Discovery, Structure-Activity Relationship, and Antiparkinsonian Effect of a Potent and Brain-Penetrant Chemical Series of Positive Allosteric Modulators of Metabotropic Glutamate Receptor 4

Article abstract of DOI:10.1021/acs.jmedchem.7b00991

The metabotropic glutamate receptor 4 (mGluR4) is an emerging target for the treatment of Parkinson's disease (PD). However, since the discovery of its therapeutic potential, no ligand has been successfully developed enough to be tested in the clinic. In the present paper, we report for the first time the medicinal chemistry efforts conducted around the pharmacological tool (-)-PHCCC. This work led to the identification of compound 40, a potent and selective mGluR4 positive allosteric modulator (PAM) with good water solubility and demonstrating consistent activity across validated preclinical rodent models of PD motor symptoms after intraperitoneal administration: haloperidol-induced catalepsy in mouse and the rat 6-hydroxydopamine (6-OHDA) lesion model. Moreover, we also describe the identification of compound 60 a close analogue of compound 40 with improved pharmacokinetic profile after oral administration. On the basis of its favorable and unique preclinical profile, compound 60 (PXT002331, now foliglurax) was nominated as a candidate for clinical development.

Full text of DOI:10.1021/acs.jmedchem.7b00991

Subscriber access provided by Columbia University Libraries
Article
Discovery, structure-activity relationship and anti-parkinsonian
effect of a potent and brain-penetrant chemical series of positive
allosteric modulators of metabotropic glutamate receptor 4
Delphine Charvin, Vincent Pomel, Millan Ortiz, Mélanie Frauli, Sophie Scheffler, Edith Steinberg, Luc Baron,
Laurène Deshons, Rachel Rudigier, Delphine Thiarc, Christophe Morice, Baptiste Manteau, Stanislas Mayer,
Danielle Graham, Bruno Giethlen, Nadia Brugger, Gaël Hedou, François Conquet, and Stephan Schann
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00991 • Publication Date (Web): 13 Sep 2017
Downloaded from http://pubs.acs.org on September 14, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery, structureꢀactivity relationship and antiꢀ
parkinsonian effect of a potent and brainꢀpenetrant
chemical series of positive allosteric modulators of
metabotropic glutamate receptor 4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
∗ #
#
#
†
†
Delphine Charvin, Vincent Pomel, Millan Ortiz, Mélanie Frauli, Sophie Scheffler, Edith
†
†
†
‡
‡
Steinberg, Luc Baron, Laurène Deshons, Rachel Rudigier, Delphine Thiarc, Christophe
‡
†
†
§
‡
Morice, Baptiste Manteau, Stanislas Mayer, Danielle Graham, Bruno Giethlen, Nadia
∗†
§
§
#
Brugger, Gaël Hédou, François Conquet, Stephan Schann
#
†
‡
§
Prexton Therapeutics, 14 Chemin des Aulx, 1228 PlanꢀlesꢀOuates, Geneva, Switzerland
Domain Therapeutics, 850 Boulevard Sébastien Brant, 67400 Illkirch ꢀ Strasbourg, France
Prestwick Chemical, 220 Boulevard Gonthier d'Andernach, 67400 Illkirch ꢀ Strasbourg, France
EMD Serono, 45A Middlesex Turnpike, Billerica, MA 0182, United States
Corresponding Authors
*
DC: phone, +41 79 833 11 32; eꢀmail, Delphine.Charvin@prextontherapeutics.com
SS: phone, +33 390 40 61 50; eꢀmail, sschann@domaintherapeutics.com
*
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
The metabotropic glutamate receptor 4 (mGluR4) is an emerging target for the treatment of
Parkinson’s disease (PD). However, since the discovery of its therapeutic potential, no ligand has
been successfully developed enough to be tested in the clinic. In the present paper, we report for
the first time the medicinal chemistry efforts conducted around the pharmacological tool (ꢀ)ꢀ
PHCCC. This work led to the identification of compound 40, a potent and selective mGluR4
positive allosteric modulator (PAM) with good water solubility and demonstrating consistent
activity across validated preclinical rodent models of PD motor symptoms, after intraperitoneal
administration: haloperidolꢀinduced catalepsy in mouse and the rat 6ꢀhydroxydopamine (6ꢀ
OHDA) lesion model. Moreover, we also describe the identification of compound 60 a close
analog of compound 40 with improved pharmacokinetic profile after oral administration. Based
on its favorable and unique preclinical profile, compound 60 (PXT002331, now foliglurax) was
nominated as candidate for clinical development.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Parkinson's disease (PD) is a chronic neurodegenerative disorder affecting more than six
1
million people worldwide . It results from the loss of dopamine neurons in the basal ganglia
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
BG), a brain region responsible for the control of motor functions. Available treatments, mainly
2
based on dopamine replenishment, are only effective at managing early PD symptoms . As the
3
ꢀ
disease progresses, these treatments become less effective and produce debilitating side effects
4
.
Amongst them, dyskinesia (or LꢀDopaꢀInduced Dyskinesia ꢀ LID), characterized by
involuntary movements appearing after several years of LꢀDopa therapy, represents a serious
challenge for lateꢀstage PD treatment.
Over the past decade, novel therapeutic strategies targeting nonꢀdopaminergic transmissions
5
have emerged . Amongst them, modulation of presynaptic glutamate receptors such as
metabotropic glutamate receptor 4 (mGluR4) has proven to be a promising approach to
6
ꢀ7
normalize the BG circuitry . To date, eight mGluRs have been cloned and classified in three
groups according to their sequence homologies, pharmacological properties and signal
transduction mechanisms: group I includes mGluR1 and mGluR5, group II mGluR2 and
8
mGluR3 and group III mGluR4, mGluR6, mGluR7 and mGluR8 . Given its expression in
desired regions of the BG motor circuit, its presynaptic localization and its physiological
function to decrease neurotransmitter release, the mGluR4 receptor has received much interest as
a therapeutic target for a LꢀDopaꢀsparing strategy in PD. Indeed, mGluR4 localizes
presynaptically at striatopallidal fibers where its activation circumvents dopamine action via the
9
indirect pathway .
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Initial modulation of mGluR4 was made with nonꢀselective group III mGluR agonists
activating mGluR4, mGluR6, mGluR7 and mGluR8) such as LꢀAP4 or ACPTꢀI (Figure 1).
(
These highly polar pharmacological tools were used to demonstrate the potential of group III
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
10ꢀ11
activation in several in vitro and in vivo paradigms of PD
. More selective tools came later
1
2
13ꢀ14
with positive allosteric modulators (PAMs) such as (ꢀ)ꢀPHCCC , ADX88178
, Lu AF21934
1
5
16ꢀ17
or VU0418506
(Figure 1). Allosteric modulation offers several advantages over
orthosteric approaches such as the possibility to obtain subtypeꢀselectivity and the access to
druggable compounds more amenable to medicinal chemistry strategies. However, despite more
than a decade of chemical optimization, none of these mGluR4 PAMs has progressed in the
clinic.
In this paper, we report for the first time the medicinal chemistry work based on (ꢀ)ꢀPHCCC,
which led to a novel chromenoneꢀoxime series with improved properties. We demonstrate that
this effort led to the identification of compounds with an improved profile, up to the discovery of
1
8
compound 60 (PXT002331) , which is currently in clinical development. We also describe
more in depth compound 40 (6ꢀ(2ꢀMorpholinꢀ4ꢀylꢀethoxy)ꢀ2ꢀpyrrolo[1,2ꢀc]pyrimidinꢀ3ꢀylꢀ
1
9
chromenꢀ4ꢀone oxime ꢀ PXT001687) , as a representative candidate of the series with
significant antiꢀparkinsonian activity in preclinical rodent models of Parkinson's disease.
ACS Paragon Plus Environment

