Discovery of N-(5-Fluoropyridin-2-yl)-6-methyl-4-(pyrimidin-5-yloxy)picolinamide (VU0424238): A Novel Negative Allosteric Modulator of Metabotropic Glutamate Receptor Subtype 5 Selected for Clinical Evaluation

Article abstract of DOI:10.1021/acs.jmedchem.7b00410

Preclinical evidence in support of the potential utility of mGlu₅ NAMs for the treatment of a variety of psychiatric and neurodegenerative disorders is extensive, and multiple such molecules have entered clinical trials. Despite some promising results from clinical studies, no small molecule mGlu₅ NAM has yet to reach market. Here we present the discovery and evaluation of N-(5-fluoropyridin-2-yl)-6-methyl-4-(pyrimidin-5-yloxy)picolinamide (27, VU0424238), a compound selected for clinical evaluation. Compound 27 is more than 900-fold selective for mGlu₅ versus the other mGlu receptors, and binding studies established a K_i value of 4.4 nM at a known allosteric binding site. Compound 27 had a clearance of 19.3 and 15.5 mL/min/kg in rats and cynomolgus monkeys, respectively. Imaging studies using a known mGlu₅ PET ligand demonstrated 50% receptor occupancy at an oral dose of 0.8 mg/kg in rats and an intravenous dose of 0.06 mg/kg in baboons.

Full text of DOI:10.1021/acs.jmedchem.7b00410

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Discovery of N-(5-fluoropyridin-2-yl)-6-methyl-4-(pyrimidin-5-
yloxy)picolinamide (VU0424238): A novel negative allosteric modulator of
metabotropic glutamate receptor subtype 5 selected for clinical evaluation
Andrew S Felts, Alice L Rodriguez, Anna L Blobaum, Ryan D. Morrison, Brittney S Bates,
Analisa Thompson Gray, Jerri M. Rook, Mohammed N. Tantawy, Frank W Byers, Sichen
Chang, Daryl F Venable, Vincent B Luscombe, Gilles D Tamagnan, Colleen M Niswender,
J. Scott Daniels, Carrie K. Jones, P. Jeffrey Conn, Craig W Lindsley, and Kyle A Emmitte
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 22 May 2017
Downloaded from http://pubs.acs.org on May 23, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Nꢀ(5ꢀfluoropyridinꢀ2ꢀyl)ꢀ6ꢀmethylꢀ4ꢀ
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
(
pyrimidinꢀ5ꢀyloxy)picolinamide (VU0424238): A
novel negative allosteric modulator of metabotropic
glutamate receptor subtype 5 selected for clinical
evaluation
†
†
†
†
Andrew S. Felts , Alice L. Rodriguez , Anna L. Blobaum , Ryan D. Morrison , Brittney S.
†
†
†
±
†
Bates , Analisa Thompson Gray , Jerri M. Rook , Mohammed N. Tantawy , Frank W. Byers ,
†
†
†
‡
Sichen Chang , Daryl F. Venable , Vincent B. Luscombe , Gilles D. Tamagnan , Colleen M.
†
†
†
†
†,§
Niswender , J. Scott Daniels , Carrie K. Jones , P. Jeffrey Conn , Craig W. Lindsley and Kyle
,
†,§,#
A. Emmitte*
†
Vanderbilt Center for Neuroscience Drug Discovery, Department of Pharmacology, Vanderbilt
University, Nashville, Tennessee 37232
§
±
Department of Chemistry, Vanderbilt University, Nashville, TN 37232
Department of Radiology and Radiological Sciences, Vanderbilt University Institute of
Imaging Science, Vanderbilt University Medical Center, Nashville, TN 37232
Molecular NeuroImaging, a division of inviCRO, New Haven, CT, 06510
‡
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
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
KEYWORDS. Metabotropic glutamate receptor subtype 5 (mGlu ), negative allosteric modulator
5
(NAM), major depressive disorder (MDD), treatmentꢀresistant depression (TRD), anxiety,
obsessiveꢀcompulsive disorder (OCD), picolinamide
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
ABSTRACT.
Preclinical evidence in support of the potential utility of mGlu NAMs for the treatment of a
5
variety of psychiatric and neurodegenerative disorders is extensive, and multiple such molecules
have entered clinical trials. Despite some promising results from clinical studies, no small
molecule mGlu NAM has yet to reach market. Here we present the discovery and evaluation of
5
Nꢀ(5ꢀfluoropyridinꢀ2ꢀyl)ꢀ6ꢀmethylꢀ4ꢀ(pyrimidinꢀ5ꢀyloxy)picolinamide (27, VU0424238),
a
compound selected for clinical evaluation. Compound 27 is more than 900ꢀfold selective for
mGlu versus the other mGlu receptors, and binding studies established a K value of 4.4 nM at a
5
i
known allosteric binding site. Compound 27 had a clearance of 19.3 and 15.5 mL/min/kg in rats
and cynomolgus monkeys, respectively. Imaging studies using a known mGlu PET ligand
5
demonstrated 50% receptor occupancy at an oral dose of 0.8 mg/kg in rats and an intravenous
dose of 0.06 mg/kg in baboons.
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
The metabotropic glutamate (mGlu) receptors are an eightꢀmembered family of class C G
proteinꢀcoupled receptors (GPCRs) that are activated by Lꢀglutamic acid, the major excitatory
neurotransmitter of the mammalian central nervous system (CNS). Further stratification of the
mGlu receptors is made based on a variety of considerations, including structure, downstream
signaling partners, and pharmacology. Whereas group I mGlu receptors (mGlu and mGlu ) are
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
5
found in postꢀsynaptic locations and couple via G to the activation of phospholipase C (PLC),
q
both group II (mGlu ) and group III (mGlu and mGlu ) receptors are found mainly in preꢀ
2
ꢀ3
4
6ꢀ8
synaptic locations and couple via G /G to the inhibition of adenylyl cylase activity. Common to
i
o
all mGlu receptors are a seven transmembrane (7TM) αꢀhelical domain that connects to a large
biꢀlobed extracellular aminoꢀterminal domain, sometimes termed the “venus fly trap (VFT)”
domain. Located within this VFT domain is the binding site for the native ligand; however,
allosteric binding sites for small molecules have thus far been confined to the transmembrane
domain. The structural and functional properties of the mGlu receptors have been reviewed
1,2
extensively.
In the search for compounds that are sufficiently selective for individual mGlu receptor
subtypes, we have found allosteric modulation to be a successful approach. In fact, in many
cases, such a strategy can be employed for the design of both positive allosteric modulators
3
(PAMs) and negative allosteric modulators (NAMs) of the same receptor. Among the most
4,5
actively investigated and advanced areas within this field is that of mGlu NAMs. Potential
5
6
therapeutic applications for small molecule mGlu NAMs are extensive and include addiction,
5
7
8
9
Alzheimer’s disease (AD), anxiety, autism spectrum disorder (ASD), fragile X syndrome
10
(
FXS), the levodopaꢀinduced dyskinesia (LID) experienced by many patients with Parkinson’s
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
11,12
13
disease (PD),
major depressive disorder (MDD), and obsessiveꢀcompulsive disorder
1
4
(
OCD). Much of the preclinical work that has served to establish mGlu antagonism as an
5
exciting new mechanism for targeting these various conditions has been conducted with two
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
15
related tool compounds, methylꢀ6ꢀ(phenylethynyl)ꢀpyridine (1, MPEP) and 3ꢀ[(2ꢀmethylꢀ1,3ꢀ
1
6
thiazolꢀ4ꢀyl)ethynyl]ꢀpyridine (2, MTEP) (Fig. 1). Both compounds are structurally similar
disubstituted alkynes, a motif that has been exploited for the design of several other highly
4
,5
optimized mGlu NAM compounds. While results from clinical studies have been reported
5
1
7,18
19,20
with nonꢀalkyne based mGlu NAMs such as 3 (fenobam),
4 (AZD9272),
and 5
5
19,20
(AZD2066),
the most highly studied and advanced clinical compounds are three alkyne
1
4,21ꢀ25
26ꢀ29
30,31
mGlu NAMs: 6 (mavoglurant),
7 (basimglurant),
and 8 (dipraglurant)
(Fig. 1).
