Synthesis and Pharmacological Characterization of C4-(Thiotriazolyl)-substituted-2-aminobicyclo[3.1.0]hexane-2,6-dicarboxylates. Identification of (1R,2S,4R,5R,6R)-2-Amino-4-(1H-1,2,4-triazol-3-ylsulfanyl)bicyclo[3.1.0]hexane-2,6-dicarboxylic Acid (LY2812223), a Highly Potent, Functionally Selective mGlu2 Receptor Agonist

Article abstract of DOI:10.1021/acs.jmedchem.5b01124

Identification of orthosteric mGlu_2/3 receptor agonists capable of discriminating between individual mGlu₂ and mGlu₃ subtypes has been highly challenging owing to the glutamate-site sequence homology between these proteins. Herein we detail the preparation and characterization of a series of molecules related to (1S,2S,5R,6S)-2-aminobicyclo[3.1.0]hexane-2,6-dicarboxylate 1 (LY354740) bearing C4-thiotriazole substituents. On the basis of second messenger responses in cells expressing other recombinant human mGlu_2/3 subtypes, a number of high potency and efficacy mGlu₂ receptor agonists exhibiting low potency mGlu₃ partial agonist/antagonist activity were identified. From this, (1R,2S,4R,5R,6R)-2-amino-4-(1H-1,2,4-triazol-3-ylsulfanyl)bicyclo[3.1.0]hexane-2,6-dicarboxylic acid 14a (LY2812223) was further characterized. Cocrystallization of 14a with the amino terminal domains of hmGlu₂ and hmGlu₃ combined with site-directed mutation studies has clarified the underlying molecular basis of this unique pharmacology. Evaluation of 14a in a rat model responsive to mGlu₂ receptor activation coupled with a measure of central drug disposition provides evidence that this molecule engages and activates central mGlu₂ receptors in vivo.

Full text of DOI:10.1021/acs.jmedchem.5b01124

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Synthesis and Pharmacological Characterization of C4-(Thiotriazolyl)-
Substituted-2-Aminobicyclo[3.1.0]hexane-2,6-Dicarboxylates.
Identification of (1R,2S,4R,5R,6R)-2-Amino-4-(1H-1,2,4-triazol-3-
ylsulfanyl)bicyclo[3.1.0]hexane-2,6-dicarboxylic acid (LY2812223),
a Highly Potent, Functionally Selective mGlu2 Receptor Agonist
James A Monn, Lourdes Prieto, Lorena Taboada, Junliang Hao, Matt R Reinhard, Steven S. Henry,
Christopher D. Beadle, Lesley Walton, Teresa Man, Helene Rudyk, Barry Clark, David Tupper, S. Richard
Baker, Carlos Lamas, Carlos Montero, Alicia Marcos, Jaime Blanco, Mark Bures, David K. Clawson, Shane
Atwell, Frances Lu, Jing Wang, Marijane Russell, Beverly A Heinz, Xushan Wang, Joan H. Carter, Brian
G. Getman, John T. Catlow, Steven Swanson, Bryan G. Johnson, David B. Shaw, and David L. McKinzie
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01124 • Publication Date (Web): 27 Aug 2015
Downloaded from http://pubs.acs.org on September 3, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Russell, Marijane; Eli Lilly and Company,
Heinz, Beverly; Eli Lilly and Company, Quantitative Biology
Wang, Xushan; Eli Lilly and Company, Quantitative Biology
Carter, Joan; Eli Lilly and Company, Quantitative Biology
Getman, Brian; Eli Lilly and Company, Quantitative Biology
Catlow, John; Eli Lilly and Company, Drug Disposition
Swanson, Steven; Eli Lilly and Company, Drug Disposition
Johnson, Bryan; Eli Lilly and Company, Neuroscience Research
Shaw, David; Eli Lilly and Company, Neuroscience Research
McKinzie, David; Eli Lilly and Company, Neuroscience Research
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Synthesis and Pharmacological Characterization of C4-(Thiotriazolyl)-Substituted-2-
Aminobicyclo[3.1.0]hexane-2,6-Dicarboxylates.
Amino-4-(1H-1,2,4-triazol-3-ylsulfanyl)bicyclo[3.1.0]hexane-2,6-dicarboxylic
LY2812223), a Highly Potent, Functionally Selective mGlu Receptor Agonist
Identification of (1R,2S,4R,5R,6R)-2-
acid
(
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ф*
ф
ф
ф
ф
James A. Monn, Lourdes Prieto, Lorena Taboada, Junliang Hao, Matthew R. Reinhard,
ф
ф
ф
ф
ф
Steven S. Henry, Christopher D. Beadle, Lesley Walton, Teresa Man, Helene Rudyk, Barry
ф
ф
ф
ф
ф
ф
Clark, David Tupper, S. Richard Baker, Carlos Lamas, Carlos Montero, Alicia Marcos,
ф
ф
‡
‡
‡
‡
Jaime Blanco, Mark Bures, David K. Clawson, Shane Atwell, Frances Lu, Jing Wang,
‡
†
†
†
†
Marijane Russell, Beverly A. Heinz, Xushan Wang, Joan H. Carter, Brian G. Getman, John
†
†
§
§
§
T. Catlow, Steven Swanson, Bryan G. Johnson, David B. Shaw, and David L. McKinzie
ф
†
‡
Discovery Chemistry Research and Technologies, Quantitative Biology, Structural Biology,
†
§
Drug Disposition and Neuroscience Research, Eli Lilly and Company, Indianapolis, IN 46285.
Abstract:
Identification of orthosteric mGlu2/3 receptor agonists capable of discriminating between
individual mGlu and mGlu subtypes has been highly challenging owing to the glutamateꢀsite
2
3
sequence homology between these proteins.
characterization of a series of molecules related to 1S,2S,5R,6Sꢀ2ꢀaminobicyclo[3.1.0]hexaneꢀ
,6ꢀdicarboxylate 1 (LY354740) bearing C4ꢀthiotriazole substituents. Based on second
Herein we detail the preparation and
2
messenger responses in cells expressing recombinant human mGlu2/3 subtypes, a number of high
potency and efficacy mGlu receptor agonists exhibiting low potency mGlu partial agonist /
2
3
antagonist activity were identified. From this, (1R,2S,4R,5R,6R)ꢀ2ꢀAminoꢀ4ꢀ(1Hꢀ1,2,4ꢀtriazolꢀ
ꢀylsulfanyl)bicyclo[3.1.0]hexaneꢀ2,6ꢀdicarboxylic acid 14a (LY2812223) was further
characterized. Coꢀcrystallization of 14a with the amino terminal domains of hmGlu and
3
2
hmGlu combined with siteꢀdirected mutation studies has clarified the underlying molecular
3
basis of this unique pharmacology. Evaluation of 14a in a rat model responsive to mGlu2
receptor activation coupled with a measure of central drug disposition provides evidence that this
molecule engages and activates central mGlu receptors in vivo.
2
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Metabotropic glutamate (mGlu) receptors are members of the family C subgroup of the sevenꢀ
transmembrane (7ꢀTM) domain receptor superfamily. The mGlu receptors have attracted
considerable attention as targets for small moleculeꢀbased therapeutic interventions in both
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
central nervous system (CNS) and nonꢀCNS disorders. Of the eight known mGlu subtypes
(
mGlu ), mGlu and mGlu receptors are among the most extensively studied owing to both the
1ꢀ8 2 3
availability of potent and selective small molecule orthosteric agonists, orthosteric antagonists
and allosteric modulators (both positive and negative) and to the robust physiologic effects
produced by many of these ligands in preclinical models of psychiatric (e.g. anxiety, depression,
2
3
schizophrenia and addiction), neurologic (e.g. neuropathic pain and neurodegeneration), and
4
proliferative (e.g. glioma) disorders. Interest in identifying molecules that elicit or enhance
mGlu2/3 receptor signaling has been the subject of particular interest owing to clinical evidence
5
6
demonstrating beneficial effects of mGlu2/3 receptor agonists 1 and 2 (Figure 1; dosed as oral
7
8
prodrug forms 3 and 4) in generalized anxiety disorder and schizophrenia patients, respectively.