Page 5 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
OH
N
HOOC
HOOC
NH2
O
O
P
H
OH
OH
HN
S
HO
N
H
N
N
NH2
N
COOH
O
N
N
H
O
L-AP4
ACPT-1
(-)-PHCCC
ADX88178
OH
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H
F
H
H
N
N
N
Cl
Cl
H
N
N
N
Cl
O
O
H
H N
O
N
2
O
Lu AF21934
VU0418506
VU0359516
Figure 1. Structure of reference mGluR4 ligands.
SAR from (ꢀ)ꢀPHCCC to compound 40
(
ꢀ)ꢀNꢀphenylꢀ7ꢀ(hydroxyimino)cyclopropa[b]chromenꢀ1aꢀcarboxamide ((ꢀ)ꢀPHCCC ꢀ Figure 1)
1
2
is the first partially selective mGluR4 PAM described in the literature . It constitutes an
unprecedented pharmacological tool that, in its racemic form, was used by multiple teams to
demonstrate the therapeutic potential of mGluR4 potentiation in several disease paradigms
2
0ꢀ21
22
23
24ꢀ25
including Parkinson's disease
, anxiety , medulloblastoma , pain
and multiple
26
sclerosis . However, (ꢀ)ꢀPHCCC does not constitute a good drug candidate as it suffers from
weak micromolar potency (EC50 = 2.25 µM on mGluR4 with the racemic form), residual
mGluR1 NAM activity, and poor brain penetration forcing pharmacologists to use central
administration or toxic DMSO vehicles. Several medicinal chemistry explorations were
conducted around (ꢀ)ꢀPHCCC in order to improve its properties, but limited successes were
reported and this scaffold was long considered as “flat” with any chemical modifications
2
7ꢀ29
resulting in a loss of mGluR4 PAM activity
. The only optimized derivative found was
3
0
VU0359516 (Figure 1), a 2ꢀpyridyl amide analog described with a 10ꢀfold improvement in
30
potency and no longer side activity on mGluR1 .
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Despite these disappointing reports, we decided to reꢀexamine the potential of (ꢀ)ꢀPHCCC
scaffold by checking importance of the cyclopropane and amide moieties. Indeed, we considered
that removal of the chiral centers of (ꢀ)ꢀPHCCC and replacement of the amide rightꢀhand side
were key in order to simplify the scaffold for an optimization effort and to increase chances of
brain penetration. Analog 1 (Table 1) was synthesized and was found to retain some level of
PAM activity on mGluR4 (EC =14.2 µM). Interestingly, this molecule 1 was previously
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
reported by a team from Vanderbilt University as being inactive at 30 µM in a Gqi5 functional
2
8
test (compound 1i in supplementary information of reference ) whereas it behaves as a full
PAM in our chimeric Gi/Gq (GqTOP) assay. This illustrates the differences that can be measured
3
1
depending on the in vitro model systems used and underlines the importance of validating the
compounds activity in animal models in the early steps of an optimization program.
Syntheses and all analytical characterizations of the different compounds described in this article
are detailed in the Supporting information (SI).
Encouraged by the result obtained with 1, we decided to further explore the styryl moiety and
prepared a library of rigidified analogs (see Table 1). First, rigidification with a naphtyl or
several 5ꢀ6 and 6ꢀ6 bicyclic heterocyclic groups lead to very weak or inactive analogs 2ꢀ8.
However, quinolinyl 9ꢀ10 and isoquinolinyl 11 were active and position of the nitrogen alpha to
the link with the chromenone oxime central core seems to be important to reach subꢀmicromolar
potency, compound 11 showing a 7ꢀfold improved potency compared with PHCCC. This was
further confirmed with good levels of potency obtained with other 6ꢀ6 and 6ꢀ5 bicyclic
heterocycles 12ꢀ19 all having a nitrogen atom positioned similarly as in isoquinolinyl 11, and
with the absence of activity of compound 20, isomer of the active compounds 14 or 17.
Interestingly, although a bicyclic heterocyle with a “rightly” positioned nitrogen is crucial for
ACS Paragon Plus Environment