5
1
4,22,23
28
Still, failures to reach primary clinical end points with 6
and 7 point to the continued need
for highly optimized mGlu NAMs. It should be noted that a recent report from Pfizer has
5
described cutaneous toxicities in cynomolgus monkey toxicology studies with three structurally
3
2
related nonꢀalkyne based mGlu NAMs. Given that other mGlu NAMs that are not structurally
5
5
related to the Pfizer compounds have been used in longꢀterm, repeat dose studies in nonꢀhuman
3
3,34
14,22ꢀ24,28
primates (NHPs)
and humans
without notes of a similar significant toxicity, it seems
reasonable that such effects are not endemic to the mechanism. Still, these results make a
compelling argument for using cynomolgus monkeys as the nonꢀrodent species for toxicology
studies in the event that our conclusion is incorrect.
Alkynes are potentially reactive functional groups. Pfizer has reported an instance of biliary
epithelial hyperplasia in a NHP toxicology study with an alkyneꢀbased mGlu NAM developed
5
3
5
by Wyeth. This toxicity was believed to be linked to the reactivity of the alkyne based on
35
metabolic studies that revealed extensive glutathione conjugation. While this type of
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
25,27
biotransformation does not always manifest in humans with alkyneꢀcontaining compounds,
many research groups have sought to design novel mGlu NAMs that are outside of this
5
4
,5
chemotype. We have actively pursued a variety of approaches toward that end over the last
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
36,37
cade.
For example, we conducted a highꢀthroughput screen (HTS) of a collection of 160,000
compounds that identified multiple distinct mGlu NAM hits. This same platform functions as
5
our primary assay for lead optimization by measuring the ability of novel compounds to block
the mobilization of calcium induced by an EC80 concentration of glutamate in HEK293A cells
that express rat mGlu5.
Figure 1. Prototypical mGlu NAM preclinical tools 1 and 2 and clinically investigated mGlu5
5
NAMs 3ꢀ8.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Initial analogs. Recently, we described the discovery of an in vivo tool compound 9
3
7ꢀ40
(VU0409106)
that resulted from the optimization of one of these HTS hits (Table 1). That
3
7
report focused exclusively on compounds with a benzamide core; however, we were also
interested in exploring picolinamide analogs as well (10ꢀ13, Table 1). Moving from the 3ꢀ
fluorophenyl benzamide core (9) to an unsubstituted picolinamide core (10) reduced potency by
approximately 8ꢀfold; however, substitution of the picolinamide core at the 6ꢀposition with a
methyl group (11) improved potency. In addition, structureꢀactivity relationships (SAR)
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
7
established in the benzamide series translated to the picolinamides in the form of enhanced
potency with the 5ꢀfluoropyridinꢀ3ꢀyl ether (12). Once again, 6ꢀmethyl substitution of the
picolinamide core (13) improved potency relative to the unsubstituted analog 12. Buoyed by the
encouraging results from this limited set of analogs and in search of a compound that represented
a step forward from 9 to a compound worthy of advancement to clinical development, we
devised a plan to prepare additional 6ꢀalkylꢀpicolinamide analogs. Herein, we report the
successful execution of that strategy.
Table 1. Initial picolinamide core analogs of 9
mGlu
(± SEM)
5
pIC50
a
a,b
No.
9
a
mGlu
5
IC50 (nM)
% Glu Max
7.62 ± 0.03
6.70 ± 0.15
7.01 ± 0.09
7.22 ± 0.23
8.01 ± 0.16
24
200
98
1.2 ± 0.1
1.2 ± 0.1
1.3 ± 0.1
1.4 ± 0.1
1.3 ± 0.1
1
1
0
1
12
60
1
3
10
a
Calcium mobilization mGlu
5
assay; values are average of n ≥ 3 independent experiments
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
b
Amplitude of response in the presence of 30 µM test compound as a percentage of
maximal response (100 µM glutamate); average of n ≥ 3 independent experiments
Synthesis of Compounds. The synthetic routes to the picolinamide analogs largely mirror those
3
7
described for the benzamide mGlu NAMs. Briefly, 4ꢀnitropyridineꢀNꢀoxides 14 were refluxed
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
in concentrated hydrochloric acid to afford 4ꢀchloropyridineꢀNꢀoxides 16 (Scheme 1).
Alternatively, 4ꢀchloropyridines 15 were oxidized with hydrogen peroxideꢀurea adduct in the
presence of trifluoroacetic acid to provide 16. Treatment of 16 with trimethylsilyl cyanide and
dimethylcarbamyl chloride generated 2ꢀcyanopyridines 17. A nucleophilic aromatic substitution
reaction between heteroaryl alcohols 18 and 17 was employed to generate ether products 19.
Hydrolysis of the nitrile moieties of 19 was accomplished via thermal heating with aqueous
sodium hydroxide solution to give the penultimate intermediate acids 20. Coupling with the
4
appropriate primary heteroaryl amines (H NR ) generated the target analogs.
2
a
Scheme 1. Synthesis of picolinamide analogs 10–13 and 21–37
a
1
2
4
Reagents and conditions: R = H, Me, CHF ; R = H, Me, C H ; A = N, CH, CF, CCl; R = 4ꢀ
2
3
5
methylthiazolꢀ2ꢀyl or substituted pyridinꢀ2ꢀyl; (a) con. HCl, reflux, 53ꢀ75%; (b) H O ·urea,
2
2
(
F CCO) O, THF, 0 ºC to r.t., 44ꢀ99%; (c) Me SiCN, CH Cl , Me NCOCl, 75ꢀ86%; (d) 18,
3 2 3 2 2 2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
4
K CO , DMF, 80 ºC, 45ꢀ77%; (e) 2N aq. NaOH, dioxane, reflux, 82ꢀ99%; (f) H NR , POCl ,
2
3
2
3
pyridine, ꢀ15 ºC, 22ꢀ81%.
