Synthesis and SAR development of novel mGluR1 antagonists for the treatment of chronic pain

Article abstract of DOI:10.1016/j.bmcl.2012.09.048

High throughput screening identified the pyridothienopyrimidinone 1 as a ligand for the metabotropic glutamate receptor 1 (mGluR1 = 10 nM). Compound 1 has an excellent in vivo profile; however, it displays unfavorable pharmacokinetic issues and metabolic

Full text of DOI:10.1016/j.bmcl.2012.09.048

Bioorganic & Medicinal Chemistry Letters 22 (2012) 7223–7226
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and SAR development of novel mGluR1 antagonists for the treatment
of chronic pain
Stephanie Brumfield a, , Peter Korakas a, Lisa S. Silverman a, Deen Tulshian a, Julius J. Matasi a, Li Qiang a,
⇑
Chad E. Bennett a, Duane A. Burnett a, William J. Greenlee a, Chad E. Knutson a, Wen-Lian Wu a,
T. K. Sasikumar a, Martin Domalski a, Rosalia Bertorelli b, Mariagrazia Grilli b, Gianluca Lozza b,
Angelo Reggiani b, Cheng Li a
a Merck Research Laboratories, 2015 Galloping Hill Road, MS 2545, Kenilworth, NJ 07033-0539, USA
b Merck Research Laboratories, San Raffaele Science Park, Via Olgettina, 58, 20132 Milan, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 June 2012
Revised 10 September 2012
Accepted 17 September 2012
Available online 2 October 2012
High throughput screening identified the pyridothienopyrimidinone 1 as a ligand for the metabotropic
glutamate receptor 1 (mGluR1 = 10 nM). Compound 1 has an excellent in vivo profile; however, it dis-
plays unfavorable pharmacokinetic issues and metabolic stability. Therefore, using 1 as a template, novel
analogues (10i) were prepared. These analogues displayed improved oral exposure and activity in the
Spinal Nerve Ligation (SNL) pain model.
Ó 2012 Elsevier Ltd. All rights reserved.
Glutamate is the principal excitatory neurotransmitter in
mammalian brains. Glutamate receptors can be divided into two
subtypes: ionotropic and metabotropic. Ionotropic receptors have
been associated with the opening of cationic channels,1 and are
responsible for rapid neuronal excitation of glutamate transmis-
sion.1b Metabotropic G-protein receptors (mGluR) indirectly regu-
late electrical signaling by influencing intracellular metabolic
processes via G-proteins.1c Eight metabotropic G-protein coupled
receptors have been reported1 which have been divided into three
groups: Group I (mGluR1 and mGluR5), Group II (mGluR2 and
mGluR3) and Group III (mGluR4, mGluR7 and mGluR8).1 It has
been reported that mGluR1 is essential for motor coordination,2
perception of pain,3 and may play an important role in seizures
and related disorders.4 mGluR1 knock out mice exhibit lower pain
sensitivity and are also more receptive to morphine than the wild
types.5 Thus an mGluR1 receptor antagonist would be beneficial
for the treatment of neuropathic pain.3 Our goal for this project
was to develop a compound that would have a potency of less than
sponse to stimulus with the von Frey filaments of the rats that
had undergone the surgery but not the ligation.
The commercially available compound 1 exhibited potent bind-
ing at mGluR1.8 However, further testing showed that the oral
exposure for this compound was relatively poor. (AUC = 679 ngh/
mL, Fig. 1) Compound 1 did show activity in the SNL disease model
with an 87% reversal of tactile allodynia at 10 mg/kg. This com-
pound also was prone to rapid demethylation of the N,N-dimethyl
amine moiety which contributed to the activity in the SNL animal
model. Two of the possible metabolites were then tested in the
binding and pharmacokinetic assays (Fig. 2). These two potential
metabolites were potent at mGluR1, but showed similar oral expo-
sure to the parent.
