Discovery of molecular switches within the ADX-47273 mGlu5 PAM scaffold that modulate modes of pharmacology to afford potent mGlu5 NAMs, PAMs and partial antagonists

Article abstract of DOI:10.1016/j.bmcl.2010.11.119

This Letter describes a chemical lead optimization campaign directed at a weak mGlu₅ NAM discovered while developing SAR for the mGlu ₅ PAM, ADX-47273. An iterative parallel synthesis effort discovered multiple, subtle molecular switches that afford potent mGlu₅ NAMs, mGlu₅ PAMs as well as mGlu₅ partial antagonists.

Full text of DOI:10.1016/j.bmcl.2010.11.119

Bioorganic & Medicinal Chemistry Letters 21 (2011) 2711–2714
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of molecular switches within the ADX-47273 mGlu5 PAM scaffold
that modulate modes of pharmacology to afford potent mGlu5 NAMs, PAMs
and partial antagonists
Jeffrey P. Lamb a,c, , Darren W. Engers a,c,d, , Colleen M. Niswender a,c,d, Alice L. Rodriguez a,c,d
,
Daryl F. Venable a,c, P. Jeffrey Conn a,c,d, Craig W. Lindsley a,b,c,d,
⇑
a Department of Pharmacology, Vanderbilt University Medical Center, Nashville, TN 37232, USA
b Department of Chemistry, Vanderbilt University, Nashville, TN 37232, USA
c Vanderbilt Program in Drug Discovery, Nashville, TN 37232, USA
d Vanderbilt Specialized Chemistry Center(MLPCN), Nashville, TN 37232, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
This Letter describes a chemical lead optimization campaign directed at a weak mGlu5 NAM discovered
while developing SAR for the mGlu5 PAM, ADX-47273. An iterative parallel synthesis effort discovered
multiple, subtle molecular switches that afford potent mGlu5 NAMs, mGlu5 PAMs as well as mGlu5 partial
antagonists.
Received 28 October 2010
Revised 23 November 2010
Accepted 29 November 2010
Available online 3 December 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
mGlu5
NAM
PAM
Allosteric
Metabotropic
Glutamate
The metabotropic glutamate receptor subtype 5 (mGlu5) has
become a prominent molecular target for a number of CNS pathol-
ogies.1,2 mGlu5 negative allosteric modulators (NAMs) are being
actively pursued for anxiety, pain, Parkinson’s disease, cocaine
addiction and Fragile X Syndrome, while mGlu5 positive allosteric
modulators (PAMs) are under development for the treatment of
schizophrenia.3–9 The prototypical mGlu5 allosteric ligand is MPEP
(1),10 a NAM, and many allosteric ligands, both PAM and NAM, bind
at the MPEP-site.1–10 Recently, we reported on the discovery of
molecular switches in a series of MPEP-site phenylethynyl pyrim-
idines in which incorporation of a single methyl group in either the
3- or 4-position converted an mGlu5 partial antagonist lead 2
(IC50 = 486 nM, 71% partial) into either a NAM 3 (IC50 = 7.5 nM)
mGlu5 PAM ADX-47273 5 series in 2009 produced potent PAMs,
such as 6 (EC50 = 240 nM, 14-fold shift), and ago-PAMs such as 7
(EC50 = 170 nM, 20-fold shift), but only one weak NAM
8
(IC50 = 8.7
l
M).15 This was the first indication that pharmacology
switching is possible in the ADX-47273 series by replacing an aryl
amide, as in 6, with a cyclobutyl amide in 8.15
While we were exploring this finding, a manuscript appeared in
2010 describing the identification of racemic mGlu5 NAM 9, closely
related to our NAM 8, from an HTS screen, and the parallel synthe-
sis of over 1300 analogs.16 However, within this manuscript, there
is little discussion of the impact of stereochemistry and no mention
of pharmacology switching. Here, we present our SAR study, devel-
oped though an iterative parallel synthesis approach, that afforded
potent mGlu5 PAMs, NAMs and partial antagonists from subtle
modifications to the ADX-47273 scaffold.
