The identification of novel orally active mGluR5 antagonist GSK2210875

Article abstract of DOI:10.1016/j.bmcl.2010.09.120

The identification of novel orally active mGluR5 antagonist GSK2210875 is described.

Full text of DOI:10.1016/j.bmcl.2010.09.120

Bioorganic & Medicinal Chemistry Letters 20 (2010) 7521–7524
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The identification of novel orally active mGluR5 antagonist GSK2210875
Maria Pilla b, Michela Andreoli b, Michela Tessari b, Sonia Delle-Fratte b, Adelheid Roth b, Sharon Butler c,
Fiona Brown c, Parita Shah c, Ezio Bettini b, Palmina Cavallini b, Roberto Benedetti b, Doug Minick c,
Paul Smith a, Ben Tehan a, Pier D’Alessandro a, Olivier Lorthioir a, Catherine Ball a, Vincenzo Garzya a,
Caroline Goodacre a, Stephen Watson a,
⇑
a GlaxoSmithKline: Neurosciences Centre of Excellence for Drug Discovery, New Frontiers Science Park, Harlow, UK
b Medicines Research Centre, Verona, Italy
c Molecular Discovery Research, New Frontiers Science Park, Harlow, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The identification of novel orally active mGluR5 antagonist GSK2210875 is described.
Received 2 September 2010
Revised 23 September 2010
Accepted 24 September 2010
Available online 2 November 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
mGluR
mGluR5
Brain free fraction
Negative allosteric modulator
Allosteric modulator
GPCR
Family C GPCR
Metabotropic glutamate receptor
F
Antagonism of the type 5 metabotropic glutamate receptor
(mGluR5) has therapeutic potential in a wide range of disease
states,1,2 including, for example, pain, anxiety, and drug depen-
dency. Since the discovery of archetypal selective mGluR5 antago-
nist MPEP, more correctly described as a Negative Allosteric
Modulator (NAM), there has been a huge effort industry wide to
identify and develop mGluR5 antagonists suitable for clinical use
(Fig. 1).1,2 Much effort has focussed on acetylenic compounds related
to MPEP itself, for example, Addex report3 a positive PoC in migraine
and gastroesophogeal reflux disease (GERD) for acetylenic/MPEP
analogue compound ADX10059, and also on compounds bearing
an acetylene equivalent or isostere, such as compound 1 exemplified
by workers at Merck.4 The previously known anxiolytic Fenobam
has been demonstrated to be an mGluR5 antagonist5 and a number
of chemotypes containing its structural features have appeared in
the patent literature.1,2
Cl
O
N
N
O
N
N
N
O
N
N
N
N
N
H
1
Fenobam
MPEP
N
Cl
O
O
S
N
N
H
N
O
N
N
H
N
N
N
GSK2210875
mGluR5 pIC50 = 7.4
2
Figure 1. mGluR5 antagonists.
filled challenge and the search for novel and potential superior
chemotypes continues. In this context, herein we report
GSK2210875, a novel orally active low molecular weight mGluR5
antagonist.
Latterly further distinct chemotypes such as the orally active
piperidyl amides6 (2) have been reported. However despite the
substantial activity in this area the identification of safe tolerated
and clinically efficacious mGluR5 antagonists remains an unful-
The GSK2210875 chemotype was identified from
a High
Throughput Screen (HTS) of the GlaxoSmithKline compound col-
lection, and represented an attractive starting point for iteration
based on its low molecular weight, relatively high polarity, ligand
efficiency7 (0.42), and selectivity over other mGluRs.
⇑
Corresponding author. Tel.: +44 1279 875004.
E-mail address: stwa@lundbeck.com (S. Watson).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.09.120

7522
M. Pilla et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7521–7524
Table 1
O
O
O
H
N
O
N
N
S
R1
S
Potency of carbamate and triazole variants
R1
a
R2
R3
N
R3
N
N
H
O
Hal
R2
O
O
N
N
4
mGluR58
pIC50
N
H
R4
R1
R1
R4
3
N
b
All racemates
O
N
S
OH
R3
R4
N
S
R1
c
N
H
7
8
9
10
11
12
H
H
H
H
H
H
Cyclohexyl
Phenyl
5.5
7.2
6.3
7.4
5.6
5.3
N
R1
R3
N
N
R2
N
2-Cl-Phenyl
3-Cl-Phenyl
4-Cl-Phenyl
3,4-di-Cl
Phenyl
R2
6
5
Scheme 1. Reagents and conditions: (a) Ethanol,
l
Wave, 130 °C, 1 h; (b) NaBH4, rt,
1 h; (c) RNCO, Wave, 80 °C, 20 min, or RNH2, CDI,
l
l
Wave, 120 °C, 20 min, 20–75%.
