Orally efficacious NR2B-selective NMDA receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(02)01061-2

A novel series of benzamidines was synthesized and shown to exhibit NR2B-subtype selective NMDA antagonist activity. Compound 31 is orally active in a carrageenan-induced rat hyperalgesia model of pain and shows no motor coordination side effects.

Full text of DOI:10.1016/S0960-894X(02)01061-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 697–700
Orally Efficacious NR2B-Selective NMDA Receptor Antagonists
Christopher F. Claiborne,a,* John A. McCauley,a,* Brian E. Libby,a Neil R. Curtis,b
b
Helen J. Diggle,b JanuszJ. Kulagowski, Stuart R. Michelson,a Kenneth D. Anderson,a
David A. Claremon,a Roger M. Freidinger,a Rodney A. Bednar,c Scott D. Mosser,c
Stanley L. Gaul,c Thomas M. Connolly,c Cindra L. Condra,c Bohumil Bednar,c
Gary L. Stump,d Joseph J. Lynch,d Alison Macaulay,e Keith A. Wafford,e
Kenneth S. Koblanc and Nigel J. Livertona
aDepartment of Medicinal Chemistry, Merck Research Laboratories, West Point, PA 19486, USA
bDepartment of Medicinal Chemistry, Merck Sharp & Dohme Research Laboratories, The Neuroscience Research Centre,
Terlings Park, Eastwick Road, Harlow, Essex CM20 2QR, UK
cDepartment of Molecular Pharmacology, Merck Research Laboratories, West Point, PA 19486, USA
dDepartment of Pharmacology, Merck Research Laboratories, West Point, PA 19486, USA
eDepartment of Pharmacology, Merck Sharp & Dohme Research Laboratories, The Neuroscience Research Centre,
Terlings Park, Eastwick Road, Harlow, Essex CM20 2QR, UK
Received 2 October 2002; accepted 8 November 2002
Abstract—A novel series of benzamidines was synthesized and shown to exhibit NR2B-subtype selective NMDA antagonist activity.
Compound 31 is orally active in a carrageenan-induced rat hyperalgesia model of pain and shows no motor coordination side effects.
# 2003 Elsevier Science Ltd. All rights reserved.
N-Methyl-d-aspartate (NMDA) receptor antagonists
have considerable therapeutic potential in the treatment
of several disease states including stroke, neuropathic
pain, and Parkinson’s disease.1 The clinical develop-
ment of NMDA antagonists however, has been encum-
bered by their small therapeutic index, since they have
unacceptable side effects including, but not limited to,
motor deficit, sedation, and psychotomimetic states.2,3
of NR2B message in the cerebellum would suggest that
NR2B-selective antagonists might not adversely affect
locomotor function.7 Ifenprodil (1, Fig. 1), a ligand
which binds selectively to the NR2B subtype, effectively
modulates ion flux8,9 and demonstrates a separation of
side effects from antinociceptive properties in animal
models.7 Thus, the ifenprodil binding site has proven an
attractive target to enhance the pharmacological profile
relative to non-selective NMDA antagonists.10,11
A
The discovery that the NMDA channel is a hetero-
oligomeric assembly composed of a minimum of two
different subunits, assigned NR1 and NR2, with at least
eight isoforms (NR1a–h) for the former and four dis-
tinct subtypes (NR2A–D) for the latter, afforded the
possibility of mapping the regional distributions of these
polypeptides in the brain.4,5 mRNA mapping studies
indicate that whereas the NR2A subunit is ubiquitously
expressed in brain, the NR2B subunit is largely confined
to structures in the forebrain including the cerebral
cortex, hippocampus, and olfactory bulb.6 The absence
variety of compounds have since shown NR2B subtype
selectivity, including CP-101,606 (2),12 CI-1041 (3)13
and aminoquinoline 4.14
Screening efforts at Merck for compounds binding at
the ifenprodil site resulted in the identification of styryl
amidine 5 (Fig. 2) which displayed significant affinity for
the NR2B receptor (Ki=9 nM vs [3H]-ifenprodil). The
potency of this compound could be enhanced by addi-
tion of a methoxy substituent on the benzyl ring (6,
Ki=0.7 nM versus [3H]-ifenprodil). The NR2B binding
data reported in this communication utilizes radio-
labeled amidine 6 (KD=1.0 nM) as the ligand rather
than ifenprodil (KD=94 nM) due to its superior signal-
to-noise ratio.15
*Corresponding author. Fax: +1-215-652-3971; e-mail: john_ mccauley@
merck.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(02)01061-2

698
C. F. Claiborne et al. / Bioorg. Med. Chem. Lett. 13 (2003) 697–700
Simply substituting the benzyl ring of compound 7 with
chloro at the 3-position (12) partially restored the
activity observed in the styryl series. Holding R2
constant as 3-Cl, it was clear that R1 substituents in the
4-position led to compounds with the greatest activity in
the NR2B binding assay (13–18). Furthermore, a vari-
ety of 4-R1 substitutions were tolerated (19–23). Reex-
amination of benzyl substituent effects in the presence
of a potency enhancing R1 group demonstrated that
both the 2- and 3-positions were accommodating of an
array of R2 substituents (24–30). Substitution of the
benzylic amine fragment at the 4-position (26) generally
led to diminished activity.
