Increased Potency and Selectivity for Group III Metabotropic Glutamate Receptor Agonists Binding at Dual sites

Article abstract of DOI:10.1021/acs.jmedchem.7b01438

A group III metabotropic glutamate (mGlu) receptor agonist (PCEP) was identified by virtual HTS. This orthosteric ligand is composed by an l-AP4-derived fragment that mimics glutamate and a chain that binds into a neighboring pocket, offering possibilities to improve affinity and selectivity. Herein we describe a series of derivatives where the distal chain is replaced by an aromatic or heteroaromatic group. Potent agonists were identified, including some with a mGlu₄ subtype preference, e.g., 17m (LSP1-2111) and 16g (LSP4-2022). Molecular modeling suggests that aromatic functional groups may bind at either one of the two chloride regulatory sites. These agonists may thus be considered as particular bitopic/dualsteric ligands. 17m was shown to reduce GABAergic synaptic transmission at striatopallidal synapses. We now demonstrate its inhibitory effect at glutamatergic parallel fiber-Purkinje cell synapses in the cerebellar cortex. Although these ligands have physicochemical properties that are markedly different from typical CNS drugs, they hold significant therapeutic potential.

Full text of DOI:10.1021/acs.jmedchem.7b01438

Subscriber access provided by READING UNIV
Article
Increased potency and selectivity for group-III metabotropic
glutamate receptor agonists binding at dual sites
Selvam Chelliah, Isabelle A. Lemasson, Isabelle Brabet, Nadia Oueslati, Berin Karaman,
Alexandre Cabaye, Amélie TORA, Bruno Commare, Tiphanie Courtiol, Sara
Cesarini, Isabelle McCort-Tranchepain, Delphine Rigault, Laetitia Mony, Thomas Bessiron,
Heather McLean, Frédéric R. Leroux, Françoise Colobert, Hervé Daniel, Anne Goupil-
Lamy, Hugues-Olivier Bertrand, Cyril Goudet, Jean-Philippe Pin, and Francine Clemy Acher
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01438 • Publication Date (Web): 04 Feb 2018
Downloaded from http://pubs.acs.org on February 5, 2018
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 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 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 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Increased potency and selectivity for groupꢀIII
metabotropic glutamate receptor agonists binding at
dual sites
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
†
,^
†,∫
‡
‡
†,∞
†
Selvam Chelliah, Isabelle A. Lemasson, Isabelle Brabet, Nadia Oueslati, Berin Karaman,
†
‡,♦
†,▲
†,◊
Alexandre Cabaye, Amélie S. Tora, Bruno Commare, Tiphanie Courtiol, Sara Cesarini,
†
†
†,┴
║
Isabelle McCortꢀTranchepain, Delphine Rigault, Laetitia Mony, Thomas Bessiron, Heather
║
▲
▲
║
#
McLean, Frédéric R. Leroux, Françoise Colobert, Hervé Daniel, Anne GoupilꢀLamy,
#
*,‡
‡
*,†
HuguesꢀOlivier Bertrand, Cyril Goudet, JeanꢀPhilippe Pin, and Francine C. Acher .
†
Laboratoire de Chimie et de Biochimie Pharmacologiques et Toxicologiques, CNRS UMR8601,
Université Paris Descartes, Sorbonne Paris Cité, 45 rue des SaintsꢀPères, 75270 Paris cedex 06,
France,
‡
IGF, CNRS, INSERM, Univ. Montpellier, Fꢀ34094 Montpellier, France,
▲
UMR7509/ CNRS/ ECPM, Université de Strasbourg, 25 Rue Becquerel, 67087 Strasbourg 02,
France
║
Pharmacologie et Biochimie de la synapse, NeuroPSI – UMR 9197, Université ParisꢀSudꢀ
CNRS, Fꢀ91405 Orsay, France,
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 80
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
┴
Institut de Biologie de l’Ecole Normale Supérieure, CNRS UMR 8197, INSERM U1024, 46
rue d’Ulm, 75005 Paris, France,
#
Biovia, Dassault Systèmes, 10 rue Marcel Dassault, CS 40501, 78946 VélizyꢀVillacoublay cedex,
France.
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
RECEIVED DATE
TITLE RUNNING HEAD: GroupꢀIII mGluR agonists.
ACS Paragon Plus Environment
2

Page 3 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
A groupꢀIII metabotropic glutamate (mGlu) receptor agonist (PCEP) was identified by virtual
HTS. This orthosteric ligand is composed by an ꢀAP4ꢀderived fragment that mimics glutamate
L
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
and a chain that binds into a neighboring pocket, offering possibilities to improve affinity and
selectivity. Herein we describe a series of derivatives where the distal chain is replaced by an
aromatic or heteroaromatic group. Potent agonists were identified, including some with mGlu4
subtype preference e.g. 17m (LSP1ꢀ2111) and 16g (LSP4ꢀ2022). Molecular modeling suggests
that aromatic functional groups may bind at either one of the two chloride regulatory sites. These
agonists may thus be considered as particular bitopic/dualsteric ligands. 17m was shown to
reduce GABAergic synaptic transmission at striatopallidal synapses. We now demonstrate its
inhibitory effect at glutamatergic Parallel fiber – Purkinje cell synapses in cerebellar cortex.
Although these ligands have physicochemical properties that are markedly different from typical
CNS drugs, they hold significant therapeutic potential.
KEYWORDS metabotropic glutamate receptors, groupꢀIII, agonist, orthosteric ligand,
selectivity, virtual screening hit, structureꢀactivity relationship, amino acid, phosphinate,
chloride.
BRIEFS. Optimization of a virtual screening hit reveals potent mGlu receptor agonists with
4
novel binding mode.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 80
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
Introduction
a
The hallmark of metabotropic glutamate (mGlu) receptors is their ability to modulate the
activity of many synapses in the central nervous system by acting either on the preꢀsynaptic, or
the postꢀsynaptic element. Accordingly, mGlu receptors appear as potential therapeutic targets
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
1
for a number of neurologic and psychiatric diseases. They allow a fineꢀtuning that has not been
achieved with the ligandꢀgated ion channel receptors that play a more essential role in
glutamatergic synaptic transmission. The mGlu receptors belong to the class C of the large Gꢀ
2, 3
protein coupled receptor (GPCR) family. Eight subtypes have been identified that fall into 3
4
groups (IꢀIII). GroupꢀI includes subtypes 1 and 5 (mGlu and mGlu ), that activate PLC in vitro
1
5
while groupꢀII (mGlu and mGlu ) and III (mGlu , mGlu , mGlu and mGlu ) receptors are
2
3
4
6
7
8
coupled to G proteins and inhibit adenylyl cyclase or to G proteins that activate PLC signaling
i/o
q
5
pathway. Since groupꢀII / III receptors are mostly presynaptic and appear to function as
autoreceptors, ligands that trigger (agonists) or enhance (positive allosteric modulators (PAMs))
their activation are of interest to treat pathologies where an excess of transmitter release is
involved (including Parkinson’s disease, anxiety, epilepsy, addiction and pain).
Class C GPCRs are characterized by a large amino terminal domain (ATD) where the agonist
binding site is located while most allosteric modulators (NAMs and PAMs) identified so far bind
6, 7
to the transmembrane domain. For many years, the groupꢀIII receptor pharmacology was the
least investigated. Recently, several allosteric modulators have been discovered, mainly mGlu4
receptor PAMs. They were shown to be neuroprotective and beneficial in animal models of
8
Parkinson’s disease, pain and anxiety. Several amino acid analogues with agonist properties and
9
, 10
11ꢀ15
similar behaviour properties have been identified such as 1 (
L
ꢀAP4, Chart 1),
2 (ACPTꢀI),
1
6, 17
18, 19
3 ((1S,2R)ꢀAPCPr)
and 4 ((S)ꢀ3,4ꢀDCPG).
While they display effects in vivo, it is
ACS Paragon Plus Environment
4