Page 7 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
activity, it seems to be sensitive to decoration as the methyl substitution of 21 prevent mGluR4
PAM activity.
Table 1: Exploration of the aromatic substituents of chromenone oxime scaffold
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OH
N
O
Aro
Aromatic mGluR4 PAM
Aromatic
group
mGluR4 PAM
Compounds
a
Compounds
a
group
EC (µM)
EC (µM)
5
0
50
*
*
*
*
*
*
PHCCC
/
2.25 ± 1.07
11
12
13
14
15
16
17
0.31 ± 0.18
1.27 ± 0.31
0.24 ± 0.07
0.11 ± 0.03
0.13 ± 0.09
0.20 ± 0.14
0.28 ± 0.08
N
*
*
N
1
2
3
4
5
6
14.2 ± 3.75
NA
N
N
N
N
N
N
*
>30
O
S
N
*
*
*
NA
S
S
>30
N
H
*
*
N
NA
N
N
N
N
*
7
8
NA
NA
18
19
0.86 ± 0.32
N
*
*
N
N
0.050 ± 0.08
N
N
H
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
*
*
*
N
9
3.96 ± 1.03
6.99 ± 0.66
20
21
NA
NA
N
N
N
N
N
*
10
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
a
Values are the mean (± SD) of a minimum of 3 independent experiments. NA: not active.
In parallel to the study of aromatic substituents of the chromenone oxime described in Table 1,
we decided to expand the exploration of the core scaffold with analogs of the representative
compound 11 (Table 2). First, we observed that modifications of the oxime function, changed for
a keto function in 22 or substituted by a methyl group in 23, resulted in inactive molecules. As a
matter of fact, this observation correlates with the results obtained by Williams et al showing that
3
0
modifications of the oxime moiety of VU0359516 always result in loss of activity . Secondly,
we explored the influence of small substituents on the core group. We observed that substitution
of position 3, close to the oxime moiety, with a methyl group (compound 24) was not tolerated.
This was not the case for position 6 (compounds 25 to 29) and position 7 (compounds 30 and 31)
that do tolerate small substituents ranging from methyl to bromine atom. However, position 8
appeared to be more sensitive as the chloro substituted 32 was inactive on mGluR4. We then
explored the possibility to add larger groups at positions 6 and 7 that are chemically more
accessible. All the tested bulkier groups resulted in compounds still potent on mGluR4 with EC50
ranging between 0,75 µM to 9,8 µM (compounds 33 to 38). Both polar and nonꢀpolar groups
were introduced illustrating the fact that positions 6 and 7 can constitute a handle to improve
different properties of the molecules.
Table 2: Exploration of the core chomenone oxime of compound 11
ACS Paragon Plus Environment

Page 9 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
N
OH
O
N
O
5
6
3
7
O
O
8
N
N
N
22
23
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
mGluR4
mGluR4
Compounds Substituents
PAM EC50
Compounds
Substituents
PAM EC50
a
a
(µM)
(µM)
11
22
23
H
/
0.31 ± 0.18
NA
30
31
32
33
34
7-Br
7-OMe
1.72 ± 0.11
1.16 ± 0.31
NA
/
NA
8-Cl
2
2
4
5
3-Me
6-Me
NA
6-cyclopropyl
1.01 ± 0.07
0.75 ± 0.40
0.98 ± 1.08
6-O-(CH ) -OMe
2 2
6
- O-(CH ) -
2
2
2
6
6-CF
3
2.47 ± 1.72
1.63 ± 1.59
35
36
9.84 ± 4.21
5.26 ± 2.85
methylpiperazine
6
- NH-(CH ) -
2
3
27
6-OCF
3
methylpiperazine
7-O-(CH -OMe
7- (CH -phenyl
2
2
8
9
6-Br
0.69 ± 0.04
1.39 ± 0.56
37
38
)
2
2
3.01 ± 2.84
1.74 ± 0.76
6-OMe
)
2 2
a
Values are the mean (± SD) of a minimum of 3 independent experiments. NA: non active.
Based on the observations made with compound 11 analogs, we next investigated the effects of
introductions of polar groups on position 6 of compound 14 that showed mGluR4 PAM activity
with a potency of 110 nM. The objective of this investigation was to increase the solubility of 14
that was rather poor (0.2 µM in MilliQ water) making the in vivo characterization of this
molecule challenging. It was found that, to a greater extent than with compound 11, introduction
of polar groups was well tolerated and even slightly improved the mGluR4 PAM activity with
compound 40 that seemed to bear an ideal substituent in terms of polarity and / or basicity (Table
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
). Water solubility was also clearly improved with a 53 000 ꢀfold increase for the HCl salt of
compound 40 compared with 14. However, all polar groups were not similarly tolerated as
illustrated with compounds 42 and 43 that exert more than 10ꢀfold decrease of activity compared
with compound 14 (Table 3).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 3: Introduction of polar groups on molecule 14
OH
N
R
O
N
N
mGluR4 PAM EC
5
0
b
Compounds
14
R
a
Water solubility (µM)
(µM)
H
0.11 ± 0.03
0.16 ± 0.08
0.046 ± 0.018
0.15 ± 0.09
2.34 ± 0.67
1.71 ± 0.98
0.2
O
O
3
4
4
4
4
9
0
1
2
3
*
16.3
10 600
ND
O
N
*
O
O
N
*
F
F
O
N
*
ND
O
N
*
ND
N
a
b
Values are the mean (± SD) of a minimum of 3 independent experiments. Thermodynamic
solubility measured in milliꢀQ water. ND: not determined.
In vitro properties of compound 40
ACS Paragon Plus Environment