Figure 1. Structures of NR2B-selective NMDA receptor antagonists.
Using the above SAR data, we designed a number of
high affinity compounds (Table 2), including amidine 35
which shows a remarkable >25,000-fold improvement
over the parent benzamidine 7 in the NR2B binding
assay. Selected compounds were screened against the
closely related NR2A subtype in the FLIPR assay. The
results demonstrated no measurable NR2A activity
(IC50>10 mM) for compounds 31–35. Furthermore,
compound 31 was tested against the four known NR2B
subtypes in a whole cell patch-clamp assay and was
shown to be selective for NR2B (Fig. 3).
Figure 2. Amidine derived NR2B/NMDA receptor antagonists.
Recognizing the possibility that styryl amidine 5 may
act as a Michael acceptor, we sought to develop a non-
styryl series compounds,16 into a novel class of selective
NR2B receptor antagonists with the goal of demonstrating
oral activity in a rodent hyperalgesia model.
We proceeded to evaluate the pharmacokinetic proper-
ties and oral activity of selected compounds (results
summarized in Table 2). The PK experiments used rats
dosed at 2 mpk iv and 10 mpk po. With the exception of
compound 34 (the most potent compound in the func-
tional assay), compounds in the benzamidine series
demonstrated good oral bioavailability and plasma half-
lives with compound 35 having a remarkably long 7-h
half-life. Efficacy was measured by scoring behavioral
responses to noxious stimuli in a carageenan-induced
It had been shown that a naphthamidine (8), combined
with appropriate substitution on the benzyl ring, gave
compounds of comparable activity to the styryl ami-
dines.16 However, benzamidine 7 had weak affinity for
the NR2B subtype (Ki=10,000 nM vs [3H]-Ifenprodil).
Therefore, initial efforts focused on developing the ben-
zamidine series into a potent and selective series of
NR2B antagonists by exploring substituent effects on
the phenyl ring. Toward this end, commercially avail-
able reagents, in addition to a few individually prepared
nitriles, were utilized to rapidly diversify the lead
structure (Scheme 1).
Table 1. Structure, binding affinity and functional activity of amidine
derived NR2B/NMDA receptor antagonists
Entry
R1
R2
NR2B
Ki (nM)
NR2B FLIPR
IC50 (nM)
Approximately 150 compounds were prepared from 10
aryl nitriles and 35 benzyl amines via the Pinner synth-
esis.18 The selection criteria for the nitriles was mostly
determined by our desire to temper the basicity of the
amidine functionality with electron withdrawing groups.
The results of the binding and functional assays
(FLIPR) for selected compounds are summarized in
Table 1.15 As compared to the parent structure (7), sig-
nificant effects on affinity were observed for the
substitutions made around the N-benzylbenzamidine
core.