Though the efficacy of 4 in chronic schizophrenia patients was not confirmed in subsequent
9a,b
9bꢀd
clinical trials, an antipsychotic response in patient subgroups was noted.
Figure 1. Clinically evaluated mGlu2/3 receptor agonists (1 and 2) and their orally bioavailable
prodrugs (3 and 4).
Owing to their high degree of sequence homology, overlapping orthosteric agonist and
antagonist pharmacology and common signal transduction mechanisms, mGlu and mGlu3
2
receptors have historically been clustered together (i.e. group II mGluRs). However, these two
1
0
signaling molecules are differentially localized in the central nervous system (CNS) and have
divergent functions. For instance, while both mGlu and mGlu receptors appear to play a role in
2
3
1
1
antiꢀstress effects produced by mGlu2/3 agonists in rodents, mGlu receptors appear to mediate
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
12
13
both the acute antipsychotic and antinociceptive responses to mGlu agonists in animal
2/3
models of psychosis and pain, while activation of glial mGlu receptors mediates mGlu2/3
3
agonistꢀelicited neuroprotection via mechanisms involving increased glutamate transporter
1
4,15c
15
and the release of trophic factors including GDNF and TGFβ.
protein expression
Finally,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
postꢀsynaptically localized mGlu receptors appear to enhance network activity and induce theta
3
16
oscillations in the hippocampus as well as to trigger long term depression in the prefrontal
1
7
cortex, mechanisms which may have therapeutic implications for the treatment of thought
disorders and the extinguishing of fear memories. Elucidation of mGlu vs. mGlu receptor
2
3
function in these studies has been primarily enabled by the use of murine lines in which
embryonic deletions of the genes encoding mGlu and/or mGlu receptors were employed.
2
3
While these biological reagents are useful for validating the roles of specific proteins in
physiologic and pathophysiologic processes and for establishing the mechanism(s) of small and
large molecule ligands, compensatory changes can occur over the course of development,
potentially confounding the interpretation of biological studies. For example, deletion of the
gene encoding mGlu receptors led to increased hippocampal expression of both mGlu and
2
3
NMDA receptor subunit NR2A and a decrease in expression of the glutamate transporter
EAAT3, while deletion of the gene encoding mGlu receptors resulted in the increased
3
hippocampal expression of both mGlu and NR2A and decreased expression of glutamate
2
18
transporters (EAAT1/2). Therefore, in order to more firmly establish the roles of mGlu and
2
mGlu receptor subtypes in CNS disorders, it is of continued interest to identify potent
3
pharmacological tools that are capable of differentiating these receptors in both in vitro and in
vivo systems.
19
The primary amino acid sequences for both mGlu and mGlu receptors are known.
2
3
Furthermore, crystal structures of the amino terminal domain (ATD) segments of both rat (with
2
0a
20b
agonists)
and human (with an antagonist)
mGlu receptors have been described. Owing to
3
the high degree of protein sequence homology between these two receptor subtypes, especially
those residues involved in glutamate binding (Table 1), identification of glutamateꢀsite
(
orthosteric) agonists or antagonists capable of differentiating between these receptor subtypes
has been an extremely challenging endeavor.
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Full length and subꢀdomain homology between mGlu and mGlu
2
3
receptors
mGlu vs. mGlu
2
3
Protein Segment
% Identity
% Homology
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Full Length Receptor
67
71
81
84
7
ꢀTransmembrane (7ꢀTM) Domain
Amino Terminal Domain
64
80
a,b
Glutamateꢀbinding site residues
100
100
a
mGlu : R57, R61, K377, D295, T168, S145, A166, S143
2
b
mGlu : R64, R68, K389, D301, T174, S151, A172, S149
3
Indeed, while many positive allosteric modulators (PAMs) which bind within the 7ꢀTM domain
2
1
of mGlu have been identified , only two examples of orthosteric agonists (5 and 6, Figure 2)
2
capable of selective mGlu ꢀreceptor activation have been described. Of these, (1S,2S,4R,5R,6S)ꢀ
2
2
ꢀaminoꢀ4ꢀmethylbicyclo[3.1.0]hexaneꢀ2,6ꢀdicarboxylic acid, 5 (LY541850) displayed unusual
2
2
mGlu agonist (EC = 161 nM) and mGlu antagonist (IC = 1 ꢀM) activity while 6 is
2
50
3
50
reported to activate mGlu with low potency (EC50 = 10 ꢀM) but excellent selectivity over
2
23
mGlu (EC > 1000 ꢀM).
Interestingly, expansion of the structureꢀactivity relationship
3
50
(
SAR) for compounds substituted at the C4ꢀposition of 1 was not successful in identifying
additional mGlu receptorꢀselective agonists such as 5, but instead led to the identification of
2
multiple highly potent, pharmacologically blended mGlu2/3 receptor agonists such as
(
1R,2S,4R,5R,6R)ꢀ2ꢀaminoꢀ4ꢀfluorobicyclo[3.1.0]hexaneꢀ2,6ꢀdicarboxylic acid 7, (LY459477,
2
4
Figure 2)
and (1R,4S,5S,6S)ꢀ4ꢀaminospiro[bicyclo[3.1.0]hexaneꢀ2,1'ꢀcyclopropane]ꢀ4,6ꢀ
2
5
dicarboxylic acid 8, (LY2934747, Figure 2)
as well as low potency mGlu2/3 antagonists (e.g.
2
4
9
, 10, Figure 2).
Moderate (2.25 – 2.3 Å) resolution coꢀcrystal structures of the human mGlu2
ATD with compounds 1 and 8, and the human mGlu ATD with compound 1 were recently used
3
to more precisely understand agonistꢀreceptor interactions and provided a rationale for the mGlu2
25
agonist, mGlu antagonist pharmacological profile of 5.
In this account, we have extended our
3
investigation of the effect of substituents attached to the C4ꢀposition of 1 on mGlu2/3 receptor
binding and function. Specifically, the present investigation has focused on the preparation and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
evaluation of compounds in which substituted and unsubstituted thiotriazoles are attached to the
C4 and C4 ꢀpositions of 1. From this work we have identified a series of highly potent and
α
β
efficacious hmGlu receptor agonists that exhibit minimal hmGlu agonist activity.
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Selective and nonꢀselective mGlu2/3 receptor agonists and antagonists
_____________________________________________________________________
Lower potency,
High potency,
Low potency,
pharmacologically-selective
mGlu2 agonists
pharmacologically-balanced
mGlu2/3 agonists
pharmacologically-balanced
mGlu2/3 antagonists
CH3
H
F
SPh
H
H
HO C
4
HO C
HO C
4
2
5
2
2
6
3
H
1
H
H
CO H
CO H
CO H
2
2
2
2
H
H
H
NH2
NH2
NH2
724
924
522
SPh
4
H
H
H
H
HO C
HO C
2
2
H
HO C
CO H
2
2
H
H
CO H
CO H
2
2
NH2
NH2
NH2
rac-623
825
1024
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Chemistry
Nonꢀracemic C4 ꢀsubstituted thiotriazoles (14a-g) were readily prepared from the previously
α
2
4
described nonꢀracemic C4 ꢀtosylate 11 (Scheme 1) while nonꢀracemic C4 ꢀsubstituted
β
β
2
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
thiotriazoles (16a-g) were prepared in an analogous manner from nonꢀracemic C4 ꢀbromide 12
α
(
Scheme 2). In each case, thiolate displacement of the C4ꢀleaving group (OTs or Br) proceeded
with inversion of configuration with yields ranging from 42ꢀ99%. Assignment of relative
stereochemistry for intermediates 13a-g and 15a-g was made on the basis of the C4 proton
1
splitting pattern in the HꢀNMR spectrum.
