Fragment-Based Optimization of Dihydropyrazino-Benzimidazolones as Metabotropic Glutamate Receptor-2 Positive Allosteric Modulators against Migraine

Article abstract of DOI:10.1021/acs.jmedchem.1c00563

Our previous scaffold-hopping attempts resulted in dihydropyrazino-benzimidazoles as metabotropic glutamate receptor-2 (mGluR2) positive allosteric modulators (PAMs) with suboptimal drug-like profiles. Here, we report an alternative fragment-based optimization strategy applied on the new dihydropyrazino-benzimidazolone scaffold. Analyzing published high-affinity mGluR2 PAMs, we used a pharmacophore-guided approach to identify suitable growing vectors and optimize the scaffold in these directions. This strategy resulted in a new fragment like lead (34) with improved druglike properties that were translated to sufficient pharmacokinetics and validated proof-of-concept studies in migraine. Gratifyingly, compound34showed reasonable activity in the partial infraorbital nerve ligation, a migraine disease model that might open this indication for mGluR2 PAMs.

Full text of DOI:10.1021/acs.jmedchem.1c00563

pubs.acs.org/jmc
Article
Fragment-Based Optimization of Dihydropyrazino-
Benzimidazolones as Metabotropic Glutamate Receptor‑2 Positive
Allosteric Modulators against Migraine
György Szabó,* Péter Erdélyi, Sándor Kolok, Mónika Vastag, Attila S. Halász, Istvánné Kis-Varga,
̋
György I. Lévay, Zoltán Béni, János Kóti, István Greiner, and György M. Keseru*
Cite This: J. Med. Chem. 2021, 64, 8607−8620
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Our previous scaffold-hopping attempts resulted in dihydropyr-
azino-benzimidazoles as metabotropic glutamate receptor-2 (mGluR2) positive
allosteric modulators (PAMs) with suboptimal drug-like profiles. Here, we report
an alternative fragment-based optimization strategy applied on the new
dihydropyrazino-benzimidazolone scaffold. Analyzing published high-affinity
mGluR2 PAMs, we used a pharmacophore-guided approach to identify suitable
growing vectors and optimize the scaffold in these directions. This strategy resulted
in a new fragment like lead (34) with improved druglike properties that were
translated to sufficient pharmacokinetics and validated proof-of-concept studies in
migraine. Gratifyingly, compound 34 showed reasonable activity in the partial
infraorbital nerve ligation, a migraine disease model that might open this indication
for mGluR2 PAMs.
subtype selectivity, PAMs have further advantages over
classical agonists. These ligands would prevent the desensitiza-
tion and internalization of the targeted receptor because they
bind only in the presence of orthosteric glutamate and
potentiate the activation of the receptor.8,9
INTRODUCTION
■
Altered regulation of glutamatergic signaling has been long
associated with the pathogenesis of schizophrenia. In addition
to ionotropic glutamate receptors and glutamate transporters,
class II metabotropic glutamate receptors (mGluR2 and
mGluR3) are considered as next-generation targets for
antipsychotic drugs. Activation of these receptors reduces the
transmission at glutamatergic synapses, which nominates this
approach as a promising treatment option for schizophrenia.1
The first preclinical proof of concept was achieved by
mGluR2/3 receptor agonists in preclinical animal models of
psychosis.2 These studies highlighted LY2140023 (1), a
selective mGluR2/3 receptor agonist (Figure 1) demonstrating
an olanzapine-like antipsychotic effect, however, without its
classical aminergic side effects (prolactin elevation, extrapyr-
amidal symptoms, and weight gain).3 Despite the encouraging
phase 2 results, the compound failed a pivotal phase 3 trial.4,5
Considering the possible reasons behind the failed phase 3
trial, we analyzed the pharmacology profile of 1 and suggested
that the limited subtype selectivity on class II mGluRs and the
agonist character of the compound might contribute to the
failure. Based on the results obtained in the phencyclidine
(PCP) and amphetamine hyperlocomotion models with
mGluR2 or mGluR3 KO mice, compounds targeting
mGluR2 are preferred.6,7
This concept has first been realized by two mGluR2 PAM
candidates: AZD8529 (2) and ADX71149 (3), also known as
JNJ-40411813 (Figure 1). Although 2 failed,10 3 showed
efficacy11 in the corresponding phase 2 schizophrenia trials. As
an initial effort toward indication expansion, Addex Ther-
apeutics initiated another trial in anxious depression12;
however, it was terminated due to the lack of efficacy.13
Another option for extending mGluR2-related indication can
be pain modulation. In the peripheral terminals of primary
sensory neurons, group II mGluRs inhibit pain sensory
transmission through cAMP-dependent regulation of TRPV1
channels and TTX-resistant sodium channels. In the dorsal
horn of the spinal cord, group II mGluRs depress pain
transmission at synapses between primary afferent fibers and
second-order sensory neurons, and further central sites, like
amygdala and periaqueductal gray, were described for their
Received: March 29, 2021
Published: June 3, 2021
Improving the mGluR2 selectivity, however, is challenging at
the highly conserved glutamate binding site, which suggests
that allosteric activation in a pocket is structurally different
from that of the other mGluRs.9 In addition to the improved
https://doi.org/10.1021/acs.jmedchem.1c00563
© 2021 American Chemical Society
J. Med. Chem. 2021, 64, 8607−8620
8607

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 1. mGluR2/3 agonists (1, 4) and mGluR2 PAMs (2, 3, 5, and 6).
Figure 2. Strategies followed for the exploration of benzimidazole hits 7 and 8.
antinociceptive effects.14 The mGlu2/3 agonist LY379268 (4)
(Figure 1) reduced allodynia in the L5/L6 spinal nerve ligation
model of neuropathic pain in rats without affecting normal
nociception.15 Less is known, however, about the role of the
8608
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 3. (A) Pharmacophore alignment of mGluR2 PAMs with the benzimidazole core optimized at Richter (10, purple), Pfizer (11, green), and
GSK (12, orange). (B) Pharmacophore alignment of mGluR2 PAMs with the benztriazole core optimized at Richter (13, blue), Richter fragment
(14, salmon), and Merck (15, yellow). (C) Pharmacophore alignment of mGuR2 PAMs discovered at Richter with dihydropyrazino-benzimidazole
(10, purple) and benztriazole (13, blue and 14, salmon) cores. (D) Potential growing vectors identified for the dihydropyrazino-benzimidazole
fragment (37a). Pharmacophore features are depicted as orange circles (aromatic feature), salmon (H-bond acceptor), and green spheres
(hydrophobic contact).
mGlu2 subtype and the effect of mGlu2 PAMs in various pain
states, including migraine. As mGlu2 receptors have been
found in the trigeminal ganglia and dura of rats, and mGlu2
PAM LY487379 (5) (Figure 1) was effective in the plasma
protein extravasation (PPE) model of migraines,16 it is
reasonable to suppose a possible role of this mGluR subtype
in migraines. Recent results with a dual acting mGlu2 PAM
and leukotriene CysLt1 receptor antagonist compound
LY2300559 (6) (Figure 1) showed efficacy in the PPE
preclinical migraine model.17 However, its clinical develop-
ment has been halted in phase 2 trials in migraine due to toxic
effects probably unrelated to the mGlu2 mechanism.
Investigating further indications for mGlu2 PAMs, we decided
to test new chemotypes in the animal models of migraine
assessing their clinical development potential.
Our starting point was a series of dihydropyrazino-
benzimidazoles reported recently from our laboratories. This
novel chemotype was derived from benzimidazole high-
throughput screening (HTS) hits by the cyclization of the
N-methyl substituent to get tricyclic compound 9.18
Interestingly, benzimidazole hits were also considered as a
starting point at Pfizer (7)19 and GSK (8);20 however, these
efforts left the core intact (11, 12). Classical optimization of 9,
which resulted in a large lead 10, combined good in vitro
activity, acceptable pharmacokinetics (PK), and pharmacody-
namics (PD) profiles with only suboptimal drug-like profiles
(Figure 2). Although 10 demonstrated efficacy in the
schizophrenia model of PCP-induced hyperlocomotion in
mice, the moderate efficacy limited its use in further
translational studies.
In order to develop powerful tools for translational mGluR2
research, we followed a fragment-based optimization strategy.
In our previous work,21 we demonstrated that the optimization
of a fragment size hit to a large lead (13) was less fruitful than
a fragment optimization strategy which yielded 14, a potent
mGluR2 PAM fragment reported for the first time (Figure 3).
Results achieved in the absence of structural information
prompted us to extend this approach for the dihydropyrazino-
benzimidazolone core. In contrast to the ligand efficiency (LE)
based-approach applied previously, we analyzed the key
features of known mGluR2 PAMs and developed a
pharmacophore-driven optimization strategy. Herein, we
report the fragment-based optimization of dihydropyrazino-
benzimidazolones, which led to our fragment-like lead 34
8609
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
a
Scheme 1. Synthesis of Final Compounds 16−36
a
Reagents and conditions: (i) NaH, DMF, 60 °C, 0.5 h, 56−81% and (ii) Pd(PPh3)4, ArB(OH)2, 1,2-dimethoxyethane/H2O, K2CO3, reflux, 2 h,
28−42%.
Table 1. SAR Summary of Dihydropyrazino-benzimidazolones Optimized at Position 6
a
b
c
clog P calculated with ChemAxon software. EC50 values were obtained from three independent experiments. Intrinsic clearance (Clint) was
determined as a rate of compound consumption under first-order kinetic conditions using human and rat liver microsomes; ND, not determined.
d
IC50 values were obtained from five independent experiments.