The goal was therefore to improve the oral exposure and main-
tain the mGluR selectivity of analogue 1 and develop molecules
with improved metabolic stability. It was our hope that by replac-
ing the A ring with a pyrimidine moiety, we would improve the
10 nM in our binding assay, oral exposure6 at six hours and activity
7
in the Spinal Nerve Ligation (SNL) animal model
.
N
N
The SNL7 animal model is done by ligating the L-5 spinal nerve
in rats and then observing the sensitivity of the hind paw of the
animal to mechanical stimulation with von Frey filaments. As a
point of comparison there is a second group of rats that undergo
a similar surgery but not the ligation of the spinal nerve. This is
done to rule out the effects of surgery in the neuropathic pain mod-
el. Our goal was to obtain an inhibitor that would equal the re-
N
O
C
A
B
S
N
O
1
hmGluR1 IC50 = 10 nM
hmGluR5 IC50 = inactive at 10µM
hmGluR2 IC50 = inactive at 10µM
RR AUC0-6h6 @ 10 mpk 679
ngh/mL
⇑
Corresponding author.
E-mail address: Stephanie.Brumfield@merck.com (S. Brumfield).
Figure 1. Initial lead from high throughput screening.
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.09.048

7224
S. Brumfield et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7223–7226
Table 1
NH2
NH
N
In vitro data for pyrimidinone N-1 modifications
N
N
O
N
O
N
N
N
S
O
N
S
O
1a
N
N R
1b
N
S
hmGluR1 IC50 = 1 nM
hmGluR1 IC50 = 12 nM
O
RR AUC0-6h6 @ 10 mpk 429
gh/mL
RR AUC0-6h6 @ 10 mpk 442 ngh/mL
9
Entry
R
hmGluR1 IC50 hmGluR5 IC50 RR AuC0–6h
(nM)
(nM)
(nM h)
Figure 2. Metabolites of Compound 1.
9a
9b
9c
2
1000
1313
O
metabolic stability of the core. Modifications to the peripheral
functional groups of the tricyclic core would also be investigated.
To access these analogues, a versatile synthetic route was devel-
oped8e, exemplified by the preparation of analogue 9a (Scheme 1)
These modifications not only generated potent, selective com-
pounds but provided analogues with improved pharmacokinetic
profiles relative to 1. Starting with readily available 4,6-dihydroxy
pyrimidine (2), Vilsmeier conditions were used to effect two trans-
formations: installation of a formyl handle at the pyrimidine 5-po-
sition and transformation of both hydroxyl moieties into chlorides.
Exposure of the formyl group to hydroxyl amine hydrochloride un-
der acidic aqueous conditions followed by thionyl chloride yielded
the nitrile 4. This step installs the expected oxime and dehydrates
to the corresponding nitrile; however, unwanted monohydroxyla-
tion also occurs. Resubjecting 4 to refluxing phosphorus oxychlo-
ride in the presence of a small amount of triethylamine yields
the requisite dichloropyrimidine. Treatment of 5 with two equiva-
lents of methyl thioglycolate followed by exposure to refluxing tri-
ethylamine provided bicycle 6. This moiety proved to be a versatile
intermediate in that the modification to the C-4 thioether allows
for the installation of a variety of functional groups to facilitate
SAR investigation. In this instance, displacement of the methyl thi-
oglycolate using a solution of dimethylamine in THF provided 7.
1
1
10,000
ND
0
0
Cl
Br
F
9d
9e
9f
1
ND
0
4
ND
0
14
16
10,000
10,000
0
CF3
O
9g
166
CF3
S
9h
9i
8
10,000
200
950
ND
ND
7
OH
Cl
Cl
CHO
Cl
CN
OH
9j
204
ND
N
c
a
N
b
N
N
N
N
2
OH
N
N
3
N
4
9k
9l
20
15
5
10,000
ND
45
Cl
MeO2C
S
N
O
NH2
CN
Cl
d,e
N
CO2Me
N
1380
0
S
N
5
6
O
F
N
N
NH2
N
9m
ND
f
N
g
N
O
CO2Me
CO2Me
F
S
S
N
N
N
7
8
9n
7
10,000
15,431
N
OMe
O
N
O
N
h
The IC50 data is an average of at least three measurements, performed on human
mGluR1/5. The standard error was 10%, and variability was less than twofold from
assay to assay.
