Surveying heterocycles as amide bioisosteres within a series of mGlu7 NAMs: Discovery of VU6019278

Article abstract of DOI:10.1016/j.bmcl.2019.03.016

This letter describes a diversity-oriented library approach to rapidly assess diverse heterocycles as bioisosteric replacements for a metabolically labile amide moiety within a series of mGlu₇ negative allosteric modulators (NAMs). SAR rapidly

Full text of DOI:10.1016/j.bmcl.2019.03.016

Accepted Manuscript
Surveying heterocycles as amide bioisosteres within a series of mGlu7 NAMs:
discovery of VU6019278
Carson W. Reed, Jordan P. Washecheck, Marc C. Quitlag, Matthew T. Jenkins,
Alice L. Rodriguez, Darren W. Engers, Anna L. Blobaum, P. Jeffrey Conn,
Colleen M. Niswender, Craig W. Lindsley
PII:
S0960-894X(19)30147-7
https://doi.org/10.1016/j.bmcl.2019.03.016
BMCL 26330
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
31 January 2019
18 February 2019
13 March 2019
Please cite this article as: Reed, C.W., Washecheck, J.P., Quitlag, M.C., Jenkins, M.T., Rodriguez, A.L., Engers,
D.W., Blobaum, A.L., Jeffrey Conn, P., Niswender, C.M., Lindsley, C.W., Surveying heterocycles as amide
bioisosteres within a series of mGlu7 NAMs: discovery of VU6019278, Bioorganic & Medicinal Chemistry
Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.03.016
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.9.0_/0
Surveying heterocycles as amide bioisosteres within a
series of mGlu7 NAMs: discovery of VU6019278
Leave this area blank for abstract info.
Carson W. Reed, Jordan P. Washecheck, Marc C. Quitalig, Matthew T. Jenkins, Alice L. Rodriguez, Darren W. Engers,
Anna L. Blobaum, P. Jeffrey Conn, Colleen M. Niswender and Craig W. Lindsley

Bioorganic & Medicinal Chemistry Letters
Surveying heterocycles as amide bioisosteres within a series of mGlu7 NAMs:
discovery of VU6019278
Carson W. Reed,a,c Jordan P. Washecheck,a,c Marc C. Quitlag,a,b Matthew T. Jenkins,a,b Alice L.
Rodriguez,a,b Darren W. Engers,a,b Anna L. Blobaum,a,b P. Jeffrey Conn,a,b,e Colleen M. Niswender,a,b,e*
and Craig W. Lindsley a,b,c,d*
aVanderbilt Center for Neuroscience Drug Discovery, Vanderbilt University Medical Center, Nashville, TN 37232, USA
bDepartment of Pharmacology, Vanderbilt University School of Medicine, Nashville, TN 37232, USA
cDepartment of Chemistry, Vanderbilt University, Nashville, TN 37232, USA
dDepartment of Biochemistry, Vanderbilt University, Nashville, TN 37232, USA
eVanderbilt Kennedy Center, Vanderbilt University School of Medicine, Nashville, TN 37232, USA
*To whom correspondence should be addressed: colleen.m.niswender@vanderbilt.edu or craig.lindsley@vanderbilt.edu
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
This letter describes a diversity-oriented library approach to rapidly assess diverse heterocycles
as bioisosteric replacements for a metabolically labile amide moiety within a series of mGlu7
negative allosteric modulators (NAMs). SAR rapidly honed in on either a 1,2,4- or 1,3,4-
oxadizaole ring system as an effective bioisostere for the amide. Further optimization of the
southern region of the mGlu7 NAM chemotype led to the discovery of VU6019278, a potent
mGlu7 NAM (IC50 = 501 nM, 6.3% L-AP4 Min) with favorable plasma protein binding (rat fu=
0.10), low predicted hepatic clearance (rat CLhep = 27.7 mL/min/kg) and high CNS penetration
(rat Kp = 4.9, Kp,uu = 0.65).
Keywords:
mGlu7
metabotropic glutamate receptor
negative allosteric modulator (NAM)
Structure-Activity Relationship (SAR)
VU6019278
2019 Elsevier Ltd. All rights reserved.
Metabotropic glutamate receptor subtype 7 (mGlu7) is an
attractive therapeutic target for a number of CNS disorders,
including ADHD, schizophrenia, depression, Rett syndrome,
epilepsy, anxiety and autism.1-13 However, of the eight mGlu
receptors (mGlu1-8), a lack of highly selective in vivo tool
compounds has severely hindered the field from assessing the
true therapeutic potential of activating or inhibiting mGlu7 in the
CNS.14,15 In order to provide subtype selectivity amongst the
mGlu family, efforts have largely focused on target modulation
through allosteric mechanisms. To date, there are no selective
mGlu7 positive allosteric modulators (PAMs), but both pan-
Group III PAMs and mGlu7-preferring PAMs have been
metabolically labile amide linker with bioisosteres. Recently, we
reported on the first approach, which led to the discovery of 6.23
Here, we will describe efforts examining a diverse array of small
heterocycles to serve as replacement bioisosteres for the amide
linker and the identification of an additional new tool compound
and potential lead.
reported.16
The literature with mGlu7 negative allosteric
modulators (NAMs) is more mature (e.g, 1-5),17-22 but only
recently has a robust in vivo tool compound, 6 (VU6012962)
been reported (Fig. 1).23
Prior to the discovery of 6, poor
physiochemical and DMPK properties precluded key in vivo
proof of concept studies. Along the path to 6, NAM 5 was an
important step,22 demonstrating robust efficacy in native tissues
(i.e., electrophysiology), but metabolic instability of the amide
linker limited in vivo studies. Further optimization of 5 followed
two parallel tracks: 1) functionalization of the eastern aryl ring to
improve DMPK properties and 2) replacement of the
Figure 1. Structures of reported mGlu7 NAMs 1-5, and the key in vivo
tool compound 6.