(Figure 2) having good mGluR2 potency and reasonable
absorption, distribution, metabolism, excretion (ADME), and
translational migraine model. Compound 34 showed reason-
PK profiles, which nominated this compound for testing in a
able activity in the partial infraorbital nerve ligation (pIONL)
8610
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 2. SAR Summary of Dihydropyrazino-benzimidazolones Optimized Along the Growing Vectors
a
b
c
clogP calculated with ChemAxon software. EC50 values were obtained from three independent experiments. Intrinsic clearance (Clint) was
determined as a rate of compound consumption under first-order kinetic conditions using human and rat liver microsomes; ND, not determined.
d
IC50 values were obtained from five independent experiments.
model of migraine that might open an additional indication for
mGluR2 PAMs.
followed by treatment of resulting carboxylic acids 41a−e
with aminoethanol, afforded amide benzimidazoles 42a−e.
Finally, cyclization of 42a−e with thionyl chloride yielded, in
one pot, target intermediates 37a−e.
RESULTS AND DISCUSSION
■
Fragment Optimization. Fragment-based optimization of
the recently discovered dihydropyrazino-benzimidazolone core
requires the identification of potential growing vectors. Due to
the lack of structural information on the binding modes of
different mGluR2 PAM chemotypes, we first analyzed a
representative set of optimized benzimidazoles and benzo-
triazoles. Benzimidazole-type mGluR2 PAMs involved 1119
(Pfizer), 1220 (GSK), and our optimized dihydropyrazino-
benzimidazolone derivative (10).18 Alignment of the available
pharmacophores revealed that the N-acceptor feature of the
benzimidazole ring was replaced by carbonyl oxygen in the
dihydropyrazino-benzimidazolone system (Figure 3A). Phar-
macophore alignment of benzotriazole chemotypes including
an optimized Merck compound (15),22 our advanced lead
Chemistry. The synthesis of dihydropyrazino-benzimida-
zolone target compounds 16−36 is outlined in Scheme 1.
They were prepared via SN alkylation of the corresponding
tricyclic core 37a−e with commercially available bromo-
derivatives (38a−k). Phenyl-substituted dihydropyrazino-
benzimidazolone derivatives 16−19 were synthesized via
Suzuki coupling of bromo-derivative 22 with commercially
available aryl boronic acids. The chemical structures of
compounds 16−35 are shown in Tables 1 and 2. Key tricyclic
intermediates 37a−e were prepared following the procedures
shown in Scheme 2.
Treatment of o-phenylenediamines 39a−e with methyl
2,2,2-trichloroacetimidate provided trichloromethyl benzimi-
dazoles 40a−e. Hydrolysis of the trichloromethyl group,
8611
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
a
Scheme 2. Synthesis of Dihydropyrazino-Benzimidazolone Building Blocks 37a−e
a
Reagents and conditions: (i) Methyl 2,2,2-trichloroacetimidate, CH3COOH, rt, 3 h, 76−91%; (ii) 2 N NaOH, 0 °C, 1 h, 78−93%; and (iii)
Aminoethanol, EDC, HOBt, DMF, rt, 3 h, 56−78%; (iv) SOCl2, DMF, 150 °C, 3 h, 34−56%.
Figure 4. Initial strategy for the fragment optimization of dihydropyrazino-benzimidazolone core 37a.
(13),21 and the optimized fragment (14)21 suggested that in
addition to the N-acceptor feature, a partial negative charge is
preferred (Figure 3B).
lipophilic substituents at position 6 of the tricyclic core also
seemed to be beneficial for potency improvement. Gratifyingly,
we introduced a phenyl ring (16), which resulted in a
significant boost to mGlu2 PAM potency (mGluR2 PAM EC50
= 0.63 μM vs 6.17 μM). Although incorporation of the phenyl
ring resulted in increased lipophilicity (clog P = 3.41 vs 1.77),
we were encouraged by the relative balance among mGlu2
PAM potency, LE, and MW compared to parent compound
36.
For our initial exploration, we kept constant the N-
cyclopropylmethyl group of the core and explored different
substituents and decoration patterns of the C-6 phenyl ring
(16−19, Table 1).
Initial SAR showed that fluoro substitutions (17,18) offered
no improvements in mGluR2 PAM potency. When the
fluorine atom in compound 18 was replaced by more lipophilic
chlorine (19), a twofold increase in potency was obtained,
which is in line with the growth vector used in our previous
work.21 Unfortunately, the improvement in potency was
associated with a significant increase in the lipophilicity (clog
P = 4.02), making the drug-like properties of 19 less than
optimal. Our goal was to systematically investigate the SAR on
the C-6 position of the tricyclic core in order to identify groups
Interestingly, this alignment confirmed that we discovered a
new growing vector different from that used by the Merck
team. Considering the beneficial N-acceptor and partial
negative features of the heterocyclic cores, we next aligned
our benzotriazoles 13 and 14 with optimized dihydropyrazino-
benzimidazolone 10 (Figure 3C). In this alignment, we fit the
aromatic N-acceptors of benzotriazoles to that of the
benzimidazolone ring that allowed carbonyl oxygen of the
dihydropyrazino ring to overlap with the halogen substituents
of benzotriazoles. This arrangement of the heterocyclic rings
clearly suggested two potential growing vectors for the
fragment-based optimization of dihydropyrazino-benzimidazo-
lone core 37a (Figure 3D).
Guided by the pharmacophore model, we investigated minor
changes to the core without growing the molecule substantially
(Figure 4). Based on our structure−activity relationship (SAR)
knowledge gained in the mGluR2 program, we quickly found a
marked improvement in mGluR2 PAM potency by introducing
the less lipophilic cyclopropylmethyl moiety on the amide
position of the tricyclic core (36). As shown in Figure 3C,
8612
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 3. Physicochemical Properties and in vitro ADME Parameters of 10 and 34
that would improve mGlu2 PAM potency maintaining the
druglike properties. Shifting our attention to less lipophilic
groups led to the synthesis of compounds 20−23. Replace-
ment of the phenyl ring by the nitro group (20) resulted in a
significant decrease in activity. Better results were obtained
with groups such as chlorine (21), bromine (22), and
trifluoromethyl (23), which resulted in a great balance
among mGluR2 PAM potency, lipophilicity, and LE. Addi-
tionally, compounds 21−23 showed acceptable metabolic
stability both in human and rat liver microsomes (HLM and
RLM, respectively) and were free from hERG channel
A significant improvement in the metabolic profile was found
with trifluoromethyl analogue 31; however, this increase in
metabolic stability was accompanied by a stronger interaction
with the hERG channel (IC50 = 3 μM). Besides this result, the
overall druglike profile of 31 is far from optimal. Apart from
improved LE from 10 to 31 (LE = 0.34 and 0.40, respectively),
we recognized that 31 is still quite lipophilic with a clog P of
4.19. We believe that the high lipophilicity of 31 and related
analogues is the major contributor to the hERG liability.
Therefore, we next focused on strategies decreasing the
lipophilicity of compounds. First, we changed the phenyl
ring to the corresponding N- heterocycle incorporating
pyridine nitrogen. As anticipated, basic pyridine nitrogen
(32) had a beneficial effect on the lipophilicity (clog P = 3.09
vs 4.08); however, 32 was found to be 18-fold less active (EC50
= 0.54 μM) than its phenyl analogue 27 (EC50 = 0.03 μM).
Continuing our efforts on the reduction of the lipophilicity, we
replaced the phenyl group of 27 with cycloalkyl substituents
(Table 2). Enlargement of the cyclopropyl ring (13) to the
four-membered cyclobutyl ring (33) was not only well
tolerated in terms of mGlu2 PAM potency, metabolic stability,
and hERG profile; it also lowered the clog P by more than a
unit (clog P = 2.98) as compared to the phenyl derivative (27,
clog P = 4.08). A trifluoromethyl substituent at the C-6
position (34) resulted in the same potency and safety profile,
but further improvement in the metabolic profile (both in
HLM and RLM) was observed. Interestingly, further enlarge-
ment of the cycloalkyl ring led to compound 35 with increased
hERG affinity. These data confirm that the overall lipophilicity
of the compounds contributes to their interaction with the
hERG channel. Compound 34 represented a promising lead
molecule with high mGluR2 PAM activity (EC50 = 0.06 μM),
low MW (323.3), high LE (0.44), and reasonable lipophilicity
(clog P = 3.09). Extended ADME characterization revealed
interaction, as confirmed in patch clamp assay (hERG IC50
30 μM) (Table 1).
=
Searching for the optimal substituents at the C-6 position
led us to expand our investigations toward bulkier and more
lipophilic groups at the amide position of the tricyclic core in
order to obtain compounds with improved potency (Table 2).
Surprisingly, compound 24, having a more lipophilic phenyl
substituent, showed only a slight improvement in potency
(EC50 = 0.25 μM) as compared to 22 (EC50 = 0.35 μM). Next,
we turned our attention to explore different substituents of the
phenyl ring (25−31). Focusing SAR knowledge on this region
suggested that aryl substitution in this series might be
preferred. In fact, the best results were obtained with more
lipophilic substituents (27, 28), reinforcing that increased
lipophilicity is advantageous in this region.
Further profiling of 27 and related analogues revealed
potential metabolic stability and cardiac safety issues. For
instance, compound 27 showed significantly improved
mGluR2 PAM activity (EC50 = 0.03 μM); however, this was
associated with extensive metabolic degradation in RLM.
Moreover, 27 showed a moderate hERG channel interaction
(IC50 = 4.7 μM) in patch clamp assay, which may result in
QTc prolongation and can lead to serious cardiac arrhythmias.