S
N
9a
Scheme 1. Reagents and conditions: (a) POCl3, DMF, reflux, 24 h; (b) H2NOH, AcOH,
reflux, 12 h. Then SOCl2 reflux, 12 h; (c) POCl3, reflux, 3 h. (d) HSCH2CO2Me, TEA,
THF, RT, 1 h; (e) TEA, toluene, reflux 24 h; (f) HNMe2, THF, reflux, 12 h; (g)
(MeO)2CHNMe2, reflux, 5 h R3NH2, HOAc, toluene, reflux, 12 h.
Reaction of 7 with dimethyl amine dimethyl acetal generated
intermediate 8. Exposure of 8 to a variety of amines under acidic

S. Brumfield et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7223–7226
7225
Table 2
Table 2 (continued)
Modification of the 4-position of the pyrimidine A-ring
Entry
R
hmGluR1
IC50 (nM)
hmGluR5
IC50 (nM)
RR AuC0–6h
(nM h)
R
N
N
N
O
O
Cl
N
S
10t
>1000
ND
ND
10
N
H
Entry
R
hmGluR1
IC50 (nM)
hmGluR5
IC50 (nM)
RR AuC0–6h
(nM h)
The IC50 data is an average of at least three measurements, performed on human
mGluR1/5. The standard error was 10%, and variability was less than twofold from
assay to assay.
10a
10b
10c
N
H
5
7
3
10,000
10,000
10,000
778
N
H
4018
934
conditions proceeded smoothly to give the target compounds (9a–
n, Table 1).
N
From the data shown in Table 1 it can be seen that most mod-
ifications at the pyrimidinone N-1 position were successful in
yielding compounds that were both potent and selective mGluR1
inhibitors. Modifications to the para position of the pendant aro-
matic ring were well tolerated (entries 9a–9h), while replacement
of the phenyl ring with a 3-pyridine substituent (entry 9j) led to a
significant decrease in potency. This activity could be restored
upon the addition of a suitably disposed methoxy. Disubstitution
H
N
10d
10e
10f
19
24
86
ND
ND
ND
1905
5405
ND
F3C
HO
NH
N
H
10g
HO
HO
N
12
ND
1174
H
10h
10i
10j
N
H
9
ND
1973
9614
ND
Table 3
Pyrimidinone N-1 modifications of 10i
H
6
10,000
ND
N
N
H
NH
N
154
N
N
O
R
N
10k
10l
5
ND
ND
443
ND
N
S
11
H
N
60
Entry
R
hmGluR1 IC50
(nM)
hmGluR5 IC50
(nM)
RR AuC0–6h
(nM h)
10m
10n
277
ND
ND
ND
ND
N
10i
6
9
3
10,000
10,000
10,000
9614
890
H
O
N
F
>1000
11a
11b
N
H
4540
10o
10p
10q
>1000
>1000
>1000
ND
ND
ND
ND
ND
ND
N
Cl
F
H
11c
11d
11e
82
70
19
ND
ND
ND
ND
N
H
NC
N
ND
N
H
F
8599
O
10r
10s
>1000
192
ND
ND
ND
ND
N
F
H
F
11f
30
ND
948
O
N
H
The IC50 data is an average of at least three measurements, performed on human
mGluR1/5. The standard error was 10%, and variability was less than twofold from
assay to assay.

7226
S. Brumfield et al. / Bioorg. Med. Chem. Lett. 22 (2012) 7223–7226
11b with a fluoro group led to 30-fold decrease in activity (11c).
F
N
N
Likewise, an analogue in which the phenyl ring had been replaced
with a cyclohexyl group was not potent enough to warrant further
investigation.