replacements for the amide linker in 5. A 1,3,4-oxadiazole 7h and a
1,2,4-oxadiazole 7i stood out as viable replacements.
The optimization plan is depicted in Figure 2, where we
envisioned surveying a diverse array of heterocycles as amide
bioisosteres to afford analogs 7. If a viable heterocyclic
bioisostere could be identified, we would then optimize the
substituents on the eastern aryl ring, and ensure that the southern
1,2,4-triazole remained the optimal moiety in the context of
analogs 5.22
Both NAMs displayed favorable profiles, with predicted hepatic
clearances (CLhep) in rat of 27.2 and 26.8 mL/min/kg for 7h and
7i, respectively, moderate plasma protein binding (rat fus of 0.14
and 0.31 for 7h and 7i, respectively) and, in the case of 7h,
favorable rat brain homogenate binding (fu = 0.022).24 Based on
these data, further optimization was pursued around both 7h and
7i.
SAR around the 1,2,4-oxadiazole core was steep, with very
few analogs displaying even modest mGlu7 NAM potency, and
no analogs were more potent than 7i. Therefore, all attention
shifted to a focus on optimization of the 1,3,4-oxadiazole 7h. A
rapid, four-step, microwave-assisted route to analogs 8 was
developed (Scheme 1) to quickly assess diverse aryl (R2)
moieties. Commercial esters 9 were treated with hydrazine under
microwave conditions to provide the analogous acyl hydrazides
10 in 81-88% yields. In parallel, commercial 2-fluoro-5-
(trifluoromethoxy)benzonitrile 11 undergoes an SNAr reaction
with 1H-1,2,4-triazole to deliver 12 in 67% yield, followed by
hydrolysis of the nitrile to the acid 13. Finally, condensation of
acyl hydrazides 10 with acid 13 with T3P® under microwave
conditions provided analogs 8.
Figure 2. Optimization plan for the labile amide containing 5. First,
heterocyclic bioisoteres for the amide in 5 would be evaluated, followed
by optimization of the southern 1,2,4-triazole and the eastern 3,4-
dimethoxy phenyl ring.
To survey a wide array of potential heterocyclic bioisosteres
for the amide linker in 5, we held the western (R1) and eastern
(R2) moieties of 5 constant, and surveyed 13 heterocycles with
varying hydrogen-bond donating and accepting capabilities in
analogs 7 (Fig. 3). SAR was steep, with many inactive analogs
(mGlu7 IC50 > 30 M, e..g, 7a, 7j-m), a number of weak NAMs
(e.g., 7b-d) and a few with modest activity (e.g., 7e-g).
Interestingly, a 1,3,4-oxadiazole analog 7h proved to be an
acceptable amide replacement (mGlu7 IC50 = 1.58 M, 7.1% L-
AP4 Min), as did a 1,2,4-oxadiazole 7i (mGlu7 IC50 = 1.32 M,
14.5% L-AP4 Min); moreover, both analogs were within two-fold
of the parent 5.22 Prior to optimizing the R1 and R2 positions of
these oxadiazole biosiosteres, we evaluated their in vitro DMPK
properties to ensure further consideration was warranted.
Scheme 1. Synthesis of putative mGlu7 NAM analogs 8.a
aReagents and conditions: (a) H2NNH2 H2O, EtOH, 160 oC, mw, 1h, 81-88%;
·
(b) 1H-1,2,4-triazole, K2CO3, DMF, 150 oC, mw, 30 min, 67%; (c) KOH,
EtOH:H2O (1:1), 160 oC, mw, 30 min, 51%; (d) 10, propylphosphonic
anhydride (T3P®), Et3N, 150 oC, mw, 1h, 27-42%.
Table 1. Structure and mGlu7 NAM activities of analogs 8.
Cmpd
Het
mGlu7
mGlu7
% L-AP4 Min
IC50 (M)a
8a
8b
1.31
1.55
13.2
10.4
8c
>30
-
Figure 3. Diversity library assessing a wide array of heterocycles 7 as

