Novel bicyclo[3.1.0]hexane analogs as antagonists of metabotropic glutamate 2/3 receptors for the treatment of depression

Article abstract of DOI:10.1016/j.bmcl.2016.10.067

Negative modulators of metabotropic glutamate 2 & 3 receptors demonstrate antidepressant-like activity in animal models and hold promise as novel therapeutic agents for the treatment of major depressive disorder. Herein we describe our efforts to prepare and optimize a series of conformationally constrained 3,4-disubstituted bicyclo[3.1.0]hexane glutamic acid analogs as orthosteric (glutamate site) mGlu_2/3receptor antagonists. This work led to the discovery of a highly potent and efficacious tool compound 18 (hmGlu₂IC₅₀46 ± 14.2 nM, hmGlu₃IC₅₀= 46.1 ± 36.2 nM). Compound 18 showed activity in the mouse forced swim test with a minimal effective dose (MED) of 1 mg/kg ip. While in rat EEG studies it exhibited wake promoting effects at 3 and 10 mg/kg ip without any significant effects on locomotor activity. Compound 18 thus represents a novel tool molecule for studying the impact of blocking mGlu_2/3receptors both in vitro and in vivo.

Full text of DOI:10.1016/j.bmcl.2016.10.067

Accepted Manuscript
Novel Bicyclo[3.1.0]hexaneAnalogs asAntagonists of Metabotropic Glutamate
2/3 Receptors for the Treatment of Depression
Bruce A. Dressman, Eric G. Tromiczak, Mark D. Chappell, Allie E. Tripp,
Steven J. Quimby, Tatiana Vetman, Adam M. Fivush, James Matt, Carlos
Jaramillo, Renhua Li, Albert Khilevich, Maria-Jesus Blanco, Stephon C. Smith,
Mercedes Carpintero, José Eugenio de Diego, Mario Barberis, Susana García-
Cerrada, José F. Soriano, Jeffrey M. Schkeryantz, Jeffrey M. Witkin, Keith A.
Wafford, Wesley Seidel, Thomas Britton, Carl D. Overshiner, Xia Li, Xu-Shan
Wang, Beverly A. Heinz, John T. Catlow, Steven Swanson, David Bedwell, Paul
L. Ornstein, Charles H. Mitch
PII:
S0960-894X(16)31105-2
DOI:
http://dx.doi.org/10.1016/j.bmcl.2016.10.067
BMCL 24370
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
2 September 2016
21 October 2016
22 October 2016
Please cite this article as: Dressman, B.A., Tromiczak, E.G., Chappell, M.D., Tripp, A.E., Quimby, S.J., Vetman,
T., Fivush, A.M., Matt, J., Jaramillo, C., Li, R., Khilevich, A., Blanco, M-J., Smith, S.C., Carpintero, M., Eugenio
de Diego, J., Barberis, M., García-Cerrada, S., Soriano, J.F., Schkeryantz, J.M., Witkin, J.M., Wafford, K.A., Seidel,
W., Britton, T., Overshiner, C.D., Li, X., Wang, X-S., Heinz, B.A., Catlow, J.T., Swanson, S., Bedwell, D., Ornstein,
P.L., Mitch, C.H., Novel Bicyclo[3.1.0]hexane Analogs as Antagonists of Metabotropic Glutamate 2/3 Receptors
for the Treatment of Depression, Bioorganic & Medicinal Chemistry Letters (2016), doi: http://dx.doi.org/10.1016/
j.bmcl.2016.10.067
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
Leave this area blank for abstract info.
R2
C4
R2
R1
C3
17 Analogs
18
hmGlu2 IC50 = 12.7-1,130 nM
hmGlu3 IC50 = 14.3-1,070 nM
R1 = OCH2Ar, R2 = NHCOR, SR,
OH, F, NH2, N3
hmGlu2 IC50 = 46 ± 14.1 nM
hmGlu3 IC50 = 46.1 ± 36.2 nM
ED60 = 0.72 mg/kg
Mouse forced swim test
ED60 = 0.72-9.5 mg/kg

Bioorganic & Medicinal Chemistry Letters
Novel Bicyclo[3.1.0]hexane Analogs as Antagonists of Metabotropic Glutamate
2/3 Receptors for the Treatment of Depression
Bruce A. Dressman*,a , Eric G. Tromiczak,a Mark D. Chappell,a Allie E. Tripp,a Steven J. Quimby,a Tatiana
Vetman,a Adam M. Fivush,a James Matt,a Carlos Jaramillo,b Renhua Li,a Albert Khilevich,a Maria-Jesus
Blanco,a Stephon C. Smith,a Mercedes Carpintero,b José Eugenio de Diego,b Mario Barberis,b Susana
García-Cerrada,b José F. Soriano,b Jeffrey M. Schkeryantz,a Jeffrey M. Witkin,a Keith A. Wafford,c Wesley
Seidel,c Thomas Britton,a Carl D. Overshiner,a Xia Li,a Xu-Shan Wang,a Beverly A. Heinz,a John T.
