3-(Imidazolyl methyl)-3-aza-bicyclo[3.1.0]hexan-6-yl)methyl ethers: A novel series of mGluR2 positive allosteric modulators

Article abstract of DOI:10.1016/j.bmcl.2008.09.026

The synthesis and structure-activity relationship (SAR) of a novel series of 3-(imidazolyl methyl)-3-aza-bicyclo[3.1.0]hexan-6-yl)methyl ethers, derived from a high throughput screening (HTS), are described. Subsequent optimization led to identification of potent, metabolically stable and orally available mGluR2 positive allosteric modulators (PAMs).

Full text of DOI:10.1016/j.bmcl.2008.09.026

Bioorganic & Medicinal Chemistry Letters 18 (2008) 5493–5496
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
3-(Imidazolyl methyl)-3-aza-bicyclo[3.1.0]hexan-6-yl)methyl ethers: A novel
series of mGluR2 positive allosteric modulators
*
Lei Zhang , Bruce N. Rogers, Allen J. Duplantier, Stanley F. McHardy, Ivan Efremov, Helen Berke, Weimin
Qian, Andy Q. Zhang, Noha Maklad, John Candler, Angela C. Doran, John T. Lazzaro Jr., Alan H. Ganong
Pfizer Global Research and Development, Groton Laboratories, Eastern Point Road, Groton, CT 06340, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 July 2008
Revised 3 September 2008
Accepted 5 September 2008
Available online 10 September 2008
The synthesis and structure–activity relationship (SAR) of a novel series of 3-(imidazolyl methyl)-3-aza-
bicyclo[3.1.0]hexan-6-yl)methyl ethers, derived from a high throughput screening (HTS), are described.
Subsequent optimization led to identification of potent, metabolically stable and orally available mGluR2
positive allosteric modulators (PAMs).
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
mGluR2
Positive allosteric modulators
Potentiators
Anxiety
Schizophrenia
The metabotropic glutamate receptors (mGluRs) belong to fam-
ily C of GPCRs characterized by a large extracellular bi-lobed ago-
nist binding site connected to a heptahelical transmembrane
(7TM) domain through a cysteine-rich linker.1 These receptors
are typically subdivided into three groups based on structural
homology and signal transduction pathways.2 Group II are com-
prised of mGluR2 and mGluR3 which are both negatively coupled
to adenylate cyclase. Efficacy observed in preclinical animal mod-
els and human clinical trials with known dual mGluR2/3 agonists
(e.g., LY-354740) provides solid support of group II mGluRs as no-
vel targets for the treatment of anxiety,3 psychosis,4 convulsive
disorders,5 Parkinson’s disease,6 neurodegeneration,7 and pain.8
Tremendous excitement in this field has been generated since
the report that LY 2140023, a pro-drug of mGluR2/3 agonist LY-
404039, demonstrated efficacy against positive and negative
symptoms in a 4-week phase IIb schizophrenia trial.9
Due to the high degree of homology in the orthosteric sites of
group II mGlu receptors, selective mGluR2 agonists have yet to
be discovered. Moreover, known direct agonists are typically ami-
no acid analogs,10 which can limit their utility in pharmacological
studies due to poor physicochemical properties and low CNS expo-
sure. Positive allosteric modulators (PAMs), in comparison, bind in
the transmembrane domain where there is much less homology
among different subtypes, thus providing an opportunity to
achieve subtype selectivity. Furthermore, the large number of
structural diverse, non-amino acid allosteric modulators of mGluRs
found to date11 hold promise of identifying mGluR2 PAMs with
suitable physicochemical properties for CNS exposure. Herein the
discovery and structure–activity relationship of a novel series of
mGluR2 PAMs, 3-(imidazolyl methyl)-3-aza-bicyclo[3.1.0]hexan-
6-yl)methyl ethers, are described.
An in-house high throughput screening campaign using a func-
tional FLIPR assay12 yielded several leads featuring a [3.1.0] azabi-
cyclic core, as represented by compound 1 (Fig. 1). Compound 1
displays potent functional activity (mGluR2 EC50 = 24 nM) and
good brain penetration (brain/plasma = 2).13 However, further
pharmacokinetic (PK) profiling revealed high in vitro clearance in
rat and human liver microsomes (r-CLh: >65.7 ml/min/kg; h-CLh:
14.2 ml/min/kg).14 Moreover, compound 1 has low absorption
(5%) and oral bioavailability (<2%),15 potentially due to poor aque-
ous solubility of 1 observed during sample formulation. In light of
these results, our chemistry efforts were focused on developing a
structure–activity relationship around the 3-aza-bicyclo[3.1.0]hex-
H
O
N
N
H
N
1
* Corresponding author. Tel.: +1 860 441 4116; fax: +1 860 715 2350.
E-mail address: Lei.Zhang3@pfizer.com (L. Zhang).
Figure 1. The original HTS lead, compound 1.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.09.026