Page 11 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Having identified compound 40, a potent mGluR4 PAM with clearly improved water solubility,
we further extended its characterization in vitro. Its mode of action on mGluR4 was studied
using the human recombinant receptor expressed in a transfected human cell line (HEK 293
cells). These cells were also transfected with a plasmid encoding a chimeric G protein that
allowed redirection of the activation signal to intracellular calcium pathway. Receptor activity
was then detected by changes in intracellular calcium, measured using a fluorescent calciumꢀ
sensitive dye (Fluo4AM, Molecular Probes). Agonist and PAM activities of compound 40 were
consecutively evaluated on the same cell plate. Agonist activity was first tested for 10 minutes
with the addition of the compound alone in the cell media. Cells were then stimulated by
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
addition of glutamate at a concentration that resulted in 20% of the maximal effect (EC ) and
20
fluorescence was recorded for an additional 3 minutes. Agonist or PAM activities were evaluated
in comparison to, respectively, basal signals evoked either by the buffer or an EC glutamate
20
concentration. For potency and efficacy determination, a concentrationꢀresponse test of
compound 40 was performed. In addition, the mode of action of compound 40 was further
characterized in experiments assessing the shift of glutamate concentrationꢀresponse curves. In
this alternative setting, 10 concentrations of glutamate (ranging over 4.5 logs) were tested alone
or in presence of 4 concentrations of compound 40 (ranging from 0.1 to 3 µM). Potency, of
glutamate alone or in presence of compound 40, was determined and used to calculate the foldꢀ
increase in the apparent affinity of glutamate. Results showed that in these cell lines, compound
40 had a low agonist activity, with an average stimulation of the receptor of 19 ± 12% at 3 µM.
However, as a PAM, compound 40 potentiated the response of human mGluR4 to glutamate with
a potency of 46 ± 18 nM and amplified the effects of the EC glutamate concentration to 77 ± 2
2
0
%
of the maximal response to glutamate (Figure 2A). For comparison, in the same cellular assay,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
PHCCC exerts mGluR4 PAM effects with a potency of 2.25 ± 1.07 µM, which represents an
approximately 50ꢀfold improvement in potency for compound 40. It should be noted here that no
clear structureꢀactivity relationship was observed with regards to efficacy modulation as most of
the active PAMs described in this study resulted in an efficacy between 75% and 100% of the
maximal response to glutamate. Moreover, increasing concentrations of compound 40
progressively produced a leftward shift in the glutamate concentrationꢀresponse relationship and
increased the potency of glutamate for mGluR4 by approximately 10ꢀfold at 3 µM (Figure 2B).
Compound 40 had a similar profile in a distinct cell assay that measured the cAMP response
following mGluR4 stimulation. Indeed, in this second cellular assay, compound 40 potentiated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the response of human mGluR4 to a glutamate EC with a potency of 70 nM (compared to
20
PHCCC, which had a potency of 2.5 µM – data not shown) and induced a leftward shift of the
glutamate concentrationꢀresponse curve. In cell lines expressing the rat mGluR4, compound 40
was slightly more potent with an EC of 27 ± 2 nM and the responses were enhanced to 75 ± 2
50
%
of the maximal response to glutamate (SI Figure S1).
ACS Paragon Plus Environment

Page 13 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2: In vitro properties of compound 40. A, Agonist and PAM activities of compound 40
were consecutively evaluated in HEK293 cells expressing human mGluR4. In these cell lines,
compound 40 potentiated the increases in intracellular calcium concentrations induced by an
EC20 glutamate (PAM effect) but had a minor agonist activity in absence of glutamate. B,
Increasing concentrations of compound 40 induced a 10ꢀfold leftward shift of the glutamate
concentrationꢀresponse curves in these cell lines. A, B, Each point represents the mean (± SD) of
duplicate determination from a representative experiment. C, Selectivity profile of compound 40
among mGluRs. Compound 40 was tested at 1 ꢁM on HEK293 cells expressing each of the
human mGluR, alone or in presence of either an EC or an EC glutamate. The corresponding
20
80
glutamate concentrations were determined for each receptor subtype. LꢀAP4 has been used
instead of glutamate on hmGluR7. Each bar represents the mean (+ SEM) of activities measured
in at least two (when not active) or three experiments.
Then, the selectivity of compound 40 for mGluR4 was investigated using human cell lines
(HEK 293) expressing each of the other human mGluRs as well as by measuring inhibition of
orthosteric ligand binding on the other glutamate receptors, namely NMDA, AMPA and kainate,
on membranes from rat cerebral cortex. Compound 40 had no effect on iGluRs, neither on group
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
I (mGluR1 and 5) nor on group II (mGluR2 and 3) mGluRs up to 10 µM, the highest
concentration tested. Among group III mGluRs, compound 40 showed 13ꢀtimes higher
selectivity for mGluR4 than for mGluR6, was not active on mGluR7 and had a very partial
agonist activity on mGluR8 with an average stimulation of the receptor of 23 + 7% at 10 µM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
(Figure 2C). It has to be noted that expression of mGluR6 is strictly restricted to the retina .
Thus, it is not expected that an activity of compound 40 on mGluR6 may be confounding with a
potential antiꢀparkinsonian effect in animal models. Finally, not only the potency but also the
selectivity for mGluR4 has been improved compared with (ꢀ)ꢀPHCCC since compound 40 has no
longer any activity on mGluR1.
Additionally, compound 40 was also evaluated on other targets of importance for Parkinson's
disease. Compound 40 at 10 µM showed no functional activity (neither agonist, nor PAM and
antagonist activity) on D1, D2 and A2A receptors. No effect was observed up to 10 µM on
L
COMT, MAOꢀA and MAOꢀB, reducing the risk for interaction with LꢀDopa metabolism in
animal models and in Parkinson's disease patients. Finally, activity of compound 40 was also
assessed on a panel of diverse kinases (e.g. Flt3, GSK3beta, IRAK4, JAK3, TAK1). Except a
64%ꢀinhibition of Flt3, no kinase inhibition was detected at 10 µM. Altogether, these results
demonstrate that compound 40 is both potent and selective for mGluR4.
Pharmacokinetic properties of compound 40
We next evaluated the pharmacokinetic properties of compound 40 in rat and mouse. Following
intravenous administration, compound 40 had a very similar pharmacokinetic profile in both
species, with a high clearance (113 and 117 mL/min/kg in rat and in mouse, respectively), a high
ACS Paragon Plus Environment