7
H
H
2-F
3-F
4-F
2-Cl
3-Cl
4-Cl
H
3-Cl
3-Cl
3-Cl
3-Cl
3-Cl
3-Cl
3-Cl
H
H
H
H
H
2-OMe
3-OMe
4-OMe
2-Cl
3-Cl
3-CF3
3,5-diMe
>15,000
4200
>15,000
2000
600
—
—
—
—
450
—
420
160
—
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
3900
800
63
1800
120
970
480
4-Ph
4-OCF3
4-Cl
46
720
360
—
4.1
9.7
—
48
32
660
8.7
4-CF3
4-SO2Me
4-OCF3
4-OCF3
4-OCF3
4-CF3
4-CF3
4-CF3
4-CF3
3500
5.7
23
1100
56
20
700
Scheme 1. Synthesis of amidine 11: (a) HCl, EtOH; (b) R2BnNH2,
EtOH.17
3.9

C. F. Claiborne et al. / Bioorg. Med. Chem. Lett. 13 (2003) 697–700
699
Table 2. Binding affinity, functional activity, rat PK and oral efficacy of compounds 31–35
Entry
Structure
NR2B Ki (nM)
NR2B FLIPR
IC50 (nM)
NR2A FLIPR
IC50 (nM)
%F
100
T1/2
(min)
ED50
(mpk)
31
72
47
>10,000
147
90
5.5
32
16
110
>10,000
44
14
33
34
35
12
57
4.2
24
>10,000
>10,000
>10,000
31
4
183
92
7.5
>30
16
1.2
0.6
68
420
Figure 3. Inhibition of NMDA receptor-activated currents at recom-
binant human NMDA receptor subtypes. NR1a +NR2A, NR2B and
NR2D were stably expressed in L(tk-) cells while NR1a+NR2C were
expressed transiently in HEK cells.19 Data represents the inhibition of
maximally evoked currents in response to a glutamate/glycine appli-
cation, and are meanÆSEM from at least five individual cells. Mean
data from NR1+NR2B was fitted using the Hill Equation.20
Figure 4. Effect of oral administration of 31 in methocel to rats in
carrageenan-induced hyperalgesia and rotarod assays. Data are
mean+SEM with n=8 for hyperalgesia, n=6 for rotarod.
assay, and virtually no separation exists between effi-
cacy and loss of motor coordination.7 In contrast, dosed
orally at 10 and 30 mpk, compound 31 exhibits sig-
nificant block of the hyperalgesic response with no
measurable effect on motor function. At 100 mpk,
compound 31 continues to be free from locomotor side
effects.
hyperalgesia assay in the rat.7 Compounds were dosed
orally, and results are reported as the dose (mpk)
required to achieve a 50% reduction in hyperalgesic
response as compared to control. All agents having
good oral pharmacokinetics provided ED50 values ran-
ging from 5.5 to 16 mpk.
In conclusion, we have developed a novel class of
NR2B-selective NMDA receptor antagonists which
demonstrate oral efficacy in a rodent hyperalgesia model
with no untoward effect on locomotor function asso-
ciated with non-selective NMDA antagonists.
Compound 31 was further evaluated in a rat rotarod
assay7 to assess any effect on locomotor function (Fig.
4). Non-selective NMDA antagonists, such as MK-801,
produce significant locomotor effects in the rotarod

700
C. F. Claiborne et al. / Bioorg. Med. Chem. Lett. 13 (2003) 697–700
Lett. See preceeding communication. doi:10.1016/S0960-
894X(02)01060-0.
References and Notes
17. Experimental procedure for the preparation of compound
31: A solution of 4-(trifluoromethoxy)benzonitrile (3.0 g, 16.0
mmol) in EtOH (40 mL) under N2 was cooled to 0 ꢀC. HCl gas
was bubbled through the solution for 10 min and the solution
was warmed to rt and stirred for 1 h. The reaction mixture was
concentrated, diluted with Et2O, sonicated for 5 min and fil-
tered. The imidate salt (2.8g, 65% yield) was recovered as a
white solid.
1. Le, D. A.; Lipton, S. A. Drugs & Aging 2001, 18, 717.
2. Max, M. B.; Byas-Smith, M. G.; Gracely, R. H.; Bennet,
G. J. Clin. Neuropharmacol. 1995, 18, 360.
3. Muir, K. W.; Lees, K. R. Stroke 1995, 26, 503.
4. Sugihara, H.; Moriyoshi, K.; Ishii, T.; Masu, M.;
Nakanishi, S. Biochem. Biophys. Res. Commun. 1992, 185,
826.
5. Cull-Candy, S.; Brickley, S.; Farrant, M. Curr. Opin. Neu-
robiol. 2001, 11, 327.
6. Monyer, H.; Burnashev, N.; Laurie, D. J.; Sakmann, B.;
Seeburg, P. H. Neuron 1994, 12, 529.
7. Boyce, S.; Wyatt, J. K.; Webb, R.; O’Donnell, R.; Mason,
G.; Rigby, M.; Sirinathsinghji, D.; Hill, R. G.; Rupniak,
N. M. J. Neuropharmacology 1999, 38, 611.
8. Williams, K. Mol. Pharmacol. 1993, 44, 851.
9. Kew, J. N. C.; Trube, G.; Kemp, J. A. J. Physiol. 1996,
497.3, 761.