5494
L. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5493–5496
Table 1
an-6-yl)methyl ether template to identify compounds with im-
proved PK characteristics.
Functional activities and in vitro human microsomal clearance data of representative
mGluR2 PAMs
A general synthetic strategy of 3-aza-bicyclo[3.1.0]hexan-6-yl)-
methyl ether analogs is illustrated in Scheme 1. Starting from N-
benzylmaleimide 2, room temperature cycloaddition with ethyl
diazoacetate followed by thermolysis16 gave cyclopropane 3,
which upon reduction yielded amino alcohol 4. Benzyl removal
via standard hydrogenation with Pd/C and subsequent Boc-protec-
tion afforded the key Boc amine intermediate 5. A facile three-step
protocol involving a Mitsunobu reaction, Boc removal and a reduc-
tive amination reaction provided rapid access to the desired prod-
ucts 6 bearing structural variations in the R1 and R2 positions.
Preliminary SAR at the R2 position of 6 (Table 1) revealed N-
methyl benzimidazole (MBI) as a key structural moiety for mGluR2
activity. Replacing MBI with alternative heteroaryls led to signifi-
cant loss of potency, even with structurally similar benzothiazole
(6a), imidazopyridine (6b), and quinoline (6c). N-methyl appeared
to be the optimal side chain as a bulkier cyclopropyl group (6d) or
a tricyclic benzimidazole moiety (6e) resulted in much lower
mGluR2 activities. On the other hand, a range of substitutions on
the terminal phenyl ring moiety of compound 1, were well toler-
ated (6f–k). Simple substituents such as halogens and small alkyls
were preferred, yielding compounds (6g–k) with improved or
comparable in vitro potency to compound 1. Subsequent pharma-
cokinetic evaluation revealed that the incorporation of para-sub-
stituents (6h–k) dramatically improved microsomal stability. In
particular, para-CF3 substitution (6k) led to a low turnover rate
in human liver microsome (h-CLh <5.3 ml/min/kg), albeit with a
threefold drop in potency.
With clearance issues effectively addressed, we turned our
attention to the low oral bioavailability associated with compound
1. Compounds such as 6g and 6f were found to have the poor aque-
ous solubility similar to that of compound 1. To address solubility
as a component of bioavailability, we explored lowering cLogP
through incorporation of pyridyl nitrogens within the R1 and R2
moieties of 6.
Pyridyl ether analogs (14a–c, Table 2) were synthesized
through the three-step protocol described in Scheme 1 using com-
mercially available hydroxyl-pyridines. The syntheses of N-methyl
aza-benzimidazoles are illustrated in Scheme 2. Starting from com-
mercially available nitropyridine 7a, the key aldehyde intermedi-
ate 9a was obtained in four steps through a cycloaddition of
Compound
R1
R2
mGluR2
h-CLh
EC50 (nM)a
(ml/min/kg)b
N
N
1
27
14.2
N
6a
6b
>10,000
>10,000
<5.3
NDc
S
N
N
6c
9750
3020
<5.3
N
N
6d
ND
N
N
N
6e
6f
6g
6h
6i
967
113
7
17.3
11
N
Ph
N
F
N
N
N
N
N
>18.7
5.8
N
N
N
N
N
F
26
26
8
OH
Cl
CO2Et
6.6
H
H
a, b
H
H
c
O
O
N
O
O
N
N
F
Cl
6j
6.9
4
3
2
CF3
R1
6k
85
<5.3
O
OH
f, g, h
d, e
H
H
H
H
a
EC50 values obtained from mGluR2 receptor transfected HEK cell line using
FLIPR method in the presence of glutamate (EC10–20); means of at least three
experiments; an EC50 value >10,000 indicates that no curve was noted in the dose–
N
N
response up to 10
l
M.
Predicted hepatic clearance (h-CLh) from human liver microsomal stability
assay.
Boc
R2
b
5
6
c
ND, not determined.
Scheme 1. Reagent and conditions: (a) ethyl diazoacetate, Et2O, rt; (b) neat, 200–
230 °C, 36% for two steps; (c) LiAlH4, THF, reflux; (d) Pd/C (10%), H2 (40 psi), MeOH,
rt; (e) Et3N, Boc2O, CH2Cl2, 74% for three steps; (f) R1-OH, PPh3, DEAD, THF, rt; (g)
TFA, CH2Cl2; (h) R2-CHO, Et3N, MgSO4, then Na(OAc)3BH, CH2Cl2.
diamino pyridine intermediate 8 and diethoxy ethanimidate 13.
Following a similar synthetic sequence, the three other regioisom-

L. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5493–5496
5495
Table 2
N
H2N
N
O2N
N
Functional activities and human microsomal clearance data of pyridyl ether and aza-
methylbenzimidazole analogs
a, b
Br
N
H
MeO
Compound R1
R2
mGluR2 EC50 h-CLh (ml/
NO2
(nM)a
min/kg)b
8
7a
7c
N
N
c, d
1
27
14.2
N
N
X1
X4
O
H
N
X2
X3
N
O2N
Cl
N
Cl
N
14a
14b
14c
15
26
511
30
12.3
NDc
<5.3
9.3
9a: X1 =N, X2, X3, X4 =CH
9b: X2 =N, X1, X3, X4 =CH
9c: X3 =N, X1, X2, X4 =CH
9d: X4 =N, X1, X2, X3 =CH
O2N
N
Cl
7d
Cl
7b
N
N
N
N
N
N
N
R1
O
R1
O
CF3
CF3
CF3
f
N
H
H
H
R2=
H
X1
X2
X3
N
N
N
X4
N
H
R2
N
11
20
10
HN
OEt
OEt
OEt
OEt
e
N
N
NC
MeO
16
251
2120
419
64
10.5
11.5
ND
N
N
N
13
12
Scheme 2. Reagent and conditions: (a) MeNH2, EtOH, microwave (105 °C, 4 h), 57%;
(b) Pd/C (10%), DME/MeOH (1:1), H2 (40 psi), rt, 97%; (c) 13, DME, AcOH, rt, then
catalytic amount TsOH, reflux, 39%; (d) 4 N HCl aq, 60 °C, 96%; (e) Na, MeOH, rt,
68%; (f) 9a–d, MgSO4, Et3N, CH2Cl2, then Na(OAc)3BH.
17
N
N
CF3
compound 1 (EC50 = 30 nM). Incorporation of N-methyl aza-benz-
imidazole proved to be a fruitful SAR direction. The 4-aza and 5-
aza-analogs, 15 and 16 (Scheme 2), displayed good potency, while
compounds derived from the other two regioisomers (17, 18) were
generally much weaker. Interestingly, these aza-benzimidazole
moieties yielded compounds generally with low to moderate clear-
ances, even in the absence of para-substituents on the top phenyl
ring (19, 20).
Based on the in vitro microsomal clearance predictions, the
in vivo pharmacokinetics of compounds 14a, 14c, and 19 were
evaluated in jugular and portal vein cannulated rats. Improve-
ments in clearance and fraction absorbed contributed to the
improvement in bioavailability (Table 3). As an example, com-
pound 14c has a rat in vivo clearance of 28 ml/min/kg compared
to 67 ml/min/kg for compound 1.18 Excellent absorption (70–90%,
18
N
N
Cl
N
N
N
19
10.5
<5.3
N
N
Cl
20
86
N
a
EC50 values obtained from mGluR2 receptor transfected HEK cell line using
FLIPR method in the presence of glutamate (EC10–20); means of at least three
experiments.
b
Predicted hepatic clearance (h-CLh) from human liver microsomal stability
assay.
c
ND, not determined.
Table 3
Rat in vivo PK evaluation
ers 9b–d were synthesized from commercially available nitropyri-
dines 7b–d, respectively. Reductive amination of 9a–d with sec-
ondary amine 10 gave the desired products 11 bearing N-methyl
aza-benzimidazoles, exemplified by compounds 15–20 in Table 2.
To this end, the phenyl moiety of 6i and 6k was replaced with
pyridyl groups (Table 2, 14a–c). 2-Pyridyl was preferred as more
than a 20-fold activity drop was observed when moving the nitro-
gen to the 3-position (14a–14b). Low to moderate predicted
clearances were maintained in pyridyl ethers with para-substitu-
ents. 14c was stable in the presence of human microsome
(h-CLh <5.3 ml/min/kg) and provided mGluR2 potency similar to
Compound
In vivo r-CLha
(ml/min/kg)
Fraction absorbedb
(%)
Oral bioavailabilityc
(%)
1
67
59
28
28
5
72
—
<2
20
27
79
14a
14c
19
84
a
Observed plasma clearance in rats following
a
single 1 mg/kg iv bolus
administration.
b
Fraction absorbed determined using portal vein concentrations from rats fol-
lowing 5 mg/kg oral dose and clearance method. Fraction absorbed = Clb*AUCpo,pv
/
17
Dosepo
.
c
Determined from oral pharmacokinetics in rats following 5 mg/kg oral dose.