Catlow,a Steven Swanson,a David Bedwell,a Paul L. Ornsteina and Charles H. Mitcha
aLilly Research Laboratories, Eli Lilly & Co., Indianapolis, IN 46285, USA, bAvenida de la Industria, 30, 28108 Alcobendas, Madrid, Spain and cErl Wood
Manor, Windlesham, Surrey GU20 6PH, UK
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
Negative modulators of metabotropic glutamate 2 & 3 receptors demonstrate
antidepressant-like activity in animal models and hold promise as novel therapeutic
agents for the treatment of major depressive disorder. Herein we describe our efforts
to prepare and optimize a series of conformationally constrained 3,4-disubstituted
bicyclo[3.1.0]hexane glutamic acid analogs as orthosteric (glutamate site) mGlu2/3
receptor antagonists. This work led to the discovery of a highly potent and efficacious
tool compound 18 (hmGlu2 IC50 46 ± 14.2 nM, hmGlu3 IC50 = 46.1 ± 36.2 nM).
Compound 18 showed activity in the mouse forced swim test with a minimal effective
dose (MED) of 1 mg/kg ip. While in rat EEG studies it exhibited wake promoting
effects at 3 and 10 mg/kg ip without any significant effects on locomotor activity.
Compound 18 thus represents a novel tool molecule for studying the impact of
blocking mGlu2/3 receptors both in vitro and in vivo.
Keywords:
Metabotropic glutamate
mGlu2/3 antagonist
Major depressive disorders
Antidepressant
2009 Elsevier Ltd. All rights reserved.
Metabotropic glutamate (mGlu) receptors belong to the class
C GPCR family and consist of eight known subtypes which have
dioxide (LY2140023)) have demonstrated efficacy in generalized
anxiety disorder1 and schizophrenia patients,2 respectively.
been historically divided into three groups (Group I: mGlu1 & ;
5
Conversely, mGlu2/3 receptor antagonists facilitate the
presynaptic release of glutamate and thereby enhance synaptic
AMPA- and NMDA-receptor activation and neurotransmission
under conditions where mGlu2/3 receptors are tonically activated.
Consistent with this, preclinical studies have demonstrated that
mGlu2/3 antagonists such as (1S,2S)-2-[(1S)-1-amino-1-carboxy-
Group II: mGlu2 & ; Group III: mGlu4, , & ) The Group II
3
6
7
8
mGlu receptors are highly expressed in prefrontal cortex,
striatum, thalamus, hippocampus, and amygdala, where they act
to regulate neuronal excitability via presynaptic, postsynaptic and
glial mechanisms. Activation of mGlu2/3 receptors is known to
inhibit the synaptic release of glutamate, leading to a reduction of
synaptic transmission. Accordingly, mGlu2/3 agonists (e.g.
(1S,2S,5R,6S)-2-aminobicyclo[3.1.0]hexane-2,6-dicarboxylic
acid (LY354740), (1R,4S,5S,6S)-4-amino-2-thiabicyclo[3.1.0]
hexane-4,6-dicarboxylic acid 2,2-dioxide (LY404039)), Figure 1)
produce beneficial effects in rodent models (anxiety, psychosis,
pain) thought to be driven by excessive glutamate
neurotransmission, and oral prodrugs of these agents
((1S,2S,5R,6S)-2-(L-Alanylamino)bicyclo[3.1.0]hexane-2,6-di
carboxylic acid (LY544344), (1R,4S,5S,6S)-4-(L-methionyl
amino)-2-thiabicyclo[3.1.0]hexane-4,6-dicarboxylic acid 2,2-
2-(9H-xanthen-9-yl)ethyl]cyclopropanecarboxylic
acid
(LY341495) and (1R,2R,3R,5R,6R)-2-amino-3-[(3,4-
dichlorobenzyl)oxy]-6-fluorobicyclo[3.1.0]hexane-2,6-
dicarboxylic acid (MGS0039) (Figure 1) elicit glutamate-driven
AMPA receptor-dependent antidepressant3 and wake-promoting4
responses in rodents.
As part of our ongoing research efforts targeting these
receptors, we explored the effect of structural modifications at
the C3- and C4-positions of LY354740 on functional

mGlu2/3-mediated activity. Literature data18 as well as our own
in-house structure activity relationship (SAR) data as exemplified
by compound (+/-)-1,5 showed that substitution at the C3-
position of LY354740 produced compounds that are functional
antagonists. Conversely, we have published an extensive SAR of
analogs of LY354740 with a range of C4-postion substituents
that exhibit agonist activity.6,7 In order to further understand the
effects of substituents at these two positions on functional
activity we set out to develop synthetic methodologies that would
enable their dual functionalization. This work has led to the
identification of a number of novel potent and selective mGlu2/3
receptor antagonists.
elimination product 4 in good yield along with unreacted alcohol
3 which could be recovered by normal phase chromatography.