8613
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Table 4. PK Parameters of 34 in Rats
b
a
b
b
b
b
c
cmpd
route
Cl (mL/min/kg)
60
Vd (L/kg)
Thalf (h)
Tmax (h)
AUC0−5h (ng/mL*h)
Cmax (ng/mL)
F (%)
B/P
34
IV
PO
1.33
0.26
0.91
833 (0−2 h)
1817
0.5
1363
67
1.02
a
Study in male Wistar rats dosed at 10 mg/kg p.o., formulated in 5% Tween 80 in d.w. at pH = 6.01, and 3 mg/kg iv, formulated in 5% DMA/
b
c
43.75% HPbCD at pH = 6.83. Plasma values are the mean of four animals. Ratios are calculated based on AUC0−5h
.
Figure 5. Effect of compound 34 (A, n = 8−10/group) and JNJ-40068782 (B, n = 7−8/group) in the pIONL model in ratssingle oral dosing. %
reversals (maximal values signed in bold); **p < 0.01 (vs Vehicle, ANOVA/Dunnett).
that 34 also showed improved properties in permeability and
plasma protein binding as compared to our previously
identified large lead, 10 (Table 3). Due to the much decreased
lipophilicity, 34 has almost three times higher aqueous
solubility and in line with the lower MW, it showed much
improved passive permeability. Improved physicochemical
properties of 34 also reflected better in the CYP profile as
compared to that of 10.
Its improved ADME profile has been translated to promising
PK data (Table 4), which include reasonably high exposure
with acceptable bioavailability and brain penetration. Although
the elimination half-life was shorter than expected, AUCmax was
8614
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
Figure 6. Plots of benzimidazole type mGluR2 PAM pEC50 values vs clog P (A) and heavy atom count (B).
achieved rapidly that is in line with the general requirement in
migraine indication.
potency comparable to our first lead (10) and to that of Pfizer
(11) but having much lower lipophilicity and smaller size.
Consequently, this optimization improved the LE and
lipophilic efficiency significantly. Considering the potency
and physicochemical profile of 34, this optimization clearly
shows the benefits of fragment-based approaches. Interestingly,
these advantages were realized successfully despite the lack of
structural information by carefully analyzing structure−activity
relationships and pharmacophore arrangements explored by
competitors and our teams.
Finally, our fragment like lead (34) and a druglike reference
compound from a different chemotype JNJ-4006878223 were
tested in a translational migraine model in rats. Both
compounds showed dose-dependent and significant antiallo-
dynic effects in the infraorbital nerve ligation model (Figure 5),
which confirmed the involvement of mGluR2 receptors in the
pathophysiology of migraine. Consequently, migraine can be
considered as a promising indication for mGluR2 PAMs.
The optimization path was analyzed in the context of other
literature mGluR2 optimization projects started from the
benzimidazole hits. For this comparative analysis, we used
trajectories depicted in the potency versus lipophilicity and the
potency versus heavy atom count plots (Figure 6).24 These
plots are useful following the changes in LE and lipophilic
ligand efficiency (LLE) during the optimization path.
First, we investigated the classical optimization of the HTS
hits identified at Pfizer (7) and GSK (8): Both optimization
programs realized a constant LLE path and provided lead
compounds 11 and 12 with improved potency. The potency
improvement was larger for the Pfizer compound, and the
lipophilic efficiency was also improved, reaching that of the
GSK compound. In terms of LE, the Pfizer optimization (red
arrow in Figure 6) kept LE almost constant, while LE was
slightly decreased during the GSK optimization (blue arrow in
Figure 6).
Turning to our optimization programs, the hit-to-lead phase
resulted in the dihydropyrazino-benzimidazole starting point
(9) obtained by scaffold hopping from the benzimidazole core
of the HTS hits. In terms of potency, lipophilicity, and size,
this compound was closer to the Pfizer hit (7) than the GSK
starting point (8). Classical optimization of 9 (yellow arrow in
Figure 6) provided our first lead compound (10) with potency
and size similar to those of the Pfizer lead (11); however, the
lipophilicity of 10 was better than that of 11 and was closer to
that of the GSK lead (12). Although this optimization slightly
decreased the LE, the lipophilic efficiency was much improved.
The optimization path discussed in this work (green arrow in
Figure 6), however, improved both LE and LLE significantly.
Fragment optimization of the dihydropyrazino-benzimidazole
starting point (9) resulted in a fragment size lead (34) with a
CONCLUSIONS
■
Here, we report the optimization of dihydropyrazino-
benzimidazoles by fragment-based drug discovery (FBDD)
principles. Our starting point was a series of dihydropyrazino-
benzimidazoles reported recently from our laboratories. This
novel chemotype was derived from benzimidazole HTS hits by
the cyclization of the N-methyl substituent to get tricyclic 9.18
Classical optimization of 9 resulted in a large and complex lead
10. Because 10 was a moderately effective compound with only
suboptimal drug-like profiles, our first attempts were focused
on improving these features. In order to maximize the potential
of dihydropyrazino-benzimidazoles, we finally changed to a
fragment-based optimization strategy focused on the tricyclic
core. The lack of structural information on the target, however,
represented a significant challenge in the FBDD program. By
implementing the pharmacophore-guided strategy, we were
able to identify the most effective growing vectors and
produced a series of dihydropyrazino-benzimidazole analogues
with reasonable potency and improved druglike properties.
Careful optimization of the initial hit (37a) finally led to 34
that still maintains the fragment features. Compound 34
represented a promising lead molecule with high mGluR2
PAM activity (EC50 = 0.06 μM), low MW (323.3), reasonable
LE (0.44), and lipophilicity (clog P = 3.09). Extended ADME
characterization revealed that 34 also showed improved
properties in permeability as compared to our previously
identified more complex lead (10). Its improved ADME
properties have been translated to the promising PK profiles
that include reasonably high exposure with acceptable
bioavailability and brain penetration. These data nominated
8615
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
STEP 4. 6-Nitro-3,4-dihydropyrazino[1,2-a]benzimidazol-1(2H)-
one (37b). To a stirred suspension of N-(2-hydroxyethyl)-7-nitro-1H-
benzimidazole-2-carboxamide (42b, 3.81 g, 15.25 mmol) in DMF (30
mL), thionyl chloride (1.34 mL, 18.3 mmol) was added dropwise at
room temperature. The mixture was stirred for 3 h at 150 °C. The
reaction mixture was poured into a mixture of 40 mL of ice water and
15 mL of saturated sodium carbonate solution. The precipitated
product was collected by filtration and washed 3 times with 20 mL of
water. Yield: 1.61 g (45%).
34 for proof of concept testing in an animal model of migraine.
Our results achieved by a new chemotype in an independent
animal model confirmed that migraine might be considered as
an additional indication of mGluR2 PAMs.
EXPERIMENTAL SECTION
■
General Experimental Methods. All reactions were carried out
using dry nitrogen. Commercially available reagents were used
without further purification. Solvents and gases were dried according
to standard procedures. Organic solvents were evaporated with
reduced pressure using a rotary evaporator. The purity of final
compounds was verified using a Merck-Hitachi Lachrom Elite HPLC
system. A linear gradient using water and 0.1% trifluoroacetyl (TFA)
(solvent A) and acetonitrile and 0.1% TFA (solvent B) (t = 0 min, 0%
B, t = 7.5 min, 70% B, t = 9.7 min, 90% B, t = 12.5 min, 90% B) (12.5
min) was employed on a Chromolith RP18e (3 × 100 mm) column.
The flow rate was set to 1.7 mL/min, and the column was maintained
at 40 °C. The purity of final compounds was assessed using UV
detection (Diode Arry Detector: Merck-Hitachi Lachrom Elite L-
2450) at 215 nm; all tested compounds were ≥95% pure. NMR
spectra were recorded at 298K on a 400 (1H: 399.8 MHz; 13C: 100.7
MHz) or 500 (1H: 499.9 MHz; 13C: 125.7 MHz) MHz Bruker
AVANCE III HD spectrometer equipped with a liquid N2-cooled
Prodigy probe or with a room temperature BBO probe, respectively.
1H and 13C chemical shifts were referenced to internal TMS, and 19F
and 15N chemical shifts are referenced to external TFA and CH3NO2,
STEP 5. 2-(Cyclopropylmethyl)-6-nitro-3,4-dihydropyrazino[1,2-
a]benzimidazol-1(2H)-one (20). To a stirred suspension of 6-nitro-
3,4-dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (37b, 0.30 g, 1.3
mmol) in N,N-dimethylformamide (15 mL), sodium hydride (60%
dispersion in mineral oil) (0.12 g, 2.86 mmol) was added in portions
at room temperature and then stirred for 30 min at 60 °C. To this
solution, (bromomethyl)cyclopropane (38a, 0.21 g, 1.56 mmol) was
added and stirred for 1 h at 100 °C. The reaction mixture was poured
into water, and the precipitated product was collected by filtration.
The product could be recrystallized from isopropanol. Yield: 0.28 g
(75%). EI-HRMS: calcd for C14H14O3N4, 286.10604; found,
1
286.10627. δ = 0.80 ppm. H NMR (400 MHz, DMSO-d6): δ 8.22
(dd, J = 1.0, 8.1 Hz, 1H), 8.13−8.18 (dd, J = 1.0, 8.1 Hz, 1H), 7.43 (t,
J = 8.1 Hz, 1H), 4.64−4.68 (m, 2H), 3.89−3.94 (m, 2H), 3.44 (d, J =
7.0 Hz, 2H), 2.34−2.43 (m, 1H), 0.48−0.53 (m, 2H), 0.29−0.35 (m,
2H); 13C NMR (101 MHz, DMSO-d6): δ 155.9, 146.1, 145.3, 133.2,
127.8, 126.0, 122.1, 121.1, 49.9, 45.0, 44.3, 9.1, 3.1.