5496
L. Zhang et al. / Bioorg. Med. Chem. Lett. 18 (2008) 5493–5496
8. Jones, C. K.; Eberle, E. L.; Peters, S. C.; Monn, J. A.; Shannon, H. E.
Neuropharmacology 2005, 49, 206.
9. Patil, S. T.; Zhang, L.; Martenyi, F.; Lowe, S. L.; Jackson, K. A.; Andreev, B. V.;
Avedisova, A. S.; Bardenstein, L. M.; Gurovich, I. Y.; Morozova, M. A.; Mosolov, S.
N.; Neznanov, N. G.; Reznik, A. M.; Smulevich, A. B.; Tochilov, V. A.; Johnson, B.
G.; Monn, J. A.; Schoepp, D. D. Nat. Med. 2007, 13, 1102.
10. For a recent review, see: Recasens, M.; Guiramand, J.; Aimar, R.; Abdulkarium,
A.; Barbanel, G. Curr. Drug Targets 2007, 8, 651.
11. For recent reviews, see: (a) Gasparini, F.; Spooren, W. Curr. Neuropharmacol.
2007, 5, 187; (b) Rudd, M. T.; McCauley, J. A. Curr. Top. Med. Chem. 2005, 5, 869.
12. The mGluR2 FLIPR functional assay uses a stable HEK cell line expressing the
14a and 19) together with low to moderate in vivo clearances led
to good oral bioavailability (20–79%), a pronounced improvement
over the original lead, compound 1.
In conclusion, a series of 3-(imidazolyl methyl)-3-aza-bicy-
clo[3.1.0]hexan-6-yl)methyl ethers has been developed as novel
mGluR2 positive allosteric modulators. The high clearance and
low bioavailability issues of the original lead have been success-
fully addressed by introducing para-substitutions and incorporat-
ing pyridyl nitrogens. A group of lead compounds (14a, 14c, and
19) have been identified with excellent potency and pharmacoki-
netic profiles. These compounds have been profiled in in vivo effi-
cacy models and the results will be reported in due course.
rat mGluR2 receptor and G
a15 G protein. Cells are plated in 384 well poly-D-
lysine coated plates and incubated in glutamine free medium. Cells are loaded
with Fluo-4 am containing Glutamic Pyruvic transaminase to reduce
endogenous glutamate. Compound is added with the FLIPR, incubated at
room temperature approximately 30 min, then tested against an approximate
EC10 to EC20 glutamate challenge. Data is analyzed as a ratio of the peak
fluorescence to the baseline after compound addition. Assay to assay
variability in PAM potency required multiple experiments to determine
potency, typically n > 3.
Acknowledgments
The authors thank Joel Schachter and Diane Lang for construct-
ing cells and Prof. Shigetada Nakanishi of Kyoto University for pro-
viding cDNA.
13. Whole brain and blood samples were collected from Sprague–Dawley rats at
various times following administration of a single 10 mg/kg subcutaneous
dose. Plasma and brain homogenate samples were quantitated using an LC/MS
method and the brain to plasma ratio reported.
14. Compound 1 (1 lM) was incubated in rat or human liver microsome and
monitored for substrate depletion using an LC/MS method. Hepatic clearance
was determined by scaling of the observed half-life using the well stirred
model.
References and notes
1. Pellicciari, R.; Constantino, G. Curr. Opin. Chem. Biol. 1999, 3, 433.
2. Conn, P. J.; Pin, J.-P. Annu. Rev. Pharmocol. Toxicol. 1997, 37, 205.
3. Levine, L.; Gaydos, B.; Sheehan, D.; Goddard, A. W.; Feighner, J.; Potter, W. Z.;
Schoepp, D. D. Neuropharmacology 2002, 43, 294.
4. Spooren, W. P. J. M.; Gasparini, F.; van der Putten, H.; Koller, M.; Nakanichi, S.;
Kuhn, R. Eur. J. Pharmacol. 2000, 397, R1.
5. Klodzinska, A.; Chojnacka-Wojcik, E.; Pilc, A. Pol. J. Pharmacol. 1999, 51, 543.
6. Kinoeczny, J.; Ossowaska, K.; Wolfarth, S.; Pilc, A. Nauyn-Schmiedeberg’s Arch.
Pharmacol. 1998, 358, 500.
7. Kingston, A. E.; O’Neill, M. J.; Lam, A.; Bales, K. R.; Monn, J. A.; Schoepp, D. D.
Eur. J. Phamacol. 1999, 377, 155.
15. Pharmacokinetic parameters were determined in rats following a 1 mg/kg
intravenous or 5 mg/kg oral administration to male Sprague–Dawley rats.
Whole blood was collected from a jugular (iv and oral) or portal vein (oral only)
catheter and plasma samples analyzed using an LC/MS method.
Pharmacokinetic parameters were calculated using
analysis of the plasma concentration versus time data.
16. Brighty, K. E.; Castaldi, M. J. Synlett 1996, 11, 1097.
17. Kwon, Y. Pharmacol. Res. 1996, 13, 566.
a non-compartmental
18. In vitro rat microsomal stability r-CLh of 14c is 41 ml/min/kg, consistent with
the in vivo measurement (within two folds).