Page 5 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
assumed that the range of effective concentrations is narrowed by their lack of subtype
selectivity among groupꢀIII receptors.
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
Chart 1. Orthosteric agonists of groupꢀIII mGlu receptors.
20
We have disclosed a new agonist (S)ꢀ5 ((S)ꢀPCEP, Chart 1) that is able to reach a new pocket
close to the glutamate binding site of group III mGlu receptors. Although this pocket is lined
with subtype selective residues, (S)ꢀ5 is not able to discriminate among mGlu , mGlu and
4
6
mGlu receptors because of a lack of interaction with the variable side chains within the binding
8
domain cleft. To discover agonists that allow such interactions, we designed, synthesized and
studied a series of aromatic derivatives of (S)ꢀ5. We identified 17m (LSP1ꢀ2111) that
21
preferentially activates mGlu receptor in comparison to mGlu and mGlu receptors and 16g
4
7
8
2
2
(
LSP4ꢀ2022) that is highly potent at mGlu and selective with respect to other mGlu receptors.
4
These two compounds have been tested in vivo and both demontrated efficacy in animal models
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 80
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
21ꢀ32
of Parkinson’s disease, anxiety, schizophrenia and pain.
In this paper, we describe the
synthesis of aromatic (S)ꢀ5 derivatives, the structureꢀactivity relationship study that led to the
discovery of 17m and 16g, as well as their potential mechanism of action. A binding mode for
2
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
1
6g was already proposed in ref. We now report on a new binding mode for 17m, other
nitrophenyl derivatives 17aꢀq and 16g. The gain of potency may be due to stabilizing
interactions of the two types of aromatic substituents at either of the two chloride binding sites
3
3
that were identified in a recent study. Compound 17m was also tested on acute cerebellar
slices, where groupꢀIII inhibition of excitatory transmission at parallel fiber (PF) ꢀ Purkinje cells
(
PC) synapses has been previously demonstrated to be exclusively mediated by presynaptic
1
7
mGlu receptors.
4
Results
Chemistry
The new ligands were prepared by two successive alkylations of hypophorous acid (H PO ). The
3
2
first step was a radical addition of H PO onto (S)ꢀNꢀCbzꢀvinylglycine methyl ester or (S)ꢀNꢀ
3
2
Cbzꢀallylglycine methyl ester yielding the common intermediates Hꢀphosphinates 6 and 7 as we
2
0, 34, 35
have previously reported.
Nucleophilic addition of the silylated P(III) form of these
2
0
intermediates onto aromatic aldehydes, imines and benzylic bromides gave the protected
phosphinates 8ꢀ13 (Table 1) according to the general procedures A, B, and E (see experimental
section). These methods varied in the workꢀup conditions depending on the sensitivity of the
products to basic and acidic conditions. Acidic deprotection and purification (general procedures
C or D) afforded the desired compounds 15ꢀ20 (Scheme 1, Table 1). Thus, series of carboxy
(
15aꢀc, 15e), 4ꢀmonosubstituted (16aꢀg, 19a), 3ꢀnitro (17a, 17cꢀh, 17jꢀm, 17oꢀq, 19b) and
differently substituted
α
ꢀhydroxybenzyl phosphinates (18aꢀh) were synthesized from aromatic
ACS Paragon Plus Environment
6

Page 7 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
aldehydes while the corresponding benzylic derivatives (15d, 17b, 17i, 17n) were obtained from
bromides and αꢀaminobenzylic derivatives (20aꢀg) from imines.
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
Table 1. Structures of compounds 8 – 13 and 15 – 20 depicted in Scheme 1ꢀ4.
1
2
3
4
5
Compounds
n
X
R
R
R
R
R
1
OH
OH
OH
H
H
CO R
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
8a R = Me, 15a R = H
2
1
CO R
H
H
8
8
8
b R = Me, 15b R = H
c R = Me, 15c R = H
d R = Me, 15d R = H
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CO R
2
CO R
H
2
OH
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
CO H
OH
8
e, 15e
a, 16a
b, 16b
9c, 16c
d, 16d
2
H
H
H
H
H
H
H
H
NO2
9
C(O)Me
9
CF3
SO Me
9
2
OH
9
e, 16e
a
OMe
9
f , 16f
g, 16g
O
CO H
2
9
NH2
9
h, 16h
10a, 17a
0b, 17b
0c, 17c
0d, 17d
0e, 17e
0f, 17f
NO2
H
NO2
H
CF3
F
1
OH
OH
OH
OH
NO2
NO2
NO2
NO2
1
1
Me
Cl
1
1
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 80
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
1
OH
H
NO2
H
H
1
0g, 17g
N
O
1
1
1
1
1
1
1
1
OH
H
H
H
NO2
NO2
NO2
NO2
OH
OH
H
H
H
H
H
1
0h, 17h
1
0i, 17i
0j, 17j
0k, 17k
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
OH
OH
OH
OH
H
OH
H
H
H
1
H
H
OH
H
1
H
NO2
OMe
OH
H
10l, 17l
H
NO2
OMe
OMe
OEt
H
1
0m, 17m
H
NO2
NO2
OH
H
1
0n, 17n
OH
H
OH
H
1
0o, 17o
0p, 17p
0q, 17q
1
1
OH
OH
H
H
NO2
NO2
OH
OH
OH
F
H
H
1
1
1
1
1
1
1
1
1
1
2
OH
OH
OH
OH
OH
OH
OH
OH
OH
H
H
CF3
NO2
H
H
H
H
H
H
1
1a, 18a
11b, 18b
1c, 18c
1d, 18d
1e, 18e
1f, 18f
1g, 18g
1h, 18h
12a, 19a
2b, 19b
3a R = Cbz, 20a R = H
OH
NO2
NO2
NO2
H
H
H
NO2
H
1
H
H
H
H
F
NO2
H
1
NO2
NO2
1
OH
H
NO2
NO2
H
1
CN
H
H
1
F
F
F
F
1
H
H
H
H
O
CO H
2
2
1
1
1
OH
H
H
H
H
NO2
NO2
NO2
OH
OMe
H
H
H
H
H
1
NHR
NHR
H
OH
H
1
OMe
H
1
3b R = Cbz, 20b R = H
a
NHCH Ph
NO2
13c , 20c
2
ACS Paragon Plus Environment
8