In all cases the C4 proton for αꢀsubstituted
thiotriazoles 13a-g is a multiplet while in βꢀsubstituted thiotriazoles 15a-g the C4 proton appears
2
6
as a doublet. Removal of the amino acid protecting groups was achieved in a single step under
a variety of conditions (e.g. ZnBr , HCl, or AcOH). The final products 14a-g, 16a-g were
2
isolated as either HCl salts or neutral zwitterions (see experimental section for details).
Confirmation of absolute and relative stereochemistry of one final product from this series (14a)
was obtained from the structure of this compound coꢀcrystalized with hmGlu and hmGlu amino
2
3
1
terminal domain (ATD) proteins (see Figure 8). All final compounds were characterized by Hꢀ
13
NMR, CꢀNMR and LCMS and judged to be >95% pure by capillary electrophoresis.
Scheme 1. Preparation of C4 ꢀsubstituted thiotriazoles 14a-g
α
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Scheme 2. Preparation of C4 ꢀsubstituted thiotriazoles 16a-g
β
N
N
N
N
N
N
HS
R
N
Br
H
S
R
S
R
H
H
H
H
N
H
H
N
H
H
tꢀBuO C
tꢀBuO C
HO C
2
2
K CO , DMF
2
2
3
Deprotection
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
or
H
CO tꢀBu
2
H
CO2tꢀBu
NHBoc
H
CO2H
N
N
NHBoc
NH2
1
2
ꢀ
15a-f
16a-f
S
R
N
H
Na+
1
1
1
6a: R=H
6b: R=CH3
DMF
6c: R=CF3
16d: R=CHF2
16e: R=NH2
6f: R=CH(CH3)2
1
N
N
N
N
N
N
NH
NH
NH
S
S
ꢀ
H
H
H
H
S
Na+
tꢀBuO C
HO C
2
2
Deprotection
DMF
H
CO tꢀBu
2
H
CO2H
NHBoc
NH2
15g
16g
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Results
hmGlu2/3 receptor binding: Compounds 14a-g, 16a-g along with comparators 1, 9 and 10
27
3
28
were assessed for their ability to displace [ H]ꢀ7 from membranes expressing recombinant
human mGlu or mGlu receptors (Table 2). Across the entire series, mGlu and mGlu Ki
2
3
2
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
values were poorly correlated with each other (r = 0.29, supporting information Figure S1).
Analogs bearing C4 ꢀthiotriazole substituents (14a-g) exhibited a limited range of binding
α
affinities at both hmGlu (Ki values ranged from 67.0 nM for 14e to 266 nM for 14f) and
2
hmGlu (Ki values ranging from 56.7 nM for 14c to 259 nM for 14g) with no remarkable (i.e.
3
within 10x) loss or improvement of affinity over either the unsubstituted analog 1 or C4αꢀ
thiophenylether 9. Furthermore, C4 ꢀsubstituted compounds 14a-g did not exhibit a notable
α
difference in binding affinities between hmGlu and hmGlu receptors (affinity ratios were
2
3
within 2ꢀfold between these two receptors across the series). Conversely, analogs bearing C4βꢀ
thiotriazole substituents (16a-g) exhibited a considerable dynamic range of binding affinities for
hmGlu (Ki values ranging from 8.55 nM for 16c to 339 nM for 16f), while the range in the
2
affinities of 16a-g for hmGlu was relatively narrow (Ki values ranging from 43.1 nM for 16c to
3
1
26 nM for 16a). Enhanced mGlu receptor affinity for certain members of the current series
2
compared to both unsubstituted analog 1 and C4 ꢀthiophenylether 10 was apparent. Noteworthy
β
in this regard was the nearly 150ꢀfold improvement in mGlu binding affinity for 16c (Ki = 8.55
2
nM) over 10 (Ki = 1,230 nM). As was observed for C4 ꢀsubstituted analogs 14a-g, individual
α
C4 ꢀsubstituted analogs 16a-g exhibited no particular binding preference for one receptor
β
subtype over the other, demonstrating mGlu2/3 affinity ratios for individual molecules within 5ꢀ
fold across the series. The effect of C4ꢀstereochemistry on binding affinity for otherwise
identically substituted analogs (i.e. comparing matched pairs of 14a-g and 16a-g) varied.
Diastereomers 14c and 16c (4ꢀCF substituted triazoles) demonstrated the greatest difference in
3
binding affinities (10ꢀfold at mGlu favoring the C4 ꢀisomer) while most other isomeric pairs
2
β
showed only modest (1ꢀ3 fold) differences in Ki values.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Binding affinities and functional agonist or antagonist responses for C4ꢀsubstituted analogs in cells expressing recombinant
a
human mGlu and mGlu receptors
2
3
3
Inhibition of Hꢀ459477
binding to human mGlu
receptors
Functional agonist or antagonist activity in cells expressing recombinant
human mGlu receptors
b
EC50 (nM) or IC50 (nM) ± SEM (Emax or Imax)
Ki (nM) ± SEM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
R
No.
mGlu2
mGlu3
mGlu2
mGlu3
2
+
c
2+
d
c
Ca
FLIPR
cAMP
Ca FLIPR
cAMP
d
7
.0 ± 1.2
34.4 ± 4.01
(94)
27.9 ± 3.0
(100%)
140 ± 4.6
(103%)
H
1
9
71.7 ± 5.9
420 ± 30.3
107 ± 12.5
138 ± 15.3
(100%)
e
>12500
>25000
1070 ± 258 (92)
>25000
7690 ± 7260 (62)
SPh
7000± 645 (80)
>
12500
1
170 ± 232 (49)
15300± 5400
>25000
12500
>25000
3370 ± 1210 (77)
>5000
6710 (n=2) (56)
SPh
10
1230 ± 139
272 ± 25.1
>
(53)
>25000
25200 ± 15600
1
29 ± 34.9 (61)
H
14a
14b
14c
144 ± 20.0
121 ± 34.3
84.4 ± 13.3
156 ± 23.5
62.4 ± 12.6
56.7 ± 8.86
5.6 ± 0.5 (99)
10.4 ± 1.8 (99)
13.9 ± 2.6 (99)
21.1 ± 2.2 (95)
39.5 ± 7.5
2
730 ± 1430 (40)
(58)
Me
CF3
103 ± 13 (77)
105 ± 78 (49)
(100)
7
5
9.5 ± 18.7 (66)
28± 99.4 (46)
54.2 ± 6.3 (82)
82.0 ± 49.6 (37)
5
3.6 ± 13.7
82)
77.7 ± 14 (60)
459 ± 209 (33)
>25000
154 ± 34.4 (71)
CHF2
NH2
14d
14e
14f
14g
95.8 ± 9.35
67.0 ± 8.11
266 ± 60.3
142 ± 18.0
63.2 ± 9.37
58.3 ± 9.11
157 ± 33.8
259 ± 20.6
11.8 ± 0.3 (98)
4.5 ± 0.2 (100)
16.5 ± 4.5 (100)
12.4 ± 1.4 (99)
(
1
3.0 ± 2.6
100)
>25000
94.7 ± 18.9 (80)
74 ± 11.1 (77)
(
2
8
81 ± 63.3 (67)
59 ± 340 (59)
>25000
>12500 (47)
iPr
87.8 ± 6.6 (89)
f
46.1 ± 17.6
235 ± 37.2 (69)
2380 ± 1270 (49)
>25000
5510 ± 1860 (63)
1
23Triaz
(92)
69.1 ± 21.9
89)
227 ± 39.1 (52)
4260 ± 4090 (54)
>25000
3100 ± 170 (72)
H
16a
16b
60.4 ± 6.20
171 ± 16.7
126 ± 14.5
79.0 ± 9.4
11.2 ± 1.3 (99)
44.3 ± 5.5 (99)
(
>25000
Me
146 ± 24.3 (84)
7.8 ± 1.6 (99)
6.3 ± 1.9 (95)
96.6 ± 28.8 (74)
30.5 ± 4.1 (83)
27.3 ± 8.8 (80)
112 ± 25.4 (68)
3
590 (57, n=1)
0
0
.65 ± 0.07
100)
>25000
624 ± 473 (61)
CF3
CHF2
NH2
16c
16d
16e
8.55 ± 1.63
13.4 ± 2.55
27.8 ± 2.92
43.1 ± 14.1
57.5 ± 12.1
69.0 ± 10.7
(
.58 ± 0.06
100)
>25000
244 ± 90.8 (79)
(
2
0.1 ± 4.2
100)
>25000
3870 ± 1490 (62)
6.4 ± 1.0 (99)
(
ACS Paragon Plus Environment

Page 11 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
>25000
350 ± 1690 (62)
iPr
16f
339 ± 41.0
111 ± 14.0
82.7 ± 3.78
74.0 ± 6.5 (99)
5.2 ± 1.5 (98)
242 ± 27.5 (71)
19.0 ± 0.9 (94)
77.9 ± 7.0 (87)
92.7 ± 21.9 (81)
2
f
>25000
2790 ± 1170 (67)
1
23Triaz
16g
65.1 ± 5.87
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