13
14
15
16
17
18
19
H
Cl
3-F-Phenyl
Phenyl
Phenyl
Phenyl
Phenyl
7.3
8.0
7.9
8.1
8.0
7.4
7.4
nb product from (a) typically progressed directly to (b) without work-up or
purification steps: Where appropriate enantiomers were separated by chiral HPLC
as final compound or alcohol precursor.
Cyclopropyl
Thiophen-2-yl
4-F-phenyl
Methyl
Phenyl
Structure–activity relationships8 (SAR) were established via
synthesis of analogues according to Scheme 1 or simple variants
thereof. 1,2,4-Triazole-thiones (3) were reacted with the appropri-
Trifluoromethyl Phenyl
ate
a-halo-diketone (4) to afford heteroarylmethylketones (5)
which were reduced to the corresponding alcohols and capped to
afford final compounds (6).
Alternative triazole isomers of type 26 were prepared according
to the chemistry of Scheme 2, where the 2-hydrazido-thiazole (28)
was cyclised with formic acid to form the key bicycle. Reduction of
the ketone and capping of the alcohol to afford final compound
proceeds as described in Scheme 1. The corresponding oxygen ana-
logue (27) was prepared according novel chemistry developed in
these laboratories.10
It was established that mGluR5 potency (relative to racemic
parent compound (8)) could be readily modulated by substitution
of the triazolo ring (e.g., enhanced potency of thiophene (16) and
cyclopropane (15)), and that whilst potency was not readily en-
hanced via substitution of the carbamate aryl ring, this region of
the molecule was sensitive to structural changes (e.g., lower
potency of 4-Cl phenyl analogue (11) (see Table 1)
Intriguingly both methyl groups of the chemotype play a key
role in interactions with the mGluR5 receptor. Removal of either
or both of the methyl groups essentially ablates activity (20–22),
such a profound role for a methyl group suggests more than a sim-
ple lipophilic binding interaction with the receptor, but also likely
a role in pre-organising the conformation of the ligand. N-Methyl-
ation of the carbamate likewise ablates activity (23). The single
enantiomers of structure 8, compounds 24 and GSK2210875, were
separated by chiral HPLC and activity shown to reside in exclu-
sively GSK2210875 (the R enantiomer as confirmed by VCD).9
The mouse neophobia/marble burying anxiety model11,12 was
chosen as the principal pharmacodynamic (PD) model to establish
oral activity. Experiments with MPEP and Fenobam had suggested
that a free brain concentration of antagonist corresponding to ca.
10 times its pIC50 (i.e., ca. 90% theoretical receptor occupancy)
would be required to achieve a significant effect in this model.
We thus sought to identify compounds which could afford this pro-
file at doses of <10 mg/kg.
The key interplay between potency and brain free fraction in
achieving receptor occupancy has been elegantly demonstrated
by Watson et al.13 in dopamine D2 antagonists. We sought to ob-
tain an optimal balance of these parameters by maximising a hy-
brid parameter pIC50eff (Eq 1) which corrects the in vitro
potency of a compound for the fraction which will be free/unbound
in the brain13 (fub) to afford a more realistic measure of potency
in vivo.
O
O
O
O
O
S
O
S
N
N
S
N
N
N
N
N
H
N
H
N
H
N
N
N
22
21
20
pIC50eff ¼ pIC50 þ log10½fubꢀ
ð1Þ
mGluR5 pIC50 < 4.5
mGluR5 pIC50 < 4.5
mGluR5 pIC50 = 5.0
O
O
The data of Table 2 demonstrate a clear positive correlation be-
tween mGluR5 potency and lipophilicity, and a negative correla-
tion between potency and brain free fraction in compounds of
type 6 such that the more polar compounds 8 and 30 afford the
highest effective potency pIC50eff, that is, the best balance between
potency and free fraction. In more potent compounds such as 16
the higher potency does not compensate for the reduction in brain
free fraction that the increased lipophilicity causes.