Page 15 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
volume of distribution (6.5 and 6 L/kg in rat and mouse, respectively) and a short halfꢀlife (44
and 12 min in rat and mouse, respectively) (Table 4). In both mouse and rat, compound
concentrations were 3ꢀfold higher in the brain than in plasma indicating that the compound is
CNS penetrant and has a preferential exposure in brain (Table 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4: Mean PK parameters of compound 40 following intravenous administration at
1
mg/kg in rat, and at 2 mg/kg in mouse.
Cmax
t1/2
AUC0ꢀ∞
CLp
Vd
Compound
Brain/Plasma
40
(
ng/mL)
(min)
(mL/min/kg)
(L/kg)
(
h.ng/mL)
1
2
mg/kg i.v.
(rat)
444
44
166
113
6.5
3.3
+
93 a
+ 9 a
+ 36 a
+ 29 a
+ 0.75 a
+ 0.04 a,c
mg/kg i.v.
753
12
+ 1.8 a
269
+ 39.8 a
123
+ 18.2 a
6
3.6
+ 0.3 b,d
80 b
+ 0.8 a
(
mouse)
+
a
b
c
ND: not determined. Values are the mean (± SEM) of 3 or 2 animals. Ratio calculated at one
hour or 30 min postꢀdose.
d
Compound 40 was administered intraperitoneally at 10 and 30 mg/kg or orally at 10 mg/kg in
rats (Table 5). Higher than doseꢀproportional increases were observed in plasma Cmax and AUC0ꢀ
following intraperitoneal administration from 10 to 30 mg/kg and plasma concentrations were
∞
at maximum 15 min after administration. Measurements made after oral administration of 10
mg/kg revealed that the compound has a medium oral bioavailability in rats (F = 54%). All
animals were exposed to compound 40 at the first sampling time, 15 min postꢀoral dose,
confirming drug absorption, with Cmax reached 20 min after administration (Table 5). Following
a single dose of compound 40, complete elimination was observed, with concentrations between
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
% and 10% of the maximum concentrations observed at 24 hours postꢀdose. The mean halfꢀlife
was 7.3 hours with an oral dose of 10 mg/kg. The pharmacokinetics of compound 40 in mouse is
qualitatively similar to those in rat.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 5: Mean PK parameters of compound 40 in rat following intraperitoneal or oral
administration, and in mouse following oral administration
Plasma
Cmax
Plasma
Tmax
Plasma
T1/2
Plasma AUC0ꢀinf [Brain]30 min
Brain
Brain
Compound
40
a
(
h.ng/mL)
(ng/g)
/Plasma
/rEC50
(ng/mL)
(min)
(h)
10 mg/kg
i.p. (rat)
1 498
15
+ 0
4.0
+ 0.16
2 245
+ 269
5 647
+ 765
3.1
+ 0.3
3.6
14.5
0.3
+
62
30 mg/kg
i.p. (rat)
9 324
745
15
+ 0
4.5
+ 0.7
11 349
+ 1 433
22 792
+ 1 111
3.1
+ 0.05
+
10 mg/kg
p.o. (rat)
135
20
+ 5
7.3
+ 1.7
903
+ 223
454
+ 28
2.3
+ 0.4
+
39
b
1
0 mg/kg
229
30
+ 0
ND
ND
390
+ 78
1.7
+ 0.34
0.2
p.o. (mouse)
+
50
ND: not determined. rEC : EC measured in cell lines overexpressing the rat mGluR4 receptor.
Values are the mean (± SEM) of 3 rats or 2 mice. Ratio calculated at 30 min postꢀdose. The
mouse EC was assumed to be the same as the rEC .
5
0
50
a
b
5
0
50
This PK profile, together with the high brain protein binding measured in rat brain homogenate
(99.24%), revealed that brain exposures are greater than the in vitro EC50 values, after
intraperitoneal administration of 10 and 30 mg/kg but not after oral administration of 10 mg/kg
in rat (Table 5). Assuming a linear PK profile after oral administration, a dose of 100 mg/kg
ACS Paragon Plus Environment

Page 17 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
would have to be administered in rats in order to reach brain exposures greater than EC and
5
0
expect to see a pharmacodynamic effect in vivo.
Based on these in vitro and in vivo parameters, compound 40 was chosen for further
characterization in rodent models of Parkinson's disease, and SAR efforts were pursued in order
to improve the oral PK profile of the series.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Parkinson's Disease (PD) models
As a preliminary model of Parkinson's disease motor symptoms, the ability of compound 40 to
reverse haloperidolꢀinduced catalepsy was assessed in mouse. Mice received an injection of
haloperidol (0.5 mg/kg, ip) to induce catalepsy and one hour later, while catalepsy was present,
animals were administered an i.p. dose of compound 40. Catalepsy was then assessed 40 minutes
after compound 40 dosing. As shown in Figure 3A, compound 40 produced a doseꢀdependent
reversal of catalepsy, with significant effects obtained between 3 and 30 mg/kg i.p., which is in
accordance with the prediction based on the PK profile of compound 40. In this study, we
measured the compound exposure in mouse brain at the end of the experiment to determine the
in vivo EC . Using the free fraction previously determined in rat brain (0.76%), we extrapolated
50
the concentrations of free compound 40 in brain from the measured total brain concentrations (SI
Table S1). The corresponding PK/PD relationship is shown in Figure 3B and indicates that the in
vivo EC in brain is around 5 nM, which corresponds well with the in vitro EC on rat mGluR4
5
0
50
(27 nM), inferring that the effect seen in this model is mGluR4ꢀrelated.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3: Reversal of haloperidolꢀinduced catalepsy in mice by compound 40. Mice were
administered haloperidol (0.5 mg/kg, ip) and one hour later, when catalepsy was present, they
received an ip dose of compound 40 or vehicle. Catalepsy was assessed 40 minutes after
compound 40 dosing. A, Each bar represents the mean (+ SD) latency to move on a vertical grid
(immobility time). *p < 0.05 when compared to Vehicle (one way ANOVA followed by Dunn's
multiple comparison) (n=8/group). B, Relationship between efficacy in reversal of haloperidolꢀ
induced catalepsy and compound 40 unbound fraction exposed in the brain.
We next evaluated the effects of compound 40 in a more elaborated rodent model of Parkinson's
disease motor symptoms. Compound 40 was administered to rats that had previously undergone
a bilateral lesion of the striatum induced by 6ꢀhydroxydopamine (6ꢀOHDA). In most models
33ꢀ34
utilizing the neurotoxin 6ꢀOHDA, animals are dopamine depleted only unilaterally
. While
dopamine loss in human PD can be asymmetrical at the earliest stages of the disease, it
3
5ꢀ36
ultimately results in dopamine loss in both hemispheres
. In order to select a context that is
closer to the human condition, we chose a model where dopamine is being depleted bilaterally.
Male SpragueꢀDawley rats received bilateral injections of 6ꢀOHDA into the striatum, which
induced parkinsonian motor deficits indicated by reduced spontaneous activity and rearing in the
open field arena (Figure 4a, SI Figure S2a). Thus, active time, slow activity (i.e. nonꢀstereotyped
activity) and rearing time were significantly reduced to 64%, 66% and 68% respectively, by the
ACS Paragon Plus Environment