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
Initial SAR. Even though compound 9 is a useful tool for acute behavioral studies in mice and
3
7,39
rats,
it has several properties that prevent its further development. In particular, 9 has
moderate to high clearance in rats (Cl 43 mL/min/kg) and cynomolgus monkeys (Cl 24
p
p
4
0
mL/min/kg). In addition, 9 has poor bioavailability following oral administration in rats (F <
5%), and is an inhibitor of CYP1A2 (IC < 100 nM). Evaluation of the picolinamide analogs of
5
0
9
indicated that the 4ꢀmethylthiazolꢀ2ꢀyl amide contributed substantially to the potent inhibition
of CYP1A2 (e.g., 10 CYP1A2 IC50 < 100 nM), and SAR for mGlu activity within the
5
benzamide series showed limited tolerance for even subtle modifications on the thiazolꢀ2ꢀyl
37
moiety. Fortunately, analogs with a pyridinꢀ2ꢀyl replacement for the thiazolꢀ2ꢀyl ring appeared
more amenable to substitution, and initial work in the picolinamide series was carried out in that
context while taking advantage of SAR knowledge gained from the benzamide work (Table 2).
Table 2. Substituted 2ꢀpyridyl amine analogs
c
d
Protein Binding (f
u
)
P450 Inhibition IC50 (µM)
mGlu
5
pIC50
mGlu
IC50 (nM)
5
% Glu
Max
No.
Series
R
a
a
a,b
(± SEM)
rat
human
3A4
>30
>30
>30
2D6
>30
>30
>30
2C9
4.1
1.3
1.8
1A2
2.4
0.3
0.9
2
2
1
2
A
B
A
B
A
B
6ꢀMe
6ꢀMe
6ꢀEt
6ꢀEt
6ꢀF
7.70 ± 0.04
8.04 ± 0.12
8.03 ± 0.07
7.49 ± 0.31
7.94 ± 0.21
8.45 ± 0.11
20
9.1
9.3
32
1.4 ± 0.1
1.4 ± 0.1
1.5 ± 0.2
1.2 ± 0.1
1.3 ± 0.1
1.4 ± 0.1
0.076
0.028
0.027
0.010
0.066
0.016
0.034
0.005
0.132
0.023
23
2
2
2
4
5
6
e
11
ND
>30
8.0
>30
>30
1.8
1.8
0.6
0.1
e
6ꢀF
3.6
ND
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
2
2
7
8
9
A
B
A
B
A
B
5ꢀF
5ꢀF
7.97 ± 0.05
8.51 ± 0.04
8.17 ± 0.14
8.02 ± 0.16
7.27 ± 0.22
7.72 ± 0.13
11
3.1
6.8
9.6
54
1.6 ± 0.1
1.2 ± 0.1
1.3 ± 0.1
1.3 ± 0.1
1.6 ± 0.3
1.3 ± 0.2
0.106
0.015
0.067
0.120
0.011
0.049
>30
4.3
>30
>30
>30
>30
>30
>30
12.1
5.4
1.0
0.2
5ꢀCl
5ꢀCl
4ꢀF
16.4
2.0
11.0
>30
>30
5.7
0.9
30
<0.1
6.0
3
3
1
2
>30
27.4
4ꢀF
19
0.4
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
a
5
Calcium mobilization mGlu assay; values are average of n ≥ 3 independent experiments
Amplitude of response in the presence of 30 µM test compound as a percentage of maximal response (100 µM glutamate); average of n ≥ 3
independent experiments
f = fraction unbound; equilibrium dialysis assay
u
Assayed in pooled HLM in the presence of NADPH with CYPꢀspecific probe substrates
Compound was unstable in rat plasma
b
c
d
e
As expected, the majority of these pyridinꢀ2ꢀyl analogs (21ꢀ32) proved to be potent
antagonists of mGlu on par with their thiazolꢀ2ꢀyl analogs (11 and 13). In a majority of
5
instances, analogs with the 5ꢀfluoropyridinꢀ3ꢀyl ether exhibited slightly superior mGlu potency
5
compared to the pyrimidinꢀ5ꢀyl ether analogs; however, the 6ꢀethylpyridinꢀ2ꢀyl analogs 23 and
2
4 and 5ꢀchloropyridinꢀ2ꢀyl derivatives 29 and 30 were an exception to this trend. Assays that
measured the degree of binding to rat and human plasma and a human liver microsomes (HLM)
4
1
cocktail assay with probe substrates for four common P450s were used to profile several
selected compounds. SAR with regard to the heteroaryl ether group was clear as the pyrimidinꢀ5ꢀ
yl ether (Series A) analogs were consistently less bound to plasma than their 5ꢀfluoropyridinꢀ3ꢀyl
ether (Series B) analogs. No activity at 30 ꢁM test compound was noted with regard to inhibition
of CYP2D6, and inhibition of CYP3A4 and CYP2C9 ranged considerably. As anticipated, the
most consistently potent P450 inhibition was observed with CYP1A2; however, there was
another clear SAR with respect to the heteroaryl ether group. In this case, the pyrimidinꢀ5ꢀyl
ether (Series A) analogs were less potent CYP1A2 inhibitors than the 5ꢀfluoropyridinꢀ3ꢀyl ether
(Series B) analogs.
Moving our attention to other portions of the chemotype, we prepared further analogs that
revealed additional SAR. Modifications made in the context of the 5ꢀfluoropyridinꢀ2ꢀyl amide
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
found in analogs 27 and 28 are depicted here (Table 3). Installation of a methyl group at the 2ꢀ
position of the pyrimidinꢀ5ꢀyl ether (33) resulted in a 3ꢀfold loss in potency relative to 27. While
this modification increased the extent of binding to plasma, it also reduced inhibition versus
CYP1A2 and CYP2C9. Moving to a larger cyclopropyl group at the 2ꢀposition of the pyrimidinꢀ
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
5ꢀyl ether (34) substantially eroded potency at mGlu , indicating minimal tolerance for steric
5
bulk at this position. Likewise, replacement of the 6ꢀmethyl group on the picolinamide core with
a difluoromethyl group (35) led to an approximately 5ꢀfold loss in mGlu potency, illustrating
5
little tolerance for modification at that position. Replacement of the 5ꢀfluoro group of the
pyridinꢀ3ꢀyl of 28 with a 5ꢀchloro group (36) resulted in only a slight drop in potency.
Replacement of the metaꢀhalogen atoms found in 28 and 36 with the paraꢀmethyl group in 6ꢀ
methylpyridinꢀ3ꢀyl ether 37 reduced mGlu potency by 50 and 20ꢀfold, respectively.
5
Table 3. Analogs of compound 27 and 28
c
d
Protein Binding (f
u
)
P450 Inhibition IC50 (µM)
mGlu
5
pIC50
mGlu
(nM)
5
IaC50
a,b
No.
27
a
% Glu Max
(± SEM)
rat
human
3A4
>30
>30
2D6
>30
>30
2C9
12.1
>30
1A2
1.0
7.97 ± 0.05
7.48 ± 0.04
6.60 ± 0.25
7.32 ± 0.02
8.51 ± 0.04
8.14 ± 0.27
6.80 ± 0.11
11
33
1.6 ± 0.1
1.5 ± 0.2
1.5 ± 0.1
1.3 ± 0.1
1.2 ± 0.1
1.0 ± 0.1
1.2 ± 0.1
0.106
0.081
0.120
0.074
3
3
3
3
4
5
>30
251
48
28
3.1
7.2
158
0.015
0.058
0.011
0.014
4.3
3.8
>30
>30
5.4
0.2
0.2
3
3
6
7
21.5
a
5
Calcium mobilization mGlu assay; values are average of n ≥ 3 independent experiments
Amplitude of response in the presence of 30 µM test compound as a percentage of maximal response (100 µM glutamate); average of n ≥ 3
independent experiments
f = fraction unbound; equilibrium dialysis assay
u
Assayed in pooled HLM in the presence of NADPH with CYPꢀspecific probe substrates
b
c
d
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
In vivo Pharmacokinetics (PK) Studies. Previous studies in our laboratories with pyrimidinꢀ5ꢀ
3
8,40
yl ethers identified this ring as a substrate for aldehyde oxidase (AO)ꢀmediated metabolism.