The addition of a fluoro substituent to the pendant aromatic
ring, as in entries 11e and 11f, gave no improvement in oral expo-
sure when compared to their non-fluoro counterpart 10i. This
stands in contrast to what had been observed in the dimethyl-
amino series outlined in Table 1.
Compound 10i, when compared to 9n, has a similar pharmaco-
kinetic profile, but carries the advantage of a more metabolically
stable cyclopropylamino substituent. Further, 10i exhibits a three-
fold increase in brain:plasma ratio (Fig. 3).
The favorable pharmacokinetic profile of 10i translates into an
orally efficacious compound in the rat SNL model for the treatment
of tactile allodynia. 10i exhibited a rat Ki of 14.5 nM and at 3 mg/kg
there is almost a complete reversal of allodynia with an ED50 of
2.0 mg/kg at the 2 h time point.
RR AUC: 15.4 µM.h
B/P Ratio: 0.5
N
N
O
O
N
S
9n
NH
N
RR AUC: 9.6 µM.h
B/P Ratio: 1.6
N
N
O
N
S
O
10i
Figure 3. Pharmacokinetic comparisons of two potent mGluR1 antagonists.
of the aromatic ring also appeared to be well tolerated. (entries 9l–
9n) The 2-fluoro-4-methoxy substitution pattern (9n) on the right
hand side aromatic ring yielded not only a potent, selective inhib-
itor of mGluR1 but also gave an analogue with oral exposure that
was superior to any of the other compounds examined. Compound
9n exhibited a rat Ki of 13.4 nM and turned out to be fairly effica-
cious in the rat SNL model. At 10 mg/kg dose there was almost a
complete reversal of allodynia with an ED50 of 3.8 mg/kg at the
2 h time point.
In conclusion, modifications to the peripheral functional groups
of the tricyclic core of 9a led to the identification of potent, selec-
tive mGluR1 inhibitors with good oral exposure. The two high-
lighted compounds (9n, 10i) also exhibited desirable
pharmacological properties, as well as efficacy in the rat SNL
model.
Subsequent metabolite studies showed that both 9n and 1 were
readily demethylated at the N,N-dimethylamino moiety. In order
to further improve the oral exposure of 9n and thus lower the
ED50, a second round of investigations focused on identifying a
suitable replacement for the N,N-dimethyl substituent (Table 2).
Amino substituents possessing small alkyl groups generally
gave rise to analogues with slightly improved potency relative to
9n (entries 10a–10d). Several of these analogues had reasonable
levels of oral exposure; curiously, none offered any improvement
compared to 9n. Likewise, analogues bearing hydroxyl alkyl substi-
tuted amines were also tolerated (entries 10f–h) Hydroxypropyl
(10g) and hydroxybutyl (10h) analogues exhibited comparable
mGluR1 inhibitory activity to 9n. Unfortunately the pharmacoki-
netic profiles for all three of these compounds were modest. Ana-
logues possessing aminocycloalkyl substituents were generally
equipotent to the dimethylamino analogue 9n (entries 10i–10l).
Notably, the cyclopropylamino species 10i exhibited exposure lev-
els similar to 9n. Arylamino analogues were all significantly less
potent that 9n (entries 10m–10t).
The SAR of the pyrimidinone ring substituent was then reinves-
tigated, using a fixed cyclopropylamino group appended to the
pyrimidine (Table 3). It was anticipated that this group would pro-
vide analogues with improved pharmacokinetic profiles relative to
the initial lead 9a. This was not the case. Although analogues pos-
sessing a 4-methylphenyl (11a) and 4-chloro substituent (11c)
were equipotent to 4-methoxyphenyl species 10i, their corre-
sponding rat AUC values were approximately 10 fold and twofold
less, respectively. Replacing the chloro substituent in compound
References and notes
1. (a) Schoepp, D. D.; Jane, D. E.; Monn, J. A. Neuropharmacology 1999, 38, 1431; (b)
Nakanishi, S. Neuron 1994, 1031, 13; (c) Knöpfel, T.; Kuhn, R.; Allegeier, H. J. Med.