Page 9 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
1
1
1
1
NH(CH ) Ph
H
H
H
H
NO2
H
H
H
H
H
H
H
H
H
H
H
H
1
3d , 20d
2 2
a
NH(CH ) Me
NO2
13e , 20e
2 5
a
NH(CH ) Me
NO2
1
3f , 20f
2 7
a
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
NHCH ꢀ1ꢀNapht
NO2
1
3g , 20g
2
a
Methyl phosphinate
a,b
Scheme 1. Synthesis of derivatives 8ꢀ12, 14ꢀ19 and 21.
a
20, 35
Reagents and conditions: (i) ref
CH Cl ; (iii) 1 M HCl; (iv) 6 M HCl 100 °C; (v) Dowex AG50WꢀX4, water elution or AG1X4,
dilute HCl elution. For R , n, X and Het see Tables 1 and 2. For 9f and 16f see Scheme 2, for
4d and 21d see Scheme 3.
; (ii) BSA, Arꢀaldehyde (X =OH) or halide (X = H), RT,
2
2
1
ꢀ5
b
1
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 80
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
We checked that nucleophilic addition of the Hꢀphosphinate could also be performed with the
36
methyl ester 6a in the presence of lithium diisopropylamide (Scheme 2, 9f). When protected
phosphinates could not withstand the harsh acidic conditions (6 M HCl, 100 °C, procedure C or
D), esters were first hydrolyzed with lithium hydroxide, then the benzyloxycarbonyl group was
removed by stirring in 6 M HCl at room temperature for several days (18g).
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
a
Scheme 2. Synthesis of derivative 16f.
O
O
O
NH
O
O
NH
OH
(
i)
(ii), (iii)
O
6
MeO C
2
P H
MeO C
2
P
MeO
MeO
6a
9f
OMe
NH2
OH
(
iv)
O
P
HO C
2
(
v)
HO
6f
OMe
1
a
Reagents and conditions: (i) EDC/MeOH; (ii) pꢀanisaldehyde, LDA, 0 °C THF; (iii) 1 M HCl, 0
°C; (iv) 6 M HCl, 100 °C; (v) Dowex AG50WꢀX4, H O elution.
2
Derivative 10i was obtained by reaction of 6 with the 4ꢀhydroxyꢀ3ꢀnitrobenzyl bromide that was
prepared by reduction of the corresponding acid to the alcohol and subsequent bromination with
PBr . 5ꢀEthoxyꢀ4ꢀhydroxyꢀ3ꢀnitrobenzaldehyde was obtained by nitration of 3ꢀethoxyꢀ4ꢀ
3
3
7
hydroxybenzaldehyde with copper nitrate, and used to synthesize 17o. 4,5ꢀDihydroxyꢀ3ꢀ
nitrobenzaldehyde was obtained by dealkylation of 4ꢀhydroxyꢀ5ꢀmethoxyꢀ3ꢀnitrobenzaldehyde
in HBr 33% solution in AcOH and reacted with 6 to afford phosphinate 17p. 4ꢀNitro 3ꢀ
ACS Paragon Plus Environment
10

Page 11 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trifluoromethylbenzaldehyde was prepared by reduction of the acid to the alcohol with BH3
followed by a limited oxidation under the Swern conditions and then used to produce 18a. 3,5ꢀ
Dinitrobenzaldehyde was prepared in a similar manner for the synthesis of 18e. Protected
primary αꢀaminobenzylic phosphinates (13a,b and 14d) were synthesized by a one pot reaction
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
3
8, 39
and hydrolyzed to 20a,b and 21d (Scheme 3).
Substituted αꢀaminobenzylic derivatives (20cꢀ
2
0g) resulted from reaction of methyl Hꢀphosphinate 6a (see above) with a series of imines (22cꢀ
2g) of mꢀnitrobenzaldehyde in the presence of BF Et O (13cꢀ13g) and subsequent hydrolysis
2
3
2
(
Scheme 4). The protected phosphinates 13cꢀ13g were purified on silica gel before acidic
hydrolysis (6 M HCl, 100 °C). Finally, we synthesized a series of heterocyclic analogues 21aꢀl
Scheme 1, Table 2) employing the suitable substituted furaldehyde, thiophenecarboxaldehyde,
(
pyridine carboxaldehyde, hydroxypyridineꢀcarboxaldehyde, aminothiazoleꢀcarboxaldehyde or
azaindolꢀcarboxaldehyde.
a
Scheme 3. Synthesis of derivatives 20a,b and 21d.
a
Reagents and conditions: (i) CH COCl/CH CO H, benzylcarbamate, ArꢀCHO, 0 °C to RT; (ii)
3
3
2
+
6
M HCl 100 °C; (iii) Dowex 50WꢀX4 (H ) 1 M NH OH elution.
4
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 80
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
a
Scheme 4. Synthesis of derivatives 20cꢀg.
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
a
Reagents and conditions: (i) EDC/MeOH; (ii) BF ·Et O, CH Cl ; (iii) 6 M HCl 100 °C.
3
2
2
2
Table 2. Structure of compounds 14 and 21 depicted in Scheme 1 and 3.
n
X
Het
Compounds
4a, 21a
1
OH
O
S
NO2
1
NO2
1
1
1
OH
OH
1
4b, 21b
NO2
1
4c, 21c
S
S
NO2
NHCbz
1
4d, 21d
4e, 21e
4f, 21f
1
1
OH
OH
O
CF3
1
1
CO H
2
O
ACS Paragon Plus Environment
12