b
c
See experimental section for assay details; Maximal response achieved compared to glutamate (100%); hmGlu B
2
max
=
21859 fmol/mg protein, hmGlu Bmax = 5271fmol/mg protein. Agonist activity assessed as an inhibition of forskolinꢀ
3
d
stimulated cAMP, antagonist activity assessed as a reversal of DCGꢀIVꢀinhibited, forskolinꢀstimulated cAMP. hmGlu
2
2+
Bmax = 1350 fmol/mg protein, hmGlu B = 6100 fmol/mg protein. Agonist activity assessed as a stimulation of Ca
3
max
2+
e
mobilization, antagonist activity assessed as an inhibition of glutamateꢀstimulated Ca mobilization. As previously
disclosed (ref. 24) in two out of eight agonist format assays, low potency and submaximal (~30%) efficacy agonist
responses were observed. The other six agonist format assays failed to demonstrate an agonist response above baseline.
In comparison, six out of six antagonist format assays displayed consistent antagonist effects with 100% maximal
f
inhibition of glutamate. 123Triaz = 1,2,3ꢀtriazolꢀ4ꢀyl
2
+
27
hmGlu receptor cAMP and Ca FLIPR assays:
Compounds 14a-g and 16a-g produced
2
potent and maximallyꢀefficacious agonist responses in cells expressing human mGlu receptors,
2
2
inhibiting forskolinꢀstimulated cAMP formation in mGlu ꢀexpressing cells with calculated EC50
values ranging from 0.58 nM (16d) to 74.0 nM (16f) (Table 2, representative concentrationꢀ
response curves shown in Figure 4a). Relative to prototypic agonist 1, mGlu agonist potency in
2
the current series ranged from an approximately 10ꢀfold improvement for C4 ꢀsubstituted
β
thiotriazoles 16c and 16d to a 13ꢀfold loss in potency for 16f. The highly potent mGlu cAMP
2
agonist responses observed for 14a-g and 16a-g are in sharp contrast to the pharmacology
exhibited by thiophenylethers 9 (mGlu antagonist IC = 7,000 nM) and 10 (mGlu agonist EC
50
2
50
2
24
=
1,170 nM (49% max response), antagonist IC = 15,300 nM (53% maximal response) when
50
tested under identical assay conditions. The effect of C4ꢀstereochemistry on mGlu cAMP
2
agonist potency was generally modest, with EC50 values for diastereomeric pairs 14a/16a,
1
4b/16b, 14e/16e, 14f/16f and 14g/16g being within 3ꢀfold of one another. In contrast, C4βꢀ
substituted analogs 16c (R = CF ) and 16d (R = CF ) exhibited a considerable potency
3
2
enhancement (20ꢀfold) over their C4 ꢀsubstituted counterparts 14c and 14d. The remarkable
α
increase in functional potency for these analogs may be related to a predicted decrease in pK of
a
the triazole NH relative to that of unsubstituted triazole 16a (calculated pK = 8.8, 5.9 and 6.8 for
a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
29
1
6a, 16c and 16d, respectively). Indeed, for C4 ꢀsubstituted analogs 16a-g, a positive
β
correlation between the calculated pK of the triazole NH and functional agonist potency at
a
2
mGlu (Figure 3b, r = 0.85) was observed, suggesting that the deprotonated triazole may be
2
preferred for mGlu receptor binding. Interestingly, this relationship was not apparent in the case
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
of either C4 ꢀsubstituted analogs 14a-g (Figure 3a, r = 0.05) or in the combined (14a-g, 16a-g)
α
2
dataset (r < 0.35, data not shown), suggesting there may be distinct molecular drivers for the
observed functional potencies between these diastereomericallyꢀrelated series.
a
b
Figure 3. Relationship of calculated pK of thiotriazole NH and functional agonist potency for
a
αꢀsubstituted thiotriazoles 14a-g (a) and βꢀsubstituted thiotriazoles 16a-g (b) in cells expressing
recombinant human mGlu receptors.
2
a.
b.
a
27
b
Calculated values from Marvin ; pEC values calculated from measured cAMP EC values
50
50
found in Table 2. See supporting information Table S1 for the actual pK and pEC values used
a
50
in these regression analyses.
In the hmGlu calcium mobilization (FLIPR) assay, potent agonist responses were also generally
2
observed across the current series (Table 2, representative compounds in Figure 4b) and highly
2
correlated with cAMP EC50 values (r = 0.91, supporting information Figure S2) though
potencies measured in the FLIPR assay were generally lower than those derived from the cAMP
format. This may be due to a number of factors associated with these two assay systems,
including (a) higher receptor expression levels in cells utilized in the cAMP assay (Bmax = 21.8
pmol/mg protein) vs. FLIPR assay (Bmax = 1.35 pmol/mg protein), (b) less efficient coupling of
2
+
hmGlu receptors to G (Ca FLIPR) compared to G (cAMP), (c) different experimental
2
αq
i/o
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conditions under which these assays are run (cAMP assay measures accumulated cAMP at 37 °C
2
+
2+
over 20 min while Ca FLIPR measures Ca transients at 24 °C over 2.5 min), or to some
combination of these variables. The net effect of these differences leads to an overall
2
+
amplification of agonist signals in the cAMP vs. Ca FLIPR formats. Human mGlu receptor
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
binding affinity was highly correlated with functional agonist potency measured in both cAMP
2
2
(
r = 0.77) and FLIPR (r = 0.65) formats (supporting information Figure S3). Relative to
hmGlu binding affinity, agonist potency in the cAMP assay was leftꢀshifted approximately 5ꢀ
2
fold (for 16a, 16b, 16e, 16f) to 20ꢀfold (for 14a, 16c, 16d, 16g). In contrast, agonist potency in
2
+
the Ca FLIPR assay was less markedly leftꢀshifted relative to affinity, with Ki / EC ratios
50
2+
ranging from 1ꢀ7. The degree of agonist efficacy in the hmGlu Ca FLIPR assay format was
2
generally high (80ꢀ100%) with the exception of 16f, which showed a somewhat blunted maximal
response (Emax = 71%). Compared to C4ꢀunsubstituted analog 1 (EC = 34.4 nM), mGlu
2
5
0
2
+
Ca FLIPR agonist potency for compounds in this series ranged from 4ꢀ to 5ꢀfold improvement
for 16c and 16d (EC = 7.8 nM and 6.3 nM respectively) to a 7ꢀfold loss in potency for 16f
5
0
(EC50 = 242 nM). In contrast to thiotriazole analogs 14a-g and 16a-g, comparator
thiophenylether analogs 9 and 10 did not produce discernible agonist responses in the mGlu2
2
+
Ca FLIPR assay.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figures 4a-d. Representative agonist concentrationꢀresponse curves for 1, 14a, 16a, 16b, 16c,
a
and 16f in cells stably expressing recombinant human mGlu and mGlu receptors.