Our initial library evaluated two dimensions: stereochemistry
at the 3-postion and replacement for the 2-pyridyl moiety while
holding the cyclobutyl amide constant. In our earlier work in the
ADX-47273 series,15 the (S)-stereochemistry at the 3-position
was essential for mGlu5 PAM activity, and it was important to
ascertain the stereochemical bias, if any, to produce NAMs. In the
event, (S)-10 was converted to the methyl ester 11, followed by
acylation to yield 12. Saponification provides 13, which is then
or PAM 4 (EC50 = 3.3 l
M, 4.2-fold shift), respectively (Fig. 1).11 Fur-
ther SAR identified additional, subtle molecular switches that
afforded centrally penetrant and in vivo active mGlu5 NAMs and
PAMs.12 After these key findings, we began to take note of pharma-
cology switches, and identified these in multiple mGlu5 allosteric
modulator scaffolds.13,14 Interestingly, our initial SAR work in the
⇑
Corresponding author.
E-mail address: craig.lindsley@vanderbilt.edu (C.W. Lindsley).
These authors contributed equally to this work.
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.11.119

2712
J. P. Lamb et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2711–2714
Table 1
N
N
Structures and activities of analogs (S)-15 and (R)-15
N
N
N
N
N
O
O
R
R
N
N
N
N
2
3
1, MPEP
(NAM)
(Partial Antagonist)
(NAM)
N
O
O
O
N
O
N
N
N
(S)-15
(R)-15
F
N
N
a
a
N
Compd
R
Pharmacology IC50
EC50
Glu
N
(l
M)
(l
M)
Maxa (%)
O
O
(S)-15a
(R)-15a
NAM
Inactive
9.3
—
NA
—
67
—
F
F
F
S
4
5, ADX-47273
(Ago-PAM)
6
(PAM)
(PAM)
F
(S)-15b
(R)-15b
Inactive
Inactive
—
—
—
—
—
—
N
N
O
O
Cl
N
N
N
S
(S)-15c
(R)-15c
NAM
NAM
>10
9.9
NA
NA
33
19
N
N
N
N
N
O
N
O
O
O
(S)-15d
(R)-15d
NAM
NAM
2.4
>10
NA
NA
31
60
F
F
7
8
9
(Ago-PAM)
(NAM)
(NAM)
(S)-15e
(R)-15e
NAM
NAM
0.2
3.1
NA
NA
2.4
18
Figure 1. Structures of selected MPEP-site allosteric ligands that display a range of
mGlu5 pharmacology with subtle modifications.
Cl
(S)-15f
(R)-15f
NAM
NAM
0.7
4.7
NA
NA
2.5
14
CH3
coupled to various (Z)-N’-hydroxylimidamides 14 and refluxed to
deliver analogs (S)-15 (Scheme 1). The analogous (R)-15 congeners
were made via the same scheme except (R)-10 was used.
(S)-15g
(R)-15g
NAM
NAM
1.8
>10
NA
NA
2.1
54
OCH3
As shown in Table 1, the stereochemical preference we identi-
fied in our earlier PAM work in this series carried over into the
NAM pharmacology with the (S)-enantiomer preferred, that is,
a
Average of at least three independent determinations. NA, not applicable.
(S)-15e (IC50 = 0.2 lM) versus (R)-15e (IC50 = 3.1 lM). Significantly,
3-substituted aryl congeners (S)-15e–f, proved most enlightening,
affording submicromolar mGlu5 NAMs, with in the case of (S)-15e,
an ꢀ41-fold increase in potency over 8.15 These data led us to con-
sider if there is stereochemical bias for pharmacological mode of
action within the 9 scaffold. Thus we prepared small, enantiopure
libraries of analogs (S)-20 and (R)-20, from either (S)-16 and (R)-16,
respectively, and evaluated them in our mGlu5 assays (Scheme 2).
As shown in Table 2, this effort found that both enantiomers afford
comparable activity and mode of pharmacology. This library pro-
vided an efficacious submicromolar PAM (S)-20c (EC50 = 730 nM,
71% Glu Max) as well as several submicromolar NAMs ((S)- and
(R)-20e–f) which also afforded a full blockade of the EC80, and in
b
a
OMe
O HCl
OH
OMe
N
H
N
N
Boc
O
O
N
O
16
17
18
HO
OH
d
c
N
N
N
+
R
O
O
N
H2N
R
O
O
19
14
(S)-20
Scheme 2. Reagents and conditions: (a) SOCl2, MeOH (99%), cylcobutane carbonyl
chloride, DIEA, DCM (95%); (c) LiOH, THF, H2O (95%); (d) EDCI, HOBt, DIEA, dioxane,
reflux, 24 h (40–55%).