Thus, given the likelihood that several of our interesting new compounds would be subject to
nonꢀP450 mediated metabolism, we decided to move directly into an in vivo setting to evaluate
the clearance of a variety of select analogs (Table 4). In these studies, the three analogs with the
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
5ꢀfluoropyridinꢀ2ꢀyl amide group (27, 28, and 33) exhibited superior profiles compared to the
other compounds that contained pyridinꢀ2ꢀyl amides with 6ꢀalkyl (21ꢀ23) or 6ꢀfluoro (26)
4
2
groups. The plasma clearance of 26 was greater than hepatic blood flow, which suggests that
the instability previously observed in rat plasma for this compound (Table 2) is not merely an in
vitro phenomenon. Blocking the 2ꢀposition of pyrimidinꢀ5ꢀyl ether (33) reduced clearance and
lengthened mean residence time relative to unsubstituted analog 27. Such a result points toward
this position as one potential site for metabolism, which is consistent with prior studies on related
3
8,40
analog 9.
Compounds 27, 28, and 33 were subsequently evaluated in tissue distribution
studies (Table 5) using oral dosing. Acceptable CNS penetration was achieved with all three
compounds, and unbound brain to unbound plasma ratios (Kp,uu) were near unity with 27 and 28,
indicating distribution equilibrium between the compartments and a low probability of these two
4
3
compounds being substrates for transporters in rats. The Kp,uu observed for 33 raises the
possibility of active transport with this compound; however, additional studies would be required
to conclusively determine if that was indeed the case. All compounds achieved unbound brain
levels in excess of their respective mGlu functional potency at the dose used in these
5
experiments (10 mg/kg). Compounds 27 and 28 were particularly attractive from this
perspective, reaching unbound brain levels of 18ꢀ and 10ꢀfold over their functional potency,
respectively.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Table 4. Preliminary rat PK results using intravenous (IV) dosing
mGlu
IC50
5
t
1/2
MRTc
Cl
p
V
SS
(L/kg)
No.
c
c
c
b
(min)
(min)
(mL/min/kg)
(nM)
d
2
1
20
105
268
404
82
39
120
80
34.2
32.5
43.7
96
1.6
4.5
6.6
11
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
e
f
2
2
9.1
9.3
3.6
11
23
d
d
d
f
2
2
2
6
7
8
157
93
390
281
229
19.3
9.2
1.8
1.9
1.2
3.1
33
208
191
33
6.2
a
Average n = 2, male SpragueꢀDawley rats; dose = 1.0 mg/kg
b
Calcium mobilization mGlu
independent experiments
5
assay; values are average of n ≥ 3
c
t
1/2 = terminal phase plasma halfꢀlife; MRT = mean residence time;
p
= plasma clearance; VSS = volume of distribution at steadyꢀstate
Cl
d
e
f
Vehicle = 10% EtOH, 70% PEG 400, 20% saline
Vehicle = 10% EtOH, 60% PEG 400, 30% saline
Vehicle = 10% EtOH, 50% PEG 400, 40% saline
a
Table 5. Tissue distribution studies using oral dosing
c
Total
plasma
Total
brain
(nM)
Unbound
plasma
(nM)
Unbound
brain
(nM)
Unbound
brain /
mGlu IC50
5
u
Protein Binding (f )
mGlu
5
IbC50
d
e
No.
K
p
K
p,uu
(nM)
plasma
0.106
0.015
0.081
brain
0.078
0.011
0.044
(
nM)
2
2
3
7
8
3
11
3.1
33
3000
2570
4820
2580
2870
3220
318
38.6
390
201
31.6
142
0.86
1.1
0.63
0.82
0.36
18
10
0.67
4.3
a
b
c
d
e
Average n = 2, male SpragueꢀDawley rats; dose = 10 mg/kg; vehicle = 10% polysorbate 80 in 0.5% methyl cellulose; t = 1 h post dose
Calcium mobilization mGlu assay; values are average of n ≥ 3 independent experiments
= fraction unbound; equilibrium dialysis assay; brain = mouse brain homogenates
= total brain to total plasma ratio
p,uu = unbound brain (brain f · total brain) to unbound plasma (plasma f
5
f
u
K
p
K
u
u
· total plasma) ratio
Considering their favorable profiles, compounds 27 and 28 were selected for oral
pharmacokinetic and bioavailability studies in rats and cynomolgus monkeys (Table 6).
Cynomolgus monkeys were chosen as a second preclinical species as it has been established that
AO activity varies across species. Specifically, AO activity is highest in humans and NHPs,
4
4
moderate in rats, and undetectable in dogs. Using in vitro methods described previously for
3
8,40
9
,
we confirmed that AO contributed significantly to the metabolism of 27 in rats,
cynomolgus monkey, and human hepatic S9 fractions (Fig. 2). Oral exposure was significant as
ACS Paragon Plus Environment

Page 13 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
bioavailability exceeded 50% with both compounds in rats. Compound 27 exhibited an earlier
and slightly higher plasma Cmax than 28; however, the lower clearance of 28 resulted in
prolonged exposure and larger area under the curve (AUC) than observed for 27. The PK results
from analogous studies in cynomolgus monkeys showed these compounds to be similar in that
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
4
2
both compounds showed clearance approximately oneꢀthird of hepatic blood flow.
Bioavailability, while lower than that observed in rats, was still greater than 30% in each case.
With largely attractive in vivo PK profiles, both compounds were advanced into further studies to
identify additional factors for differentiation.
a
Table 6. Rat and cynomolgus monkey pharmacokinetics of compounds 27 and 28
b
b
Rat IV PK
Rat Oral PK
c
c
c
c
d
d
d
d
No.
t
1/2 (min)
MRT (min)
Cl
p
(mL/min/kg)
V
SS (L/kg)
T
T
max (min)
C
max (ꢀM)
AUC (ꢀM·h)
14.6
% F
53
2
2
7
8
390
93
19.3
1.8
30
90
4.34
281
208
9.2
1.9
3.33
38.0
71
e
e
Cynomolgus monkey IV PK
Cynomolgus monkey Oral PK
c
c
c
c
d
d
d
d
No.
t
1/2 (min)
MRT (min)
163
Cl
p
(mL/min/kg)
15.5
V
SS (L/kg)
max (min)
C
max (ꢀM)
AUC (ꢀM·h)
7.32
% F
44
2
2
7
8
302
2.5
180
1.72
563
303
13.3
4.0
120
1.36
6.43
35
a
Oral vehicle = 10% polysorbate 80 in 0.5% methyl cellulose; IV vehicle = 10% EtOH, 70% PEG 400, 20% saline
average n = 2 male SpragueꢀDawley rats
b
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
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
c
d
e
t
T
1/2 = terminal phase plasma halfꢀlife; MRT = mean residence time; Cl
max = time at which maximum concentration occurs; Cmax = maximum concentration; AUC = area under the curve; F = bioavailability
p
= plasma clearance; VSS = volume of distribution at steadyꢀstate
average n = 2 male cynomolgus monkeys
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. Metabolite (38) of 27 identified in rat, cynomolgus monkey, and human hepatic S9
fractions that is produced predominately by AO.