O
O
S
N
N
O
S
O
S
N
N
N
N
H
N
N
N
H
N
N
N
23
mGluR5 pIC50 < 4.5
24
mGluR5 pIC50 < 5.0
GSK2210875
mGluR5 pIC50 = 7.4
Modifications to the triazole region of compounds of type 6
afforded modest gains in potency in the case of the corresponding
imidazole (25), a modest decrease in potency in the case of the cor-
responding triazole isomer (26) and complete ablation of activity
in the corresponding oxygen analogue (27).
O
H
N
O
O
R1
O
O
S
O
N
S
a
b
S
N
H
R1
NH2
N
R2
R3
N
O
R3
N
N
H
N
H
R3
O
O
Hal
R1
R2
S
N
R2
O
O
O
O
S
N
N
N
28
N
H
N
H
N
H
N
N
N
N
27
26
mGluR5 pIC50 = 6.9
25
Scheme 2. Reagents and conditions: (a) EtOH,
120 °C, 8 h, ꢁ50%.
lwave, 130 °C, 1 h; (b) HCO2H, PPA,
mGluR5 pIC50 < 4.5
mGluR5 pIC50 = 7.9

M. Pilla et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7521–7524
7523
Table 2
Relationship between potency, brain free fraction (fub) and pIC50eff
Table 3
Profile of 31 and GSK2210875
Compound (all as racemates)
mGluR58
pIC50
log D
(pH 7.4)
fub14
pIC50eff
Compound
Human8
mGluR5
pIC50
Murine15
mGluR5
pIC50
mGluR1 Solubility
pIC50
(lg/mL)
O
O
N
N
N
S
O
F
N
H
O
S
N
6.5
7.2
7.1
7.3
7.4
7.9
1.0
2.0
2.3
2.6
2.9
2.4
28
5.9
N
N
H
N
7.4
7.4
7.5
7.5
<4.3
31
N
29
O
O
F
31
N
S
O
N
N
H
O
S
N
6.8
8
6.0
6.0
5.5
5.4
5.4
N
N
H
N
4.5
166
N
30
O
O
GSK2210875
N
N
S
N
H
N
N
N
N
8
and 4 in agonist and antagonist mode) and against another 40 tar-
gets in the GSK in-house selectivity panel (data not shown). In
addition, in anticipation of progression to the neophobia model,
both compounds demonstrated essentially indistinguishable po-
tency at mouse14 and human receptors.
O
F
O
N
N
S
S
S
N
H
1.7
0.9
0.3
13
O
O
Cl
Both GSK2210875 and 31 demonstrated excellent oral exposure
and bioavailability in the rat (Table 3). Dissappointingly 31 showed
poor brain penetration, despite no evidence for PGP efflux hence we
hypothesize some other efflux mechanism, however GSK2210875
demonstrated excellent brain exposure. Indeed the values in
Tables 3 and 4 would suggest a 1 mg/kg oral dose affords a brain
Cmax of ca. 250 nM and a free brain Cmax of 30 nM equivalent to
the compound’s pIC50, and hence that a 5–10 mg dose would afford
the 5–10ꢃ pIC50 or 90% theoretical receptor occupancy appropriate
for activity in the neophobia model.
N
N
N
H
10
O
O
N
N
H
25
O
O
N
N
S
N
H
N
8.1
3.7
<0.1
<5
S
In the mouse neophobia model11,12 GSK2210875 did indeed
demonstrate oral activity (Fig. 3) with a minimum effective dose
of 5 mg/kg consistent with that predicted from potency/free frac-
tion/exposure calculations above.
16
In summary a novel, selective, orally active MPEP competitive
mGluR5 negative allosteric modulator GSK2210875 has been iden-
tified. Functional studies clearly demonstrated the insurmountable
interaction with the native mGluR5 receptor and binding studies
with the [3H] MPEP radioligand evidenced the allosteric nature of
this interaction. However our initial molecular modeling could
Based on their effective potency 8 and 30 were progressed for
further evaluation as their corresponding single enantiomers
GSK2210875 and 31.