Page 13 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
1
1
1
1
1
1
OH
2
1g
OH
OH
OH
OH
OH
21h
N
N
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
OH
2
1i
OH
N
2
1j
S
NH2
1
4k, 21k
N
N
1
4l, 21l
N
The hydroxybenzylic phosphinates 15ꢀ19, 21 (X = OH, Scheme 1, Table 1) are mixtures of two
diastereomers that result from the addition of (S)ꢀHꢀphosphinates 6 and 7 onto aldehydes. The
mixtures could be separated by chiral HPLC on a Crownpak CR(+) column. Pure diastereomers
of selected 3ꢀnitrophenyl derivatives (17a, 17h, 17l, 17m, 17o, 17q) were isolated on a semiꢀ
preparative scale eluting with a pH 2 HCl solution. In some cases the separation was improved
by cooling the column. The first eluting diastereomer was named 17ꢀ1, the second 17ꢀ2 and
1
absolute configurations were assigned based on optical rotations and H NMR signals of the
35
20
methylene protons according to Commare et al. Optical rotations ([α]D in H O) were
2
compared for each pair of diasteromers: 17hꢀ1 (+1.5) / 17hꢀ2 (+21.9), 17lꢀ1 (ꢀ4.5) / 17lꢀ2
(
+19.4), 17oꢀ1 (+6.4) / 17oꢀ2 (+18.3) , 17qꢀ1 (ꢀ5.9) /17qꢀ2 (+18.6). The second diasteromer 17ꢀ2
2
0
1
consistantly displayed a larger positive [α]D value compared to that of 17ꢀ1. The H NMR
chemical shifts of the two Hꢀ1 protons of the pairs of diastereomers appear as a single multiplet
for the first eluting diastereomer 17ꢀ1, and as two distinct multiplets for the second isomer 17ꢀ2.
Accordingly, 17ꢀ1 were assigned (3S,1’S) configuration and 17ꢀ2 (3S,1’R), based on 17aꢀ1 / 17aꢀ
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 80
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
35
2
and 17mꢀ1 / 17mꢀ2 absolute configurations. The chiral stability of the pure diastereomers
was checked in a pH 7.3 buffered solution after 4 days at RT and showed no epimerization.
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
Pharmacological activity at mGlu receptors
Agonist activity of 5 and its aromatic derivatives was tested at the four groupꢀIII mGlu receptor
subtypes (Table 3). To that aim, the groupꢀIII mGlu receptors (mGlu , mGlu , mGlu and mGlu )
4
6
7
8
were transiently transfected in HEK293 cells together with a chimeric Gq/i protein that allows
2+
their efficient coupling to the phospholipaseꢀC/inositol phosphate/Ca pathway. Receptor
activation was determined by two different functional assays, one measuring the production of
2
+
total radiolabelled inositol phosphates (IP) and the other measuring the intracellular Ca release
using Fluoꢀ4 as a fluorescent probe. The results are summarized in Table 3.
Table 3. GroupꢀIII mGlu receptor agonist activities of compounds 15ꢀ21 and comparison with 1ꢀ5
mGlu4
EC50
mGlu8
EC50
mGlu8/4
mGlu6
EC50
mGlu7
EC50
Compounds
EC ratio
50
µM (n)
µM (n)
µM (n)
µM (n)
a
b
a
b
a
a
b
1
0.10±0.13 (66)
0.29±0.06 (50)
0.29±0.07 (44)
2.5
2.2
2.4±0.2 (12)
337±69 (10)
249±107 (3)
b
0
.13±0.02 (34)
1.0±0.3 (7)
a
a
a
a
2
3
4
1.7±0.2 (10)
5.1±0.8 (5)
3.0
0.5
11±3 (6)
280±85 (2)
a,c
a,c
a,c
a,c
0.72±0.21
0.34±0.01
1.95±0.1
602±312
a
a
b
b
b
1.96±0.26 (16)
0.023±0.03 (12)
0.044±0.01 (12)
0.012
0.014
6.2±0.4 (3)
30±8 (2)
b
b
b
3
.2±0.9 (13)
3.2±0.2 (3)
>100 (2)
a
a
a
(
S)-5
6.4±1.4 (4)
7.8±0.5 (4)
1.2
1.1
31±6 (4)
-
b
b
b
b
5
.94±0.22 (4)
6.5±1.7 (4)
6.7±3.3 (3)
89±15 (2)
a
a
a
1
5a
22±3 (8)
125±8 (3)
5.8
>100 (1)
-
ACS Paragon Plus Environment
14

Page 15 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
5b
N.E. at
-
-
-
a
1
00 µM
a
a
a
a
1
1
5c
8.0± 0.4 (4)
141±10 (3)
18
31±3 (3)
-
169±31 (3)
a
a
5d
212±120 (2)
511 (1)
2.4
-
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
b
>100 (1)
b
b
15e
16a
15.8±6.0 (3)
40±5 (2)
2.5
-
-
-
-
a
b
a
b
0.38±0.10 (4)
0.53±0.12 (2)
1.4
0.8
b
b
0
.64±0.22 (3)
0.49±0.10 (3)
4.6±0.8 (3)
>100 (2)
a
a
1
1
6b
6c
7.8±1.2 (3)
14±3 (3)
1.8
2.3
-
-
-
-
-
-
a
a
15±1 (4)
34±9 (2)
a
16d
>100 (2)
-
b
b
b
b
5
9±28 (4)
>100 (4)
>100 (4)
>100 (4)
b
b
b
1
1
1
1
6e
6f
7.7±2.1 (3)
6.5±0.8 (3)
25±10 (3)
3.2
9.3
256
0.9
-
-
-
b
b
61±15 (3)
23.0 (1)
b
a
b
b
b
6g
6h
0.11±0.02 (30)
8.8±3.4 (3)
29.2±4.2 (27)
8.4±1.0 (2)
4.2±0.6 (14)
-
11.6±1.9 (19)
-
a
a
a
17a
0.54±0.05 (19)
1.5±0.5 (6)
2.8
2
8.8±0.7 (5)
>1000 (7)
b
b
0
.6±0.3 (2)
1.2±0.9 (2)
-
-
b
b
b
b
1
1
1
7a-1
1.0±0.4 (5)
1.25±0.3 (5)
1.2
1.5
2.4±1.3 (2)
>100 (1)
b
b
b
b
7a-2
7b
0.32±0.10 (8)
1.3±0.5 (5)
3.2±1.3 (4)
>100 (3)
a
a
0.8±0.2 (5)
4.4±2 (6)
5.4
2.5
-
-
b
a
b
b
b
0.42±0.08 (7)
0.38±0.13 (6)
1.0±0.2 (6)
3.2±2.0 (3)
>100 (2)
a
1
7c
-
-
5.7 ±2.2 (3)
-
b
b
b
b
0
.44 ± 0.06 (3)
1.54 ± 0.72 (3)
3.5
11.9±0.7 (3)
>100 (3)
a
a
a
17d
0.6±0.1 (6)
1.3±0.3 (2)
2.1
1.5
25±15 (2)
-
b
b
b
b
0
.44±0.05 (4)
0.66±0.19 (3)
4.0±1.8 (4)
>100 (3)
a
a
a
a
1
7e
2.8±1.0 (9)
5.6±2.5 (4)
1.6
0.5
>100 (3)
N.E. (3)
b
b
3.3±0.5 (3)
1.6±0.4 (3)
-
-
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 80
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
a
b
a
1
1
1
7f
2.0±0.3 (3)
.7±1.9 (3)
2.5±1.5 (3)
1.2
1.0
-
-
b
b
b
b
4
4.9±1.3 (3)
18±6 (2)
>100 (3)
a
a
7g
7h
241±35 (4)
>100 (1)
-
-
b
a
-
-
>100 (2)
>100 (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
a
b
a
a
b
2.3±0.3 (20)
42±5 (14)
18
48
6.0±0.9 (3)
6.3±2.1 (4)
69±8 (3)
b
b
b
1
.5±0.4 (8)
73±8 (6)
125±55 (4)
b
b
b
b
1
1
1
1
7h-1
7h-2
7i
0.5±0.05 (4)
31.4±6.1 (4)
39.8±6.5 (4)
61
3.6±1.4 (5)
62±27 (3)
b
b
b
4.5±0.2 (3)
8.5
11
68.4±14.5 (3)
295±178 (3)
b
b
b
b
24±4 (3)
277±82 (3)
>100 (6)
>100 (4)
a
a
7j
23.0±6.6 (4)
21.2 ± 8.8 (2)
0.92
4.6
1.7
1
-
-
-
-
-
-
-
-
a
a
17k
17l
0.90±0.25 (6)
4.2±2.3 (2)
a
a
1.6±0.7 (5)
2.7±1.0 (6)
b
b
3.5±1.3 (3)
3.8±1.0 (3)
b
b
b
b
17l-1
2.0±0.4 (3)
2.5±0.8 (3)
1.2
1.3
4.5 (1)
>100 (1)
b
b
b
b
1
1
7l-2
7m
1.7±0.6 (3)
2.2±0.6 (3)
3.9 (1)
>100 (1)
a
b
a
a
a
2.2 ± 0.3 (26)
66±12 (16)
30
1.7±0.1 (3)
53±20 (3)
b
b
b
0
.9±0.13 (13)
19.4±7 (12)
21.5
3.2±1.4 (4)
101±33 (3)
b
b
b
b
1
7m-1
0.72±0.08 (6)
18±5 (7)
25
6
1.1±0.4 (4)
63±35 (3)
b
b
b
b
17m-2
17n
4.4±0.5 (6)
26±9 (9)
2.4±0.8 (4)
100±49 (3)
b
b
19±6 (3)
28±12 (3)
1.5
2.4
4.4
1.0
-
-
-
-
-
-
-
-
-
-
b
b
1
1
1
1
7o
8.5±1.1 (3)
20±13 (3)
b
b
7o-1
7o-2
7p
5.1±1.2 (3)
22±21 (3)
b
b
6.4±2.4 (3)
6.5±3.1 (3)
a
a
4.75±1.25 (2)
5.4±2.9 (2)
1.1
7.6
b
b
b
b
4.3±1.0 (6)
32.8±6.5 (7)
26±2 (2)
>100 (2)
b
b
b
b
b
17q
7q-1
1.95±0.51 (6)
2.42±0.62 (4)
23.4±7.6 (6)
12
14
1.07±0.22 (6)
51.2±7.3 (7)
b
b
b
34.3±19.3 (4)
0.76±3.5 (3)
39±5 (4)
1
ACS Paragon Plus Environment
16

