Discovery and biological profile of isoindolinone derivatives as novel metabotropic glutamate receptor 1 antagonists: A potential treatment for psychotic disorders

Article abstract of DOI:10.1016/j.bmcl.2009.07.145

We describe here the discovery and biological profile of a series of isoindolinone derivatives as developed mGluR1 antagonists. Our combined strategy of rapid parallel synthesis and conventional medicinal optimization successfully led to N-cyclopropyl 22 and N-isopropyl isoindolinone analogs 21 and 23 with improved in vivo DMPK profiles. Moreover the most advanced analog 23 showed an oral antipsychotic-like effect at a dose of 1 mg/kg in an animal model.

Full text of DOI:10.1016/j.bmcl.2009.07.145

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5310–5313
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery and biological profile of isoindolinone derivatives as novel
metabotropic glutamate receptor 1 antagonists: A potential treatment
for psychotic disorders
*
Satoru Ito , Yukari Hirata, Yasushi Nagatomi, Atsushi Satoh, Gentaroh Suzuki, Toshifumi Kimura,
Akio Satow, Shunsuke Maehara, Hirohiko Hikichi, Mikiko Hata, Hisashi Ohta, Hiroshi Kawamoto
Tsukuba Research Institute, Banyu Pharmaceutical CO., Ltd, 3 Okubo, Tsukuba 300-2611, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We describe here the discovery and biological profile of a series of isoindolinone derivatives as developed
mGluR1 antagonists. Our combined strategy of rapid parallel synthesis and conventional medicinal opti-
mization successfully led to N-cyclopropyl 22 and N-isopropyl isoindolinone analogs 21 and 23 with
improved in vivo DMPK profiles. Moreover the most advanced analog 23 showed an oral antipsy-
chotic-like effect at a dose of 1 mg/kg in an animal model.
Received 2 April 2009
Revised 19 July 2009
Accepted 30 July 2009
Available online 3 August 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
mGluR1
Antagonist
Glutamate
Psychotic disorder
Glutamate is one of the major excitatory neurotransmitters in
the central nervous system (CNS) and it acts on ionotropic gluta-
mate receptors including NMDA and non-NMDA receptors and
on G-protein coupled metabotropic glutamate receptors (mGluRs).
The mGluRs are classified into eight subtypes (three subclasses)
based on sequence homology, coupling mechanisms to G-protein,
and pharmacological properties. mGluRs are considered to be drug
targets for modulating glutamate transmission in the treatment of
various neurological and psychiatric diseases including pain, epi-
lepsy, Parkinson’s disease, cognitive disorders, drug abuse, anxiety,
ture at the left hand part was required to develop a more appropri-
ate mGluR1 antagonist for entering pre-clinical phase.
At first, we decided to construct a 1-phenyl-5-methyl-1,2,3-tri-
azole library to explore the SAR of the left hand part of the mole-
cule, which was a metabolic soft spot in lead compound 1. The
fluoropyridine unit at the right hand part was tentatively replaced
with a synthetically feasible phenyl group. Both important starting
materials were prepared according to reported methods.9 Most
of the coupling partners were picked up from the company reagent
,10
and the others were prepared by ourselves according to the litera-
1
–6
11,12
and schizophrenia.
ture.
This library was rapidly prepared in two different man-
In our previous work, we demonstrated that a potent and selec-
tive mGluR1 allosteric antagonist, FTIDC 1, inhibited psychostimu-
lant methamphetamine (MAP)-induced behavioral alterations such
as hyperlocomotion and disruption of prepulse inhibition (PPI) at
intraperitoneal doses of 10 and 30 mg/kg.7 These results suggest
that blockage of mGluR1 mimics some effects of antipsychotics
and that mGluR1 antagonists have therapeutic potential for the
treatment of psychotic disorders. However, FTIDC 1 itself was not
taken forward due to its unacceptable DMPK profile, namely its
insufficiently short half-life (0.2 h) and oral bioavailability (18%)
ners as described in Scheme 1.
The compounds listed here were tested for antagonistic activity
2+
on human mGluR1a expressing CHO cells by measuring [Ca ] i
7
with a FLIPR. The assay results are summarized in Table 1. Resem-
,8
bling lead compound 1, many of the carbonylated benzene analogs
showed moderate mGluR1 antagonistic activity. Substitution of the
9
in rats. A major metabolite in rat hepatocytes was obtained as
des-methylated analog 2 (Fig. 1). Therefore, an alternative struc-
*
Corresponding author. Tel.: +81 29 877 2000; fax: +81 29 877 2029.
E-mail address: satoru_ito@merck.com (S. Ito).
Figure 1. Lead compound and its major metabolite.
0
960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.07.145