Page 19 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
6
ꢀOHDAꢀinduced bilateral lesion compared with preꢀsurgery (Figure S2a, "Post" versus "Pre").
When administered alone, compound 40 (30 mg/kg, i.p.) induced an increase in active time, slow
activity and rearing time compared with vehicle administration (Figure S2a and 4a). The time
course analysis reveals that effects of compound 40 on slow activity were observable during the
30ꢀ120 min postꢀadministration period (Figure 4a). Comparison of pharmacokinetics and
pharmacodynamic effects demonstrates an excellent correlation between the increase of
compound 40 exposure in brain (Figure 4b) and the efficacy in improving the slow activity
(Figure 4a). Indeed, compound 40 effects become observable from 30 min postꢀadministration
and at 30 and 60 min after i.p. dosing at 30 mg/kg, brain exposures were respectively 14.5ꢀ and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7.6ꢀ fold greater than the in vitro EC at rat mGluR4 (Figure 4b), supporting that the effect seen
50
in this model is mGluR4ꢀrelated. Assuming a linear brain/plasma ratio over time, brain exposure
would barely reach the mGluR4 EC at 120 min after administration, which is in accordance
50
with the loss of effect of compound 40 at that time point in this rat model (Figure 4a and 4b).
3
7
In a recent study, Jenkins et al.
have shown that treatment of rodents with 6ꢀOHDA may
slightly decrease the expression of mGluR4 in brain. Although we have used a different model,
where 6ꢀOHDA induced a milder bilateral lesion, it was important to do this PK/PD correlation
in order to support target engagement in the antiꢀparkinsonian effects of compound 40.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4: Antiparkinsonian effect of standꢀalone compound 40 treatment on locomotor
activity of 6ꢀOHDAꢀlesioned rats. A, Timeꢀcourse of spontaneous locomotor activity in adult
male SpragueꢀDawley rats before ("Preꢀsurgery, vehicle", dotted line) or after 6ꢀOHDAꢀinduced
striatal lesions. Lesioned rats were treated either with vehicle ("Postꢀsurgery, vehicle", grey
curve) or with compound 40 at 30 mg/kg ip ("Postꢀsurgery, Cpd 40", black curve). Each timeꢀ
point represents the mean counts (+/ꢀ SEM) for every 5 minutes. Compound 40 was administered
ip at T0. (n = 10). B, Compound 40 plasma and brain exposures in rats dosed with 30 mg/kg ip.
Brain exposure was measured 30 and 60 min after administration. Brain levels at 120 and 240
min (c) were extrapolated from measured plasma exposure by using an average brain/plasma
ratio of 3.25 as determined at 30 and 60 min. Unbound fraction of compound 40 in brain (a) was
calculated by using the brain protein binding previously measured (99.24%, i.e. 0.76% unbound
fraction). Brain unbound over rEC50 ratio (b) was then calculated by using the in vitro rat
mGluR4 EC obtained in HEK293 cells. (n=3).
50
Then, parkinsonian rats received LꢀDopa either alone or in combination with compound 40. As
expected, LꢀDopa was effective in improving motor performance, inducing a doseꢀdependent
increase in locomotor activity compared with vehicle administration (Figure 5a). Doseꢀresponse
curves performed with LꢀDopa (3, 6 and 20 mg/kg) showed that 20 mg/kg LꢀDopa represented
an optimal antiꢀparkinsonian dose, restoring the motor activity of the rats to normal levels
(Figure 5a) while the dose of 6 mg/kg LꢀDopa was selected as an ineffective subꢀthreshold dose
for use in subsequent experiments. When compound 40 was coꢀadministered with the subꢀ
threshold low dose of LꢀDopa, we observed a significant improvement in motor performance that
showed dose dependence for compound 40 (Figure 5a). Thus, the combination of 6 mg/kg Lꢀ
Dopa and 10 mg/kg compound 40 was able to fully reverse the motor deficit of parkinsonian rats
ACS Paragon Plus Environment

Page 21 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and reached the same level of efficacy as the high dose of 20 mg/kg LꢀDopa (Figure 5a).
Consequently, in this rat model of Parkinson's disease, administration of compound 40 (10
mg/kg, i.p.) allowed a decrease by 70% of the dose of LꢀDopa while maintaining the same antiꢀ
parkinsonian efficacy. As shown in Figures 5a and 5b, compound 40 induced a doseꢀ and timeꢀ
dependent antiꢀparkinsonian effect, with the highest improvement obtained at the dose of 10
mg/kg (Figure 5a), from 30 to 120 min postꢀadministration of compound 40 (Figure 5b). At the
highest dose of 30 mg/kg, the antiꢀparkinsonian effect of compound 40 in combination with 6
mg/kg of LꢀDopa has reached a plateau with a maximal efficacy equivalent to the normal activity
of preꢀlesion rats (Figure 5a).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5: Antiparkinsonian effect of the combined treatment with compound 40 and a low
subꢀoptimal dose of LꢀDopa on locomotor activity of 6ꢀOHDAꢀlesioned rats. A, Total
spontaneous locomotor activity values are presented. Data are expressed as mean of group
(+SEM) and analyzed using oneꢀway analysis of variance (ANOVA) repeated measures
followed by Dunnett's multiple comparisons. ***p < 0.001, **p < 0.01, * p < 0.05 vs Post Veh.
The dotted horizontal lines represent the mean locomotor activity of the control conditions in 6ꢀ
OHDAꢀlesioned rats: optimal LꢀDopa (upper line) and vehicle (lower line). Pre Veh: rats activity
before the surgery, after vehicle ip administration ; Post Veh: activity of the same animals after
the surgery and stabilization of the 6ꢀOHDAꢀinduced striatal lesion, after vehicle ip
administration; compound 40 ip administration. (n=10). B, Timeꢀcourse of spontaneous
locomotor activity of 6ꢀOHDAꢀlesioned rats treated either with a low subꢀoptimal dose of Lꢀ
Dopa alone ("6mg/kg LꢀDOPA", grey curve) or with addition of compound 40 at 10 mg/kg ip
("6mg/kg LꢀDOPA +10mg/kg compound 40", black curve). Each timeꢀpoint represents the mean
counts (+/ꢀ SEM) for every 5 minutes. (n = 10).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In this model consisting of a partial bilateral lesion of the striatum induced by 6ꢀOHDA, the dose
of 20 mg/kg LꢀDopa did not induce dyskinesia. However, this dose produced an increase of
rearing activity that is above normal activity measured before the lesion (SI Figure S3a). This
abnormally high rearing activity is considered as an early sign of potential dyskinesia
development in this model. Thus, it is noteworthy that none of the conditions tested with
compound 40 induced an overactive rearing behavior, neither the high dose of compound 40
alone nor any dose of compound 40 in combination with LꢀDopa (SI Figure S3b). Consequently,
in this model of Parkinson's disease, compound 40 completely reversed the motor deficits
without increasing the risk of developing dyskinesia.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As a first conclusion, in the 6ꢀOHDA rat model of Parkinson's disease motor symptoms,
administration of compound 40 allowed a decrease by 70% of the dose of LꢀDopa while
maintaining the same antiꢀparkinsonian efficacy, without inducing any adverse event such as
overactive rearing, stereotyped behavior or dyskinesia. These results show supportive evidence
that compound 40 could be developed as a potential LꢀDopaꢀsparing treatment in PD, preserving
maximal antiꢀparkinsonian benefits without the need to use high LꢀDopa dosage. Importantly,
compound 40 demonstrated consistent antiꢀparkinsonian activities in two rodent models and
exhibited robust PK/PD relationship.
Further improvements of PK properties, identification of candidate 60
ACS Paragon Plus Environment

