A novel series of metabotropic glutamate receptor 5 negative allosteric modulators based on a 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine core

Article abstract of DOI:10.1016/j.bmcl.2013.06.044

A series of potent non-acetylinic negative allosteric modulators of the metabotropic glutamate receptor 5 (mGlu5 NAMs) was developed starting from HTS screening hit 1. Potency was improved via iterative SAR, and physicochemical properties were optimized to deliver orally bioavailable compounds acceptable for in vivo testing. A lead molecule from the series demonstrated dose-dependent activity in the second phase of the rat formalin test from 30 mg/kg, and a preliminary PK/PD relationship was established.

Full text of DOI:10.1016/j.bmcl.2013.06.044

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4523–4527
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A novel series of metabotropic glutamate receptor 5 negative
allosteric modulators based on a 4,5,6,7-tetrahydropyrazolo
[1,5-a]pyridine core
a
a
a
Guillaume Duvey a, Benjamin Perry a, , Emmanuel Le Poul , Sonia Poli , Beatrice Bonnet ,
Nathalie Lambeng a, Delphine Charvin a, Tansy Donovan-Rodrigues a, Hasnaà Haddouk a,
Stefania Gagliardi b, Jean-Philippe Rocher a
⇑
a Addex Therapeutics, 12 Chemin des Aulx, Plan-les-Ouates 1228, Geneva, Switzerland
b NiKem Research s.r.l., Via Zambeletti 25, 20021 Barazante (MI), 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 potent non-acetylinic negative allosteric modulators of the metabotropic glutamate receptor 5
(mGlu5 NAMs) was developed starting from HTS screening hit 1. Potency was improved via iterative SAR,
and physicochemical properties were optimized to deliver orally bioavailable compounds acceptable for
in vivo testing. A lead molecule from the series demonstrated dose-dependent activity in the second
phase of the rat formalin test from 30 mg/kg, and a preliminary PK/PD relationship was established.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 30 May 2013
Revised 12 June 2013
Accepted 16 June 2013
Available online 25 June 2013
Keywords:
mGlu5
Metabotropic
Pain
Glutamate
Allosteric
mGlu5 is a family I Gq-coupled metabotropic glutamate recep-
tor expressed both peripherally and within the CNS, primarily
postsynaptically in the limbic cortex, hippocampus, amygdala, ba-
sal ganglia, thalamus and olfactory tubercule.1 mGlu5 has been the
target of significant drug discovery efforts due to its implications in
numerous, varied indications such as migraine, Fragile X syn-
drome, chronic pain, gastroesophageal reflux disease (GERD) and
Parkinson’s disease.2 The majority of this work has focused on neg-
ative allosteric modulators (NAMs) of mGlu5 receptor such as
MTEP, mavoglurant and dipraglurant (Fig. 1).3–5 With the intention
of identifying novel mGlu5 receptor NAM pharmacophores, a high
throughput screen of the Addex corporate library was performed
(ca. 70,000 molecules) using a FLIPR-based Ca2+ release assay.6
Amongst the hits identified was 4,5,6,7-tetrahydropyrazolo[1,5-
a]pyridine 1, which afforded full inhibitory modulation with an
the compound suffered high intrinsic clearance in both human
and rat microsomes (97 and 118 L/min/mg prot. respectively).
As such, it was decided to further investigate this chemotype with
view to identifying compounds displaying improved potency and
in vitro microsomal stability.
l
Initial investigation focused on the nature of the link between
the pyridine ring and 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine
core, in order to find the optimal distance between these two mo-
tifs in terms of potency. Various linkers were investigated in this
position, the majority of which were inactive. Those that were
found active are shown in Table 1; these optimal linkers were of
2-atom lengths bearing either a carbonyl or ether functionality at
the linking position with the 4,5,6,7-tetrahydropyrazolo[1,5-a]pyr-
idine core (ether 2, ketone 3 and amide 4). Reversing the amide or
ether resulted in inactive compounds, as did extending or reducing
chain length. It is postulated that an element of conjugation be-
tween the two aromatic systems due to linker tautomerism may
be a significant potency driver; amide 4, capable of a tautomeric
form where the linker is a double bond, is active whereas substi-
tuted amide 5 is inactive. Likewise ketone 3, which is observed
as a 2:3 mixture between tautomeric enol ether and ketone in
1D 1H NMR, is rather potent. It is plausible that these pseudo-
conjugated linkers may be occupying the same area of the mGlu5
receptor NAM pharmacophore as the classic acetylene linker
IC50 of 1.3
receptor rat cortex binding assay (3H-MPEP) showed a binding IC50
of 1.2 M. Further profiling showed this hit compound to have a
lM. Subsequent validation of this molecule in an mGlu5
l
good solubility in kinetic solubility assays (0.17 and 0.18 mg/mL
at pH 1.0 and 7.4 respectively), no major issue in CYP inhibition
(no inhibitory IC50 >10 lM on 4 major CYP isoforms), however
⇑
Corresponding author.