6-Chloro-2-(cyclopropylmethyl)-3,4-dihydropyrazino[1,2-a]-
benzimidazol-1(2H)-one (21). It was prepared according to the
procedure described for 20 using 3-chlorobenzene-1,2-diamine (39c)
in step 1. ESI-HRMS: calcd for C14H15ON3Cl [M + H]+, 276.08982;
found, 276.08941; δ = −1.47 ppm. 1H NMR (400 MHz, DMSO-d6):
δ 7.75 (dd, J = 0.9, 8.2 Hz, 1H), 7.43 (dd, J = 0.9, 7.8 Hz, 1H), 7.30
(t, J = 7.7 Hz, 1H), 4.78−4.86 (m, 2H), 3.96−4.04 (m, 2H), 3.44 (d,
J = 7.1 Hz, 2H), 1.04−1.14 (m, 1H), 0.48−0.54 (m, 2H), 0.28−0.36
(m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 155.5, 143.8, 143.1,
129.1, 124.7, 123.1, 119.4, 115.4, 49.5, 44.5, 41.8, 8.6, 2.6.
1
respectively. Standard H and 13C NMR spectra were recorded in all
cases, while in some cases HSQC, HMBC, 15N-HMBC, and 19F
spectra were collected as well, using the pulse sequences available in
the Topspin 3.5 sequence library. For data analysis and reporting,
ACD/Labs NMR Workbook Suite 2017.1.3 was used.
Electrospray ionization-high-resolution mass spectrometry (ESI-
HRMS) analyses were performed in the positive ion mode on a
Thermo Velos Pro Orbitrap Elite (Thermo Fisher Scientific, Bremen,
Germany) system. The protonated molecular ion peaks were
fragmented by collision-induced dissociation, using helium as the
collision gas, at a normalized collision energy of 35%. The samples
were dissolved in methanol prior to analysis. Electron impact high-
resolution MS measurements (EI-HRMS) were performed on a
Thermo Q-Exative GC Orbitrap mass spectrometer (Thermo Fisher
Scientific, Bremen, Germany). The ion source temperature was set at
250 °C, and the applied ionization energy was 70 eV. A resolving
power of 30,000 (full width at half-maximum) at m/z 400 was used.
Data acquisition and analysis were accomplished with Xcalibur
software version 4.0 (Thermo Fisher Scientific) in all cases.
General Procedure for the Synthesis of Target Compounds
(20−36). STEP 1. 7-Nitro-2-(trichloromethyl)-1H-benzimidazole
(40b). To a stirred solution of 3-nitrobenzene-1,2-diamine (39b, 10
g, 65.2 mmol) in acetic acid (120 mL), methyl 2,2,2-trichloroaceti-
midate (8.08 mL, 65.2 mmol) was added at ambient temperature. The
as-obtained mixture was stirred for 3h at ambient temperature and
then poured into water (100 mL). The precipitated product was
collected by filtration and washed 2 times with 50 mL of water. Yield:
16 g (87%).
STEP 2. 7-Nitro-1H-benzimidazole-2-carboxylic Acid (41b). 7-
nitro-2-(trichloromethyl)-1H-benzimidazole (40b, 16 g, 57 mmol)
was added in portions at 0−5 °C to a stirred solution of 200 mL of
2N NaOH. The as-obtained solution was stirred for 1 h at 0−5 °C,
and then, the pH value was adjusted to 2 with 6 N HCl solution. The
precipitated product was collected by filtration and washed 2 times
with 50 mL of water. Yield: 10.4 g (88%).
STEP 3. N-(2-Hydroxyethyl)-7-nitro-1H-benzimidazole-2-carbox-
amide (42b). A mixture of 7-nitro-1H-benzimidazole-2-carboxylic
acid (41b, 12.4 g, 60 mmol), aminoethanol (5.1 mL, 60 mmol), 1-
hydroxybenzotriazole (10.1 g, 66 mmol), and N-(3-dimethylamino-
propyl)-N′-ethylcarbodiimide hydrochloride (12.56 g, 66 mmol) in
120 mL of N,N-dimethylformamide was stirred for 3 h at room
temperature and then concentrated in vacuo. The residue could be
crystallized from ethanol (70 mL). Yield: 7.7 g (51%).
6-Bromo-2-(cyclopropylmethyl)-3,4-dihydropyrazino[1,2-a]-
benzimidazol-1(2H)-one (22). It was prepared according to the
procedure described for 20 using 3-bromobenzene-1,2-diamine (39d)
in step 1. ESI-HRMS: calcd for C14H15ON3Br [M + H]+: 320.03930;
found: 320.03980; δ = 1.56 ppm. 1H NMR (400 MHz, DMSO-d6): δ
7.79 (dd, J = 0.8, 8.1 Hz, 1H), 7.59 (dd, J = 0.8, 7.7 Hz, 1H), 7.24 (t, J
= 8.0 Hz, 1H), 4.86 (dd, J = 5.3, 6.7 Hz, 2H), 4.00 (dd, J = 5.3, 6.8
Hz, 2H), 3.44 (d, J = 7.1 Hz, 2H), 3.27−3.30 (m, 1H), 2.07−2.12 (m,
1H), 1.17−1.31 (m, 1H), 1.09 (s, 1H), 0.48−0.54 (m, 2H), 0.29−
0.35 (m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 155.5, 143.7,
143.2, 130.3, 128.1, 123.6, 119.9, 102.3, 49.4, 44.4, 41.7, 8.6, 2.6.
2-(Cyclopropylmethyl)-6-(trifluoromethyl)-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (23). It was prepared according to
the procedure described for 20 using 3-(trifluoromethyl)benzene-1,2-
diamine (39e) in step 1. ESI-HRMS: calcd for C15H15ON3F3Cl [M +
1
H]+, 310.11617; found, 310.11571; delta = −1.49 ppm. H NMR
(400 MHz, DMSO-d6): δ 8.13 (d, J = 8.2 Hz, 1H), 7.81 (d, J = 7.6
Hz, 1H), 7.49 (t, J = 7.9 Hz, 1H), 4.51 (t, J = 5.9 Hz, 2H), 4.02 (dd, J
= 5.3, 6.79 Hz, 2H), 3.45 (d, J = 7.0 Hz, 2H), 1.00−1.17 (m, 2H),
0.41−0.62 (m, 2H), 0.26−0.37 (m, 2H); 13C NMR (101 MHz,
DMSO-d6): δ 155.4, 144.0, 143.5, 128.7, 125.4, 122.1, 121.7, 122.9,
112.3, 49.5, 44.3, 42.2, 28.4, 8.6, 2.6.
2-Benzyl-6-bromo-3,4-dihydropyrazino[1,2-a]benzimidazol-
1(2H)-one (24). It was prepared according to the procedure described
for 22 using benzyl bromide (39b) in step 5. ESI-HRMS: calcd for
C17H115ON3Br [M + H]+, 356.03930; found: 356.03917; δ = −0.37
ppm. H NMR (400 MHz, DMSO-d6): δ 7.81 (dd, J = 0.8, 8.1 Hz,
1H), 7.59 (dd, J = 0.9, 7.7 Hz, 1H), 7.36−7.40 (m, 4H), 7.25 (t, J =
8.0 Hz, 1H), 7.22−7.27 (m, 1H), 4.85 (dd, J = 5.3, 6.7 Hz, 2H), 4.78
(s, 2H), 3.86 (dd, J = 5.3, 6.7 Hz, 2H); 13C NMR (101 MHz, DMSO-
d6): δ 156.3, 144.2, 143.6, 136.5, 130.9, 128.6, 128.5, 127.6, 127.3,
124.2, 120.5, 102.8, 49.0, 44.8, 42.2.
6-Bromo-2-[(2-fluorophenyl)methyl]-3,4-dihydropyrazino[1,2-
a]benzimidazol-1(2H)-one (25). It was prepared according to the
8616
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
procedure described for 22 using 1-(bromomethyl)-2-fluorobenzene
(38c) in step 5. EI-HRMS: calcd for C17H13ON3BrF, 373.02205;
found, 373.02258. δ = 1.41 ppm. 1H NMR (400 MHz, DMSO-d6): δ
7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.59 (dd, J = 0.9, 7.7 Hz, 1H), 7.47 (dt,
J = 1.7, 7.7 Hz, 1H), 7.33−7.41 (m, 1H), 7.22−7.27 (m, 1H), 7.25 (t,
J = 8.0 Hz, 1H), 7.18−7.23 (m, 1H), 4.85−4.90 (m, 2H), 4.82 (s,
2H), 3.83−3.99 (m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 160.3,
156.4, 144.2, 143.5, 130.8, 129.9, 129.5, 128.7, 124.5, 124.2, 123.3,
120.5, 115.3, 102.8, 45.1, 43.4, 42.2.
6-Bromo-2-[(4-fluorophenyl)methyl]-3,4-dihydropyrazino[1,2-
a]benzimidazol-1(2H)-one (26). It was prepared according to the
procedure described for 22 using 1-(bromomethyl)-4-fluorobenzene
(38d) in step 5. EI-HRMS: Calcd. for C17H13ON3BrF: 373.02205;
found: 373.02258. δ = 1.41 ppm. 1H NMR (400 MHz, DMSO-d6): δ
7.81 (dd, J = 0.9, 8.1 Hz, 1H), 7.59 (dd, J = 0.9, 7.7 Hz, 1H), 7.37−
7.50 (m, 2H), 7.25 (t, J = 8.0 Hz, 1H), 7.16−7.22 (m, 2H), 4.80−
4.88 (m, 2H), 4.76 (s, 2H), 3.76−3.97 (m, 2H); 13C NMR (101
MHz, DMSO-d6): δ 161.4, 156.3, 144.2, 143.5, 132.8, 129.7, 128.7,
124.2, 120.5, 115.3, 102.8, 48.3, 44.7, 42.2.