Page 23 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Encouraged by the promising results in PD models obtained with ip administration of compound
40, we next decided to pursue the optimization of the PK profile with the objective of obtaining
an optimized compound with increased brain exposure over time and thus improved potential
target engagement after oral administration. Table 6 and SI Figure S4 show the brain exposures
of diverse molecules with good mGluR4 PAM activities (EC between 4 nM and 165 nM). As a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
comparator, concentrations of 454 and 262 ng/g of compound 40 were measured in rat brains 30
and 60 min postꢀadministration of an oral dose of 10 mg/kg (Tables 5 and 6). Within the same
pyrrolopyrimidine series (HET=A in Table 6), we did not identify substituent that give rise to a
better brain exposure. Indeed, replacements of the morpholine moiety of compound 40 by a 4ꢀ
(dimethylamino)ꢀpiperidine (compound 44), a close amide (compound 45) or urea (compound
46), or pyridine moieties (compounds 47 & 48) were detrimental to brain exposures (Table 6).
As a consequence, none of these first analogs of compound 40 had brain exposure at
concentrations higher than 100 ng/g at 30 or 60 min when given orally at 10 mg/kg. By contrast,
changing the rightꢀhand side heterocycle moiety for a thienopyridine (HET=B in Table 6)
seemed rather beneficial for brain exposure. Indeed, compound 49 showed much higher brain
concentrations 60 min postꢀdosing compared with the close compound 40 and was even detected
with concentrations above 200 ng/g 90 min after oral administration (Table 6). This encouraged
us to explore further thienopyridine with similar basic leftꢀhand side chains (compounds 50 to
52), but these two compounds did not show improvement compared with 49 despite a similar or
more potent activity. Similarly to what was observed in pyrrolopyrimidine series, introduction of
an aromatic group (compounds 53 to 56) or modifications of the morpholine moiety for close
piperidine or azetidine analogs (compounds 56 to 59) also resulted in very low brain exposures
(Table 6). Interestingly however, some of these compounds showed a clear increase in their
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
mGluR4 PAM potency as illustrated with molecules 55 to 57. Finally, the only chemical change
in 49 resulting in an increased brain exposure was the replacement of the oxygen atom linking
the chromenone oxime scaffold with the leftꢀhand side chain by a ꢀCH ꢀ group. Indeed,
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
compound 60 was found at concentrations wellꢀabove 400 ng/g (~1 µM) up to 90 min after
dosing (Table 6). This isosteric modification did not alter the mGluR4 PAM activity since
compound 60 exerts a PAM activity on mGluR4 with an EC that is similar to the ones of close
50
analogs 49 or 40 (Table 6).
Table 6: Further optimization of brain penetration
OH
N
*
*
R
S
HET =
N
N
N
O
HET
A
B
mGluR4
[Brain]3
(ng/g)
[Brain]
[Brain]
(ng/g)
0min
b
60min
90min
Compounds HET
R
PAM EC50
b
b
a
(ng/g)
(nM)
O
N
*
4
4
0
4
A
A
46 ± 18
165 ± 7
454
262
ND
ND
O
O
N
*
2.8
2.9
N
O
*
*
N
45
A
A
A
A
10 ± 1
35 ± 8
50 ± 16
19 ± 7
24.1
25.5
3.4
12.1
8.8
ND
ND
ND
ND
O
N
O
N
4
4
6
7
O
O
*
2.3
N
O
*
48
N
77.3
82.5
ACS Paragon Plus Environment