E-mail address: benjamin.perry@addexpharma.com (B. Perry).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.06.044

4524
G. Duvey et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4523–4527
N
N
S
H
HO
N
N
N
COOMe
N
H
N
N
N
F
MTEP
Mavoglurant
Dipraglurant
1
Figure 1. Known mGlu5 NAMs and HTS screening hit 1.
present in many of the known mGlu5 receptor NAMs. It should be
noted that only one stereoisomer of the enol was observed by 1D
1H NMR (E/Z stereochemistry not determined).
Table 1
Investigation of the linker space between aromatic rings
Representative synthesis of the various linked compounds is
shown in Scheme 1. Direct-linked analogues 1, 1a–e and 6a–g
can be synthesized in 4 steps from substituted 2-acetylpyridines
or acetylbenzenes via pyrazoles 9 followed by ring closure. In an
alternative route to generate 1f, key intermediate 4,5,6,7-tetrahy-
dro-[1,2,3]oxadiazolo[3,4-a]pyridin-8-ium-3-olate (13) is gener-
ated in two steps from piperidine-2- carboxcylic acid 12.7
Reaction of this intermediate with various 2-acetylinyl-pyridines
or phenylacetylenes generates the direct-linked compounds via a
[3+2]cycloaddition. Reaction of key intermediate 13 with methyl
propiolate generates ester 14, which can either be reacted with
methylpyridines under basic conditions to give ketone-linked 3
and 3a, or saponified and coupled with amines to generate amides
4, 5 and 4a–o. Ester-linked compounds are generated using an
N
N
L
N
Compound
Linker
1
2
3
4
5
H
N
nO
n
N
O
O
O
FLIPR rmGlu5
IC50 (nM)
1292
2498
390
364
NA
O
a,b,c
X = CH or N
NH
N
R
R
X
X
X
8
9
Cl
d
e,f
Cl
R
R
X
11
10
i
O-
N
H
N
R
X
N
CO2H
g,h
1, 1a-1e, 1g-1h, 6a-6g (via 9)
1f (via 11)
O
N+
N
12
13
j
O
k
O
N
N
N
N
N
O
R
R
X
3, 3a
14
15
l
O
NH
O
m,n
N
N
N
HO
X
4, 5, 4a-o
O
q,r
s or t
o,p
HO
O
O
NH
N
N
N
OH
N
R
N
HO
X
16
17
18
2, 2a-f, 7a-f
Scheme 1. Reagents and conditions: (a) Me2NNH2, EtOH, reflux, 37–92%; (b) LDA, THF, À78 °C then 5-chlorovaleroylchloride, À78 to 0 °C; (c) hydrazine mono-hydrate, EtOH;
reflux 35–66% over 2 steps; (d) NaH, THF, 0 °C, 30–51%; (e) TMSCCH, Pd(PPH3))2Cl2, CuI, TEA, DCM, 80 °C, 84%; (f) KOH, MeOH, DCM, rt, 34%; (g) NaNO2, HCl, 0 °C; (h) TFAA,
THF, rt; 63% over 2 steps (i) xylene, reflux, 67%; (j) methyl propiolate, xylene, reflux, 85%; (k) substituted 2-methyl pyridine, KHMDS, THF, À78 °C to rt, 64–88%; (l) KOH,
MeOH, RT, 78%; (m) (COCl)2, DMF, DCM, RT; (n) substituted 2-aminopyridine, pyridine, RT, 60–80% over 2 steps; (o) EtOAc, Na, EtOH, RT; (p) hydrazine mono-hydrate, EtOH,
40 °C, 77% over 2 steps; (q) HBr, H2SO4, reflux; (r) NaOH, rt, 57% over 2 steps; (s) substituted 2-halomethylpyridine/benzyl halide, NaH, DMF, 0 °C to rt 5–51%; (t) substituted
2-hyroxymethyl pyridine/benzyl alcohol, PBu3, THF, (pipCON)2.