Page 17 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
b
b
b
b
1
7q-2
9.77±3.2 (3)
17.1±7.7 (4)
1.8
2.4
8.7
3.1
1. 7
4.5
1.9
5.3
10.2±4.9 (3)
53.8±8.9 (3)
a
a
b
b
18a
18b
0.40±0.09 (4)
0.98±0.12 (2)
2.9±0.4 (3)
>100 (3)
a
a
2.3±0.3 (5)
19.8±3.1 (3)
-
-
-
-
-
-
-
-
-
-
-
-
-
-
a
a
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
1.5 ± 0.4 (5)
4.6±2.3 (3)
1
1
1
1
1
1
8c
8d
8e
8f
a
a
1.76±0.45 (5)
2.9±2.1 (2)
a
a
1.4±0.3 (3)
6.6±1.8 (2)
a
a
2.4 ± 0.3 (5)
4.5±2.8 (2)
a
a
8g
8h
41±7 (4)
219±47 (2)
a
a
4.9±0.7 (8)
18.5±4.5 (3)
3.7
1.4
b
b
b
b
5
.6±2.1(3)
7.6±1.7 (3)
30±14 (3)
>100 (3)
b
b
1
1
2
2
2
9a
9b
0a
0b
0c
N.E. (2)
N.E. (2)
N.E. (2)
-
-
-
-
-
-
b
b
N.E. (2)
a
a
0.54±0.14 (3)
2.2±0.9 (4)
4.1
3
b
b
b
b
11.6±2.6 (3)
33±12 (6)
>100 (3)
N.E. (2)
b
b
25.3±2.0 (2)
29.7±9.7 (4)
1.2
1.6
0.9
6.3
4.3
2.2
3.9
3.2
2
-
-
-
-
-
-
-
-
-
-
b
b
20d
20e
20f
4.41±1.76 (3)
7.2±1.4 (2)
b
b
25.0±8.6 (3)
22.8±4.9 (3)
b
b
6.1±2.5 (3)
38±21 (3)
b
b
2
2
2
2
0g
1a
1b
1c
7.9±0.3 (3)
35.4±3.7 (3)
a
a
a
a
a
0.15±0.03 (9)
0.49±0.06 (7)
0.16±0.03 (8)
0.3±0.1 (3)
6.1±0.6 (3)
1970±320(3)
a
b
b
1.9±0.3 (4)
4.2±1.2 (3)
>100 (2)
a
b
a
a
a
0.51±0.12 (7)
3.6±1.2 (6)
418±136 (4)
b
0
.19±0.03 (3)
0.38±0.21(3)
-
-
-
-
-
-
a
a
2
2
1d
1e
0.8±0.4 (3)
1.6±0.4 (4)
1.8±0.3 (3)
2.2
3.1
a
a
5.2±4.1 (2)
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 80
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
a
a
2
1f
2.0±0.4 (6)
9.2±1.3 (3)
4.6
1.2
1.0
0.5
1.9
1.9
2.0
-
-
b
a
21g
21h
30±9 (3)
36±20 (3)
-
-
b
b
28±13 (2)
26.2±2.8 (2)
-
-
b
b
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
33±10 (3)
17±5 (3)
-
-
2
2
2
2
1i
1j
1k
1l
b
b
42±9 (2)
81±15 (2)
-
-
b
b
b
b
15±4 (3)
28±9 (3)
>100 (1)
N.E. (3)
b
b
3.4±0.8 (3)
6.9±2.8 (3)
-
-
a
b
2+
c
16
IP production assay. Ca release and fluorescence assay. from reference . N.E. no effect at 300 µM
Structure – Activity Relationship
We have previously shown that several (S)ꢀ5 derivatives bearing a carboxyꢀalkyl chain are able
2
0
to activate mGlu receptors. We also showed that the length between the
αꢀcarbon and the
4
phosphinate moiety could be neither reduced nor extended in this series. We thus thought of
replacing the alkyl chain of (S)ꢀ5 with a benzylic group (Table 1). We first prepared the 15aꢀe
derivatives, among which 15c best activated the mGlu receptor but remained weaker than (S)ꢀ5.
4
Interestingly, this compound displayed an 18ꢀfold preference for mGlu over mGlu and mGlu
7
4
8
receptors (Table 3). We then varied the para benzylic substituent in the 16aꢀg series. All
compounds activated the receptor with diverse efficiency. Replacing the carboxylic group of 15c
with a bulky substituent such as trifluoromethyl (16c) or methylsulfonyl (16d) was unfavourable,
whereas replacement with an acetyl (16b), a hydroxyl (16e) or a methoxy (16f) group had almost
no effect on potency, but reduced or abolished the preference for mGlu over mGlu receptor. In
4
8
contrast, a nitro group (16a) significantly increased the mGlu and mGlu receptor potency but
4
8
with no subtype discrimination. Optimum mGlu receptor activity and selectivity was obtained
4
ACS Paragon Plus Environment
18