O
O
O
a
b
OH
OMe
HCl
OMe
N
H
N
H
N
the case of (S)-20f, an 77 nM NAM. Based on these data, our next
round of library synthesis employed both the 20e NAM scaffold
and the 20c PAM scaffold, and focused on evaluating other amide
moieties beyond the cyclobutyl amide. These analogs 21 and 22
were readily prepared following a variation of Scheme 2.
The library of 20e analogs, 21a–i, afforded both NAMs and par-
tial antagonists,17 with no evidence of PAM activity (Table 3). Inter-
estingly, the three and five-membered saturated ring amides 21a
and 21c, afforded partial antagonists, while the four and six-mem-
bered saturated ring amides 21b and 21d afforded full non-com-
petitive antagonists (NAMs). In contrast, the library of 20c
analogs, 22a–i, afforded predominantly PAMs and ago-PAMs. For
example, 22a proved to be a potent (EC50 = 78 nM, 70% Glu Max)
mGlu5 PAM, more potent than the previous PAMs 6 and 7 we
O
N
10
11
12
N
O
O
R
OH
HO
N
c
d
N
+
N
H2N
R
O
O
13
14
(S)-15
Scheme 1. Reagents and conditions: (a) SOCl2, MeOH (99%), cylcobutane carbonyl
chloride, DIEA, DCM (96%); (c) LiOH, THF, H2O (95%); (d) EDCI, HOBt, DIEA, dioxane,
reflux, 24 h (45–59%).

J. P. Lamb et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2711–2714
2713
Table 2
Table 4
Structures and activities of analogs (S)-20 and (R)-20
Structures and activities of analogs 23
N
N
N
N
R
N
N
R
R
O
N
O
O
O
N
N
O
O
23
(S)-20
(R)-20
a
a
Compd
R
Pharmacology IC50
M)
EC50
M)
Glu Maxa
a
a
Compd
R
Pharmacology IC50
EC50
Glu
(l
(l
(%)
64
72
25
(l
M)
(
l
M)
Maxa
(%)
23a
23b
23c
PAM
PAM
NAM
NA
NA
10
1.5
2.7
NA
(S)-20a
(R)-20a
NAM
NAM
2.6
3.5
NA
NA
11
7
F
S
N
F
(S)-20b
(R)-20b
NAM
Inactive
10
—
NA
—
33
—
Cl
(S)-20c
(R)-20c
PAM
PAM
NA
NA
0.7
0.6
71
37
F
23d
NAM
2.4
NA
34
(S)-20d
(R)-20d
NAM
NAM
0.9
10
NA
NA
6
38
CH3
N
N
N
Cl
23e
23f
Inactive
Inactive
NA
NA
NA
NA
NA
NA
F
(S)-20e
(R)-20e
NAM
NAM
0.5
0.3
NA
NA
3
6
CH3
(S)-20f
(R)-20f
NAM
NAM
0.08
0.3
NA
NA
1.8
0.5
23g
Inactive
NA
NA
NA
N
OCH3
a
Average of at least three independent determinations. NA, not applicable.
a
Average of at least three independent determinations. NA, not applicable.
Table 3
Table 5
Structures and activities of analogs 21 and 22
Structures and activities of analogs 24
CH3
F
F
N
N
N
N
N
N
O
O
N
O
N
N
R
O
R
O
R
O
24
21
22
EC50
Glu Maxa
(%)
a
a
Compd
R
Pharmacology IC50
M)
EC50
M)
a
a
Compd
R
Pharmacology IC50
M)
Glu Maxa
(%)
(l
(
l
(l
(lM)
24a/
23a
21a
21b
Part. Antag.
Ago-PAM
0.28
NA
NA
0.08
16
70
PAM
PAM
PAM
NA
NA
NA
2.9
68
63
75
24b
1.5
1.0
21b/20c
22b/20e
NAM
PAM
0.37
NA
NA
0.65
5.8
38
24c
21c
22c
Part. Antag.