Alkene 4 was oxidized using catalytic osmium tetroxide (OsO4)
and N-methylmorpholine-N-oxide (NMO) as a co-oxidant to give
the expected cis C3,C4-diol 5 in high diastereoselective (>50:1)
and synthetic yield through oxidant approach from the less
sterically hindered face of the bicyclo[3.1.0]hexane ring system.
Regioselective alkylation of the C3 carbinol was achieved using
silver oxide (Ag2O) and tetrabutyl ammonium iodine (TBAI) and
a benzyl halide to give a mixture of the mono C3 and C4 benzyl
ethers
6 and 7 that were separated by normal phase
chromatography. In this way, the diol 5 was selectively alkylated
with 3,4-dichlorobenyl bromide to give a 3:1 mixture of the 3,4-
dichlorobenzyl ether intermediates 6a and 7a that were readily
separated by normal phase column chromatography to give pure
intermediate 6a in 72% yield. Structural confirmation of 6a was
established by proton NMR.8 Exhaustive deprotection of
intermediate 6a in glacial acetic acid and water using our
previously reported procedure6 at elevated temperatures in a
microwave provided the C4-hydroxyl-C3-3,4-dichlorobenzyl
ether 11 as a white solid in 74% yield and in high chemical purity
mGlu2/3 Agonist
(as determined by proton NMR and
LCMS) following
LY354740
LY404039
concentration of the reaction mixture and trituration of the
recovered solid with water and ether. Using a similar process a
series of C4-hydroxyl-C4-benzyl ethers 8-15 were prepared and
their functional hmGlu2 and hmGlu3 activity was evaluated.
mGlu2/3 Antagonist
a
b, c
2
3
LY341495
MGS0039
e
d
5
4
(+/-)
f
1
6
8-15
7
LY3020371
hmGlu2 IC50 = 312 nM
6a, R = 3,4-diCl-Ph
7a, R = 3,4-diCl-Ph
hmGlu3 IC50 = 76.1 nM
Scheme 1: Reagents and Conditions: (a) L-Selectride®(1M in
THF, 1.5 eq.), THF, 0°C then 30% H2O2, NaHCO3(quantitative);
(b) MsCl, TEA; (c) TBAF, THF, reflux (76% 2 steps); (d) OsO4
(0.05 eq.), NMO (2.5 eq.), acetone, water [>50:1dr] (91%); (e)
RCH2Br (1.5 eq.), Ag2O (1.5 eq.), TBAI (1 eq.), DMF [generally
3:1 regioselectivity for 6 vs 7] (R = 3,4-diCl-Ph, 94%); (f)
Acetic acid, water, 140°C, microwave (20-84%).
Current Work
R2
C4
R2
R1
C3
Based on our initial biochemical results of the C4-hydroxyl-
C4-benzyl ethers 8-15 (Table 1), further investigation of the C4
position was conducted using the C4-hydroxyl-C3-3,4-
dichlorobenzyl ether intermediate 6a (Scheme 2). Inversion of
the C4 alcohol of 6a was readily achieved via a Mitsunobu
reaction using p-nitrobenzoic acid followed by hydrolysis of the
Figure 1: Literature and in-house mGlu2/3 receptor ligands.
We began our work employing the previously reported
enantiomerically pure ketone 26 which can be synthesized on
multi-kilogram scale (Scheme 1). Ketone 2 was reduced to give
resulting ester to give the C4-alcohol 16.
Exhaustive
deprotection using aqueous acetic acid as previously described
provided the C4-hydroxyl-C3-benzyl ether 18. In a similar
manner the C4-fluoro compound 19 was prepared from 6a in
moderate yield using an excess of Deoxo-Fluor®.
the C4-alcohol
3
in high yield and with excellent
The alcohol was
diastereoselectivity using L-Selectride®.