Page 19 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with a para oxyacetic acid in 16g. This compound is as potent as 1 at the mGlu receptor but is
4
2
2
more than 250ꢀfold weaker at the mGlu receptor.
8
We next examined various nitrobenzylic and nitroheteroaromatic derivatives (17aꢀq, 18aꢀf, 19b,
0aꢀg, 21aꢀd). The meta isomer 17a behaves quite similarly to the para isomer 16a at groupꢀIII
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
2
mGlu receptors. Adding a substituent at position 4 (17cꢀh, 17l) did not increase the activity and a
large group such as morpholino (17g) almost abolished the agonist effect. However, the
nitrophenol derivative 17h exhibited a surprising preference for mGlu over mGlu receptor.
4
8
This preference was lost in 17l when the phenol was methylated but kept by addition of a
methoxy or fluorine substituent at position 5 (17m and 17q). Nevertheless, elongating this 5ꢀ
methoxy group to ethoxy (17o) or replacing it by a polar hydroxyl group (17p), decreased the
agonist potency and selectivity. The nitro and hydroxyl substituents of 17h need to be
respectively in the 3ꢀ and 4ꢀ positions of the phenyl ring in order to keep the mGlu preference,
4
since for other regioisomers 17j, 17k, 18b this property is less pronounced. Additionally, the
presence of the hydroxyl group at the 2ꢀposition (17j) affected significantly the activation in
comparison to the 6ꢀposition (17k). Dinitroꢀsubstitution (18cꢀf) provided no improvement.
Finally, the 3ꢀnitro group of 17a could not be replaced by a cyano group (18g) but the
pentafluoro derivative was reasonably active (18h). We also confirmed that the chain length
between the
αꢀcarbon and the phosphinate moiety could not be extended regardless of the
phosphinate substituent (19a,b).
We next examined whether the presence and the chirality of the benzylic hydroxyl in 15ꢀ19, 21
(X = OH, Scheme 1, Tables 1 and 2) had any effect on their activity. Comparison of the mGlu4
receptor EC values of these compounds with those of the corresponding desꢀhydroxy analogues
50
(15a vs 15d, 17a vs 17b, 17h vs 17i, 17m vs 17n) revealed that the deletion of the benzylic OH
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 80
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
3
4
2
3
4
in the 3ꢀcarboxy (R = CO , R = H, R = H) and substituted 3ꢀnitrophenyl (R = NO , R ≠ H, R
2
2
≠
H) derivatives partially reduced their potency. On the other hand, it had no effect on the
2
3
4
unsubstituted 17a (R = NO , R =H, R = H). Interestingly among the separated diastereomers
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
of 17a, 17h, 17l, 17m, 17o, 17q, only those from mixtures that displayed preference for mGlu4
over mGlu receptor (17h, 17m and 17q) showed a 4ꢀ9 fold difference in activity between the
8
(
3S,1’S) and (3S,1’R) isomers. Moreover, that difference was more pronounced for mGlu4
activation than for mGlu receptor. Diastereomers (3S,1’S) of 17hꢀ1 and 17mꢀ1 displayed the
8
highest potency and largest mGlu preference (Table 3, Fluoꢀ4 assay).
4
To expand the SAR of aromatic derivatives of (S)ꢀ5, we prepared an amino benzylic series (20,
6
X = NHR , Table 1). As noted with the desꢀhydroxy compounds, changing from X=OH in 17a to
X = NH in 20a had no effect on pharmacological activity while the same modification in 17m
2
that provided 20b, caused a 13ꢀfold decrease. Surprisingly substitution of the benzylic amino
group in 20a retained an agonist activity in the 4 – 40 µM range, even with large groups such as
benzyl (20c), phenylethyl (20d), nꢀhexyl (20e), nꢀoctyl (20f) and 1ꢀnaphtylmethyl (20g).
Finally, a series of heteroaromatic derivatives was assayed (21, Table 2). The nitrofuran (21a)
and nitrothiophene (21b, 21c) derivatives turned out to be among the most potent compounds of
the (S)ꢀ5 series. However, no improved subtype selectivity was observed. Furan derivatives
bearing other substituents than nitro, such as CF (21e) and CO H (21f), as well as pyridine
3
2
(
21g,h), hydroxypyridine (21i, 21j), aminothiazole (21k) and azaindole (21l) derivatives retained
µM agonist activities which demonstrates some promiscuity of the distal binding site.
Continuing the hit optimization process, the first aim of this work was to improve the potency of
the derivatives. The hit compound (R)ꢀ5 was initially found to be in the same range as glutamate
2
0
on the mGlu receptor. Its most active enantiomer (S)ꢀ5 (EC 5.94 ± 0.22 µM) was improved
4
50
ACS Paragon Plus Environment
20