2H), 4.50 (t, J = 5.8 Hz, 2H), 3.85−3.93 (m, 2H). 13C NMR (101
MHz, DMSO-d6): δ 155.8, 143.8, 143.5, 135.1, 131.4, 129.1, 128.8,
128.0, 125.5, 122.2, 121.8, 122.9, 112.3, 47.9, 44.2, 42.1.
6-Bromo-2-[(6-chloropyridin-3-yl)methyl]-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (32). It was prepared according to
the procedure described for 22 using 5-(bromomethyl)-2-chloropyr-
idine (38i) in step 5. EI-HRMS: calcd for C16H12ON4BrCl,
1
389.98775; found, 389.98794. δ = 0.48 ppm H NMR (400 MHz,
DMSO-d6): δ 8.47 (dd, J = 0.4, 2.6 Hz, 1H), 7.89 (dd, J = 2.6, 8.3 Hz,
1H), 7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.59 (dd, J = 0.9, 7.7 Hz, 1H),
7.52 (dd, J = 0.4, 8.3 Hz, 1H), 7.25 (t, J = 8.0 Hz, 1H), 4.82−4.92 (m,
2H), 4.80 (s, 2H), 3.89−3.96 (m, 2H); 13C NMR (101 MHz,
DMSO-d6): δ 156.6, 149.3, 149.2, 144.2, 143.5, 139.3, 132.1, 130.8,
128.7, 124.2, 124.1, 120.5, 102.8, 46.2, 45.0, 42.2.
6-Bromo-2-(cyclobutylmethyl)-3,4-dihydropyrazino[1,2-a]-
benzimidazol-1(2H)-one (33). It was prepared according to the
procedure described for 22 using (bromomethyl)cyclobutene (38j) in
step 5. EI-HRMS: calcd for C15H16ON3Br, 333.04713; found,
1
333.04758. δ = 1.36 ppm. H NMR (500 MHz, DMSO-d6): δ 7.78
(dd, J = 0.9, 8.2 Hz, 1H), 7.58 (dd, J = 0.9, 7.7 Hz, 1H), 7.24 (t, J =
8.0 Hz, 1H), 4.77−4.85 (m, 2H), 3.85−3.91 (m, 2H), 3.58 (d, J = 7.4
Hz, 2H), 2.58−2.71 (m, 1H), 1.97−2.10 (m, 2H), 1.82−1.92 (m,
2H), 1.70−1.82 (m, 2H); 13C NMR (126 MHz, DMSO-d6): δ 156.0,
144.2, 143.7, 130.8, 128.5, 124.1, 120.4, 102.7, 50.8, 45.2, 42.2, 33.4,
25.7, 17.9.
6-Bromo-2-[(4-chlorophenyl)methyl]-3,4-dihydropyrazino[1,2-
a]benzimidazol-1(2H)-one (27). It was prepared according to the
procedure described for 22 using 1-(bromomethyl)-4-chlorobenzene
(38e) in step 5. ESI-HRMS: calcd for C17H14ON3BrCl [M + H]+,
390.00033; found, 390.00084; delta = 1.31 ppm. 1H NMR (400 MHz,
DMSO-d6): δ 7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.59 (dd, J = 0.9, 7.7 Hz,
1H), 7.41−7.44 (m, 2H), 7.43 (s, 2H), 7.25 (t, J = 8.0 Hz, 1H),
4.82−4.88 (m, 2H), 4.77 (s, 2H), 3.82−3.90 (m, 2H); 13C NMR
(101 MHz, DMSO-d6): δ 159.7, 155.9, 143.7, 142.9, 131.1, 130.3,
128.2, 127.1, 123.7, 122.6, 120.2, 120.0, 118.2, 102.3, 44.7, 42.7, 41.7.
6-Bromo-2-[(4-bromophenyl)methyl]-3,4-dihydropyrazino[1,2-
a]benzimidazol-1(2H)-one (28). It was prepared according to the
procedure described for 22 using 1-(bromomethyl)-4-bromobenzene
(38f) in step 5. ESI-HRMS: calcd for C17H141ON3Br2 [M + H]+:
433.94981; found: 433.94981; δ = −1.37 ppm. H NMR (400 MHz,
DMSO-d6): δ 7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.59 (dd, J = 0.9, 7.7 Hz,
1H), 7.54−7.58 (m, 2H), 7.36 (d, J = 8.4 Hz, 2H), 7.25 (t, J = 8.0 Hz,
1H), 4.81−4.89 (m, 2H), 4.75 (s, 2H), 3.83−3.90 (m, 2H); 13C
NMR (101 MHz, DMSO-d6): δ 155.9, 143.7, 143.0, 135.6, 130.9,
130.3, 129.4, 128.2, 123.7, 120.0, 119.9, 102.3, 48.0, 44.3, 41.7.
6-Bromo-2-[(4-chloro-2-fluorophenyl)methyl]-3,4-
dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (29). It was pre-
pared according to the procedure described for 22 using 1-
(bromomethyl)-4-chloro-2-fluorobenzene (38g) in step 5. ESI-
HRMS: calcd for C17H13ON3BrClF [M + H]+, 407.99091; found,
2-(Cyclobutylmethyl)-6-(trifluoromethyl)-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (34). It was prepared according to
the procedure described for 23 using (bromomethyl)cyclobutene
(38j) in step 5. EI-HRMS: calcd for C16H16ON3F3, 323.12400; found,
323.12350. δ = −1.54 ppm. 1H NMR (500 MHz, DMSO-d6): δ 8.13
(1H, d, J = 8.1 Hz, H-3), 7.80 (1H, d, J = 7.6 Hz, H-6), 7.49 (1H, t, J
= 7.8 Hz, H-5), 4.47 (2H, t, J = 5.9 Hz, H-12), 3.85−3.97 (2H, m, H-
13), 3.59 (2H, d, J = 7.5 Hz, H-16), 2.67 (1H, spt, J = 7.7 Hz, H-17),
1.98−2.08 (2H, m, H-19, 20), 1.83−1.92 (2H, m, H-18), 1.75−1.85
(2H, m, H-19, 20); 13C NMR (125 MHz, DMSO-d6): δ 156.0 (C-
10), 144.4 (C-1), 144.0 (C-8), 129.2 (C-2), 125.9 (C-3), 122.6 (q,
JC−F = 5.4 Hz C-6), 122.2 (C-5), 123.4 (q, JC−F = 271.3 Hz, C-14),
112.8 (q, JC−F = 33.3 Hz, C-4), 50.8 (C-16), 45.1 (C-13), 42.7 (q,
JC−F = 3.6 Hz, C-12), 33.4 (C-17), 25.7 (C-19), 17.9 (C-18); 19F
NMR (470 MHz, DMSO-d6): δ = −54.9 (1F, s, F-21); partial 15N
NMR (50.7 MHz, DMSO-d6): δ = −261.5 (N-11), −227.2 (N-9).
2-[(Bicyclo[2.2.1]heptan-2-yl)methyl]-6-(trifluoromethyl)-3,4-
dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (35). It was pre-
pared according to the procedure described for 23 using 2-
(bromomethyl)bicyclo[2.2.1]heptane (38k) in step 5. ESI-HRMS:
calcd for C19H21ON3F3 [M + H]+: 364.16312; found: 364.16316; δ =
0.10 ppm. The sample is a mixture of four diastereomers.
1
407.99045; delta = −1.12 ppm. H NMR (400 MHz, DMSO-d6): δ
7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.60 (dd, J = 0.9, 7.7 Hz, 1H), 7.51 (t, J
= 8.3 Hz, 1H), 7.48 (dd, J = 2.1, 10.1 Hz, 1H), 7.25−7.32 (m, 1H),
7.25 (t, J = 8.0 Hz, 1H), 4.82−4.92 (m, 2H), 4.79 (s, 2H), 3.88−3.96
(m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 159.7, 155.9, 143.7,
142.9, 132.3, 130.8, 130.3, 128.2, 124.2, 123.7, 122.2, 120.0, 115.5,
102.3, 44.7, 42.6, 41.7.