G. Duvey et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4523–4527
4525
Table 2
Table 4
Scanning of aryl ring in direct-linked series
Scanning of aryl ring in amide and ketone-linked series
4
3
4
3
3
2
5
R
5
R
Y
4
R
N
N
N
N
N
N
6
N
6
N
O
1a-h
6a-g
Compound
Y
R
FLIPR rmGlu5
IC50 (nM)
Compound
R
FLIPR rmGlu5
IC50 (nM)
4
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
CH2
CH2
H
364
NA
NA
1
H
1292
NA
3075
675
3003
669
>10,000
1243
1516
1895
2648
77
4232
3994
4503
1573
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
3
3-CN
3-Me
3-OMe
3-NH2
3-OH
4-CN
4-Me
5-CN
5-F
5-NH2
5-Cl
6-Me
6-OMe
6-CF3
6-F
1a
1b
1c
1d
1e
1f
1g
1h
6a
6b
6c
6d
6e
6f
3-Me
4-Me
4-Cl
5-Me
5-Cl
5-CN
6-Me
6-Cl
H
2-Cl
3-Cl
4-Cl
3-CF3
3-OMe
3-F
NA
>10,000
>10,000
NA
NA
90
60
NA
262
NA
NA
NA
4744
390
53
6g
H
5-CN
3a
Table 3
Scanning of aryl ring in ether-linked series
O
4
3
3
2
5
4
N
N
NC
NH
R
R
O
O
a
N
20
6
N
N
N
N
N
N
NC
NH2
O
b
N
2a-f
7a-f
19
N
N
NC
NC
NC
NC
NH
Compound
R
FLIPR rmGlu5
IC50 (nM)
21
Cl
2
H
2498
>10,000
316
1548
NA
5897
NA
NA
O
O
O
2a
2b
2c
2d
2e
2f
7a
7b
7c
7d
7e
7f
4-Cl
5-CN
5-Cl
5-Ac
6-Cl
6-NH2
H
N
N
N
N
14
23
N
N
N
c
X=Cl, d
F3C
X
O
O
O
(X=I) e,d
N
N
2-CN
3-Me
3-Cl
3-CN
4-CN
NA
N
24
N
769
22 X = Cl, Br, I
332a
451a
1612
NC
(X=Br) d,f
O
N
a
N
Compounds displayed partial negative allosteric modulatory activity, with max
% inhibition between 60% and 70%.
25
Scheme 2. Reagents and conditions: (a) pyrazolo[1,5-a]pyridine-2-yl carbonyl
chloride, pyridine, rt, 34%; (b) 1,5-dimethyl-1H-pyrazole-3-carbonyl chloride,
pyridine, RT, 45%; (c) NXS, MeCN, 80–100 °C, 55–95%; (d) 5-cyano-2-methyl
pyridine, KHMDS, THF, À78 °C to rt, 5–14%; (e) MeO2CCF2SO2F, CuI, DMF, 100 °C,
68%; (f) CuI, KI, NaCN, MeNHCH2CH2NHMe, PhMe, 150 °C, 17%.
alternative pathway. Ring opening of delta lactone 16 with ethyl
acetate in the presence of LiHMDS, followed by hydrazine conden-
sation gives 3-hydroxy-(5-butan-4-ol)-pyrazole 17. This product
was cyclized to give 2-hydroxy-4,5,6,7-tetrahydropyrazolo[1,5-
a]pyridine 18. Alkylation of this hydroxyl group led to ether-linked
products 2, 2a–f and 7a–f.
Table 5
Investigation of substitution and modulation of the pyridine
ring was undertaken for the direct linker, ether linker, amide linker
and ketone linker (Tables 2–4). A variety of substitution of the pyr-
idine ring, along with switching from pyridine to substituted phe-
nyl ring, is tolerated. For the direct-linked and ether linked
compounds, the m-chloro-phenyl analogue such as in compound
6c led to greater potency while the 5-cyano pyridinyl left hand side
as in 2b is preferred. This SAR observed in the pyridinyl series can
be transferred to the amido linker subseries. Interestingly, certain
of the ether linked compounds bearing a 3-substituted phenyl
ring displayed only partial modulatory activity (7d, 7e). From
Substitution of the 3-position of 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine
Compound
20
21
23
24
25
FLIPR rmGlu5
IC50 (nM)
1806
2620
108
7178
5377
comparison of 2a, 2e and 7d it can be seen that SAR was not en-
tirely transferable between phenyl and pyridyl left-hand side. For
the amide-linked compounds, 5-cyano and 5-fluoro substituted
pyridines were most favored in terms of potency (4h, 4i), whilst

4526
G. Duvey et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4523–4527
Table 6
[3H]MPEP binding and ADMET profiling of key compounds
Cpd ID
1
6c
2b
4h
4i
3a
21
23
FLIPR rat mGlu5 IC50 (nM)a
1292
1180
118/97
0.17/0.18
>10 all
isoforms
77
401
384/386
0.11/0.02
<0.75 on 1A2
316
299
42/85
0.16/0.10
>10 all
isoforms
90
60
166
53
139
108/46
0.08/0.03
>10 all
isoforms
2620
n.d.