2
3
2
+
Figure 4a. Inhibition of forskolinꢀstimulated
Figure 4b. Stimulation of intracellular Ca
cAMP production in cells expressing
release in cells expressing recombinant
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
recombinant human mGlu receptors by
human mGlu receptors by representative
2
2
representative compounds from the current
investigation
compounds from the current investigation
2
+
Figure 4c. Inhibition of forskolinꢀstimulated
Figure 4d. Stimulation of intracellular Ca
cAMP production in cells expressing
release in cells expressing recombinant
recombinant human mGlu receptors by
human mGlu receptors by representative
3
3
representative compounds from the current
investigation
compounds from the current investigation
a
See experimental section for details
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2+
27
hmGlu receptor cAMP and Ca FLIPR assays: Compounds 14a-g and 16a-g produced
3
agonist responses in the hmGlu cAMP agonist assay with calculated EC values ranging from
3
50
27.3 nM (16d) to 281 nM (14f) (Table 2, representative compounds in Figure 4c). As was
2
observed for mGlu , mGlu cAMP agonist potency correlated well with binding affinity (r =
2
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
.67, supporting information Figure S4. Relative to prototypic agonist 1 (EC = 27.1 nM),
50
mGlu cAMP agonist activity in the current series varied from being essentially equivalent to 1
3
for C4 ꢀsubstituted thiotriazoles 16c and 16d (EC values = 30.5 nM and 27.3 nM, respectively)
β
50
to a 10ꢀfold loss in agonist potency for 14f (EC = 281 nM). The consistent mGlu cAMP
50
3
agonist responses observed across compounds 14a-g and 16a-g are in sharp contrast to the
antagonist pharmacology exhibited by thiophenylethers 9 (IC = 1,070 nM) and 10 (IC = 3,370
5
0
50
2
4
nM) in this assay.
The effect of C4ꢀstereochemistry on mGlu cAMP agonist potency was
3
generally modest, with EC values for each of the diasteriomeric pairs being within 4ꢀfold of
50
one another. As was observed in the case of mGlu , functional agonist potency measured in the
2
cAMP assay format was highly correlated with calculated pKa of the thiotriazole NH for βꢀ
2
2
substituted (16aꢀg, r = 0.67) but not αꢀsubstituted (14aꢀg, r = 0.14) analogs (Figure 5).
a
b
Figure 5. Relationship of calculated pK of thiotriazole NH and functional agonist potency for
a
αꢀsubstituted thiotriazoles 14a-g (a) and βꢀsubstituted thiotriazoles 16a-g (b) in cells expressing
recombinant human mGlu receptors.
3
a.
b.
a
27
b
Calculated values from Marvin ; pEC values calculated from measured cAMP EC values
5
0
50
found in Table 2. See supporting information Table S1 for the actual pK and pEC values used
in these regression analyses.
a
50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In contrast to what was observed in the mGlu cAMP agonist assay and in spite of the amplified
2
agonist responses typically observed using the cAMP assay format, agonist responses in the
mGlu cAMP assay for this series were not consistently maximal (100% defined by maximal
3
glutamate response). None of the analogs exhibited an Emax value above 90%, four (16c, 16d,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
16f and 16g) produced Emax values ≥ 80%, eight between 65ꢀ80% and two (14a and 16a)
elicited maximal agonist efficacies below 65% (52% and 61%, respectively). For certain
compounds (14a, 14c, 14d, 14f, 14g, and 16a), mGlu cAMP antagonist IC values could be
3
50
determined (Table 2, 14a exemplified in Figure 6a), confirming the partial mGlu agonist /
3
antagonist characteristics of these compounds in this assay format. Functional agonist potencies
in the mGlu cAMP assay were not substantially shifted relative to mGlu affinity, with EC and
3
3
50
Ki values being within 3ꢀfold of one another. The relatively small leftꢀshift (vs. Ki) in this assay
compared with that seen in the case of hmGlu (vide supra) is likely the result of lower
2
3
expression of mGlu receptors in this cell line (Bmax= 5.27 pmol/mg protein) compared to mGlu2
receptor expression (Bmax= 21.8 pmol/mg protein), resulting in lower receptor reserve.
In the hmGlu calcium mobilization (FLIPR) assay, agonist responses (with submaximal
3
efficacies) were observed for only two of the fourteen analogs of this series (14b: EC = 105
50
nM, Emax = 49% and 14c: EC = 82 nM, Emax = 37%, Table 2). Representative concentrationꢀ
50
response curves for members of this series are shown in Figure 4d. With the exception of 14b
and 14c, none of the other compounds from this series produced an agonist response in this assay
up to the highest concentration tested (25 ꢀM); rather, many produced discernible though
generally low potency antagonist responses. This pharmacological profile is shared by
thiophenylethers 9 (IC = 7,690 nM) and 10 (IC = 6,710 nM) but is notably different when
5
0
50
2+
compared to the potent and maximally efficacious agonist response produced by 1 (mGlu Ca
3
FLIPR EC = 140 nM, Emax = 95%). The lack of robust agonist responses for compounds in
5
0
2
+
this series in the mGlu Ca FLIPR assay cannot be readily explained by receptor expression
3
levels, Gꢀprotein coupling or experimental assay considerations as (a) mGlu receptor expression
3
2+
levels in the Ca FLIPR assay (Bmax = 6.1 pmol/mg protein) are comparable to those observed in
the cells used for the mGlu cAMP assay (Bmax = 5.27 pmol/mg protein) and higher than mGlu2
3
2+
levels in the mGlu Ca FLIPR assay (B
= 1.35 pmol/mg protein) ꢀꢀ two assays in which
2
max
consistent agonist responses were observed for these molecules; and (b) compound 1 exhibits
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
full agonist responses in both mGlu and mGlu assays, regardless of format. Rather, it appears
2
3
that compounds in the current series, while binding with comparable affinity to hmGlu and
2
hmGlu receptor proteins, exhibit differential functional responses at these targets, being potent
3
and maximally (or near maximally) efficacious agonists in cells expressing mGlu receptors and
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
partial agonists (with varying levels of maximal effectiveness) in cells expressing mGlu3
2
2
receptors. This is a pharmacological profile similar to that observed for compound 5, though
with significantly enhanced mGlu receptor potency. From this series, compound 14a
2
3
0
(LY2812223), owing to its potent, maximally efficacious agonist effects in cells expressing
mGlu receptors and low potency and efficacy in mGlu ꢀexpressing cells (Figure 6), was selected
2
3
for additional investigation.
Figure 6. Agonist and antagonist concentrationꢀresponse relationship for 14a in cells
expressing recombinant human mGlu receptors.
3
a. Agonist (open squares) and antagonist b. Agonist responses elicited by 14a in hmGlu2
(
closed circles) responses elicited by 14a in (closed triangles) and hmGlu (open circles)
3
2
+
hmGlu ꢀexpressing cells employing cAMP expressing cells employing Ca
FLIPR
3
format.
format.