converted to the corresponding mesylate using methyl sulfonyl
chloride (MsCl) and triethyl amine (TEA) and then subjected to
tetrabutylammonium fluoride (TBAF) to cleanly give the

Additional C4-substituted variants were accessible from
mesylate 20, itself prepared from 6a in good yield. Owing to the
high steric congestion of groups around the C4 position as well
as the cup-shape of the bicyclo[3.1.0]hexane ring system the
approach of nucleophiles to the alpha (concave) face of these
molecules is hindered and displacement of the mesylate with
various nucleophiles required large excesses of the nucleophile,
long reaction times and elevated reaction temperatures to obtain
sufficient synthetic yields of products. For example, a large
stoichiometric excess (>5 equiv.) of sodium azide and heating in
DMF over 6 days with a catalytic amount of 15-crown-5
mesylate 20 was successfully converted to the C4-azide in good
yield (71%). Reduction of the azide to the amine was
accomplished using trimethylphosphine in wet THF to cleanly
give the primary amine intermediate 21. Global deprotection of
intermediate 21 and the corresponding azide intermediate using
our standard deprotection conditions successfully provided the
C4-primary amine 22 as well as the C4-azide 23 (Table 2; not
expressing recombinant human mGlu2 or mGlu3 receptors along
with the recombinant human glutamate transporter EAAT1. An
EC90 concentration of the mGlu2/3 receptor agonist (1R,2R)-3-
[(1S)-1-amino-2-hydroxy-2-oxo-ethyl]cyclopropane-1,2-
dicarboxylic acid (DCG-IV)9 was used in these experiments.7
Using our described synthetic route, we prepared a number of
C3-benzyl ether analogs and assessed their functional activity
with a common C4-hydroxyl group (Table 1). For analogs 8-15,
substitution of the aryl ring had only modest effects on antagonist
potency (Table 1). For example, 2,3-disubstituted benzyl ether
(13) and 3,4-disubstituted benzyl ethers (11, 14, 15) showed
modestly better activity (3 fold) as exemplified by the 2,3-
dichlorobenzyl and 3,4-dichlorobenzyl ethers 11 and 13 verses
2,5-disubstituted (12), monosubstituted (9, 10) or unsubstitued
benzyl ethers (8). With these data in hand, we next explored the
effect of altering substituents at the C4-position while holding the
C3 substituent constant as a 3,4-diclorobenzyloxy moiety.
shown in Scheme 2).
intermediate 21 with acetyl chloride or methyl chloroformate was
achieved using slight excess of acylating agent in
Acylation of the primary amine
a
R
dichloromethane with triethylamine as a base to give the acylated
amine adducts 24 in good yield. Deprotection as previously
described provided the C4-methyl amide and methyl carbamate
25 and 26. In a similar manner the C4-heteroaryl thio ethers 28-
30 were obtained from mesylate 20 in low to moderate synthetic
yield using a slight excess of thiol (1.3 eq.) and potassium
carbonate in DMF at elevated temperatures over 1-2 days
followed by deprotection using our standard conditions.
hmGlu2
hmGlu3
IC (nM) ±
SEM19
IC (nM) ±
Compds.
R
50
50
SEM19
Ph
306 ± 36.3
239 ± 139
8
3-Cl-Ph
280 ± 142
210 ± 22.7
169 ± 56
178 ± 136
145 ± 18.4
153 ± 130
185 ± 68.7
83.2 ± 80.2
127 ± 84.8
85.0 ± 16.9
9
4-Cl-Ph
10
11
12
13
14
15
f
d
3,4-diCl-Ph
2,5-diCl-Ph
2,3-diCl-Ph
3-F-4-Cl-Ph
3-Me-4-F-Ph
387 ± 94.3
149 ± 31.4
225 ± 56.9
215 ± 65.4
21
20
6a, R = 3,4-diCl-Ph
g
e
a or b
R2
SR1
Table 1: cAMP hmGlu2 and hmGlu3 receptor functional
activity for C4-hydroxyl-C3-benzyl ethers.
24
16, X = OH
17, X = F
27
Our initial focus was to invert the hydroxyl group at the C4
position to the C4-hydroxy analog based on data associated with
agonist-associated SAR.6 Structurally analogous agonist isomers
31 and 32 with a C4 hydroxy group that were devoid of a C3
group showed a slight preference in binding affinity of the 
isomer (Ki = 166 ± 28.1 nM) vs the  isomer (Ki = 230 ± 20.2
nM) at mGlu2 receptors (Figure 2). We hypothesized that a
similar enhancement in potency of antagonist analogs might also
be observed on inversion of OH stereochemistry from  to .
Gratifyingly the C4-hydroxy analog 18 showed a 3.7-fold
improvement in hmGlu2 functional antagonist potency compared
to the C isomer 11 (Table 2).
c
c
c
R2
SR1
25, R2 = Me
26, R2 = OMe
18, X = OH
19, X = F
28-30
Scheme 2: Reagents and Conditions: (a) X = OH, (i) p-
nitrobenzoic acid (2 eq.), PPh3 (2 eq.), DEAD (2 eq.), THF
(88%); (ii) K2CO3, MeOH (87%); (b) X = F, Deoxo-Fluor®
(2.5 eq.), DCM, -78°C to rt (58%); (c) Acetic acid, water,
140°C, microwave (20-99%); (d) MsCl, TEA, THF, 0°C–rt
(92%); (e) R1SH (1.3 eq.), K2CO3 (3 eq.), DMF, 80°C, 1-2 days
(11-62%); (f) (i) NaN3 (5 eq.), cat. 15-crown-5 (0.1 eq.), DMF,
85°C, 6 days (71%); (ii) PMe3 (1.5 eq.), THF, water, rt (87%);
(g) R2COCl, TEA, DCM (92-99%).