10. Chizh, B. A.; Headley, P. M.; Tzschentke, T. M. Trends
Pharmacol. Sci. 2001, 22, 636.
11. (a) Nikam, S. S.; Meltzer, L. T. Curr. Pharm. Des. 2002, 8,
845. (b) Chenard, B. L.; Menniti, F. S. Curr. Pharm. Des.
1999, 5, 381.
12. Chenard, B. L.; Bordner, J.; Butler, T. W.; Chambers,
L. K.; Collins, M. A.; DeCosta, D. L.; Ducat, M. F.; Dumont,
M. L.; Fox, C. B.; Mena, E. E.; Menniti, F. S.; Nielsen, J.;
Pagnozzi, M. J.; Richter, K. E. G.; Ronau, R. T.; Shalaby,
I. A.; Stemple, J. Z.; White, W. F. J. Med. Chem. 1995, 38,
3138.
13. Wright, J. L.; Kesten, S. R.; Upasani, R. B.; Lan, N. C.
PCT Int. Appl., WO 2000000197, 2000. Chem. Abstr. 2000,
132, 64255.
14. Pinard, E.; Alanine, A.; Bourson, A.; Buttelmann, B.;
Heitz, M.-P.; Mutel, V.; Gill, R.; Trube, G.; Wyler, R. Bioorg.
Med. Chem. Lett. 2002, 12, 2615.
15. For NR2B Binding and FLIPR assay protcols see:
McCauley, J. A.; Theberge, C. R.; Liverton, N. J.; Claremon,
D. A.; Claiborne, C. F. US Patent 6,316,474 B1, 2001. Chem.
Abstr. 2001, 134, 340508.
16. Curtis, N. R.; Diggle, H. J.; Kulagowski, J. J.; London,
C.; Grimwood, S.; Hutson, P. H.; Murray, F.; Richards,
P.; Macaulay A.; Wafford, K. A. Bioorg. Med. Chem.
To a solution of the imidate salt (1.0 g, 3.7 mmol) in EtOH
(15 mL) at rt was added 2-(trifluoromethyl)benzyl amine (0.52
mL, 3.7 mmol) and triethylamine (1.54 mL, 11.1 mmol). After
4 h, the reaction mixture was concentrated and purified by
reverse-phase HPLC. The TFA salt of 31 was partitioned
between EtOAc/NaHCO3, dried and evaporated to give the
free base. HCl/Et2O (1 equiv) was then added to a CH2Cl2
solution of the free base. The mixture was sonicated and fil-
tered to give the HCl salt of 31 as a white solid (650 mg, 44%
yield): 1H NMR (400 MHz, MeOH-d4) d 7.87 (d, 2H), 7.83 (d,
1H), 7.73 (dd, 1H), 7.66 (d, 1H), 7.60 (dd, 1H), 7.52 (d, 2H),
4.85 (s, 2H) ppm; HRMS (FT/ICR) m/z 363.0896 [(M+H)+;
calcd for C16H13F6N2O: 363.0927].
18. Dox, A. W. Org. Synth. 1932, CV1, 5.
19. Priestley, T.; Laughton, P.; Myers, J.; Le Bourdelle
Kerby, J.; Whiting, P. J. Mol. Pharmacol. 1995, 48, 841.
`
s, B.;
20. Whole-cell patch-clamp experiments were performed
using an Axoclamp 200B on coverslip-plated L(tk-) cells stably
transfected with NR1a/NR2A, NR1a/NR2B or NR1a/NR2D
receptors and HEK cells transiently transfected with NR1a/
NR2C receptors. The external recording solution consisted of
(in mM): NaCl (149), KCl (3.25), CaCl2 (2), HEPES (10),
glucose (11), sucrose (21.9); pH 7.4. Ringer was superfused at
a rate of 0.7–2.0 mL/min. The recording pipette solution was
composed of (in mM): CsF (120), CsCl (10), EGTA (10),
CaCl2 (0.5); pH 7.25. Agonists and antagonist were applied
locally by fast perfusion from a double-barreled pipette and
experiments were performed at room temperature. Inhibition
curves for NMDA antagonists were obtained in the presence
of 10mM glutamate and 1 mM glycine (10 mM for NR1a/
NR2A). Increasing concentrations of antagonist was pre-
applied for 30 s followed by two co-applications of glu/gly for
5 s. Inhibition of the second glu/gly response in the presence of
antagonist was determined.