Ancillary pharmacology. To judge the selectivity of these analogs versus the other members of
the mGlu receptor family, the effects of both 27 and 28 at 10 ꢁM on the orthosteric agonist
45,46
concentration response curve (CRC) was measured in foldꢀshift experiments.
No significant
effects on the agonist CRC were noted with either compound, indicating excellent selectivity for
mGlu . To assess potential ancillary pharmacology, both compounds were screened at 10 ꢁM in
5
47
a commercial radioligand binding assay panel of 68 diverse and clinically relevant targets.
While 27 had no significant responses at any target, where significant responses are defined as
inhibition of more than 50% of radioligand binding, one significant response of 86% inhibition
was noted for 28 at the dopamine transporter (DAT). For comparison, 27 only inhibited DAT at a
level of 34%. A subsequent CRC was obtained, and 28 was found to inhibit radioligand binding
48
with an IC of 1.54 ꢁM. Since inhibition of DAT is associated with abuse potential, we next
5
0
4
9
evaluated 28 in a functional cellꢀbased assay that measured inhibition of dopamine uptake.
Indeed, consistent with results obtained from the binding assay, 28 exhibited an IC of 1.38 ꢁM
5
0
in this assay. While 28 is more than 400ꢀfold selective for mGlu versus DAT, the lack of
5
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
concern regarding this target in the case of 27 was viewed as a positive for that compound. With
its preferable P450 inhibition profile (Table 2) and the higher unbound brain concentrations
achieved in rats also in its favor, the decision was made to proceed with additional detailed
characterization of 27.
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
Detailed characterization of 27. Having narrowed our interest to analog 27, we carried out a
more thorough characterization of its interaction with mGlu (Fig. 3). First, to ensure that there
5
was no appreciable species difference in the mGlu activity of 27, we carried out our calcium
5
mobilization assay in HEK293A cells expressing the human receptor (Fig. 3A). As anticipated,
the potency difference between rat (IC50 = 11 nM) and human mGlu (IC = 14 nM) was
5
50
negligible. To study the effects of 27 on mGlu activity in a native system, experiments in rat
5
cortical astrocytes were conducted (Fig. 3B). As expected, 27 induced a concentrationꢀdependent
rightward shift of the glutamate CRC and decreased the maximal response to glutamate. To
investigate whether 27 interacted with the mGlu allosteric site that is common to many mGlu5
5
NAMs such as 1 and 2 (Fig. 1), we also carried out a competition radioligand binding assay with
3
50
[ H]ꢀ3ꢀmethoxyꢀ5ꢀ(pyridinꢀ2ꢀylethynyl)pyridine (39) (Fig. 3C) in membranes prepared from
HEK293A cells expressing rat mGlu . This study confirmed the interaction of 27 with this
5
known binding site (mGlu K = 4.4 nM). We also performed saturation binding experiments
5
i
3
with increasing concentrations of [ H]ꢀ38 in the presence and absence of multiple concentrations
of 27 (Fig. 3D/E). Increasing concentrations of 27 had no effect on the Bmax of the radioligand,
but increased the K as shown (Fig. 3D), a finding that suggests a competitive interaction
d
3
between 27 and [ H]ꢀ39. Scatchard analysis (Fig. 3E) of the results plotting bound/free versus
bound ligand demonstrated that addition of 27 induced a change in slope of the regression line
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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
(
change in K ) but had no effect on the xꢀintercept (no change in Bmax). These data are consistent
d
3
with 27 competitively displacing [ H]ꢀ39 at the aforementioned known allosteric binding site. A
final set of radioligand kinetic binding experiments was conducted that examined the effect of 27
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
on the dissociation of [ H]ꢀ38. In these studies, neither concentration (300 nM or 100 nM) of 27
had any effect on the dissociation rate of the radioligand (data not shown), a result consistent
with prior studies and a competitive interaction. Lastly, we examined the effects of increasing
5
1
concentrations of the mGlu neutral site ligand 5ꢀmethylꢀ6ꢀ(phenylethynyl)ꢀpyridine (40) on
5
the calcium mobilization CRC of 27 in HEK293A cells expressing rat mGlu (Fig. 3F). In these
5
experiments, 40 induced a parallel rightward shift in the 27 concentration response relationship,
which is consistent with 40 competing for the known allosteric binding site with 27.
Furthermore, a Schild analysis (data not shown) of the effects of 40 on the 27 concentration
response yielded a linear regression with a slope of 0.96, indicating the relationship between the
two compounds was competitive. Extrapolation of the line to the xꢀintercept established a K of
i
3
approximately 240 nM, a result consistent with the K value for 40 at the [ H]ꢀ39 site as
i
51
described in the literature (388 nM). Each of these studies combine to conclusively establish
that 27 binds in a competitive manner to the aforementioned known mGlu allosteric binding
5
site.
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2+
Figure 3. (A) Ca mobilization assay CRC for 27 in HEK293A cells expressing human mGlu ;
5
(
B) Effects of 27 on glutamate CRC in rat cortical astrocytes; (C) Competition radioligand
3
binding CRC with 27 and [ H]ꢀ3ꢀmethoxyꢀ5ꢀ(pyridinꢀ2ꢀylethynyl)pyridine (39); (D) Saturation
3
binding experiment: increasing concentrations of [ H]ꢀ39 in the presence and absence of multiple
concentrations of 27; (E) Saturation binding experiment: Scatchard analysis of bound/free versus
bound ligand; (F) Effects of the neutral mGlu site ligand 5ꢀmethylꢀ6ꢀ(phenylethynyl)ꢀpyridine
5
(
40) on the functional inhibition of glutamate responses induced by 27.
With a detailed understanding of the pharmacology of 27 in hand, we moved the compound
into a host of additional assays designed to identify potential risks for clinical development
(Table 7). Permeability and potential for Pꢀglycoprotein (Pꢀgp) mediated efflux was assessed in
52
both MadinꢀDarby canine kidney (MDCK) cells transfected with the human MDR1 gene and
5
3
human colon carcinoma derived Cacoꢀ2 cells. Permeability was moderate, and efflux ratios
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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
were near unity in both cell lines, indicating a low potential for transporter mediated efflux. A
definitive and expanded assessment of the ability of 27 to reversibly inhibit the metabolism of
isoformꢀselective probe substrates of P450s was conducted. Results were generally consistent
with the preliminary results (Table 2) with moderate inhibition of CYP1A2 observed. In parallel
to these studies, the potential for 27 to timeꢀdependently inhibit each P450 was assessed with no
evidence of timeꢀdependent inhibition noted in any case. To shed further light on the likely
human PK profile of 27, the stability of the compound was assessed in human cryopreserved
hepatocytes, a system which is potentially useful for evaluating compounds with nonꢀP450
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
54
mediated routes of metabolism such as AO. Gratifyingly, the predicted hepatic clearance in
humans using this method was low with a halfꢀlife of more than 5 h. To assess any potential
5
5
cardiac liabilities, 27 was subjected to a commercial ion channel panel. Fortunately, no
significant risks were identified in this panel. Finally, the potential mutagenic activity of 27 was
5
6
investigated in the bacterial reverse mutation assay (Ames assay), and no mutagenic responses
were noted in any of the five strains either in the presence or absence of S9 metabolic activation.