Incorporation of other small groups such as fluoro (19) or
azide (22) at the C4 position led to similar potency gains
compared to 11 with the exception of a primary amino group (23)
which lost approximately 7-fold in potency and displayed only
micromolar potency (hmGlu2 IC50 = 1130 ± 340 nM, hmGlu3
IC50 = 1070 ± 400 nM). Acylation of the amine 23 recaptured
and further improved potency and the C4 N-methyl carbamate
26 (hmGlu2 IC50 = 18.8 ± 4.95 nM, hmGlu3 IC50 = 20.5 ± 20.4
nM) and N-acetyl 25 (hmGlu2 IC50 = 26.1 ± 4.69 nM, hmGlu3
IC50 = 14.3 ± 9.96 nM) showed excellent antagonist potencies at
both receptor subtypes.
We assessed hmGlu2 and hmGlu3 functional antagonist activity
by measuring the ability of test molecules to block mGlu2/3
agonist-inhibited, forskolin-stimulated cAMP formation in cells

primary ammonium group (pKa = 9.8813) which would be
expected to be fully protonated at physiological pH and therefore
exhibit repulsive interactions with the protonated Arg57 residue.
R
hmGlu2
hmGlu3
IC (nM) ± SEM19 IC (nM) ± SEM19
Cpds.
R
50
50
31
11
-OH
-F
46 ± 14.1
46.1 ± 36.2
18
19
22
23
25
26
Ki = 230 ± 20.2 nM
Ki = 76 ± 3.99 nM
66.9 ± 15.4
34.5 ± 4.68
1130 ± 340
26.1 ± 4.69
18.8 ± 4.95
71.0 ± 58.1
54.1 ± 19.9
1070 ± 400
14.3 ± 9.96
20.5 ± 20.4
-N
3
-NH
2
-NHAc
-NHCO2Me
32
18
Ki = 166 ± 28.1 nM
Ki = 20.1 ± 7.23 nM
12.7 ± 3.3
19.9 ± 3.65
16.3 ± 11.7
23.5 ± 17.4
29.6 ± 21.1
38.5 ± 13.4
28
29
30
Figure 2: Competitive hmGlu2 binding data with [3H]-
LY45947710 and C4-hydroxyl agonist and antagonist
compounds.
Table 2: cAMP hmGlu2 and hmGlu3 receptor functional data
for C4-hydroxyl-C3-benzyl ethers.
Next we assessed the impact of sterically larger C4-
substituents on antagonist potency. We prepared the C4-
sulfanyl-1H-1,2,4-triazole 28 based on similar analogs with this
C4 group from an agonist SAR effort that showed excellent
binding affinity with mixed functional agonist (mGlu2) and
partial antagonist (mGlu3) activity.7 This substituent further
improved binding affinity (hmGlu2 Ki = 7.43 ± 4.16 nM; hmGlu3
Ki = 2.70 ± 1.25 nM) which nicely translated to improved
antagonist functional potency (hmGlu2 IC50 = 12.7 ± 3.3 nM,
hmGlu3 IC50 = 23.5 ± 17.4 nM). Evaluation of the 1-methyl-
1,2,4-triazole 29 as well as the 2-imidazole analog 30 showed
slightly less potency compared to the potency observed for the
1H-1,2,4-triazole 28.
We evaluated all of our compounds for agonist functional
activity as it was uncertain what functional activity 3,4-
disubstituted analogs would possess (full agonist, partial agonist
or full antagonist). At concentrations up to 25 M, compounds
8-15 and 18-30 were found to be devoid of agonist activity at
either mGlu2 or mGlu3 receptors; rather, functional antagonist
activity was observed for each of these analogs.
Figure 3:
Compound 28 (blue) overlaid with
LY3020371(orange) in the hmGlu2 receptor suggest possible
interactions of the C4 triazole group with an Arg57 in the
binding pocket.
Compound 18 was further profiled for selectivity against the
other human metabotropic receptor subtypes using both agonist
and antagonist testing modes (Table 3). Testing under conditions
to assess antagonist functional activity it showed modest
selectivity vs mGlu6 (70-fold) and the mGlu8 (40-fold) receptor
subtypes. However, it showed greater than 200-fold selectivity
against the other receptor subtypes (mGlu1, 4, and 7). No
functional agonist activity at any of the metabotropic glutamate
receptor subtypes was observed at concentrations up to 25 M.