2
7
Recombinant mGluR selectivity assays. Compounds 14a was tested for its ability to elicit
agonist or antagonist responses in cells expressing recombinant human mGlu receptor subtypes
(Table 3). At the concentrations evaluated (25 ꢀM for agonist, 12.5 ꢀM for antagonist), no
functional responses were observed for this molecule in cells expressing recombinant human
mGlu , mGlu , mGlu or mGlu receptors. In contrast, agonist responses were observed for 14a
1
4
5
7
in cells expressing hmGlu (cAMP EC = 3,450 nM, Emax = 97%)) and hmGlu (cAMP EC =
6
50
8
50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
0,100 nM, Emax = 86%). Thus, the hmGlu agonist selectivity ratio for 14a over hmGlu
2
6
receptors is 616 fold (based on hmGlu cAMP EC ) or, more conservatively, 163 fold (based on
2
50
2
+
hmGlu Ca FLIPR EC ), while the selectivity for 14a over hmGlu receptors is 1,800 fold
2
50
8
2+
(based on hmGlu cAMP EC ) or, more conservatively, 480 fold (based on hmGlu Ca FLIPR
2 50 2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EC ). In addition to its selectivity over these closely related targets, compound 14a was
50
evaluated against a panel of CNS receptor, transporter and ion channel targets, and against the
hERG channel at a screening concentration of 10 ꢀM. No significant effect of this molecule was
observed at any of these additional targets (supporting information, Table S2).
Table 3. Functional agonist or antagonist responses of 14a in cells expressing
recombinant human mGlu receptor subtypes: EC (nM) ± SEM (Emax) or IC (nM) ±
5
0
50
a
SEM (Imax)
mGlu1
mGlu4
mGlu5
mGlu6
cAMP
mGlu7
mGlu8
cAMP
2
+
2+
2+
2+
2+
Ca
FLIPR
Ca
Ca
Ca
Ca
FLIPR
FLIPR
FLIPR
FLIPR
>
25000 >25000 >25000 3450 ± 165
12500 >12500 >12500 (97%)
See experimental section for methods
>25000
10100 ± 2150
(86%)
>25000
>
a
>12500
>12500
Molecular Modeling Studies. Previous work by our group and others has provided a detailed
2
0a,25
20b
understanding of the binding interactions of orthosteric agonists
and antagonists to the
glutamate binding site of mGlu and mGlu receptors and the associated impact of ligand
2
3
structure on the overall topology of the ligandꢀbound ATDs. Specifically, orthosteric agonists
and antagonists, while binding to common hinge residues (see Table 1), stabilize distinct ATD
conformations (closed for agonists, open for antagonists). Thus, with the established mGlu2ꢀ
agonist, mGlu partial agonist/antagonist pharmacology of 14a in mind, we performed modeling
3
experiments with this molecule to the relevant crystal structures (antagonist form of hmGlu3,
agonist form of hmGlu ) in an attempt to rationalize this unique pharmacology. Compound 14a
2
31
was built in MOE directly from and superimposed onto antagonist ligand 17 (LY341495,
3
2
20b
Figure S5) in the mGlu ATD structure (3SM9) and from agonist ligand 1 in the recently
3
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
25
disclosed mGlu ATD structure (4XAQ). In each case, the newly built 14a was minimized with
2
the MMFF94x force field in MOE. No additional minimization of ligand with protein was
performed to ensure no movement of ligand or residues from the original crystal structure
coordinates. As can be seen in Figure 7a, modeling 14a into the published (3SM9) antagonistꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
bound crystal structure of hmGlu led to a proposed complex in which key contacts of the
3
glutamate backbone of 14a with lobe 1 (LB1) residues associated with the hinge domain could
be readily established with no apparent steric clashes between ligand and protein. In contrast,
attempts to dock 14a into the closed, agonistꢀstabilized form of the mGlu ATD derived from its
2
2
5
complex with 1 resulted in predicted severe steric clashes between the C4 ꢀthiotriazolyl
α
functionality and two residues, Y144 (LB1) and R271, lobe 2 (LB2) that are situated close to the
C4ꢀposition of the bicyclic amino acid ring system (Figure 7b). We have previously ascribed
this possible negative steric interaction as the underlying reason why relatively large (e.g. NHAc,
SPh) substituents attached to this position lack agonist activity while smaller ones (e.g. F, OH,
2
4
N , NH ) retain this pharmacology.
3
2
Figure 7. Modeling of 14a into the antagonist bound crystal structure of hmGlu (a) and the
3
agonist bound crystal structure of hmGlu (b).
2
a. 14a modeled into the open (antagonist b. 14a modeled into the closed form of the
25
derived) form of the hmGlu ATD (PDB: hmGlu ATD (PDB: 4XAQ). Predicted
3
2
2
0b
3SM9).
14a binding to this form of the steric clashes (depicted as orange cylinders)
hmGlu ATD appears to be feasible as no between this structure and Y144 and R271
3
close contacts between the thiotriazole suggest that this molecule would not
substituent and the mGlu2 protein are effectively bind to and stabilize the closed
observed.
(agonist) state of the hmGlu ATD.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Co-crystallization studies with recombinant human mGlu and mGlu amino terminal
2
3
domain proteins. Clearly, the predicted binding modality of 14a to the agonist recognition site
of hmGlu ATD (Figure 7b) is incompatible with the observed potent agonist activity for this
2
molecule in cells expressing these receptors. In an effort to understand the molecular basis for
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the unique pharmacological profile exhibited by 14a, coꢀcrystallization of this compound with
2
7
the ATDs of both hmGlu and hmGlu was performed.
In the case of hmGlu , a dimeric
2
2
3
protein was observed (supporting information, Figure S5). The following analysis pertains to
protomer A for hmGlu (as depicted in Figure S5) which possessed the highest quality electron
2
density data. The structures of 14a bound to the hmGlu (2.8 Å resolution) and hmGlu (2.9 Å
2
3
resolution) ATDs are depicted in Figure 8. Inspection of the overall topology of these complexes
revealed a more closed state of the mGlu ꢀ14a ATD complex (open angle = 27.5°, Figure 8a)
2
3
3
compared to the hmGlu ꢀ14a bound structure (open angle = 45.1°, Figure 8b). In each case, the
3
αꢀamino acid functionality of 14a is situated in the hinge region of the protein and involved in
Hꢀbond interactions with LB1ꢀassociated residues S165, A166 and T168 (hmGlu numbering)
2
while the distal (C6)ꢀcarboxylic acid of 14a shows close interactions with LB1 residues R61 and
K377 in hmGlu (Figures 8c) and R68 in hmGlu (Figure 8d). These interactions are identical to
2
3
those we observed for coꢀcrystals of glutamate (supporting information Figures S7, S8) and other
2
5
compounds from the 2ꢀaminobicyclo[3.1.0]hexaneꢀ2,6ꢀdicarboxylate scaffold bound to the
hmGlu and hmGlu ATDs. In contrast, while the primary ammonium functionality of 14a
2
3
appears to form a direct Hꢀbond interaction with lower lobe (LB2) D295 in the mGlu complex
2
(Figure 8c), this functionality is remote (5.5 Å) from D301 in the mGlu ꢀ14a structure, though
3
possibly interacting remotely with the D301 side chain carboxylate via an intermediary water
molecule (Figure 8d). In addition to these glutamate pharmacophore interactions, the C4αꢀ
thiotriazole group forms an Hꢀbond interaction with LB2 residue E273 and a πꢀH interaction
with backbone NH of S272 in the mGlu structure (Figure 8e) while in the mGlu complex, the
2
3
thiotriazole appears to extend into the open, solvent accessible, cleft making LB1 interactions
with Y150 and R64 and Hꢀbonding through an intermediary water molecule and chloride anion
to LB2 residues D221, Y222 and R277 (Figure 8f).
An overlay of hmGlu ATDꢀ14a and hmGlu ATDꢀ14a structures (protomer A in each case) was
2
3
generated by aligning the atoms comprising the backbone chain (carbonyl carbon, C and N) of
α
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
34
each residue in the two crystal structures and is shown in Figures 9a and 9b. Pronounced
movement of LB2 relative to LB1 in the hmGlu ATDꢀ14a complex (purple) relative to that in
2
the hmGlu structure (gold, Figure 9a) is evident.