<5/<5
n.d.
108
n.d.
649/79
n.d.
>10 all
isoforms
Rat cortex [3H]MPEP binding (nM)
151
CLint Rat/Human (
Solubility (mg/mL)c pH 1/pH 7.4
CYP inhibition ( M)
3A4/2C9/2D6/1A2
l
L/min/mg.prot)
—b/83
n.d.
—b/31
0.10/0.04
>10 all
isoforms
l
>10 all
isoforms
n.d.
a
b
c
All compounds displayed full (>99%) inhibition of mGlu5 at 30
Compound unstable in rat plasma matrix.
Kinetic solubility measurement using DMSO solution.
lM.
many of the other changes attempted resulted in complete loss of
activity; such moieties and steep SAR and are often observed in
mGlu5 receptor NAM chemical series, with fluoro-, chloro- and cy-
ano-substituted pyridines being regularly found amongst the most
potent molecules in other mGlu5 receptor NAM series.3–5 This fur-
ther reinforces the likely overlap of pharmacophoric space be-
tween this series and those of the acetylene-containing series.
Transfer of these active motifs to the ketone linker resulted in
the most potent compound in this series, 3a.
Finally, investigation of the 4,5,6,7-tetrahydropyrazolo[1,5-
a]pyridine core was conducted (Scheme 2, Table 5). Substitution
of the 3-position of the 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine
for ketone linked compounds showed that whilst substitution of
this position was tolerated, a general trend towards decreased
activity was observed for such compounds (23–25). It is also
shown that aromatization or ring opening of the 6 membered sat-
urated ring of the 4,5,6,7-tetrahydropyrazolo[1,5-a]pyridine is det-
rimental to potency (20, 21).
Plasma exposure of 2b (Phase II)
Dose (mg/kg)
1
3
10
46
30
501
Mean plasma exposure (ng/mL)
2.4
5.3
Figure 2. Effect of 2b (1, 3, 10 and 30 mg/kg in 80% PEG400, p.o., 3 mL/kg, 30 min
pre-treatment time) in the formalin test in Sprague–Dawley rats (n = 10/group).
Each point represents the observed mean (+SEM). ⁄⁄p <0.01 compared with 80%
PEG400 in Phase II.
Rat cortex binding studies were performed on key compounds
using known mGlu5 allosteric ligand [3H]MPEP.6 Similar levels of
potency, were seen between functional activity in the FLIPR Ca2+
release assay (Table 6), suggesting that this series of compound
occupies or perturbs the classical allosteric MPEP site onmGlu5.
Profiling of key compounds from this SAR investigation in pre-
liminary in vitro ADME screens showed that the amide, keto and
direct branched linker compounds suffered significant metabolic
instability in rat microsomes (Table 6). Furthermore, certain
amide-linked compounds were shown to be unstable in the assay
medium, suggesting rapid hydrolysis of the amide bond. Stability
of the compounds profiled in human microsomes suggests discrep-
ancy between the two species, with most compounds, amide-
linked included, being significantly more stable in human. The
much improved stability of 21 demonstrates that the saturated
ring of the core heterocycle is likely a major site of metabolism.
Only compound 2b of the ether linked compounds displayed a suf-
ficiently acceptable in vitro intrinsic microsomal clearance in rat to
progress to PK studies. All compounds displayed good solubilities
and with the exception of direct-linked compound 6c there were
no recurring major flags in CYP inhibition on 4 major isoforms.