dimethoxybenzohydrazide, propylphosphonic anhydride (T3P®), Et3N, 150
oC, mw, 1h, 22-62%.
8d
>30
-
8e
8f
>30
>30
4.38
-
-
Table 2. Structure and mGlu7 NAM activities of analogs 14.
8g
13.5
8h
8i
>30
>30
-
-
Cmpd
mGlu7
mGlu7
% L-AP4 Min
IC50 (M)a
8j
>30
>10
-
14a
14b
>30
-
8k
68.9
5.86
65.4
8l
>30
>10
-
14c
14d
1.18
>30
23.1
-
8m
72.8
aFor SAR determination, calcium mobilization assays with rat mGlu7/Gα15/HEK
cells performed in the presence of an EC80 fixed concentration of glutamate;
values represent (n = 1) independent experiments performed in triplicate.
14e
14f
0.57
2.04
6.3
50.6
As shown in Table 1, SAR amongst analogs 8 was very steep,
affording few active mGlu7 NAMs. Of the very few active
compounds, 8a (mGlu7 IC50 = 1.31 M, 13.2% L-AP4 Min) was a
1,3,4-oxadiazole analog of 6, and a cyclobutyl ether congener 8b of
5 was also of comparable activity (mGlu7 IC50 = 1.55 M, 10.4% L-
AP4 Min). Alternate functionality and/or incorporation of a
heteroatom (8g, 8j and 8k) all led to inactive or a significant
diminution in NAM activity. To compare with 7h, we evaluated the
DMPK profile of 8a. NAM 8a displayed very low predicted hepatic
clearance in rat (CLhep = 2.9 mL/min/kg) in combination with
favorable plasma protein binding (fu = 0.038). Moreover, a rat
plasma:brain level (PBL) study demonstrated that 8a was CNS
penetrant (Kp = 2.0, Kp,uu = 0.43). This positive disposition data
supported further optimization efforts to enhance mGlu7 NAM
potency, while hopefully, maintaining an attractive DMPK
profile.
aFor SAR determination, calcium mobilization assays with rat mGlu7/Gα15/HEK
cells performed in the presence of an EC80 fixed concentration of glutamate;
values represent (n = 1) independent experiments performed in triplicate.
As shown in Table 2, analogs 14 proved more tractable
towards producing mGlu7 NAMs. Saturated azacines, such as
pyrrolidine (14a) lacked mGlu7 NAM activity, but ring expansion
to the piperidine derivative (14b) afforded a weak partial NAM,
while a morpholine congener (14c) proved as potent (mGlu7 IC50
= 1.18 M, 23.1% L-AP4 Min) as 7h. In contrast, a basic
piperazine analog was inactive. Replacement of the 1,2,4-
triazole with an imidazole ring (i.e., 14e, VU6019278) was
highly successful, affording the most potent and efficacious
mGlu7 NAM (mGlu7 IC50 = 571 nM, 6.3% L-AP4 Min) discovered
during this campaign, and more potent than 5. When assayed in
triplicate, the potency of 14e further improved (IC50 = 501 nM,
pIC50 = 6.30±0.11, 6.3±1.0 L-AP4 min). Moreover, NAM 14e
displayed low predicted hepatic clearance in rat (CLhep = 27.7
mL/min/kg) in combination with favorable plasma protein binding
(fu = 0.10) and rat brain homogenate binding (fu = 0.013).
Finally, a rat plasma:brain level (PBL) study demonstrated that
14e was highly CNS penetrant (Kp = 4.9, Kp,uu = 0.65), offering
improvements over NAM 5.22 Finally, NAM 14e was also
selective versus the other seven mGlu receptors (>10 M vs.
mGlu1-6,8).
Based on these data, we elected to hold the 3,4-
dimethoxyphenyl moiety of 7h constant and surveyed saturated
and unsaturated nitrogen-containing heterocycles as replacements
for the 1H-1,2,4-triazole in N-linked analogs 14. Once again, an
SNAr with 11, and various saturated and unsaturated nitrogen-
containing heterocycles 15, provided N-linked derivatives 16.
Hydrolysis of the nitrile delivers acid 17, which undergoes a
T3P®-mediated microwave–assisted condensation with 3,4-
dimethoxybenzohydrazide to provide putative mGlu7 NAMs 14.
Scheme 2. Synthesis of putative mGlu7 NAM analogs 14.a
Since the N-linked southern azacines 14 proved attractive, we
elected to explore the viability of C-linked heterocyclic
congeners 18 as mGlu7 NAMs. Analogs 18 were readily
prepared in two steps from commercial benzoic acid 19 (Scheme
3). A T3P®-mediated microwave–assisted condensation with
3,4-dimethoxybenzohydrazide and 19 delivers 20 in acceptable
yield (43%). Next, a Suzuki coupling with various heteroaryl
boronic acids provides putative mGlu7 NAMs 18 in yields
ranging from 36-60%.
aReagents and conditions: (a) 11, K2CO3, DMF, 150 oC, mw, 30 min, 54-
Representative structures and activities for tricyclic analogs 18
are highlighted in Table 3. While not as productive as the N-
72%; (b) KOH, EtOH:H2O (1:1), 160 oC, mw, 30 min, 38-51%; (c) 3,4-