Examined in closer detail, distinct
3
conformational states of both the C4 ꢀthiotriazole of 14a and a LB1 tyrosine residue (Y144/150)
α
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
in the hmGlu and hmGlu structures are apparent (Figure 9b). Comparing the 14a-hmGlu2
2
3
25
ATD structure with the analogous 1ꢀhmGlu ATD complex 4XAQ (Figure 9c) it appears that a
2
nearly 90° rotation of the C ꢀC bond of Y144 (Figure 10a) from its position in the previously
α
β
2
5
described structure with 1 occurred. This provides an expanded agonist binding pocket
adjacent to the C4ꢀposition of 1 and enables the C4 ꢀthiotriazole ring of 14a to both avoid steric
α
clashes predicted from previous agonistꢀbound structures (Figure 7b) and engage in productive
LB2 interactions, specifically with residues S272 and E273. It is noteworthy that the latter
residue is not conserved in the mGlu receptor, where it is instead a oneꢀcarbon shortened acidic
3
amino acid, D279. We therefore propose that the ability of 14a to stabilize the closed form of
mGlu and produce an agonist response is due to energetically productive interactions with S272
2
and E273, and that this cannot be effectively achieved with mGlu owing to the inability of the
3
thiotriazole functionality of 14a to reach and form stabilizing LB2 Hꢀbond contacts with D279.
Interestingly, in protomer B (see supporting information Figure S6) of the 14aꢀmGlu crystal
2
structure similar stabilizing interactions between compound 14a thiotriazole, S272 and E273 are
evident, but the movement of Y144 is far less pronounced than in protomer A, and the overall
position of this residue is comparable to that found in the crystal structure of 1 with hmGlu2
(
Figure 9d). Specifically, in protomer B, no apparent rotation about the C ꢀC bond has
α β
occurred. Rather, a slight (20°) rotation about the C ꢀC bond (Figure 10b) compared to that
β Ar
2
5
present in previously described structure with 1 is evident. This rather subtle movement is
sufficient to both minimize steric clashes with the bound ligand and allow the C4 ꢀthiotriazole
α
substituent access to the LB2 interacting residues S272 and E273, perhaps owing to this structure
exhibiting a slightly more open angle of the binding cleft (28.6° vs. 27.5° found in protomer A).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 8. Structure of 14a bound to the amino terminal domains (protomer A) of human mGlu2
(a, c) (residues 2ꢀ493 with C121S / C234S mutation and CꢀHis tag, PDB ID: 5CNJ) and human
2
7
mGlu (b, d) (residues 2ꢀ507 with C240S mutation and CꢀHis tag, PDB ID: 5CNM).
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a. 2.8 Å hmGlu ATD ꢀ14a crystal structure.
b. 2.9 Å hmGlu ATDꢀ14a crystal structure.
2
3
The proximity of both the upper (LB1) and
lower (LB2) domains to bound ligand results
in an open angle of approximately 28°.
While close association of the bound ligand
with the upper (LB1) domain is evident, no
direct interaction with the lower (LB2) domain
with ligand is apparent, leading to an open
33
3
3
angle of approximately 45°.
c. Site view of hmGlu ATDꢀ14a complex.
d. Site view of hmGlu ATDꢀ14a complex. The
2
3
Note direct interaction between C2ꢀ
ammonium functionality in 14a with side
chain carboxylate of LB2 residue D295.
C2 ammonium functionality of 14a is distant
from LB2 residue D301 and interacts with a
water molecule positioned between the ligand
and D301 (not shown).
ACS Paragon Plus Environment

Page 23 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
e. Site view of the C4 ꢀthiotriazole
f. Site view of the C4 ꢀthiotriazole interactions
α
α
in the hmGlu ATDꢀ14a complex. No direct
interactions in the hmGlu ATDꢀ14a complex.
3
2
interactions with LB2 are apparent. Indirect
interaction of the thiotriazole with LB2 through
a water molecule and chloride ion can be
observed. Indirect interactions with LB1
residue R64 and LB2 R277 may also be
occurring, though no specific water molecule
densities could be resolved in this region.
An Hꢀπ interaction (cyan dashed line) with
LB2 residue S272 and an Hꢀbond with LB2
residue E273 is observed, which are proposed
to provide additional stabilization of the
closed ATD structure.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 9. Overlay of protomer A crystal structures of hmGlu ꢀ14a and hmGlu ꢀ14a
2
3
a. Overlay of hmGlu ATDꢀ14a (purple) and
b. Detail in overlay of hmGlu ATDꢀ14a (purple
2
2
hmGlu ATDꢀ14a (gold) structures. Fixed
protein and ligand) and hmGlu ATDꢀ14a (gold
3
3
alignment of upper lobe (LB1) resides in the
two proteins highlights the relative movement
protein and ligand) structures. Note the distinct
conformations of thiotriazole substituents and
tyrosine residues in these structures.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of the hmGlu lower lobe (LB2) domain
2
relative to that for hmGlu3.
c. Overlay hmGlu ꢀ14a protomer A crystal
d. Overlay hmGlu ꢀ14a protomer B crystal
2
2
structure (purple) with hmGlu ꢀ1 crystal
structure (4XAQ, cyan). Note the nearly 90°
structure (orange) with hmGlu ꢀ1 crystal
structure (4XAQ, cyan). Note that in this
2
2
2
5
25
rotation of Y144 from its position in structure 1. protomer, Y144 adopts a similar overall
This movement opens a large cavity into which conformation to that found in the mGlu
ꢀ1
2
the triazole ring of 14a projects, allowing
structure.
interaction with LB2 residues including E273.
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Figure 10. Y144 rotations observed in the individual hmGlu ATDꢀ14a protomer
2
crystal structures relative to the position of this residue in the hmGlu ꢀ1 crystal
2
structure (PDB: 4XAQ).
a. Protomer A. 90
Y144 CαꢀCβ bond in the hmGlu ꢀ14a
structure (cyan) relative to the
°
rotation of the
b. Protomer B. 20
Ar bond in the hmGlu
relative to the hmGlu ꢀ1 structure (purple).
°
rotation of the Y144 Cβꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C
ꢀ14a structure (cyan)
2
2
2
hmGlu ꢀ1 structure (purple).
2
Site-Directed Mutants of mGlu and mGlu . Given the pharmacological selectivity for 14a for
2
3
activation of mGlu vs. mGlu receptors and the observed close proximity of the triazole ring to
2
3
mGlu residues (E273 and L300) that are different in mGlu (D279 and Q306), we constructed
2
3
cells transiently transfected with hmGlu wild type (WT), E273D and L300Q receptors as well as
2
cells expressing hmGlu WT, D279E and Q306L receptors. In these experiments, we compared
3
the agonist responses of 14a to glutamate (Figure 11 and Table 4) and to 18 (LY379268, Figure
3
5
S5), a previously reported potent mGlu2/3 receptor agonist (data included in the supporting
information section, Figure S9). Neither single (E273D, L300Q) or double (E273D + L300Q)
hmGlu mutants had any impact on the potency elicited by glutamate when compared to WT
2
25
mGlu expressing cells (Figure 11a, Table 4). In contrast, the response to 14a appears to be
2
affected by these point mutations (Figure 11b), with agonist potency, though not maximal
efficacy (100% defined as the maximal response obtained with glutamate), being slightly rightꢀ
shifted in cells expressing mGlu E273D (EC = 46.3 nM) and E273D + L300Q (EC = 48.2
2
50
50
nM) mutants compared to mGlu WT receptors (EC = 20.1 nM), while maximal efficacy, but
2
50
not potency, was modestly altered in the L300Q mutant. Neither potency nor efficacy of 18 was
altered by either single or double mutations (Figure S4).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In comparison, cells expressing mGlu receptor mutants D279E and D279E + Q306L, (but not
3
Q306L alone) negatively impacted the potency of glutamate compared to cells expressing WT
mGlu receptors (EC values = 1ꢀM, 4.2 ꢀM and 5.1 ꢀM, for WT, D279E and D279 + Q306L,
3
50
2
5
respectively; Table 5, Figure 11c). In sharp contrast, we observed a clear beneficial effect of
mGlu mutations on both agonist potency (EC values = 5.7 ꢀM, 2.2 ꢀM, 2.8 ꢀM and 1.2 ꢀM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
50
for WT, D279E, Q306L and D279 + Q306L, respectively) and maximal efficacy (Emax values
20%, 47%, 41% and 66% for WT, D279E, Q306L and D279 + Q306L, respectively) of 14a
=
(Table 5, Figure 11d). These effects are also distinct from those observed for 5, another
functionally mGlu ꢀselective agonist which appeared to gain agonist function in cells expressing
2
the hmGlu (Q306L) mutation, but not in the other hmGlu mutant cell lines (D279E and
3
3
25
combined Q306L + D279E).