1
Consequently, two sets of signals were detected in the H and 13C
1
NMR spectra in a ca. 3 (major) to 2 (minor) ratio. Major: H NMR
(400 MHz, DMSO-d6): δ 8.13 (d, J = 8.2 Hz, 1H), 7.80 (d, J = 7.7
Hz, 1H), 7.49 (t, J = 7.9 Hz, 1H), 4.48 (t, J = 5.9 Hz, 2H), 3.85−4.00
(m, 2H), 3.52−3.64 (m (AB), 2 H), 2.17−2.21 (m, 2H), 1.65−1.78
(m, 2H), 1.50−1.54 (m, 1H), 1.32−1.35 (m, 2H), 1.26−1.28 (m,
1H), 1.05−1.20 (m, 2H), 0.74 (ddd, J = 2.1, 5.5, 12.1 Hz, 1H);
6-Bromo-2-[(4-bromo-2-fluorophenyl)methyl]-3,4-
dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (30). It was pre-
pared according to the procedure described for 22 using 1-
(bromomethyl)-4-bromo-2-fluorobenzene (38h) in step 5. ESI-
HRMS: calcd for C17H13ON3Br2F [M + H]+, 451.94039; found:
1
minor: H NMR (400 MHz, DMSO-d6): δ 8.13 (d, J = 8.2 Hz, 1H),
7.80 (d, J = 7.7 Hz, 1H), 7.49 (t, J = 7.9 Hz, 1H), 4.48 (t, J = 5.9 Hz,
2H), 3.85−4.00 (m, 2H), 3.47 (dd, J = 8.7, 13.3 Hz, 1H), 3.25 (dd, J
= 7.1, 13.3 Hz, 1H), 2.24−2.31 (m, 1H), 2.21−2.24 (m, 1H), 2.04−
2.09 (m, 1H), 1.82−1.89 (m, 1H), 1.47−1.49 (m, 2H), 1.41−1.47
(m, 1H), 1.35−1.40 (m, 1H), 1.05−1.20 (m, 3H); major: 13C NMR
(101 MHz, DMSO-d6): δ 155.7, 144.5, 144.0, 129.2, 125.9, 123.4,
122.5, 122.2, 112.8, 47.5, 44.5, 42.7, 39.4, 38.2, 38.0, 36.3, 33.8, 29.6,
22.2; minor: 13C NMR (101 MHz, DMSO-d6): δ 156.1, 144.5, 144.0,
129.2, 125.9, 123.4, 122.5, 122.2, 112.8, 50.1, 44.9, 42.7, 39.7, 38.4,
35.9, 34.9, 34.5, 29.2, 28.4.
1
451.94075; delta = 0.79 ppm. H NMR (400 MHz, DMSO-d6): δ
7.81 (dd, J = 0.9, 8.2 Hz, 1H), 7.56−7.63 (m, 2H), 7.39−7.47 (m,
2H), 7.25 (t, J = 8.0 Hz, 1H), 4.83−4.91 (m, 2H), 4.77 (s, 2H),
3.87−3.96 (m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 159.7 d,
155.9, 143.7, 142.9, 131.1 d, 130.3, 128.2, 127.1 d, 123.7 d, 122.6 d,
120.2, 120.0, 118.2 d, 102.3, 44.7, 42.7 d, 41.7.
2-[(4-Chlorophenyl)methyl]-6-(trifluoromethyl)-3,4-
dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (31). It was pre-
pared according to the procedure described for 23 using 1-
(bromomethyl)-4-chlorobenzene (38e) in step 5. ESI-HRMS: calcd
for C18H14ON3ClF3 [M + H]+, 380.07720; found, 380.07731; δ =
2-(Cyclopropylmethyl)-3,4-dihydropyrazino[1,2-a]benzimidazol-
1(2H)-one (36). It was prepared according to the procedure described
for 20 using benzene-1,2-diamine (39a) in step 1. EI-HRMS: calcd
1
0.29 ppm. H NMR (400 MHz, DMSO-d6): δ 8.15 (d, J = 8.2 Hz,
1
1H), 7.81 (d, J = 7.6 Hz, 1H), 7.50 (t, J = 8.0 Hz, 1H), 7.42−7.45 (m,
1H), 7.42−7.45 (m, 1H), 7.42−7.45 (m, 1H), 7.43 (s, 1H), 4.78 (s,
for C14H15ON3: 241.12096; found: 241.12102. δ = 0.23 ppm. H
NMR (500 MHz, DMSO-d6): δ 7.76 (td, J = 0.9, 8.1 Hz, 1H), 7.67
8617
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(td, J = 0.9, 8.2 Hz, 1H), 7.40 (t, J = 7.6 Hz, 1H), 7.31 (ddd, J = 1.1,
7.1, 8.2 Hz, 1H), 4.46−4.50 (m, 2H), 3.96−4.02 (m, 2H), 3.44 (d, J =
7.0 Hz, 2H), 1.06−1.14 (m, 1H), 0.47−0.53 (m, 2H), 0.28−0.37 (m,
2H); 13C NMR (126 MHz, DMSO-d6): δ 156.5, 142.4, 142.3, 133.5,
124.2, 122.8, 120.5, 111.2, 50.1, 45.2, 40.0, 9.2, 3.1.
using cells seeded in standard 96-well microplates at a density
of 30,000 cells/well and maintained overnight in a tissue
culture incubator at 37 °C under an atmosphere of 95% air/5%
CO2. The composition of the plating medium was DMEM,
10% FBS, 1× MEM nonessential amino acids (Sigma M 7145),
and 1× PSA. The growth medium was removed, and the cells
were washed with assay buffer [without glutamate-pyruvate
transaminase (GPT)/pyruvate]; both actions were performed
with a cell washer (BioTek Elx405UCVWS). At the end of the
washing step, 50 μL/well assay buffer (without GPT/pyruvate)
was left in the wells, and 50 μL/well Calcium 5 kit diluted
fourfold in assay buffer (with 6 μg/mL GPT/4 mM pyruvate)
was added manually using an eight-channel pipette. After an
incubation period (30 min, 37 °C), the dye was removed, and
the cells were washed with the cell washer using the same
protocol as mentioned above. After washing, the plate was
incubated for 10 min at 37 °C. Then, 50 μL/well assay buffer
(with 6 μg/mL GPT/4 mM pyruvate) containing vehicle
DMSO or the test compounds or reference mGluR2 positive
modulators was added manually, and the cells were incubated
for an additional 30 min at 37 °C. Baseline and agonist-evoked
[Ca2+]i changes were monitored with FlexStation II (Molec-
ular Devices, Sunnyvale, CA), a plate reader fluorometer with
integrated eight-channel fluid addition capability. Fluorescence
measurements were carried out at 37 °C. The dye was excited
at 485 nm, and emission was sampled at 525 nm at 1.4 s
intervals. Baseline was recorded for 20 s, followed by agonist
stimulation. The 50 μL 3× concentrated agonist solution was
added to the cells using the pipettor of FlexStation II, and
fluorescence was monitored for an additional 40 s. Agonist was
glutamate administered at an EC5-EC20 concentration. The
final DMSO concentration was 1% for all treatments. To
achieve this, a series of DMSO stock solutions were prepared
from all test compounds. These stocks were stored under 0 °C
and were further diluted in assay buffer without GPT/pyruvate
to obtain the desired final concentration immediately before
the measurement. The results were expressed as ΔF/F values
using SoftMax Pro software (Molecular Devices), where F was
the resting fluorescence preceding agonist application and ΔF
was the increase in fluorescence at a given time (ΔF =
maximum fluorescence intensity values after stimulation minus
average fluorescence intensity values before stimulation). In all
experiments, all treatments were measured in multiple wells in
parallel, and the mean ΔF/F values were used for analysis.
Where appropriate, sigmoid curves were fitted to the
concentration−response data to obtain EC50 values.
General Procedure for the Synthesis of Target Compounds
(16−19). 2-(Cyclopropylmethyl)-6-phenyl-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (16). 6-Bromo-2-(cyclopropylmeth-
yl)-3,4-dihydropyrazino[1,2-a]benzimidazol-1(2H)-one (22, 895 mg,
2.86 mmol) and phenylboronic acid (378 mg, 3.1 mmol) in toluene/
ethanol (15 mL/5 mL) were refluxed for 2 h with 5.6 mmol of
Na2CO3 and 2.24 mmol (8 n/n %) of Pd(PPh3)4. After cooling, 18
mL of water was added, and the separated organic layer was
evaporated to dryness. The residue was crystallized from ether. Yield:
680 mg (75%). ESI-HRMS: calcd for C20H20ON3 [M + H]+:
1
318.16009; found: 318.15924; δ = −2.67 ppm. H NMR (400 MHz,
DMSO-d6): δ 7.79 (dd, J = 1.0, 8.2 Hz, 1H), 7.52−7.57 (m, 2H),
7.51−7.55 (m, 2H), 7.46−7.54 (m, 2H), 7.36 (dd, J = 7.3, 8.2 Hz,
1H), 7.22 (dd, J = 1.0, 7.3 Hz, 1H), 3.88−3.92 (m, 2H), 3.74−3.78
(m, 2H), 3.35−3.40 (m, 2H), 1.04 (tt, J = 4.8, 7.9 Hz, 1H), 0.36−
0.57 (m, 2H), 0.22−0.33 (m, 2H). 13C NMR (101 MHz, DMSO-d6):
δ 156.0, 142.9, 142.5, 136.8, 130.4, 129.0, 127.7, 127.3, 126.9, 125.6,
122.1, 119.4, 49.3, 44.4, 42.5, 8.7, 2.6.
2-(Cyclopropylmethyl)-6-(2-fluorophenyl)-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (17). It was prepared according to
the procedure described for 16 using 2-fluorophenylboronic acid. EI-
HRMS: Calcd. for C20H18ON3F: 335.14284; found: 335.14190. δ =
−2.81 ppm. 1H NMR (400 MHz, DMSO-d6): δ 7.82 (dd, J = 1.0, 8.1
Hz, 1H), 7.55 (dt, J = 6.2, 7.9 Hz, 1H), 7.41−7.46 (m, 1H), 7.39 (td,
J = 1.2, 7.6 Hz, 1H), 7.37 (dd, J = 7.4, 8.1 Hz, 1H), 7.28−7.34 (m,
1H), 7.23−7.27 (m, 1H), 3.89−3.98 (m, 2H), 3.73−3.82 (m, 2H),
3.35−3.41 (m, 2H), 0.99−1.07 (m, 1H), 0.43−0.51 (m, 2H), 0.28
(m, 2H); 13C NMR (101 MHz, DMSO-d6): δ 161.7, 156.5, 143.2,
143.1, 139.6, 139.6, 131.4, 130.7, 130.1, 126.1, 126.0, 125.9, 122.6,
120.4, 116.3, 114.6, 49.8, 44.9, 43.0, 9.2, 3.1.