Encouraged by the overall profile of 27, we advanced this compound to behavioral assays in the
rat.
Table 7. Further in vitro characterization of 27
Permeability assays
Cell line
MDCKꢀMDR1
Cacoꢀ2
Papp (AꢁB)
Papp (BꢁA)
Efflux Ratio
ꢀ6
ꢀ6
6.2 x 10 cm/s 7.2 x 10 cm/s
1.2
1.2
ꢀ6
ꢀ6
6.5 x 10 cm/s 7.9 x 10 cm/s
a
Definitive P450 inhibition assays
P450
probe
IC50 (ꢀM)
CYP3A4
CYP3A4
CYP2D6
CYP2C9
CYP2C19
midazolam
testosterone
133 ± 23
>200
(R)ꢀbufuralol
diclofenac
>200
29.9 ± 7.3
106 ± 10
(S)ꢀmephenytoin
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CYP2C8
CYP2B6
CYP1A2
amodiaquine
efavirenz
>200
>200
phenacetin
2.90 ± 0.72
Stability in human cryopreserved hepatocytes
intrinsic clearance (CLINT 1.43 ꢁL/min/million cells
halfꢀlife (t1/2
)
)
322 min
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
predicted hepatic clearance (CLHEP
)
3.59 mL/min/kg
b
Cardiac ion channel panel
ion channel
hERG
IC50 (ꢀM)
73.6
ion channel
IC50 (ꢀM)
>100
hCav1.2
hCav3.2
hKir2.1
hHCN2
hHCN4
hNav1.5 (tonic)
hNav1.5 (phasic)
hKv4.3
>100
>100
>100
>100
>100
>100
hKv1.5
>100
>100
c
Bacterial reverse mutation assay
TA98
TA100
WP2 uvrA
TA135
negative
TA1537
negative
negative
negative
negative
a
Assayed in pooled HLM (with and without NADPH) using CYPꢀ
specific probe substrates
Evaluated using IonWorks Quattro system
b
c
TM
With and without S9 metabolic activation
Behavioral pharmacology. It is well known that mice will bury foreign objects such as glass
marbles in deep bedding. Likewise, pretreatment with low doses of benzodiazepines and multiple
5
7
mGlu NAM tool compounds have been shown to inhibit this behavior. It has been argued
5
recently that the marble burying assay models the repetitive and perseverative behavior
5
8
associated with OCD as opposed to noveltyꢀinduced anxiety. We have also employed this rapid
assay that utilizes naïve mice as a frontline behavioral screen for our discovery programs
36,37
directed toward mGlu NAMs.
Using intraperitoneal (IP) dosing as a convenient route for
5
rapidly and accurately dosing large numbers of animals, we examined the effects of 27 in this
test (Fig. 4A). Doseꢀdependent effects on marble burying were observed, and all doses greater
than or equal to 0.3 mg/kg produced significant effects relative to the vehicle control. The 3.0
mg/kg dose produced an effect on marble burying on par with the positive control 2 (15 mg/kg)
in this study. A satellite experiment was conducted using the 3.0 mg/kg dose, and brain samples
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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
were collected and analyzed for exposure of 27. At 15 min post dose of 27, which correlates to
the initiation time point of the marble burying study, the total brain concentration of 27 was 1.48
ꢁ
M. Calculation of the unbound brain concentration using equilibrium dialysis data (Table 5)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
yields a level of 115 nM, which is approximately 10ꢀfold more than the functional mGlu5
activity.
Figure 4. (A) Doseꢀdependent inhibition of marble burying with 27; vehicle = 10% polysorbate
80 (IP); male CDꢀ1 mice; n = 10 per treatment group; 15 min preꢀtreatment; 30 min burying
time; *, p < 0.05 versus vehicle control group, Dunnett’s test; bars denote marbles buried; (B)
Doseꢀdependent inhibition of immobility in rats with 27; vehicle = 10% polysorbate 80 (27,
intraperitoneal) and sterile saline (ketamine, subcutaneous); male SpragueꢀDawley rats; n = 8ꢀ10
per treatment group; 30 min preꢀtreatment; 6 min testing session; *, p < 0.05 versus vehicle
control group, Dunnett’s test; bars denote duration of immobility in s.
Having observed efficacy in this model of anxiety/OCD, we next studied 27 in an animal
model of depression as well. The forced swim test (FST) measures the duration of immobility
5
9
observed in rats suspended in a tank of water from which they cannot escape. Many clinical
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
antidepressants are able to reduce the time that the animals are immobile while increasing the
time that they are active. Furthermore, certain mGlu NAMs have also shown efficacy in this
5
6
0
model of depression. Examination of 27 in this assay showed a doseꢀdependent ability of the
compound to reduce the duration of immobility indicative of antidepressant activity (Fig. 4B).
This result compares favorably to the literature report for 7 in the same assay, where that
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
6
0
compound had a minimum efficacious dose of 10 mg/kg (PO). The high dose used in this study
30 mg/kg IP) produced an efficacy equivalent to that of the positive control, ketamine, a known
antagonist of the Nꢀmethylꢀ ꢀaspartate (NMDA) receptor. Though ketamine is an older drug and
(
D
produces undesirable psychotomimetic side effects, more recent clinical studies have shown that
it can induce a rapid antidepressant effect in patients with treatmentꢀresistant depression
6
1
(TRD). A report from last year implicated the activation of the αꢀaminoꢀ3ꢀhydroxyꢀ5ꢀmethylꢀ4ꢀ
isoxazolepropionic acid (AMPA) receptor by a specific ketamine metabolite rather than direct
antagonism of the NMDA receptor by ketamine as the potential source of these rapid
6
2
antidepressant effects. Interestingly, prior studies have shown that the antidepressant activity in
the FST with mGlu NAM 2 involved signaling through the NMDA but not the AMPA
5
6
3
receptor. Thus, the mechanisms by which ketamine and mGlu5 NAMs illicit their
antidepressant effects may well be distinct.
Imaging studies. Recognizing that an ability to study receptor occupancy (RO) in the clinic
would represent a potential advantage for advancing 27, we decided to investigate RO studies in
1
8
18
18
64
preclinical species. [ F]ꢀ3ꢀFluoroꢀ5ꢀ[(pyridinꢀ3ꢀyl)ethynyl]benzonitrile ([ F]ꢀFPEB, [ F]ꢀ41)
3
39
is a close structural analog of radioligand [ H]ꢀ38 that has been used both preclinically and in
64
the clinic as a positron emission topography (PET) ligand for studying mGlu RO. Thus, we
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
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
18
decided to carry out [ F]ꢀ41 ꢁPET studies with 27 in rats (Fig. 5). In these studies, 27 was dosed
1
8
orally 30 min prior to IV dosing of [ F]ꢀ41. PET imaging was initiated and continued for 1 h.
18
Representative images demonstrating doseꢀdependent displacement of [ F]ꢀ41 by 27 are
depicted in Fig. 5A. Maximum occupancy of 80% was achieved at 10 mg/kg 27 (Fig. 5B).