To understand the observed C4 SAR for compounds of the
present series, we manually superimposed and minimized the
structure of compound 28 with the crystal structure of antagonist
(1S,2R,3S,4S,5R,6R)-2-amino-3-[(3,4-difluorophenyl)
sulfanylmethyl]-4-hydroxy-bicyclo[3.1.0]hexane-2,6-
dicarboxylic acid (LY3020371, Figure 1) bound to the human
mGlu2 amino terminal domain of the receptor11 using MOE
modeling software.12 These manual docking studies suggest
potential interaction of the C4 triazole group with Arg57 in the
receptor binding pocket that is hypothesized to contribute to the
excellent potency of this compound (Figure 3). Additionally,
having Arg57 proximal to the C4 site explains the relatively poor
potency observed with compound 23 which possesses a basic C4
Broader profiling of 18 revealed it to have no significant
activity at over 50 other biologic targets including biogenic
amine receptors, ion channels (including ionotropic glutamate
receptors 1-6), nuclear hormone receptors and kinases.
Importantly, it showed low risk of cardiac toxicities related to

activity at the human ether-à-go-go-related gene (hERG) as
assessed through competitive binding with [3H]Astemizole in a
membrane binding assay (13% inhibition at 100 M).
as well as an acceptable half-life (T1/2 = 4.61 ± 0.394 h)
following iv administration at 1 mg/kg. IP administration of 18
at 10 mg/kg produced a maximal plasma concentration of 48,100
nM and 24 h exposure of 116,000 nM*h. Measured brain to
plasma (B/P) concentrations at 30 minutes in mouse for
compound 18 following this dose showed a B/P ratio of 0.015, or
447 nM (n = 3). Considering the physicochemical attributes of
this molecule that includes high aqueous solubility, low cLogD (-
4.6613) and high measured free fraction in mouse (56%) it is
assumed that most of the drug which partitions into the brain
remains in the extracellular compartment where the antagonist
binding site resides.
IC
EC
50
Receptor
hmGlu1a
hmGlu2b
hmGlu3b
hmGlu4a
hmGlu5a
hmGlu6b
hmGlu7a
hmGlu8
50
>12.5 M
46 ± 14.1 nM
46.1 ± 36.2 nM
>12.5 M
>25 M
>25 M
>25M
>25M
>25 M
>12.5M
>25 M
>12.5 M
3.3 M
Dose
1 mg/kg20
5 mg/kg20 10 mg/kg20
10.5M
1.9 Ma
>50 Mb
iv, mg ±
SDa
oral, nM
± SDa
ip, nM ±
SDb
Parameters
Units
nMh
nM
h
Table 3: Human metabotropic glutamate receptor subtype
functional activity of compound 18 in AV12 cell lines
AUC(0-
24h)
32165 ±
2012
1783 ±
471
116000 ±
37300
a
expressing the recombinant human receptor. Ca2+ FLIPR
b
testing format. cAMP testing format.
33892 ±
1026
48100 ±
4700
Co or Cmax
Tmax
262 ± 26
In preparation for pharmacokinetic studies, 18 was evaluated
for aqueous solubility as well as profiled in a number of in vitro
assays to assess ADME attributes (Table 4). It showed high
aqueous solubility at pH 7.4 (0.778 mg/mL), low rodent, monkey
and human microsomal metabolism (<20%) and no significant
cytochrome P450 enzyme isoform inhibition (<5% at 10 M) as
well as a high plasma unbound fraction (>40%). It showed
moderate permeability (A-B transport, 2%) in a Madin-Darby
canine kidney (MDCK) epithelial cell line, an in vitro model of
blood brain permeability, and it did not act as a substrate or
inhibitor of P-glycoprotein (Pgp) efflux in engineered MDCK
cell lines expressing the human ABCB1 (MDR1) transporter.14
1.67 ±
0.58
-
0.38 ± 0.13
mL
min-1
kg-1
3.67 ±
0.225
CL
-
-
-
-
0.589 ±
0.020
Vdss
T1/2
L/kg
h
4.61 ±
0.394
5.49 ±
2.50
4.48 ±
0.410
Estimated
%
-
6
-
bioavailability
Aqueous solubility (pH 7.4)
MDCK permeability (A-B transport)a
0.778 mg/mL
2%
Table 5: Mean plasma pharmakinetic attributes of compound
18 in rat following intravenous, oral and intraperitoneal dosing.
b
aMean determination based on n = 3. Average determination
% inhibition of human Pgp transport in
based on n = 2.