liked congeners 14, interesting SAR resulted. C-linked 6-
membered heterocycles, both pyridines (18a,b) and a pyridazine
(18c) were inactive. A 4-pyrazole derivative (18d) was a modest
mGlu7 NAM (mGlu7 IC50 = 2.24 M, 10.2% L-AP4 Min),
whereas the N-Me congener (18e) lost all mGlu7 NAM activity.
However, an N-Me, 5-pyrazole analog (18f), a regioisomer of
inactive 18e, proved to be the most potent mGlu7 NAM in this
subseries (mGlu7 IC50 = 1.32 M, 20.6% L-AP4 Min).
microsomal incubations (rat CLhep = 27.7 mL/min/kg) and high
CNS penetration (rat Kp = 4.9, Kp,uu = 0.65). Not only was 14e
more potent than 5, at the 0.2 mg//kg IV PBL dose, 14 achieve
total brain levels of 456 nM, comparable to the in vitro IC50 (501
nM), whereas 5 was relegated as an in vitro tool due to low total
brain concentrations. However, free brain levels for 14e from the
PBL cassette study, based on rat brain homogenate binding,
suggest ~6 nM free brain concentration at 0.2 mg/kg (if dose
linear, doses of 10-30 mg/kg would afford free brain levels of
14e above the in vitro IC50). NAM 14e represented an
improvement over the prototypical NAMs 1-5, and argued for
additional exploration of this subseries. Efforts in this vein are
underway, and results will be reported in due course.
Scheme 3. Synthesis of putative mGlu7 NAM analogs 18.a
Acknowledgments
We thank the Warren Family and Foundation for establishing the
William K. Warren, Jr. Chair in Medicine (C.W.L.). The authors
also acknowledge funding by CDMRP Grant W81XWH-17-1-
0266 (to C.M.N.).
aReagents
and
conditions:
(a)
3,4-dimethoxybenzohydrazide,
propylphosphonic anhydride (T3P®), Et3N, 150 oC, mw, 1h, 43%; (b) R-
B(OH)2, Cs2CO3, Pd(dppf)Cl2, 1,4-dioxane:H2O (9:1), 90 oC, 16h, 36-60%.
Table 3. Structure and mGlu7 NAM activities of analogs 18.
AUTHOR INFORMATION
Corresponding Authors
*C.M.N.: phone, 615-343-4303; fax, 615-343-3088; e-mail,
colleen.niswender@vanderbilt.edu.
*C.W.L.: phone, 615-322-8700; fax, 615-343-3088; e-mail,
craig.lindsley@vanderbilt.edu.
ORCID 279 Craig W. Lindsley: 0000-0003-0168-1445
Cmpd
R
mGlu7
mGlu7
% L-AP4 Min
IC50 (M)a
References
18a
>30
>10
-
(1) N.M. Fisher, M. Seto, C.W. Lindsley, et al. Front. Mol. Neurosci. 11
(2018) article 387.
18b
18c
27.9
(2) G. Sansig, T.J. Bushell, V.R. Clarke, et al. J. Neurosci. 21 (2001) 8734.
(3) H. Goddyn, Z. Callaerts-Vegh, S. Stroobants, et al. Neurobiol. Learn.
Mem. 90 (2008)103.
(4) A. Palucha, K. Klak, P. Branski, et al. Psychopharmacology 194 (2007),
>10
27.6
555.
(5) Z. Callaerts-Vegh, T. Beckers, S.M. Ball, et al. J. Neurosci. 26 (2006)
6573.
18d
18e
2.24
>30
10.3
-
(6) K. Mitsukawa, C. Mombereau, E. Lötscher, E.;