Plasma samples were collected throughout the study. Plotting the maximum unbound plasma
concentration of 27 versus RO resulted in an unbound level Occ of 13 nM (Fig. 5C). Plotting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
1
8
the dose of 27 versus the RO normalized to the maximum displacement of [ F]ꢀ41 gave an RO
5
0
of 0.8 mg/kg (Fig. 5D).
Figure 5. Imaging studies in rodents; n = 4ꢀ8 SpragueꢀDawley rats; 27 vehicle = 10%
18
polysorbate 80 in 0.5% methyl cellulose; 27 dosed 30 min prior to [ F]ꢀ41 (IV); (A)
18
Representative images of [ F]ꢀ41 by increasing oral doses of 27 (B) mGlu RO obtained with
5
18
[
F]ꢀ41 and increasing oral doses of 27; (C) Unbound maximum plasma concentration of 27
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
versus mGlu RO; (D) Dose of 27 versus mGlu RO; *normalized to the maximum displacement
5
5
18
of [ F]ꢀ41.
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
Based on the RO results in rats, we next investigated whether similar results would be
obtained in NHPs. As such, we advanced 27 into studies in baboons using the same radiotracer
18
[
F]ꢀ41 (Fig. 6). The design for this experiment was very similar to the aforementioned study in
rats; however, in this case both 27 and the radiotracer were delivered via IV administration.
1
8
There was a clear reduction in the uptake of [ F]ꢀ41 relative to baseline that followed a doseꢀ
dependent relationship with 27 (Fig. 6A, 6D, 6E). The maximum dose resulted in 89% RO of
mGlu . Plasma samples were again taken throughout the study to analyze 27 levels, and RO was
5
plotted against the plasma maximum to arrive at a total plasma Occ of 56 nM (Fig. 6B). If one
5
0
assumes that human and baboon plasma protein binding is similar, this equates to a plasma
unbound Occ50 of 6.7 nM, which is within twoꢀfold of that observed in the rodent study. The
effective IV dose that produces 50% RO (RO ) was determined to be 0.06 mg/kg (Fig. 6C).
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Imaging studies in female baboons; subject A received three doses (0.01, 0.1, and 0.18
mg/kg); subject B received two doses (0.03 and 0.56 mg/kg); 27 vehicle = 10% ethanol, 30%
PEG 400, 20% hydroxypropylꢀβꢀcyclodextrin in water; 27 dosed over 20 min, 30 min prior to
18
[ F]ꢀ41 (IV, 3 min bolus) for all doses except the 0.18 mg/kg dose, which was dosed over 2 min,
18
18
5
0 min prior to [ F]ꢀ41; (A) mGlu RO obtained in two female baboons with [ F]ꢀ41 and
5
increasing IV doses of 27; (B) Total maximum plasma concentration of 27 versus mGlu RO;
5
(C) Dose of 27 versus mGlu RO; (D) Images for subject A averaged from 0ꢀ150 min; (E)
5
Images for subject B averaged from 0ꢀ180 min.
CONCLUSION
Continuation of our mGlu NAM discovery program that began with a cellꢀbased HTS has
5
6
5
culminated in the identification of the highly optimized compound 27 (VU0424238).
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 27 is a highly potent and selective antagonist of mGlu with a favorable preclinical
5
profile. The molecule has lowꢀmoderate clearance and good oral bioavailability in two species.
Compound 27 displays CNS penetration in rodents, and we believe that the compound is a lowꢀ
risk for transporterꢀmediated efflux and drugꢀdrug interactions. While the compound is a
moderate inhibitor of CYP1A2, this aspect is considered manageable. Importantly, the
compound is active in preclinical models of anxiety/OCD and depression, both of which are
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
assays where mGlu NAMs have shown activity in the past. Moreover, a clinically useful
5
radiotracer has been used to demonstrate mGlu RO by 27 in both rodents and NHPs. Based on
5
the totality of data, we have selected 27 for advancement to clinical development including
studies designed to enable an investigational new drug (IND) application. Here is the public
disclosure of the chemical structure of 27, our candidate for clinical studies. Future publications
describing detailed metabolic profiling, chemical process development and formulation, and
toxicology studies are under preparation and will be communicated in the near future.
EXPERIMENTAL SECTION
Synthetic Procedures and Characterization Data
1
General. All NMR spectra were recorded on a 400 MHz AMX Bruker NMR spectrometer. H
1
3
and C chemical shifts are reported in δ values in ppm downfield with the deuterated solvent as
the internal standard. Data are reported as follows: chemical shift, multiplicity (s = singlet, d =
doublet, t = triplet, q = quartet, b = broad, m = multiplet), integration, coupling constant (Hz).
High resolution mass spectra were obtained on an Agilent 6540 UHD QꢀTOF with ESI source.
MS parameters were as follows: fragmentor: 150, capillary voltage: 3500 V, nebulizer pressure:
60 psig, drying gas flow: 13 L/min, drying gas temperature: 275 ºC. Samples were introduced via
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
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
an Agilent 1200 UHPLC comprised of a G4220A binary pump, G4226A ALS, G1316C TCC,
and G4212A DAD with ULD flow cell. UV absorption was observed at 215 nm and 254 nm with
a 4 nm bandwidth. Column: Agilent Zorbax Extend C18, 1.8 µm, 2.1 x 50 mm. Gradient
conditions: 5% to 95% CH CN in H O (0.1% formic acid) over 1 min, hold at 95% CH CN 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
3
2
3
0.1 min, 0.5 mL/min, 40 ºC.
Purity. Low resolution reverse phase LCMS analysis was used to assess compound purity at two
wavelengths: 215 and 254 nm. All analogs were at least 95% pure according to this analysis.
Low resolution reverse phase LCMS analysis was performed using an Agilent 1200 system
comprising a binary pump with degasser, highꢀperformance autosampler, thermostatted column
compartment, diodeꢀarray detector (DAD) and a C18 column. Flow from the column was split to
a 6130 SQ mass spectrometer and Polymer Labs ELSD. The MS detector was configured with an
electrospray ionization source. Data acquisition was performed with Agilent Chemstation and
Analytical Studio Reviewer software. Method 1: Samples were separated on a ThermoFisher
Accucore C18 column (2.6 um, 2.1 x 30 mm) at 1.5 mL/min, with column and solvent
temperatures maintained at 45 ºC. The gradient conditions were 7% to 95% acetonitrile in water
(0.1% TFA) over 1.4 min. Lowꢀresolution mass spectra were acquired by scanning from 135 to
700 atomic mass units (AMU) in 0.25 s with a step size of 0.1 AMU and peak width of 0.03
min. Drying gas flow was 11 L/min at a temperature of 350 ºC and a nebulizer pressure of 40
psi. The capillary needle voltage was 3000 V, and the fragmentor voltage was 100V. Method 2:
Samples were separated on a ThermoFisher Accucore C18 column (2.6 um, 2.1 x 30 mm) at 1.5
mL/min, with column and solvent temperatures maintained at 45 ºC. The gradient conditions
were 7% to 95% acetonitrile in water (0.1% TFA) over 1.1 min. Lowꢀresolution mass spectra
were acquired by scanning from 135 to 700 AMU in 0.25 s with a step size of 0.1 AMU and
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
peak width of 0.03 min. Drying gas flow was 11 L/min at a temperature of 350 ºC and a
nebulizer pressure of 40 psi. The capillary needle voltage was 3000 V, and the fragmentor
voltage was 100V. Method 3: Samples were separated on a Restek Aqueous C18 column (3 um,
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
3
.2 x 30 mm) at 1.25 mL/min, with column and solvent temperatures maintained at 45 ºC. The
gradient conditions were 10% to 100% acetonitrile in water (0.1% TFA) over 2 min. Lowꢀ
resolution mass spectra were acquired by scanning from 100 to 1500 AMU in 0.25 s with a step
size of 0.1 AMU and peak width of 0.03 min. Drying gas flow was 11 L/min at a temperature of
350 ºC and a nebulizer pressure of 40 psi. The capillary needle voltage was 3000 V, and the
fragmentor voltage was 70V.