<1%
MDCK cell lines expressing human MDR1b
% loss in microsomes (rodent, dog, monkey
and human)c
<20%
We next assessed compound 18 in the mouse forced swim test
(mFST),15
a rodent model capable of detecting known
antidepressant drugs16 and previously described mGlu2/3 receptor
antagonists.3a Using ip dosing and a 30 minute pretreatment
time, compound 18 demonstrated excellent potency and robust
efficacy in the mFST with a calculated ED60 value of 0.72 mg/kg
and MED of 1 mg/kg (Figure 4). A companion study in mouse to
assess plasma level concentrations using the aforementioned B/P
ratio of 1.5% showed doses of 18 of 1 mg/kg and 3 mg/kg are
predicted to exceed the IC50 values for both mGlu2 and mGlu3 in
brain (1 mg/kg, C30min = 4811 ± 2051 nM in plasma; calculated
brain = 72 nM. 3 mg/kg C30min = 16525 ± 6337 nM in plasma;
calculated brain = 248 nM). While the 0.3 mg/kg dose which
was ineffective in reducing immobility time in the mFST
produced plasma drug levels (C30min = 1288 ± 550 nM) that
would not be expected to achieve mGlu2/3 antagonist-relevant
concentrations in the central compartment (predicted brain = 19
nM). Testing of compounds 25, 26, 28-30 similarly using ip
dosing showed them to be less active in this model (ED60 2.1 –
9.5 mg/kg).17
% inhibition of human Cyp isoforms (3A4,
2C9, 2D6)d
hERG binding Kie
<5%
>100 M
>40%
Plasma unbound fraction (rodent, monkey
and human)f
Table 4: Solubility, in vitro ADME and toxicology attributes of
compound 18. aTested at 20 M. bTested at 5 and 25 M using
Cimetidine as a positive control. cTested at 4 M with a 30
minute incubation time. dTested at 10 M. eDetermined using
f
[3H]Astemizole. Tested at 1 M.
The pharmacokinetic profile of compound 18 was assessed in
rodent using intravenous (iv), oral and intraperitoneal (ip)
administration (Table 5). It showed low bioavailability using
oral dosing at 5 mg/kg (F = 6%) and achieved only modest Cmax
plasma concentration levels at this dose (262 ± 26 nM).
However, it showed low clearance (CL = 3.67 ± 0.225 mL min-1
kg-1), a low volume of distribution (Vdss = 0.589 ± 0.020 L/kg),
We further evaluated the wake promoting effects of 18 in a rat
EEG sleep study. Previous work has demonstrated a wake-

promoting effect elicited by LY341495 (Figure 1) an mGlu2/3
antagonist.4 Using ip administration, compound 18 produced a
dose dependent wake promoting effect in rats at 3 and 10 mg/kg
as measured by total increases in accumulated wake over 7 hours
(41 minutes and 131 minutes respectively vs vehicle control)
without any significant increases in recorded locomotor activity.
There was no significant effect on the percentage of NREM sleep
however there was a significant increase in the percentage of
REM sleep consistent with previously reported sleep studies with
LY341495.4
Monn, James A.; Schoepp, Darryle D.; Nat. Med. 2007,
13(9), 1102.
(3) (a) Chaki, Shigeyuki; Yoshikawa, Ryoko; Hirota, Shiho;
Shimazaki, Toshiharu; Maeda, Maoko; Kawashima, Naoya;
Yoshimizu, Takao; Yasuhara, Akito; Sakagami, Kazunari;
Okuyama, Shigeru; Nakanishi, Shigetada; Nakazato, Atsuro
Neuropharmacology 2004, 46(4), 457-467. (b) Karasawa,
Jun-Ichi; Shimazaki, Toshiharu; Kawashima, Naoya; Chaki,
Shigeyuki; Brain Res. 2005, 1042(1), 92.
(4) Feinberg, I; Schoepp, D. D.; Hsieh, K.-C.; Darchia, N.;
Campbell, I. G.; J. Pharmacol. Exp. Ther. 2005, 312(2),
826.
182
200
(5) Dominguez-Fernandez, Carmen; Helton, David Reed;
Massey, Steven Marc; Monn, James Allen; Eur. Pat.
Appl. (1997), EP 774455 A1.
142
150
98
(6) Monn, James A.; Valli, Matthew J.; Massey, Steven M.;
Hao, Junliang; Reinhard, Matthew R.; Bures, Mark G.;
Heinz, Beverly A.; Wang, Xushan; Carter, Joan H.; Getman,
Brian G.; Stephenson, Gregory A.; Herin, Marc; Catlow,
John T.; Swanson, Steven; Johnson, Bryan G.; McKinzie,
David L.; Henry, Steven S.; J. Med. Chem. 2013,
56(11), 4442.
*
100
60
*
50
0
veh
0.3
1
3
mg/kg17
(7) Monn, James A.; Prieto, Lourdes; Taboada, Lorena;
Pedregal, Concepcion; Hao, Junliang; Reinhard, Matt R.;
Henry, Steven S.; Goldsmith, Paul J.; Beadle, Christopher
D.; Walton, Lesley; Man, Teresa; Rudyk, Helene; Clark,
Barry; Tupper, David; Baker, S. Richard; Lamas, Carlos;
Montero, Carlos; Marcos, Alicia; Blanco, Jaime; Bures,
Mark; Clawson, David K.; Atwell, Shane; Lu, Frances;
Wang, Jing; Russell, Marijane; Heinz, Beverly A.; Wang,
Xushan; Carter, Joan H.; Xiang, Chuanxi; Catlow, John T.;
Swanson, Steven; Sanger, Helen; Broad, Lisa M.; Johnson,
Michael P.; Knopp, Kelly L.; Simmons, Rosa M. A.;
Johnson, Bryan G.; Shaw, David B.; McKinzie, David L.; J.