Page 21 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
by a factor 2.6 and the most potent linear derivatives of (S)ꢀ5, such as (3S)ꢀ3ꢀ[((3ꢀaminoꢀ3ꢀ
carboxy)propyl)(hydroxy)phosphinyl]propenoic acid (LSP1ꢀ3154, EC50 1.0 ± 0.2 µM), were
2
0
only better by a factor 6 as compared to (S)ꢀ5. The potency at the mGlu receptor has now been
4
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
markedly improved thanks to aromatic derivatives of 5. Compound 17a (EC 0.54 ± 0.05 µM)
50
displays a potency below the µM range, and as such is more potent than 2 (EC 1.7 ± 0.2 µM) a
5
0
wellꢀknown groupꢀIII mGluR agonist (Figure 1). The most potent compounds at mGlu and
4
mGlu are 21a and 21c (Table 3). Their EC values are 0.15 ± 0.03 µM (n = 9) and 0.16 ± 0.03
8
50
µM (n = 8) at the mGlu receptor and 0.3 ± 0.1 µM (n = 3) and 0.51 ± 0.12 µM (n = 7) at the
4
mGlu receptor in the IP production assay. Their potency values are comparable to that of 1
8
(EC value 0.10 ± 0.01 µM (n = 66) for mGlu receptor and 0.29 ± 0.06 µM (n = 50) for mGlu
8
5
0
4
receptor). Altogether, the potency at mGlu receptor has been increased by more than a 100 fold,
4
from the hit compound to the most potent cyclic derivatives (Figure 2).
Figure 1. Optimization of the potency of 5 derivatives on mGlu receptor. Rat cloned mGlu4
4
receptors were transiently transfected in HEK293 cells together with a chimeric Gꢀprotein alpha
subunit and the high affinity glutamate transporter EAAC1. Receptor activity was determined by
the production of inositol phosphates (IP) resulting from receptor activation by various
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 80
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
concentrations of (S)ꢀ5 derivatives. Each point corresponds to the mean ± S.E.M. of triplicate.
Data presented are representative of at least 3 experiments.
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 2. Optimization of mGlu versus mGlu receptor selectivity. Ratios of mGlu /mGlu
4
4
8
8
EC ’s for selected compounds are displayed as bars to illustrate the selectivity improvement
50
determined in the Fluoꢀ4 assay.
A second objective was to improve the selectivity of mGlu versus mGlu receptor. As with (S)ꢀ5
4
8
2
0
itself (ref and Table 3), most of the hit derivatives do not display potency differences between
mGlu and mGlu receptors, like 1 and 2 the classical agonists used as controls in our
4
8
experiments (Figure 2). However, two compounds, 17h (LSP1ꢀ2109) and 17m, showed a
preferential agonist activity for mGlu compared to mGlu receptor (selectivity defined by EC
50
4
8
ratio > 100 and preference with lower ratio values). The EC values of 17h and 17m at the
5
0
ACS Paragon Plus Environment
22

Page 23 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mGlu receptor were 1.5±0.4 µM (n = 8) and 0.9±0.1 µM (n = 13) and at the mGlu receptor
4
8
7
3 ± 8 µM (n = 6) and 19 ± 7 µM (n = 12) in the Fluoꢀ4 assay. They are respectively 48ꢀfold and
21.5ꢀfold more potent at mGlu than at mGlu receptor. Potency and preference are even more
4
8
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
pronounced with the S,S isomers 17hꢀ1 and 17mꢀ1 (Table 3). Their respective EC values are
5
0
0
.5 ± 0.05 µM and 0.72 ± 0.08 µM at the mGlu receptor and 31.4 ± 6.1 µM and 18 ± 5 µM at
4
the mGlu receptor in the Fluoꢀ4 assay with mGlu /mGlu EC ratios of 61 and 25. Interestingly,
8
8
4
50
the compound 16g is a selective mGlu agonist as compared to mGlu and mGlu receptors, and
4
8
7
22
preferential as compared to mGlu receptor. The potency of 16g at mGlu receptor (0.11 ± 0.02
6
4
µM, n = 30) is comparable to that of 1, while its potency at mGlu is ~ 250ꢀfold lower
8
2
2
(29.2 ± 4.2 µM, n = 27).
Molecular Modeling
Metabotropic glutamate receptors, as other members of class C GPCRs are characterized by a
large extracellular amino terminal domain, which is referred to as the Venus FlyTrap (VFT)
domain. The VFT domain consists of two lobes, Lobe 1 (LB1) and Lobe 2 (LB2) connected by a
flexible hinge, which form a cleft where glutamate binds. The VFT domain oscillates between an
4
0, 41
open and a closed conformation, the latter triggering receptor activation.
This closed state is
stabilized by agonists through interactions with both lobes. Homology models of the mGlu4
receptor binding domain have been generated earlier based on mGlu and mGlu receptor Xꢀray
1
3
2
0, 40, 42
structures.
We have previously shown that (S)ꢀ5 is composed of two parts, an
glutamate site at groupꢀIII mGlu receptors, and a distal carboxyethyl chain that binds to a new
LꢀAP4 part that binds to the
2
0
pocket adjacent to the glutamate site. Moreover, the potent mGlu agonist, 16g was previously
4
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 80
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
shown to bind similarly to (S)ꢀ5 but with an increased selectivity towards mGlu , and thus by
4
2
2
taking advantage of the nonꢀconserved residues lining the new pocket.
Furthermore, our team has recently identified the presence of two chloride (Cl ) binding sites in
ꢀ
3
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
the cleft of the VFT domain in mGlu receptor. The Cl ion in binding site 1 (Site 1) is
4
embedded in LB1 and makes Hꢀbond interactions with the side chain hydroxyl group of Thr108
ꢀ
and the backbone NH of Ser157 and Gly158 residues. On the other hand, the Cl ion in binding
site 2 (Site 2) is located in the active cleft between LB1 and LB2 and bridges these two lobes in
the closed active conformation by establishing Hꢀbond interactions between LB1 residue Ser160
and LB2 residue Asn286. The chloride location of Site 1 in LB1 of the VFT is well conserved
among mGlu receptors and other VFT folded receptors (e.g. Natriuretic Peptide, Calcium
43ꢀ47
Sensing receptors and Taste receptors),
suggesting a structural role for the chloride ion bound
at this site. In contrast, the chloride ion at Site 2 binds to both lobes and is assumed to modulate
3
3
receptor activation by stabilizing the closed VFT conformation. Site 1 overlaps with the new
pocket where the distal carboxylate of 16g was shown to mimic a chloride ion and allows similar
33, 43
interactions with the receptor (Figure SIꢀ1).
also occur.
We now show that agonist binding at Site 2 may
To rationalize the unique pharmacology of the newly identified orthosteric agonists of groupꢀIII
mGlu receptors, we carried out molecular docking studies. As there is still no crystal structure
available, a new homology model of the mGlu receptor was generated as a first step based on
4
the recently released Xꢀray structure of mGlu bound to the orthosteric agonist 23 (LY354740)
2
48
(
PDB ID: 4XAQ). This structure includes a chloride ion at Site 1 that was not present in the
templates of the first model. All compounds were then docked into the glutamate binding pocket
of the VFT domain of this newly generated mGlu homology model (mGlu ꢀHM) using the
4
4
ACS Paragon Plus Environment
24