3 mg/kg iv administration of compound 2b to rat showed
this compound to have a high clearance and moderate volume of
distribution resulting in a short half-life (Table 7). 30 mg/kg sc with
continuous CSF sampling suggested a one compartmental behav-
iour between plasma and CSF, with a moderate-high CSF/plasma
ratio similar to the in vitro measured plasma free fraction (rat
PPB fu = 32%). Therefore, it was decided to profile compound 2b
in the biphasic rat formalin model of nociception.8 mGlu5 receptor
is implicated in the processing of nociceptive behavior,9 and
mGlu5 receptor NAM MTEP has been shown to be active in reduc-
ing nociceptive behavior induced by formalin.10 Indeed, orally
dosed compound 2b (1, 3, 10, 30 mg/kg) showed dose dependent
reduction of nociceptive behaviour in Phase II (1 h post injection)
but not Phase I of the formalin test (Fig. 2), with significant effect
being observed at 30 mg/kg dose, the maximum effect being of
equal magnitude to that seen with positive control MTEP. Terminal
plasma concentrations of compound 2b showed dose-proportional
exposure of the compound, with only the 30 mg/kg dose demon-
strating an exposure of sufficient magnitude for compound 2b to
be exposed at levels close to the IC50 value in the brain.11 This con-
firms that compound 2b is orally bioavailable and strongly sug-
gests the compound is capable of blocking mGlu5 receptor
in vivo. It is postulated that the anti-nociceptive trend observed
at each dose may be due to exposure of the compound during
the early stages of the experiment, and that mGlu5 blockade
has a delayed-response anti-nociceptive effect in the formalin test
in rat.
Table 7
Pharmacokinetics of 2b in rats
Rat PK 3 mg/kg iva
Rat PK 30 mg/kg scb
CL (mL/min/kg)
Vss (L/kg)
72
1.1
0.2
2896
416
14.4
T
(h)
½
CMax (ng/mL) Plasma
CMax (ng/mL) CSF
CMax CSF/Plasma (%)
In conclusion, we have identified a novel mGlu5 receptor NAM
chemotype with postulated overlap with the MPEP/MTEP pharma-
cophoric space. Optimization of each area of the hit molecule re-
sulted in identification of several sub-100 nM compounds.
Molecules from this series are highly soluble and brain penetrant,
a
Results are mean of 3 animals.
Results are mean of 4 animals.
b

G. Duvey et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4523–4527
4527
and the potential for in vivo efficacy from this series has been
References and notes
demonstrated by lead molecule 2b in the rat formalin test. Further
investigation and optimization of this series is ongoing focused on
improving in vivo PK profile (improving clearance via reduction of
plasma free fraction, improving exposure) and improving potency.
1. Spooren, W.; Ballard, T.; Gasparini, F.; Amalric, M.; Mutel, V.; Schreiber, R.
Behav. Pharmacol. 2003, 14, 257.
2. Jaeschke, G.; Wettstein, J. G.; Nordquist, R. E.; Spooren, W. Expert Opin. Ther. Pat.
2008, 18, 123.
3. Rocher, J. P.; Bonnet, B.; Bolea, C.; Lutjens, R.; Le Poul, E.; Poli, S.; Epping-Jordan,
M.; Bessis, A. S.; Ludwig, B.; Mutel, V. Curr. Top. Med. Chem. 2011, 11, 680.
4. Emmitte, K. A. ACS Chem. Neurosci. 2011, 2, 411.
Acknowledgments
5. Célanire, S.; Duvey, G.; Poli, S.; Rocher, J. P. Annual Reports in Medicinal
Chemistry In Desai, M. C., Ed.; ; Academic Press Elsevier, 2012; 47, p 71.
6. See Supplementary data for assay conditions.
7. Venkatesan, A. M.; Agarwal, A.; Abe, T.; Ushirogochi, H.; Yamamura, I.; Ado, M.;
Tsuyoshi, T.; Dos, S. O.; Gu, Y.; Sum, F. W.; Li, Z.; Francisco, G.; Lin, Y. I.;
Petersen, P. J.; Yang, Y.; Kumagai, T.; Weiss, W. J.; Shlaes, D. M.; Knox, J. R.;
Mansour, T. S. J. Med. Chem. 2006, 49, 4623.
We would like to acknowledge the indispensable technical con-
tributions of V. Nhem, R. Furnari, H. Faure, A. Bessif, S. Charreton, B.
Mingard, A. Riu, P. Bazzy, and C. Pesenti, and thank M. Kalinichev
and I. Urios for data analysis.
8. Tjölsen, A.; Berge, O. G.; Hunskaar, S.; Rosland, J. H.; Hole, K. Pain 1992, 51, 5.
9. Kolber, B. J.; Montana, M. C.; Carrasquillo, Y.; Xu, J.; Heinemann, S. F.; Muglia, L.
J.; Gereau, R. W. J. Neurosci. 2010, 30, 8203.
Supplementary data
10. Sevostianova, N.; Danysz, W. Neuropharmacology 2006, 51, 623.
11. Exptrapolation of brain concentration of 2b based on observed plasma
exposure during experiment and CSF/plasma ratio from sc PK experiment
suggest brain concentration for 30 mg/kg dose would be 95 ng/mL = 375 nM.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.
06.044.