Ago-PAM
0.6
NA
NA
0.31
24
80
24d
24e
PAM
PAM
NA
NA
10
0.43
8
21d
22d
NAM
PAM
2.5
NA
NA
3.5
9.2
58
S
60
S
21e
22e
NAM
NAM
0.73
10
NA
NA
0.9
34
F
24f
PAM
PAM
NA
NA
3.7
62
47
F
F
21f
22f
NAM
PAM
10
NA
NA
1.4
46
93
F
24g
2.6
3.6
F
F
21g
22g
NAM
PAM
4.7
NA
NA
0.46
16
59
24h
24i
F
F
PAM
PAM
NA
NA
59
61
F
10
21h
22h
NAM
PAM
10
NA
NA
2.3
47
91
F
F
a
Average of at least three independent determinations. NA, not applicable.
F
21i
22i
Inactive
PAM
—
NA
—
2.9
—
95
a
Average of at least three independent determinations. NA, not applicable.
amides 22a and 22c affording ago-PAMs, while the 4- and 6-mem-
bered saturated ring amides 22b and 22d displaying pure PAM
activity. Again, very subtle perturbations to the core scaffold
engender opposing modes of mGlu5 pharmacology.
developed in the ADX-47273 series.15 In addition, we observed an
interesting trend here with the 3- and 5-membered saturated ring

2714
J. P. Lamb et al. / Bioorg. Med. Chem. Lett. 21 (2011) 2711–2714
F
F
studies with these ligands, as well as their metabolites, are in pro-
gress and will be reported in due course.
N
N
N
N
O
O
N
N
O
O
Acknowledgements
(S)-23a
(R)-23a
The authors thank NIDA (DA023947-01) for support of our pro-
grams in developing mGlu5 NAMs and partial antagonists for the
treatment of addiction. The authors would like to thank Nathan
Kett, Chris Denicola and Sichen Chang for the purification of com-
pounds utilizing the mass-directed HPLC system.
mGlu5 EC50 7,000 nM
57.3% Glu Max
mGlu5 EC50 = 530 nM
68.4% Glu Max
Figure 2. Resolved enantiomers of 23a. All the mGlu5 PAM activity resides in the
(R)-enantiomers, (R)-23a.
References and notes
At this point, we elected to examine the impact of contracting
the piperidine ring to a pyrrolidine ring while maintaining the
original cyclobutyl amide and surveying a diverse group of sub-
stituents on the oxadiazole ring. This initial library employed
racemic proline to afford racemic analogs 23, following a varia-
tion of Scheme 2. As shown in Table 4, this modification afforded
1. (a) Schoepp, D. D.; Jane, D. E.; Monn, J. A. Neuropharmacology 1999, 38, 1431–
1476; (b) Conn, P. J.; Pin, J.-P. Annu. Rev. Pharmacol. Toxicol. 1997, 37, 205–237.
2. (a) Lea, P. M. I. V.; Faden, A. I. CNS Drug Rev. 2006, 12, 149–166; (b) Kew, J. N. C.
Pharmacol. Ther. 2004, 104, 233–244.
3. Lindsley, C. W.; Emmitte, K. A. Curr. Opin. Drug Disc. Dev. 2009, 12, 446–457.
4. Conn, P. J.; Lindsley, C. W.; Jones, C. Trends Pharm. Sci. 2009, 30, 25–31.
5. Conn, P. J.; Christopolous, A.; Lindsley, C. W. Nat. Rev. Drug Disc. 2009, 8, 41–54.
6. Lindsley, C. W.; Wisnoski, D. D.; Leister, W. H.; O’Brien, J. A.; Lemiare, W.;
Williams, D. L., Jr.; Burno, M.; Sur, C.; Kinney, G. G.; Pettibone, D. J.; Tiller, P. R.;
Smith, S.; Duggan, M. E.; Hartman, G. D.; Conn, P. J.; Huff, J. R. J. Med. Chem.
2004, 47, 5825–5828.
7. Hammond, A. S.; Rodriguez, A. L.; Niswender, C. M.; Lindsley, C. W.; Conn, P. J.
ACS Chem. Neuro. 2010, 1, 702–716.
8. Zhao, Z.; Wisnoski, D. D.; O’Brien, J. A.; Lemiare, W.; Williams, D. L., Jr.;
Jacobson, M. A.; Wittman, M.; Ha, S.; Schaffhauser, H.; Sur, C.; Pettibone, D. J.;
Duggan, M. E.; Conn, P. J.; Hartman, G. D.; Lindsley, C. W. Bioorg. Med. Chem.
Lett. 2007, 17, 1386–1391.