2-(Cyclopropylmethyl)-6-(4-fluorophenyl)-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (18). It was prepared according to
the procedure described for 16 using 4-fluorophenylboronic acid. EI-
HRMS: Calcd. for C20H18ON3F, 335.14284; found, 335.14300. δ =
1
0.47 ppm. H NMR (400 MHz, DMSO-d6): δ 7.77−7.82 (m, 2H),
7.56−7.62 (m, 2H), 7.33−7.39 (m, 2H), 7.31−7.37 (m, 4H), 7.21−
7.21 (m, 1H), 7.19−7.23 (m, 1H), 7.19−7.23 (m, 1H), 7.19−7.23
(m, 1H), 3.86−3.93 (m, 2H), 3.74−3.79 (m, 2H), 3.38 (d, J = 7.0 Hz,
2H), 1.00−1.07 (m, 1H), 0.45−0.50 (m, 2H), 0.26−0.29 (m, 2H);
13C NMR (101 MHz, DMSO-d6): δ 161.9, 156.5, 143.1, 143.0, 133.6,
131.6, 131.0, 126.3, 126.2, 122.6, 120.0, 115.1, 49.8, 44.9, 43.0, 9.2,
3.1.
2-(Cyclopropylmethyl)-6-(4-chlorophenyl)-3,4-dihydropyrazino-
[1,2-a]benzimidazol-1(2H)-one (19). It was prepared according to
the procedure described for 16 using 4-chlorophenylboronic acid. EI-
HRMS: calcd. for C20H18ON3Cl, 351.11329; found, 351.11259. δ =
−2.00 ppm. 1H NMR (400 MHz, DMSO-d6): δ 7.81 (dd, J = 1.0, 8.2
Hz, 1H), 7.57 (d, J = 1.2 Hz, 4H), 7.37 (dd, J = 7.4, 8.1 Hz, 1H), 7.22
(dd, J = 1.0, 7.3 Hz, 1H), 3.90−3.95 (m, 2H), 3.72−3.81 (m, 2H),
3.35−3.40 (m, 2H), 0.99−1.09 (m, 1H), 0.44−0.51 (m, 2H), 0.25−
0.31 (m, 2H) 13C NMR (101 MHz, DMSO-d6): δ 157.1, 143.8,
143.6, 136.7, 133.3, 132.0, 131.4, 128.8, 126.7, 123.2, 120.8, 50.4,
45.5, 43.6, 9.7, 3.7.
Human and Rat Liver Microsomal Stability Assay. In
vitro metabolic stability was assessed using human (Xenotech,
USA) and Wistar rat (In Vitro Metabolism Research, Gedeon
Richter Plc, Hungary) liver microsomes. Test compounds at
2.5 μM initial test concentration were incubated for various
lengths of time with the liver microsomes (0.5 mg/mL). In
vitro intrinsic clearance (Clint mL/min·g liver) was calculated
using the basic concept of clearance prediction according to
PHARMACOLOGY
■
the following equations: Clint = Vmax/KM or, if S ≪ KM, Clint
=
Animal experiments were approved by the Local Ethical
Committee of Gedeon Richter Plc. and were carried out in
strict compliance with the European Directive 2010/63/EU
regarding the care and use of laboratory animals for
experimental procedures.
V/S, where Vmax is the maximal rate of enzyme reaction; KM =is
the affinity constant of substrate concentration; and V is the
actual rate of enzyme reaction under first-order conditions.
pIONL Model in Rats. For modeling facial allodynia, a
method described for injury of the infraorbital nerve in mice25
was adapted for rats. In pentobarbital anesthetized male SPRD
rats (Harlan, Germany), approximately one-third to one-half of
the diameter of the infraorbital nerve was tightly ligated. The
Human mGluR2 in Vitro Functional Assay. Chem-1
cells stably expressing human mGlu2 receptors were cultured
in the medium detailed below. Fluorometric measurements of
cytosolic Ca2+ ion concentration ([Ca2+]i) were carried out
8618
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
sensitivity of the facial region was tested with von Frey
filaments (0.4−14.8 g). The response was considered as
positive if at least 2/3 tests with a given filament expressed
head withdrawal were observed. The 50% mechanical head
withdrawal thresholds (50% HWTs, the stimulus strength
eliciting response with 50% probability, expressed in grams)
were determined by the modified up and down method as
described in the study by Chaplan et al.26 The experiment
started with one week habituation, and only rats with normal
(>10 g) and stable 50% HWTs were subjected to the pIONL
operation. Those rats that did not develop stable allodynia
(40% HWTs decrease in 2/3 measurements on days 9−11
post-surgery) were excluded from the study. Rats (280−480 g)
were repeatedly used for the drug tests between 10 and 40 days
post-surgery with at least 3-day resting periods.
Istvánné Kis-Varga − Gedeon Richter Plc., Budapest 1103,
Hungary
György I. Lévay − Gedeon Richter Plc., Budapest 1103,
Hungary
Zoltán Béni − Gedeon Richter Plc., Budapest 1103, Hungary
János Kóti − Gedeon Richter Plc., Budapest 1103, Hungary
István Greiner − Gedeon Richter Plc., Budapest 1103,
Hungary; orcid.org/0000-0002-5336-2828
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.jmedchem.1c00563
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
The drug effect was characterized by percentage reversal of
50% HWT values, calculated as
Notes
The authors declare no competing financial interest.
50% HWT
− 50% HWT
vehicle,post‐dose
drug,post‐dose
reversal % =
50% HWTmean‐drug+vehicle,pre‐OP − 50% HWT
× 100
vehicle,post‐dose
ACKNOWLEDGMENTS
■
The authors acknowledge the contribution to this work of all
members of the Mitsubishi Tanabe Pharma Corporation team.
The authors are also grateful to Csilla Horváth for
pharmacological support, Attila Potor for chemistry support,
and Ildikó Galambos for skillful technical assistance. The
support of the National Office for Research, Development and
Innovation (National Brain Research Program KTIA_-
NAP_13-1-2013-0001 and 2017-1.2.1-NKP-2017-00002) is
also acknowledged.
For statistics, one-way ANOVA/Dunnett’s multiple compar-
ison tests were used.
PHARMACOPHORE MODELING
mGluR2 ligand structures with their ionic states were predicted
■
27
̈
by Schrodinger LigPrep with default settings. Pharmaco-
̈
phore models were generated by Phase (Schrodinger 2020-
1)28,29 using the best alignment and common features. Every
standard pharmacophore feature was utilized to create the
pharmacophore sites. All active compounds were required to
find and score hypotheses using a maximum of 50 conformers
for each ligand. The hydrophobic and aromatic rings were
treated as equivalent, and vectors were replaced with projected
points during the model generation for 10, 13, and 14.
ABBREVIATIONS
■
ADME, absorption, distribution, metabolism, excretion; Clint,
intrinsic clearance; DMF, N,N-dimethylformamide; EDC, 1-
ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride;
FBDD, fragment-based drug discovery; h, hours; HOBt, 1-
hydroxybenzotriazole; HTS, high-throughput screening; LE,
ligand efficiency; LLE, lipophilic ligand efficiency; mGlu2,
metabotropic glutamate 2; MeOH, methanol; MW, molecular
weight; ND, not determined; NMR, nuclear magnetic
resonance; PAM, positive allosteric modulator; PCP, phency-
clidine; PD, pharmacodynamics; pIONL, partial infraorbital
nerve ligation; PK, pharmacokinetics; PPE, plasma protein
extravasation; SAR, structure−activity relationship; SEM,
standard error of mean; HPLC, high-performance liquid
chromatography; PAG, periaqueductal gray; SD, standard
deviation; THF, tetrahydrofuran; tPSA, topological polar
surface area; TRPV1, transient receptor potential cation
channel subfamily V member 1; TTX, tetrodotoxin
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.jmedchem.1c00563.
■
sı
Molecular formula strings (CSV)
Spectroscopic data and analytical HPLC trace of
compound 34 (PDF)
AUTHOR INFORMATION
Corresponding Authors
György Szabó − Gedeon Richter Plc., Budapest 1103,
Hungary; orcid.org/0000-0002-9255-4486;
Phone: +36-1-4315662; Email: gy.szabo@richter.hu
■
REFERENCES
■
̋
György M. Keseru − Medicinal Chemistry Research Group,
(1) Moghaddam, B.; Adams, B. W. Reversal of phencyclidine effects
by a group II metabotropic glutamate receptor agonist in rats. Science
1998, 281, 1349−1352.
(2) Schoepp, D.; Marek, G. Preclinical pharmacology of mGlu2/3
receptor agonists: novel agents for schizophrenia? Curr. Drug Targets:
CNS Neurol. Disord. 2002, 1, 215−225.
Research Centre for Natural Sciences, Budapest 1117,
Hungary; orcid.org/0000-0003-1039-7809;
Phone: +36-1-3826821; Email: gy.keseru@ttk.mta.hu
Authors
Péter Erdélyi − Gedeon Richter Plc., Budapest 1103, Hungary
(3) Patil, S. T.; Zhang, L.; Martenyi, F.; Lowe, S. L.; Jackson, K. A.;
Andreev, B. V.; Avedisova, A. S.; Bardenstein, L. M.; Gurovich, I. Y.;
Morozova, M. A.; Mosolov, S. N.; Neznanov, N. G.; Reznik, A. M.;
Smulevich, A. B.; Tochilov, V. A.; Johnson, B. G.; Monn, J. A.;
Schoepp, D. D. Activation of mGlu2/3 receptors as a new approach to
treat schizophrenia: a randomized Phase 2 clinical trial. Nat. Med.