As in the case of mGlu , neither potency nor efficacy of 18 was
2
altered by either single or double mutations of mGlu (Figure S9).
3
The observed modest loss of agonist potency of 14a in mGlu cells expressing the E273D
2
mutant, combined with loss of agonist maximal efficacy in the mGlu L300Q mutant, taken
2
together with the reciprocal gain of agonist potency and efficacy in the mGlu mutants (D279E,
3
Q306L, D279E + Q306L) is in agreement with the hypothesis that these amino acids are
involved in the pharmacological profile exhibited by 14a in cells expressing WT receptors.
However, it is also clear that these amino acids alone do not fully explain the observed mGlu2
receptor agonist, mGlu receptor antagonist pharmacological profile of this molecule.
A
3
limitation of this investigation should be noted. Because transient transfections result in a
population of cells expressing varying levels of receptor, it was not possible to determine how, if
at all, differing receptor expression levels in each transient cell line may have impacted ligand
potency or efficacy. However, the fact that the potency of glutamate was unaffected in any of
the mGlu mutants vs. WT, and the potency of glutamate was only modestly (and negatively)
2
affected in the mGlu mutant (D279E) which led to enhance functional agonist activity for 14a
3
provides some evidence that any differences in receptor expression levels were not the cause of
the differential effects of 14a in these experiments. This is further supported by the lack of
mutation effects on either the potency or efficacy of 18 (supporting information, Figure S9).
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2+
Table 4. Effect of Lꢀglutamate (LꢀGlu) and 14a on second messenger (Ca , FLIPR)
responses in cells transiently transfected with wild type (WT) and mutant human mGlu2
a,b
receptors.
WT
E273D
L300Q
E273D + L300Q
EC ± Emax ± EC ± Emax ± EC ± Emax ±
EC ±
Emax ±
SD
50
50
50
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SD
SD
SD
SD
SD
SD
SD
(nM)
(%)
(nM)
(%)
(nM)
(%)
(nM)
(%)
1827
2351 ±
689
1767 ±
518
2575 ±
763
L-Glu
94 ± 2
91 ± 4
93 ± 6
88 ± 6
±
378
20.1 ±
46.3 ±
22.1 ±
3.7
48.2 ±
6.2**
1
4a
2.3
108 ± 5
107 ± 4
100 ± 3*
105 ± 5
8.2**
a
b
See experimental section for details; Two way ANOVA followed by Dunnett’s test
performed in statistical analysis; *p<0.05 vs. mGlu WT; **p<0.0001 vs mGlu WT
2
2
2
+
Table 5. Effect of Lꢀglutamate (LꢀGlu) and 14a on second messenger (Ca , FLIPR)
responses in cells transiently transfected with wild type (WT) and mutant human mGlu3
a,b
receptors.
WT
D279E
Q306L
D279E + Q306L
EC ± Emax EC ± Emax ±
EC ±
Emax ± EC ±
SD
(%)
Emax ±
SD
(%)
50
50
50
50
SD
± SD
(%)
SD
(nM)
SD
(%)
SD
(nM)
SD
(nM)
(nM)
1002 ± 111 ± 4182 ±
1484 ±
271
5118 ±
L-Glu
**
104 ± 3
105 ± 2
**
104 ± 2*
1
20
2
997
1030
5
2
676 ±
080
2181 ±
208**
2831 ±
1175 ±
253**
1
4a
20 ± 2
47 ± 2**
#
41 ± 2**
66 ± 1**
310
a
b
See experimental section for details; Two way ANOVA followed by Dunnett’s test used in
statistical analysis; p<0.05 vs mGlu WT; p = 0.0006 vs. mGlu WT; p<0.0001 vs mGlu
*
#
**
3
3
3
WT
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 11. Effect of single and double amino acid mutations of mGlu and mGlu receptors
on functional activity of glutamate and 14a.
2
3
a. Agonist responses produced by Lꢀglutamate b.
Agonist responses produced by
in hmGlu wild type and mutant receptors compound 14a in hmGlu wild type and
2
2
a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
transiently expressed in AV12 cells.
mutant receptors transiently expressed in
a
AV12 cells.
c. Agonist responses produced by Lꢀglutamate d.
Agonist responses produced by
in hmGlu wild type and mutant receptors compound 14a in hmGlu wild type and
3
3
a
transiently expressed in AV12 cells.
mutant receptors transiently expressed in
a,b
AV12 cells.
a
b
See experimental section for details. Note that the ordinate axis is expanded to show detail
of these concentration response curves.
Pharmacokinetic Attributes of 14a. Owing to its remarkable mGlu vs. mGlu receptor
2
3
agonist attributes and selectivity profile, 14a was chosen for further assessment. In vitro
evaluation of 14a (Table 6) in assays that are informative regarding in vivo pharmacokinetics
revealed this molecule to possess high aqueous solubility (> 100 ꢀM at pH 7.4), no measurable
microsomal metabolism across species, no significant inhibition of cytochrome P450 enzyme
isoforms and high (> 98%) plasma unbound fraction.
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 6. In vitro solubility, permeability, metabolism and protein binding of 14a
a
Assay
Result
Aqueous Solubility (pH 7.4)
DMSO Solubility
> 100 ꢀM
> 10 mM
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
c
%
Loss Hepatocytes (mouse, rat, dog, human)
< 10% for each
< 10% for each
%
Inhibition Human Cyp Isoforms
d
(3A4, 2D6, 2C9)
Plasma Unbound Fraction (rat, human)
> 98%
a
b
c
See experimental section for methods; tested at 20 ꢀM; tested at 2 ꢀM;
d
tested at 10 ꢀM
The in vivo pharmacokinetic attributes of 14a were assessed in the rat, and plasma
pharmacokinetic parameters are provided in Table 7. Intravenous (iv) administration of 14a (1
mg/kg) resulted in mean peak plasma levels of approximately 9.5 ꢀM. The rate of clearance
from plasma was low (12.3 mL/min/kg) as was volume of distribution (1 L). The plasma halfꢀ
life for 14a following the 1 mg/kg iv dose was determined to be 3.3 h. Oral (po) administration
of 14a at 5 mg/kg led to relatively low plasma levels (Cmax = 0.22 ꢀM), and an estimated oral
bioavailability of approximately 4%. In contrast, intraperitoneal (ip) administration of 14a led to
doseꢀrelated increases in plasma Cmax and AUC for this compound over the range of 1ꢀ10
mg/kg with mean peak plasma concentrations being observed within 30 min of dosing and
ranging from 5 ꢀM (1 mg/kg) to 40 ꢀM (10 mg/kg). A comparison of exposures following ip
and iv dosing suggests high apparent bioavailability (73ꢀ100%) when 14a is dosed by this route.
ACS Paragon Plus Environment