Chem. 1995, 38, 1417; (d) Nakanishi, S. Science 1992, 258, 597; (e) Masu, M.;
Tanabe, Y.; Tsuchida, K.; Shigemoto, R.; Nakanishi, S. Nature 1991, 349, 760.
2. (a) Nakao, H.; Nakao, K.; Kano, M.; Aiba, A. Neurosci. Res. 2007, 57, 538; (b) Ichise,
T.; Kano, M.; Hashimoto, K.; Yanagihara, D.; Nakao, H.; Shigemoto, R.; Katsuki,
M.; Aiba, A. Science 1832, 2000, 288.
3. (a) Dolan, S.; Kelly, J.; Monteiro, A.; Nolan, A. Pain 2004, 110, 369; (b)
Neugebauer, V. Pain 2002, 98, 1.
4. Shannon, H.; Peters, S.; Kingston, A. Neuropharmacology 2005, 49, 188.
5. (a) Fundytus, M.; Yashpal, K.; Chabot, J.-G.; Osborne, M.; Lefebvre, C.; Dray, A.;
Henry, J.; Coderre, T. Brit. J. Pharm. 2001, 132, 354; (b) Joshi, S.; Hernandez, G.;
Miskusa, J.; Zhu, C.; Zhong, C.; Salyers, A.; Wismer, C.; Chandran, P.; Decker, M.;
Honore, P. Neuroscience 2006, 143, 587.
6. Mei, H.; Korfmacher, W.; Morrison, R. AAPS J. 2006, 8, E493.
7. Kim, S.; Chung, J. Pain 1992, 50, 355.
8. For related series of mGluR1 antagonists, see (a) Zheng, G. Z.; Bhatia, P.; Daanen,
J.; Kolasa, T.; Patel, M.; Latshaw, S.; ElKouhen, O. F.; Chang, R.; Uchic, M. E.;
Miller, L.; Nakane, M.; Lehto, S. G.; Honore, M. P.; Moreland, R. B.; Brioni, J. D.;
Stewart, A. O. J. Med. Chem. 2005, 48, 7374; (b) Zheng, G. Z.; Bhatia, P.; Kolasa, T.;
Patel, M.; El Kouhen, O. F.; Chang, R.; Uchic, M. E.; Miller, L.; Baker, S.; Lehto, S.
G.; Honore, P.; Wetter, J. M.; Marsh, K. C.; Moreland, R. B.; Brioni, J. D.; Stewart,
A. O. Bioorg. Med. Chem. Lett. 2006, 16, 4936; (c) Wu, W.-L.; Burnett, D. A.;
Domalski, M.; Greenlee, W. J.; Li, C.; Bertorelli, R.; Fredduzzi, S.; Lozza, G.; Veltri,
A.; Reggiani, A. J. Med. Chem. 2007, 50, 5550; (d) Sasikumar, T. K.; Qiang, L.;
Burnett, D. A.; Greenlee, W. J.; Li, C.; Heimark, L.; Pramanik, B.; Grilli, M.;
Bertorelli, S.; Lozza, G.; Reggiani, A. Bioorg. Med. Chem. Lett. 2009, 19, 3199;
Additional synthetic route: (e) Sasikumar, T. K.; Qiang, L.; Burnett, D. A.;
Greenlee, W. J.; Li, C.; Grilli, M.; Bertorelli, S.; Lozza, G.; Reggiani, A. Bioorg. Med.
Chem. Lett. 2010, 20, 2474; (f) Bennett, C. E.; Burnett, D. A.; Greenlee, W. J.;
Knutson, C. E.; Korakas, P.; Li, C.; Tulshian, D.; Wu, W.-L.; Bertorelli, R.;
Fredduzzi, S.; Grilli, M.; Lozza, G.; Reggiani, A.; Veltri, A. Bioorg. Med. Chem. Lett.
2012, 22, 1575.