Page 25 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
49
50
CDOCKER® program implemented in Discovery Studio® (see Computational Studies
section).
As expected, the amino acid entity and the phosphinic acid of 3ꢀnitrophenyl derivatives 17 bind
22
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
to the glutamate binding site in a fashion similar to 16g whereas, except for 17b, the nitro
ꢀ
substituted benzylic part of the ligands 17 goes into the Cl binding Site 2 (Figure 3). Among 3ꢀ
nitrophenyl derivatives 17, the potent mGlu agonist 17m was intensively characterized and
4
tested in several in vivo models and thus chosen as a reference compound for docking. In Site 2,
the nitro group makes direct Hꢀbond and electrostatic interactions between LB1 residues Ser160
side chain hydroxyl group and Arg60 side chain guanidinium group and LB2 residue Asn286
side chain amide group. In addition, the phenyl ring is oriented and stabilized by van der Waals
contacts with Gly158 (Figure 3). Both S,S (17mꢀ1) and S,R (17mꢀ2) diastereomers of 17m were
docked onto the mGlu ꢀHM. Comparison of the docking poses retrieved for 17mꢀ1 and 17mꢀ2
4
revealed that the two diastereomers exhibit a similar binding pattern with the residues in the
active cleft of mGlu except for the benzylic hydroxyl group. While the αꢀhydroxyl group of
4
1
7mꢀ1 is involved in Hꢀbond interactions with the LB2 residues Asn286 and Ser313 (Figure 3),
in contrast, the αꢀhydroxyl group of 17mꢀ2 accepts Hꢀbonds from the LB1 residues Lys74 and is
in close contact with Glu287 side chain (Figure SIꢀ2). We suppose that direct additional Hꢀbond
interactions with the lower lobe LB2 might provide additional stabilization of the closed active
conformation inducing higher potency in case of the (S,S) isomer (EC : 0.72 ± 0.08 µM) in
5
0
comparison to the (S,R) isomer (EC : 4.4 ± 0.5 µM) of 17m. A similar observation is made with
5
0
the 3ꢀnitrophenol derivatives 17h and 17q.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 80
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. Docking pose derived for 17mꢀ1. A) Site view of the mGlu ꢀHM (blue ribbon) in
4
complex with 17mꢀ1 (sticks). 17mꢀ1 binds in the active cleft formed between the two lobes, LB1
and LB2, in the VFT domain of mGlu . The glutamate entity is circled with a black line and the
4
chloride Site 2 displayed as a green filled circle. B) Site view of the interactions observed
ACS Paragon Plus Environment
26

Page 27 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
between the residues in the active cleft of mGlu and the 3ꢀnitrophenyl and hydroxybenzyl
4
groups of ligand 17mꢀ1. Van de Waals spheres of G258 hydrogens and nitrophenyl carbons are
shown as blue and grey surfaces respectively. C) Two dimensional (2D) ligand interaction
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
diagram of 17mꢀ1 with the residues in the active site of mGlu . Hydrogen bond and ionic
4
interactions are shown as dashed green and orange colored lines, respectively. Ligands are
colored by atom type (C: grey, O: red, N: blue, P: orange).
To study the impact of the αꢀhydroxyl group on potency, we generated compounds 17b, 17i and
1
7n. Removal of the benzylic hydroxyl group in ligands 17i (EC : 24 ± 4 µM) and 17n (EC :
50 50
19 ± 6 µM) led to a significant decrease in their agonist activity against mGlu receptor. With the
4
example of 17hꢀ1 and 17i (Figure SIꢀ3), we demonstrate that the ability of the benzylic hydroxyl
group to form additional close contacts with the residues in LB1 and/or LB2 enhances its agonist
activity for mGlu receptor.
4
In contrast to 17i and 17n, ligand 17b (EC : 0.42 ± 0.08 µM) still shows robust agonist activity
5
0
for mGlu . Surprisingly, docking experiments also suggested a distinct binding mode for the
4
ligand 17b (Figure 4). While the amino acid and the phosphinic acid groups form a similar web
of interactions with the residues in the mGlu active site as derived for other 3ꢀnitrophenyl
4
ꢀ
derivatives (Figure 4A), the nitro group extends towards LB1 instead of the Cl binding Site 2
and forms Hꢀbond and ionic interactions with LB1 residue Ser110 and LB2 residue Glu287,
ꢀ
respectively (Figure 4B). We further checked the probability of a Cl ion being in Site 2 in the
presence of ligand 17b.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 80
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. Docking pose derived for 17b. A) 2D ligand interaction diagram of 17b with the
residues in the active site of mGlu . Hydrogen bond and ionic interactions are shown as dashed
4
green and orange colored lines, respectively. B) Site view of the interactions observed between
the residues in the active cleft of mGlu and the 3ꢀnitrophenyl group of 17b. C) Comparison of
4
ꢀ
the docking solution of 17b with the location of Cl ions in the previously reported mGlu4
3
3
homology model. Chloride van der Waals spheres are displayed as green spheres. The residues
ꢀ
that form Cl binding Site 1 (Thr108 and Ser157) and Site 2 (Ser160 and Asn286) were shown as
ACS Paragon Plus Environment
28

Page 29 of 80
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ꢀ
blue colored balls and sticks. No steric clashes were predicted between 17b and the Cl ion at Site
2
. Ligands are colored by atom type (C: grey, O: red, N: blue, P: orange, Cl: green).
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
We aligned the mGlu ꢀHM/17b complex to our previous mGlu homology model bound to the
4
4
ꢀ
33
glutamate and the two predicted Cl ions (Figure 4C). It appears that, the docking solution of
ꢀ
1
7b does not block the Cl ion from binding to Site 2. To check this hypothesis, we compared the
EC ratios at high (154 mM) and low (2mM) chloride concentrations as described in Tora et al
50
33
2
015 of 17a and 17b to the same ratios for 1 and 16g (Table SIꢀ2). We observed that these
ratios separate in two groups: one with a ratio of 42 (17b and 1), the other with lower ratios of
16ꢀ19 (17a and 16g). The high ratio reveals a strong chloride sensitivity which is line with two
chlorides bound at the mGlu4 binding site wheras a lower ratio is consistent with a weaker
2
2
ꢀ
chloride sensitivity and only one chloride bound. Therefore, we suppose that the two Cl ions
may still bind in the presence of 17b and stabilize the closed active conformation. Positive
ꢀ
33, 51, 52
allosteric behavior of the Cl ion against mGlu receptors has been already shown.
We
ꢀ
suggest that binding of the Cl ion at Site 2 potentiates the agonist activity of 17b and might
explain the unique pharmacology of this ligand among other 3ꢀnitrophenyl derivatives.
Interestingly, replacement of the hydroxyl group at position 4 of the phenyl moiety (in 17h) with
lipophilic substituents such as trifluoromethyl (17c; EC : 0.44 ± 0.06 µM), fluorine (17d; EC :
50
50
0.44 ± 0.05 µM), methyl (17e; EC : 3.3 ± 0.5 µM) or chlorine (17f; EC : 4.7 ± 1.9 µM) is also
5
0
50
wellꢀtolerated. Docking experiments suggested that substituents at position 4 of the phenyl
moiety occupy a small pocket formed by side chains of LB1 residues Val161 and Ser110 and the
backbone of Arg111 (e.g. 17d; Figure 5). Hydrophobic contacts between lipophilic substituents
ACS Paragon Plus Environment
29