Med. Chem. 2015, 58(4), 1776 and references therein.
Figure 4: Activity of compound 18 in th mFST following ip
administration.
*Significant difference vs vehicle with a
Dunnett's test value < 0.05.
In conclusion, we developed a chemistry route that has allowed
for the dual functionalization of both the C3 and C4 positions of
the bicyclo[3.1.0]hexane ring system.
In particular, our
regioselective silver mediated alkylation of diol 5 provided a
highly efficient way to place benzyl ethers at the C3 position in a
-face stereochemical orientation, a group and configuration that
is known to produce functionally active antagonist in the
bicyclo[3.1.0]hexane ring system,18 while providing a hydroxyl
substituent at the C4 position for further elaboration. In the
context of the SAR study we successfully leveraged binding data
from our agonist research efforts to drive C4 group optimization
to obtain antagonists with high potency at both the mGlu2 and
mGlu3 receptors. In rodents, an optimized analog from this
investigation, 18, was shown to demonstrate good
pharmacokinetic characteristics with low clearance and volume
of distribution as well as robust efficacy in in mGlu2/3 antagonist-
sensitive animal models. Therefore, 18 should be a useful tool
for studying mGlu2/3 receptor function both in vitro and in vivo.
(8) Observed nOes used to establish the structure of 6a.
(9) Genazzani A. A.; Casabona G.; L'Episcopo M. R.;
Condorelli D. F.; Dell'Albani P.; Shinozaki H.; Nicoletti F.;
Brain Res. 1993, 622(1-2), 132.
Acknowledgements
(10) Kuo, Fengjiun; Kulanthaivel, Palaniappan; Rener, Gregory
A.; Yi, Ping; Wheeler, William; J. Labelled Compd.
Radiopharm. 2004, 47, 571.
James Monn is acknowledged for his extensive assistance with
editing of this manuscript. Jon Erickson and Mark Bures are
acknowledged for their assistance with manual docking studies.
(11) J. Med. Chem. in press. PDB deposition code 5KZQ
contains the supplementary crystallographic data for this
paper.
References and Notes
(1) Dunayevich, Eduardo; Erickson, Janelle; Levine, Louise;
Landbloom, Ronald; Schoepp, Darryle D.; Tollefson, Gary
D. Neuropsychopharmacology, 2008, 33(7), 1603.
(12) Molecular Operating Environment (MOE), 2013.08;
Chemical Computing Group Inc., 1010 Sherbooke St. West,
Suite #910, Montreal, QC, Canada, H3A 2R7, 2014.
(2) Patel, Sandeep T.; Zhang, Lu; Martenyi, Ferenc; Lowe,
Stephen L.; Jackson, Kimberley A.; Andreev, Boris V.;
Avedisova, Alla S.; Bardenstein, Leonid M.; Gurovich, Issak
Y.; Morozova, Margarita A.; Mosolov, Sergey N.;
Neznanov, Nikolai G.; Reznik, Alexander M.; Smulevich,
Anatoly B.; Tochilov, Vladimir A.; Johnson, Bryan G.;
(13) MarvinSketch version 6, ChemAxon LLC Cambridge
Innovation Center, One Broadway Cambridge, MA 02142.
(14) Feng, Bo; Mills, Jessica B.; Davidson, Ralph E.; Mireles,
Rouchelle J.; Janiszewski, John S.; Troutman, Matthew D.;

de Morais, Sonia M.;
Drug Metab. Dispos.
2008, 36(2), 268.
(15) Can, A.; Dao, D. T.; Arad, M.; Terrillion, C. E.;
Piantadosi, S. C.; Gould, T. D.; J. Vis. Exp. 2012, 59,
3638.
(16) Petit-Demouliere, Benoit; Chenu, Franck; Bourin, Michel;
Psychopharmacology 2005, 177(3), 245.
(17) See supplementary data for additional mouse forced swim
data and methods.
(18) Nakazato, Atsuro; Sakagami, Kazunari; Yasuhara, Akito;
Ohta, Hiroshi; Yoshikawa, Ryoko; Itoh, Manabu;
Nakamura, Masato; Chaki, Shigeyuki; J. Med. Chem.
2004, 47(18), 4570.
(19) In all cases the maximum % inhibition achieved in reversing
an EC90 dose of DCG-IV was greater than 94%. Reported
values are an average of >5 testing runs.
(20) The following formulations were used for compound dosing:
iv, A pH adjusted solution in saline; po, A solution in water;
ip, A suspension in water modified with hydroxyethyl-
cellulose and antifoam 1510.