Preparation of compound 27.
4ꢁChloroꢁ2ꢁmethylpyridineꢁNꢁoxide (16a). 2ꢀmethylꢀ4ꢀnitropyridine 1ꢀoxide 14a (5.0 g, 32
mmol, 1.0 eq) was dissolved in concentrated HCl (80 mL) and refluxed for 3 days. The reaction
was cooled, and the excess concentrated HCl was removed in vacuo. The viscous oil was
neutralized with 10% K CO and extracted with CH Cl (5x). The combined organics were dried
2
3
2
2
(MgSO ), filtered, and concentrated in vacuo. Purification by flash chromatography on silica gel
4
afforded 3.49 g (75%) of the title compound as a yellow solid.
(Alternative Procedure) 4ꢀChloroꢀ2ꢀmethylpyridine 15a (5.0 g, 39 mmol, 1.0 eq) and hydrogen
peroxideꢀurea adduct (7.37 g, 78.4 mmol, 2.0 eq) were dissolved in THF (196 mL) and cooled to
0
ºC. Trifluoroacetic anhydride (12 mL, 86.3 mmol, 2.2 eq) was added dropwise over 15 min,
and the reaction was allowed to warm to room temperature. After determination of the
completion of the reaction by LCMS (approximately 45 min), the reaction was cooled to 0 ºC
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
and quenched with a 10% aqueous solution of sodium thiosulfate. The reaction was extracted
with EtOAc (3x), dried (MgSO ), filtered and concentrated in vacuo. Purification by flash
4
1
chromatography on silica gel afforded 5.63 g (99%) of the title compound as a white solid. H
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
NMR (400 MHz, DMSOꢀd ) δ 8.23 (d, J = 7.0 Hz, 1H), 7.66 (d, J = 3.0 Hz, 1H), 7.40 (dd, J =
6
1
3
7.0, 3.0 Hz, 1H), 2.32 (s, 3H); C NMR (100 MHz, DMSOꢀd ) δ 149.53, 139.78, 128.73,
6
+
1
26.41, 124.17, 17.09 ppm; LCMS (Method 1): R = 0.302 min, m/z = 144.2 [M+H] ; HRMS,
T
calc’d for C H ClNO [M], 143.0138; found 143.0139.
6
6
4ꢁChloroꢁ6ꢁmethylpicolinonitrile (17a). Compound 16a (8.97 g, 62.5 mmol, 1.0 eq) was
dissolved in CH Cl and dried over MgSO . The solution was added to a flameꢀdried roundꢀ
2
2
4
bottom flask and CH Cl was added to give a total volume of 188 mL. Trimethylsilyl cyanide
2
2
(10 mL, 75 mmol, 1.2 eq) was added, and the reaction stirred for 15 min. Dimethylcarbamyl
chloride (6.9 mL, 75 mmol, 1.2 eq) was added dropwise over 20 min, and the reaction was
stirred for 24 h. An additional one equivalent each of trimethylsilylcyanide and
dimethylcarbamyl chloride were added, and the reaction was stirred for another 72 h. The
reaction was made basic with 10% K CO and extracted with CH Cl (3x). The combined
2
3
2
2
organics were dried (MgSO ), filtered and concentrated in vacuo. Purification by flash
4
1
chromatography on silica gel afforded 7.2 g (76%) of the title compound as a white solid. H
NMR (400 MHz, DMSOꢀd ) δ 8.13 (d, J = 1.6 Hz, 1H), 7.82 (d, J = 1.7 Hz, 1H), 2.52 (s, 3H);
6
13
C NMR (100 MHz, DMSOꢀd ) δ = 161.99, 144.13, 133.01, 127.65, 126.40, 116.60, 23.57 ppm;
6
+
LCMS (Method 1): R = 0.715 min, m/z = 153.2 [M+H] ; HRMS, calc’d for C H ClN [M],
T
7
5
2
152.0141; found 152.0139.
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
6
ꢁMethylꢁ4ꢁ(pyrimidinꢁ5ꢁyloxy)picolinonitrile (19a). Compound 17a (4.0 g, 26 mmol, 1.0 eq),
5ꢀhydroxypyrimidine 18a (5.56 g, 57.9 mmol, 2.2 eq), K CO (7.24 mg, 52.4 mmol, 2.0 eq) and
2 3
º
DMF (66 mL) were added to a reaction vessel and heated at 80 C for 16 h. The reaction was
filtered and concentrated on silica gel (25 g). The silica gel was loaded on top of a fresh bed of
silica gel and washed with 50% EtOAc/hexane. The solvents were removed in vacuo and the
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
crude solid was purified by flash chromatography on silica gel to afford 4.31 g (77%) of the title
1
compound as a paleꢀyellow solid. H NMR (400 MHz, DMSOꢀd ) δ 9.17 (s, 1H), 8.85 (s, 2H),
6
1
3
7.74 (d, J = 2.4 Hz, 1H), 7.32 (d, J = 2.3 Hz, 1H), 2.48 (s, 3H); C NMR (100 MHz, DMSOꢀd )
6
δ = 164.04, 162.28, 155.51, 150.02, 148.75, 133.54, 117.05, 115.59, 115.04, 23.82 ppm; LCMS
+
(Method 1): R = 0.535 min, m/z = 213.2 [M+H] ; HRMS, calc’d for C H N O [M], 212.0698;
T
11
8
4
found 212.0697.
6
ꢁMethylꢁ4ꢁ(pyrimidinꢁ5ꢁyloxy)picolinic acid (20a). Compound 19a (4.31 g, 20.3 mmol, 1.0
eq) was dissolved in dioxane (90 mL), and 2N NaOH (45 mL) was added. The mixture was
refluxed for 18 h and after cooling the reaction was neutralized with 2N HCl (45 mL). The water
and solvent were removed in vacuo, and the crude reaction was dissolved in 10%
MeOH/CH Cl . The undissolved salt was filtered off and the solvents were removed in vacuo to
2
2
afford 4.65 g (99%) of the title compound as a white solid which was used without further
purification.
Nꢁ(5ꢁFluoropyridinꢁ2ꢁyl)ꢁ6ꢁmethylꢁ4ꢁ(pyrimidinꢁ5ꢁyloxy)picolinamide (27). Compound 20a
(16.84 g, 72.83 mmol, 1.0 eq) and 5ꢀfluoroꢀ2ꢀaminopyridine (16.33 g, 0.15 mol, 2.0 eq) were
dissolved in pyridine (485 mL) in a flameꢀdried roundꢀbottom flask. The reaction was cooled to ꢀ
ACS Paragon Plus Environment