9. Felts, A. S.; Lindsley, S. R.; Lamb, J. P.; Rodriguez, A. L.; Menon, U. N.; Jadhav, S.;
Conn, P. J.; Lindsley, C. W.; Emmitte, K. A. Bioorg. Med. Chem. Lett. 2010, 20,
4390–4394.
10. Gasparini, F.; Lingenhohl, K.; Stoehr, N.; Flor, P. J.; Heinrich, M.; Vranesic, I.;
Biollaz, M.; Allgeier, H.; Heckendorn, R.; Urwyler, S.; Varney, M. A.; Johnson, E.
C.; Hess, S. D.; Rao, S. P.; Sacaan, A. I.; Santori, E. M.; Veliocelebi, G.; Kuhn, R.
Neuropharmacology 1999, 38, 1493–1503.
inactive compounds, weak NAMs (IC50s ꢀ10
lM) and two low
micromolar PAMs (23a and 23b). Based on these data, we made
a second generation library holding constant the 3-fluorobenzene
moiety of 23a, and surveyed a diverse collection of amide moie-
ties to replace the cyclobutyl group. As shown in Table 5, this
effort afforded predominantly pure PAMs 24 with a range of
potencies and efficacies. To address the role of stereochemical
preference, we separated racemic 23a into pure enantiomers
(S)-23a and (R)-23a by chiral SFC. In this case, (R)-23a is a potent
mGlu5 PAM (EC50 = 530 nM) while (S)-23a is a very weak PAM
(EC50 = 7000 nM) (Fig. 2). Note, this is the opposite stereochemical
preference observed within the 3-piperidinyl-based mGlu5 PAMs
5–7.15
11. Sharma, S.; Rodriguez, A. L.; Conn, P. J.; Lindsley, C. W. Bioorg. Med. Chem. Lett.
2008, 18, 4098–5101.
In summary, an iterative parallel synthesis optimization ap-
proach for our weak mGlu5 NAM 8, identified multiple regioiso-
meric and stereochemical ‘molecular switches’ that modulated
modes (NAM, partial antagonist, PAM, ago-PAM) of mGlu5 pharma-
12. Sharma, S.; Kedrowski, J.; Rook, J. M.; Smith, J. M.; Jones, C. K.; Rodriguez, A. L.;
Conn, P. J.; Lindsley, C. W. J. Med. Chem. 2009, 52, 4103–4106.
13. Rodriguez, A. L.; Williams, R.; Zhou, S.; Lindsley, S. R.; Le, U.; Conn, P. J.;
Lindsley, C. W. Bioorg. Med. Chem. Lett. 2009, 19, 3209–3212.
14. Zhou, Y.; Rodriguez, A. L.; Williams, R.; Weaver, C. D.; Conn, P. J.; Lindsley, C. W.
Bioorg. Med. Chem.Lett. 2009, 19, 6502–6506.
15. Engers, D. W.; Rodriguez, A. L.; Oluwatola, O.; Hammond, A. S.; Venable, D. F.;
Williams, R.; Sulikowski, G. A.; Conn, P. J.; Lindsley, C. W. ChemMedChem 2009,
4, 505–511.
cology. From 8 (IC50 = 8.7 lM), potent PAMs (EC50 = 78–200 nM)
and NAMs (IC50 = 77–400 nM), were developed. In many cases,
the perturbations in structure were subtle which led to opposing
modes of pharmacology and suggests subtle conformational
changes within the GPCR either facilitate or prohibit coupling to
the G-protein. These data, coupled with our earlier work in a struc-
turally distinct MPEP-site scaffold 2, suggests that metabolites of
MPEP-site allosteric ligands must be characterized, as metabolites
may engender opposing modes of mGlu5 modulation. Additional
16. Wágner, G.; Wéber, C.; Nyéki, O.; Nógrádi, K.; Bielik, A.; Molnár, L.; Bobok, A.;
Horváth, A.; Kiss, B.; Kolok, S.; Nagy, J.; Kurkó, D.; Gál, K.; Greiner, I.;
}
Szombathelyi, Z.; Keseru, G. M.; Dómany, G. Bioorg. Med. Chem. Lett. 2010, 20,
3737–3741.
17. Rodriguez, A. L.; Nong, Y.; Sekaran, N. K.; Alagille, D.; Tamagnan, G. D.; Conn, P.
J. Mol. Pharmacol. 2005, 68, 1793–1802.