In mouse astrocytes19 concentration response curves (CRC) of
L-quisqualate generated against increasing concentrations of com-
pound 8 were indicative of a non-competitive interaction with
the receptor as the dextral displacement of the agonist CRC oc-
curred with a concomitant decrease of the maximal response are
best described by an insurmountable profile16 (Fig. 2). The potency
of 7.3 0.2 was determined using the operational model for non-
Table 4
In vivo pharmacokinetic data in rat (1.0 mg/kg po 0.5 mg/kg iv)
Compound
Clb
(mL/min/Kg)
Oral blood
Cmax (ng/mL)
AUC
brain:blood
F (%)
competitive antagonists (
a
ꢂ1). Indeed MPEP radioligand17 bind-
GSK2210875
31
48
33
63
210
2.0
0.2
28
100
ing studies performed on the corresponding demonstrated that
compound 8 binds to the classical MPEP allosteric site.18
Both GSK2210875 and 31 showed good solubility and selectiv-
ity over other metabotropic glutamate receptors (e.g., mGluRs1,2
120
L-Quisqualate
+ 0.1 uM cpd
100
+ 0.3 uM cpd
+ 1 µM cpd
80
60
40
20
0
-10
-9
-8
-7
-6
-5
-4
log [M]
Figure 3. Effect of GSK2210875 in mouse neophobia after oral administration
(statistical analysis: ANOVA F(3,43) = 21.07, p <0.001; Dunnett’s post hoc test, * = p
<0.05, ** = p <0.001).
Figure 2. Concentration response curves of orthosteric mGluR5 agonist quisqualic
acid raised to increasing concentrations of compound 8.15

7524
M. Pilla et al. / Bioorg. Med. Chem. Lett. 20 (2010) 7521–7524
11. Njung’e, K.; Handley, S. L. Pharmacol. Biochem. Behav. 1991, 38, 63–67.
12. Fifteen minutes after oral administration of GSK2210875 (5, 20, 50 mg/kg) or
vehicle (0.5% HPMC in water, 10 mL/kg), mice were individually placed in
separate plastic cages (42.5 ꢃ 26.6 ꢃ 18.5 cm) containing 5 cm thick sawdust
bedding. Twenty-four glass marbles were arranged evenly spaced onto the
bedding (three rows of eight 1.5 cm marbles each). Thirty minutes later, mice
were removed from the cage and the number of marbles buried (by at least
two-thirds) was registered.
13. Watson, J.; Wright, S.; Lucas, A.; Clarke, K. L.; Viggers, J.; Cheetham, S.; Jeffrey,
P.; Porter, R.; Read, K. D. Drug Metab. Dispos. 2009, 37, 753.
14. Brain fractions unbound were determined using equilibrium dialysis in a 96-
well format as described in: Ferrari, L.; Crestan, V.; Sabattini, G.; Vinco, F.;
Fontana, S.; Gozzi, A. J. Neurosci. Methods 2010, 186, 143.
generate no plausible conformations which afforded a convincing
overlap with known ligands and we thus postulate GSK2210875
exhibits a novel pharmacophoric interaction with the receptor.
References and notes
1. Jaeschke, J.; Wettstein, J. G.; Nordquist, R. E.; Spooren, W. Expert. Opin. Ther.
Patents 2008, 18, 123.
2. Lindsley, C. W.; Emmitte, K. A. Curr. Opin. Drug Discov. Dev. 2009, 12, 446.
3. Reported for example at 6th International Symposium on metabatropic
Glutamate Receptors, Taormina, Italy, September, 2008.
4. Huang, A. D.; Poon, S. F.; Eastman, B. W.; Chapman, D. F.; Chung, J.; Cramer, M.;
Holtz, G.; Cosford, N. D. P.; Smith, N. D. Bioorg. Med. Chem. Lett. 2004, 14, 5477.
