, 6, 687.
22. Layton, M. E.; Kelly, M. J., III; Rodzinak, K. J. Curr. Top. Med. Chem. 2006, 6, 697.
23. Liverton, N. J.; Bednar, R. A.; Bednar, B.; Butcher, J. W.; Claiborne, C. F.;
Claremon, D. A.; Cunningham, M.; DiLella, A. G.; Gaul, S. L.; Libby, B. E.; Lyle, E.
A.; Lynch, J. J.; McCauley, J. A.; Mosser, S. D.; Nguyen, K. T.; Stump, G. L.; Sun, H.;
Wang, H.; Yergey, J.; Koblan, K. S. J. Med. Chem. 2007, 50, 807.
24. Mony, L.; Kew, J. N.; Gunthorpe, M. J.; Paoletti, P. Br. J. Pharmacol. 2009, 157,
1301.
25. Mosley, C. A.; Myers, S. J.; Murray, E. E.; Santangelo, R.; Tahirovic, Y. A.;
Kurtkaya, N.; Mullasseril, P.; Yuan, H.; Lyuboslavsky, P.; Le, P.; Wilson, L. J.;
Yepes, M.; Dingledine, R.; Traynelis, S. F.; Liotta, D. C. Bioorg. Med. Chem. 2009,
17, 6463.
lowing a subcutaneous 3 mg (free base)/kg target dose with an
estimated Tmax of 0.8 h and terminal half-life of 1.1 h. Further
study defined that, in contrast to the in-silico predictions, phenol
1 achieved significant CNS penetration with a brain to blood ratio
of 2.7:1. This may reflect masking of the acidic OH through intra-
26. Sang, C. N.; Weaver, J. J.; Jinga, L.; Wouden, J.; Saltarelli, M. D. Society for
Neuroscience, 2003, Meeting. Program #814.9.
27. Preskorn, S. H.; Baker, B.; Kolluri, S.; Menniti, F. S.; Krams, M.; Landen, J. W. J.
Clin. Psychopharmacol. 2008, 28, 631.
28. Kew, J. N.; Trube, G.; Kemp, J. A. J. Physiol. 1996, 497, 761.