Page 25 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
N
*
4
5
9
0
B
B
35 ± 8
408
521
207
281
194
O
O
N
*
F
90 ± 14
239
157
F
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
*
B
F
106 ± 20
201
197
100
5
1
F
F
N
*
5
2
3
B
B
25 ± 5
17 ± 4
275
20
137
ND
F
O
N
5
N
*
9.5
N
NH
*
N
5
4
5
B
B
53 ± 20
4 ± 0
ND
BLQ
26.7
ND
ND
N
N
O
*
N
5
33.4
N
*
N
56
B
4 ± 1
7.4
BLQ
6.5
ND
N
O
N
5
5
7
8
B
B
6 ± 2
10.7
ND
ND
O
*
O
*
N
53 ± 6
BLQ
BLQ
BLQ
O
O
N
59
60
B
B
60 ± 5
BLQ
458
ND
818
N
N
*
79 ± 19
785
O
a
b
Values are the mean (± SD) of a minimum of 3 independent experiments. Values of brain
concentrations are mean of 3 rats following oral administration at 10 mg/kg. BLQ: concentration
below lower limit of quantification (1 ng/mL); ND: not determined.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Having identified compound 60 as a potent mGluR4 PAM with clearly improved brain exposure
following oral administration, we further extended its characterization. Key characteristics of
compound 60 are summarized in Table 7. The in vitro properties of compound 60 are very close
to those of compound 40. Indeed, compound 60 had minor agonist activity, with an average
stimulation of the receptor of 32 + 8% at 15 µM, and acted as a PAM on mGluR4, with a
potency of 79 + 19 nM, amplifying the effects of the EC glutamate concentration to 70 + 5 %
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
0
of the maximal response to glutamate (SI Figure S5). Compound 60 increased the apparent
affinity of glutamate for mGluR4 by 10ꢀfold (Table 7, Figure S5) and had no effect on NMDA,
AMPA, kainate, group I (mGluR1 and 5) or group II (mGluR2 and 3) mGluRs up to 10 µM, the
highest concentration tested. Among group III mGluRs, compound 60 showed more than 15ꢀ,
110ꢀ and 50ꢀ times higher selectivity for mGluR4 than for mGluR6, 7 and 8 respectively (Table
7, Figure S5). Compound 60 had no effect on COMT, MAOꢀA or MAOꢀB at 10 ꢁM in
enzymatic assays (Table7), reducing the risk for interaction with LꢀDopa metabolism.
Table 7: Summary of key parameters for compound 60
OH
N
N
O
S
O
N
Compound 60 (PXT002331)
a
In vitro Pharmacology
mGluR4 hum ago EC50
hum PAM EC50
hum fold shift
rat PAM EC50
ACS Paragon Plus Environment

Page 27 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
> 15
µ
M
79 nM
10 (at 3
µ
M)
47 nM
Selectivity
Specificity
Group I
Not active
COMT
Group II
Not active
MAOꢀA
mGlu6
15x
mGlu7
110x
mGlu8 NMDA, AMPA, kainate
50x
Not active
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
MAOꢀB
Not active
Not active
Not active
b
In vivo Pharmacokinetics (rat)
c
1
mg/kg
Plasma Cmax
(ng/mL)
t1/2
Plasma AUC0ꢀinf
(h.ng/mL)
CLp
Vd
Brain/Plasma
i.v.
(h)
(mL/min/kg) (L/kg)
3
+
25
30
0.90
163
+ 15
104
+ 10
8.02
8.4
+ 0.15
+ 1.28
+ 0.7
d
1
0 mg/kg
Plasma Cmax
(ng/mL)
Plasma Tmax Plasma AUC0ꢀinf
Brain AUC0ꢀinf
(h.ng/g)
Brain/Plasma
p.o.
(h)
(h.ng/mL)
9
4
1.17
432
2713
+ 544
6.6
+
21
+ 0.44
+ 126
+ 0.6
a
Activities were measured at least in duplicates. Values are the mean of at least three
experiments. Values are the mean (± SEM) of 3 rats. Ratio calculated at 1 hour postꢀdose.
Ratio of AUC0ꢀinf
b
c
d
.
PK characteristics of compound 60 in rats are given in Figure 6 and Table 7. Compound 60
displayed excellent CNS penetration after oral dosing, with more than 6ꢀfold (and up to 13.5ꢀfold
depending on the dose) higher exposure in brain than in plasma (Table 7). Moreover, the close
correlation between plasma and brain concentration curves over time indicates that plasma
concentrations are good predictors of compound 60 concentration levels and kinetics in the brain
(Figure 6).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 87
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6: PK profile of compound 60 in plasma and brain following oral administration at
25 mg/kg (in water) to SpragueꢀDawley rats.
Altogether, these characteristics demonstrate that compound 60 is a potent and selective
mGluR4 PAM, with high brain exposure measured after oral administration. The compound 60
will be extensively characterized in another article, to be published in due course.
Conclusion
After a thorough development from (ꢀ)ꢀPHCCC and extensive knowledge gained on the
medicinal chemistry and pharmacology of mGluR4 PAMs, it has been possible to generate a
novel series of compounds with an improved profile and with significant antiparkinsonian
activity demonstrated in validated rodent models of PD motor symptoms. Based on its complete
preclinical profile, compound 60 has been identified as the most promising candidate, which
could be further evaluated in the clinic.
ACS Paragon Plus Environment

Page 29 of 87
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Experimental section
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemistry. All reagents were commercial grade and used without further purification.
Commercially available anhydrous solvents were used for reactions conducted under inert
atmosphere. Silica gel generally used for column chromatography was SDS silica gel (60AAC
1
4
0ꢀ63 ꢁM). Thin layer chromatography was carried out using preꢀcoated silica gel Fꢀ254plate. H
13
NMR spectra were recorded on a Bruker AMXꢀ400 operating at 400.33 MHz. C NMR spectra
were recorded on a Bruker AVANCE I 400 Fourier transform spectrometer, operating at 100.67
MHz, using a 5 mm QNP probe operating at 300 K. Proton chemical shifts are listed relative to
residual CDCl (7.27 ppm), DMSOꢀD6 (2.51 ppm) or D O (4.60 ppm). Splitting patterns are
3
2
designated as s (singlet), d (doublet), dd (doubleꢀdoublet), t (triplet), tt (tripletꢀtriplet), td (tripletꢀ
doublet), q (quartet), quint (quintuplet), sex (sextuplet), sept (septuplet), m (multiplet), b (broad).
13
For C NMR spectrum: The chemical shifts (δ) were measured in ppm and referenced from
the central peak of DMSOꢀd (39.5 ppm). Electrospray MS spectra were obtained on a Waters
6
micromass platform LCMS spectrometer. All mass spectra were fullꢀscan experiments (mass
range 100ꢀ800 amu). Mass spectra were obtained using electro spray ionization. The HPLC
system was a Waters platform with a 2767 sample manager, a 2525 pump, a photodiode array
detector (190ꢀ400 nM). The column used was an XBridge C18 3.5 ꢁM (4.6 x 50 mm) in
analytical mode and an XBridge C18 OBD 5 ꢁM (30 x 100 mm) in preparative mode. The
mobile phase in both cases consisted in an appropriate gradient of A and B. A was water with
0.05 % of TFA and B was MeOH with 0.05 % of TFA. Flow rate was 1 mL per min in analytical
mode and 25 mL min in preparative mode. All LCMS were performed at room temperature. At
ACS Paragon Plus Environment