5. Porter, R. H.; Jaeschke, G.; Spooren, W.; Ballard, T. M.; Büttelmann, B.;
Kolczewski, S.; Peters, J.; Prinseen, E.; Wichmann, J.; Vieira, E.; Muhlemann,
A.; Gatti, S.; Mutel, V.; Malherbe, P. J. Pharmacol. Exp. Ther. 2005, 315, 711.
6. Spanka, C.; Glatthar, R.; Desrayaud, S.; Fendt, M.; Orain, D.; Troxler, T.; Vranesic,
I. Bioorg. Med. Chem. Lett. 2010, 20, 184.
15. Astrocyte cultures were plated for 2 days in 384 well plates (Greiner, black
well, clear bottom). For mode of action studies cells were pre-incubated for
30 min with increasing concentrations of antagonist compounds. mGluR5
receptors were then challenged with concentration response curves of L-
quisqualate and the intracellular Calcium release monitored in real-time on a
FLIPRTETRAÒ system. For fpKi determination cells were pre-incubated with
antagonist CRCs and then challenged with the EC80 of
determined from the agonist response curve of that day.
L-quisqualate,
7. Ligand efficiency = 1.36pIC50/#heavy atoms, see: Abad-Zapatero, C.; Metz, J. T.
Drug Discovery Today 2005, 10, 464.
16. Kenakin, T.; Jenkinson, S.; Watson, C. J. Pharmacol. Exp. Ther. 2006, 319, 710.
17. Binding assay was performed as previously described (Malherbe et al. Mol.
Pharm. 2003, 64, 823–832) for native mGluR5 receptors. Brain cortex
membranes from mice or rats were incubated for 1 h at 4 °C, with 1 nM
[3H]MPEP and increasing concentrations of compounds to determine the
affinities (pKi values) of mGluR5 inhibitors. Data represented are the mean of
n P 2 experiments generated from seven point dose–response curves carried
out in duplicate.
8. mGluR5 pharmacology characterised using an aequorin calcium mobilisation
assay on the Lumilux. Assay utilises CHO cells stably expressing apo-aequorin
and induced to express the human mGluR5 receptor using doxycycline. Eleven
point concentration-response curves were generated in the presence of an EC80
of glutamate. pIC50 values are average of at least two curves.
9. The absolute configuration of GSK2210875 was determined by ab initio
Vibrational Circular Dichroism (VCD). Solution-phase VCD and IR spectra were
measured in CDCl3 using a BioTools BOMEM Chiralir™ FT-VCD spectrometer
operating at 4 cmꢄ1 resolution and 50 scans per minute. Spectral data were
acquired in the 2000–950 cmꢄ1 region of the mid-infrared spectrum. The
absolute configuration was assigned by comparing the experimental VCD
spectrum to the VCD spectrum calculated for a full structure model with (R)-
configuration. The model VCD spectrum was largely coincident with
GSK2210875, indicating that this molecule has the same absolute
stereochemistry as the model. Based on these data, GSK2210875A was
assigned as the (1R)-enantiomer. See for example: Stephens, P. J., Devlin, F. J.,
Pan, J.-J. Chirality 2008, 20, 643–663.
18. Pagano, A.; Ruegg, D.; Litschig, S.; Stoehr, N.; Stierlin, C.; Heinrich, M.;
Floersheim, P.; Prezèau, L.; Carroll, F.; Pin, JP.; Cambria, A.; Vranesic, I.; Flor, PJ.;
Gasparini, F.; Kuhn, R. J. Biol. Chem. 2000, 275, 33750.
19. CD1 mice were obtained from Charles River, Lecco, Italia. Astrocyte cultures
were prepared from mouse pups aged 1–3 days. Mixed primary cultures were
established from mouse cortices as previously described by Miller S. et al. (J.
Neurosci. 1995, 6103–6109) with the following modifications. Mouse pups
were used instead of rat pups. After 5 days in culture cells were detached by
trypsinization and re-plated to eliminate neurons from the mixed culture.
Cultures were kept at 37 °C in a humidified CO2 atmosphere until they reached
confluence (10–14 d in vitro) prior being used in experiments.
10. Ball, C.; Dean, D. K.; Lorthioir, O.; Page, L. W.; Smith, C. L.; Watson, S. P.
Tetrahedron Lett. 2010, 51, 3907.