2007, 13, 1102−1107.
Sándor Kolok − Gedeon Richter Plc., Budapest 1103,
Hungary
Mónika Vastag − Gedeon Richter Plc., Budapest 1103,
Hungary
Attila S. Halász − Gedeon Richter Plc., Budapest 1103,
Hungary
8619
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620

Journal of Medicinal Chemistry
pubs.acs.org/jmc
Article
(4) Adams, D. H.; Kinon, B. J.; Baygani, S.; Millen, B. A.; Velona, I.;
Kollack-Walker, S.; Walling, D. P. A long-term, phase 2, multicenter,
randomized, open-label, comparative safety study of pomaglumetad
methionil (LY2140023 monohydrate) versus atypical antipsychotic
standard of care in patients with schizophrenia. BMC Psychiatr. 2013,
13, 143.
(5) Adams, D. H.; Zhang, L.; Millen, B. A.; Kinon, B. J.; Gomez, J. C.
Pomaglumetad methionil (LY2140023 monohydrate) and aripipra-
zole in patients with schizophrenia: a phase 3, multicenter, double-
blind comparison. Schizophr. Res. Treat. 2014, 2014, 758212.
(6) Spooren, W. P.; Gasparini, F.; van der Putten, H.; Koller, M.;
Nakanishi, S.; Kuhn, R. Lack of effect of LY314582 (a group 2
metabotropic glutamate receptor agonist) on phencyclidine-induced
locomotor activity in metabotropic glutamate receptor 2 knockout
mice. Eur. J. Pharmacol. 2000, 397, R1−R2.
(7) Fell, M. J.; Svensson, K. A.; Johnson, B. G.; Schoepp, D. D.
Evidence for the role of metabotropic glutamate mGlu2 not mGlu3
receptors in the preclinical antipsychotic pharmacology of the
mGlu2/3 receptor agonist (-)-(1R,4S,5S,6S)-4-amino-2-
sulfonylbicyclo[3.1.0]hexane-4,6-dicarboxylic acid (LY404039). J.
Pharmacol. Exp. Ther. 2008, 326, 209−217.
(8) Rowe, B. A.; Schaffhauser, H.; Morales, S.; Lubbers, L. S.;
Bonnefous, C.; Kamenecka, T. M.; McQuiston, J.; Daggett, L. P.
Transposition of three amino acids transforms the human
metabotropic glutamate receptor (mGluR)-3-positive allosteric
modulation site to mGluR2, and additional characterization of the
mGluR2-positive allosteric modulation site. J. Pharmacol. Exp. Ther.
2008, 326, 240−251.
(9) Rudd, M.; McCauley, J. Positive allosteric modulators of the
metabotropic glutamate receptor subtype 2 (mGluR2). Curr. Top.
Med. Chem. 2005, 5, 869−884.
(10) Litman, R. E.; Smith, M. A.; Doherty, J. J.; Cross, A.; Raines, S.;
Gertsik, L.; Zukin, S. R. AZD8529, a positive allosteric modulator at
the mGluR2 receptor, does not improve symptoms in schizophrenia:
A proof of principle study. Schizophr. Res. 2016, 172, 152−157.
(11) Addex Therapeutics. Addex Reports Top-line Data from a
Successful Phase 2a Clinical Study with ADX71149 in Schizophrenia
Patients, 2015. https://www.addextherapeutics.com/en/news-and-
events/press-releases/addex-reports-top-line-data-successful-phase-2a-
clinical-study-adx71149-schizophrenia-patients/ (accessed on 26
April, 2021).
(12) Kent, J. M.; Daly, E.; Kezic, I.; Lane, R.; Lim, P.; De Smedt, H.;
De Boer, P.; Van Nueten, L.; Drevets, W. C.; Ceusters, M. Efficacy
and safety of an adjunctive mGlu2 receptor positive allosteric
modulator to a SSRI/SNRI in anxious depression. Prog. Neuro-
psychopharmacol. Biol. Psychiatr. 2016, 67, 66−73.
receptor-2 (mGluR2) positive allosteric modulators (PAMs). Eur. J.
Med. Chem. 2020, 186, 111881.
(19) Zhang, L.; Brodney, M. A.; Candler, J.; Doran, A. C.;
Duplantier, A. J.; Efremov, I. V.; Evrard, E.; Kraus, K.; Ganong, A. H.;
Haas, J. A.; Hanks, A. N.; Jenza, K.; Lazzaro, J. T.; Maklad, N.;
McCarthy, S. A.; Qian, W.; Rogers, B. N.; Rottas, M. D.; Schmidt, C.
J.; Siuciak, J. A.; Tingley, F. D., III; Zhang, A. Q. 1-[(1-Methyl-1H-
imidazol-2-yl)methyl]-4-phenylpiperidines as mGluR2 Positive Allos-
teric Modulators for the Treatment of Psychosis. J. Med. Chem. 2011,
54, 1724−1739.
(20) D’Alessandro, P. L.; Corti, C.; Roth, A.; Ugolini, A.; Sava, A.;
Montanari, D.; Bianch, F.; Garland, S. L.; Powney, B.; Koppe, E. L.;
Rocheville, M.; Osborne, G.; Perez, P.; La Fuente, J.; Los Frailes, M.;
Smith, P. W.; Branch, C.; Nash, D.; Watson, S. P. The identification
of structurally novel, selective, orally bioavailable positive modulators
of mGluR2. Bioorg. Med. Chem. Lett. 2010, 20, 759−762.
́
(21) Szabó, G.; Turós, G. I.; Kolok, S.; Vastag, M.; Sánta, Zs.;
Dékány, M.; Lévay, G. I.; Greiner, I.; Natsumi, M.; Tatsuya, W.;
̋
Keseru, G. M. Fragment based optimization of metabotropic
glutamate receptor 2 (mGluR2) positive allosteric modulators in
the absence of structural information. J. Med. Chem. 2019, 62, 234−
246.
(22) Merck Sharp & Dohme Corp.. Ether benzotriazole derivatives
U.S. Patent 20,120,135,977 A1, 2012.
(23) Lavreysen, H.; Langlois, X.; Ahnaou, A.; Drinkenburg, W.; te
Riele, P.; Biesmans, I.; Van der Linden, i.; Peeters, I.; Megens, A.;
Wintmolders, C.; Cid, J. M.; Trabanco, A. A.; Andrés, J. I.;
Dautzenberg, F. M.; Lutjens, R.; Macdonald, G.; Atack, J. R.
̈
Pharmacological characterization of JNJ-40068782, a new potent,
selective, and systemically active positive allosteric modulator of the
mGlu2 receptor and its radioligand [3H]JNJ-40068782. J. Pharmacol.
Exp. Ther. 2013, 346, 514−527.
(24) Young, R. J.; Leeson, P. D. Mapping the efficiency and
physicochemical trajectories of successful optimizations. J. Med. Chem.
2018, 61, 6421−6467.
(25) Krzyzanowska, A.; Pittolo, S.; Cabrerizo, M.; Sánchez-López, J.;
Krishnasamya, S.; Venero, C.; Avendano, C. Assessing nociceptive
sensitivity in mouse models of inflammatory and neuropathic
trigeminal pain. J. Neurosci. Methods 2011, 201, 46−54.
(26) Chaplan, S. R.; Bach, F. W.; Pogrel, J. W.; Chung, J. M.; Yaksh,
T. L. Quantitative assessment of tactile allodynia in the rat paw. J.
Neurosci. Methods 1994, 53, 55−63.
̈ ̈
dinger Release 2020-1: LigPrep; Schrodinger, LLC: New
(27) Schro
York, NY, 2020.
̈
̈
(28) Schrodinger Release 2020-1: Phase; Schrodinger, LLC: New
York, NY, 2020.
(29) Dixon, L. S.; Smondyrev, M. A.; Rao, N. S. PHASE: A novel
approach to pharmacophore modeling and 3D database searching.
Chem. Biol. Drug Des. 2006, 67, 370−372.
(13) Addex Reports top-line data from ADX71149 phase 2a study in
patients with major depressive disorder (MDD) with significant
anxiety symptoms. https://www.fiercebiotech.com/biotech/addex-
reports-top-line-data-from-adx71149-phase-2a-study-patients-major-
depressive-disorder (accessed on 26 April, 2021).
(14) Chiechio, S.; Nicoletti, F. Metabotropic glutamate receptors
and the control of chronic pain. Curr. Opin. Pharmacol. 2012, 12, 28−
34.
(15) Simmons, R. M. A.; Webster, A. A.; Kalra, A. B.; Iyengar, S.
Group II mGluR receptor agonists are effective in persistent and
neuropathic pain models in rats. Pharmacol. Biochem. Behav. 2002, 73,
419−427.
(16) Innovative Drug Development for Headache Disorders; Olesen, J.,
Ed.; Oxford University Press, 2008.
(17) Johnson, M. P.; Schkeryantz, J.; Swanson, S.; Perkins, A. N.;
Johnson, K. W. Preclinical characterization of LY2300559, a dual
acting mGlu2 receptor potentiator and leukotriene CysLt1 receptor
antagonist, for potential use in migraine. Curr. Neuropharmacol. 2011,
9, 31.
(18) Szabó, G.; Kolok, S.; Orgován, Z.; Vastag, M.; Béni, Z.; Kóti, J.;
̋
Sághy, K.; Lévay, G. I.; Greiner, I.; Keseru, G. M. Discovery of
dihydropyrazino-benzimidazole derivatives as metabotropic glutamate
8620
https://doi.org/10.1021/acs.jmedchem.1c00563
J. Med. Chem. 2021, 64, 8607−8620