Neuropsychopharmacology 31 (2006) 1112.
(7) C. Hölscher, S. Schmid, P.K. Pilz, et al. Learn. Mem 12 (2005), 450.
et al.
(8) C. Hölscher, S. Schmid, P.K. Pilz, et al. Behav. Brain Res. 154 (2004),
473.
(9) T.J. Bushell, G. Sansig, V.J. Collett, et al. ScientificWorldJournal 2
18f
1.32
3.53
20.6
49.2
(2002), 730.
(10) M. Masugi, M. Yokoi, R. Shigemoto, et al. J. Neurosci. 19 (1999) 955.
(11) N.M. Fisher, R.G. Gogliotti, S.A.D. Vermudez, et al. ACS Chem.
18g
Neurosci. 9 (2018) 2210.
(12) R.G. Gogliotti, R.K. Senter, N.M. Fisher, et al. Sci. Trans. Med. 9
aFor SAR determination, calcium mobilization assays with rat mGlu7/Gα15/HEK
cells performed in the presence of an EC80 fixed concentration of glutamate;
values represent (n = 1) independent experiments performed in triplicate.
(2017) eaai7459.
(13) N. Jalan-Sakrikar, J.R. Field, R. Klar, et al. ACS Chem. Neurosci. 5
(2014) 1221.
(14) C.M. Niswender and P.J. Conn. Ann. Rev. Pharmacol. Toxicol. 50
(2010) 295.
In summary, we detailed a diversity-oriented library approach
to rapidly assess diverse heterocycles as bioisosteric
replacements for a metabolically labile amide moiety within a
series of mGlu7 negative allosteric modulators (NAMs),
exemplified by 5. SAR rapidly identified a 1,3,4-oxadiazole ring
system as an effective bioisostere for the amide linker. Further
optimization of the southern heterocycle of this new chemotype
led to the discovery of VU6019278 (14e), a potent mGlu7 NAM
(IC50 = 501 nM, 6.3% L-AP4 Min) with favorable plasma protein
binding (rat fu= 0.10), low predicted hepatic clearance in
(15) C.W. Lindsley, K.A. Emmitte, C.R. Hopkins, et al. Chem. Rev. 116
(2016) 6707.
(16) M. Abe, M. Seto, R.G. Gogliotti, et al. ACS Med. Chem. Lett. 8 (2017)
1110.
(17) G. Suzuki, N. Tsukamoto, H. Fushiki, et al. J. Pharmacol. Exp. Ther.
232 (2007) 147.
(18) M. Nakamura, H. Kurihara, G. Suzuki, et al. Bioorg. Med. Chem. Lett.
20 (2010) 726.
(19) M. Kalinichev, M. Rouillier, F. Girard, et al. J. Pharmacol. Exp. Ther.
344 (2013) 624.

(20) C.E. Gee, D. Peterlik, C. Neuhäuser J. Biol. Chem. 289 (2014) 10975.
(21) C.M. Niswender, K.A. Johnson, N.R. Miller, et al. Mol. Pharmacol. 77
(2010) 77, 459.
(22) C.W. Reed, K.M. McGowan, P.K. et al. ACS Med. Chem. Lett. 8 (2017)
1326.
(23) C.W. Reed, S. E. Yohn, J.P. Washecheck, et al. J. Med. Chem. doi:
10.1021/acs.jmedchem.8b018810
(24) J.M. Rook, J.L. Bertron, H.P. Cho, et al. ACS Chem. Neurosci. 9 (2018)
2274.

HIGHLIGHTS



Novel series of selective and CNS
penetrant mGlu7 NAMs
Identified a 1,3,4-oxadiazole as
biosiostere for a labile amide linker
Iterative libraries quickly optimized this
novel tricyclic series.