S. Ito et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5310–5313
5311
Table 1 (continued)
No.
R
hmGluR1
IC50 ± SEMa (nM)
1
1
1
1
1
3
4
5
6
7
11 ± 1.5
2.6 ± 0.12
250 ± 35
Scheme 1. Preparation of the triazole library. Reagents and conditions: (a) Ar–Br or
I, Pd(PPh
3
4
) , DMF, 115 °C; (b) Ar–B(OH)
2 2
, PdCl (dppf)
, K
2 2
CO
3
, DMF 80 °C.
Table 1
In vitro antagonistic activity of the triazole library
4.9 ± 1.1
300 ± 140
1500 ± 240
No.
R
hmGluR1
IC50 ± SEM (nM)
a
3
470 ± 91
18
a
The IC50 value is the mean of multiple results (at least three independent
determinations performed in duplicate) with standard error of the means.
4
5
170 ± 44
22 ± 2.6
carbonyl group at the
a position was quite important to enhance
the activity. As shown in Table 1, acetophenone analog 3 only
showed 470 nM mGluR1 antagonistic activity, but replacement of
methyl group of acetophenone with ethyl and cyclopropyl groups
significantly improved their activities (170 and 22 nM). In contrast,
benzophenone derivative 6 resulted in a complete loss of activity,
which meant that relatively small substitutions were tolerable.
para-Substituted isopropyl ester analog 7 showed 33 nM antago-
nistic activity, but the corresponding meta analog 8 had decreased
activity (2400 nM). Conversion of the ester to an amide did not
have a positive effect in terms of mGluR1 antagonistic activity
6
3900 ± 2100
7
8
9
33 ± 5.9
2400 ± 770
210 ± 33
(9–11). Cyclized compounds, indanes 12, and 13, and the isoindoli-
none derivative 14, had significantly improved activity. Quinoline
derivative 16 also showed 4.9 nM activity, but no activity was ob-
served in naphthalene analog 17. These results suggested that a
bicyclic ring system having a kind of hydrogen acceptor, such as
a carbonyl group or a nitrogen atom, was necessary to elicit single
digit mGluR1 antagonist activity in this lead class.
We rapidly identified two unique and potent hits, isoindolinone
1
4 and quinoline 16, from this library. At first the isoindolinone
analog 14 was given priority for further chemical modification on
account of better in vitro mGluR1 activity and synthetic feasibility.
We next investigated the influence of isoindolinone N substitu-
tions on mGluR1 antagonistic activity. The synthetic scheme is
shown in Scheme 2. All coupling precursors were prepared by
the same method according to a literature. The target molecules
were obtained with corresponding bromides and a stannous
reagent by using the Stille reaction. The N-aryl part was again
replaced with a fluoropyridine for preferable lipophilicity.
1
1
0
1
880 ± 280
1
1
1800 ± 440
Group I mGluRs include mGluR1 and mGluR5; thus, the synthe-
sized compounds were also evaluated on human mGluR5 to con-
firm their subtype selectivity. As a result, many isoindolinone
derivatives showed single digit mGluR1 antagonistic activity, suffi-
cient selectivity for mGluR5 (over 300-fold), and also a preferable
1
2
69 ± 12

5
312
S. Ito et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5310–5313
Scheme 2. Preparation of isoindolinone derivatives. Reagents and conditions: (a) concd H
NH , TEA, toluene, reflux; (d) Pd(PPh , DMF, 115 °C.
2 4 4
SO , MeOH, reflux; (b) N-bromosuccinimide, benzoylperoxide, CCl , reflux; (c) R–
2
3 4
)
Table 2
mGluRs activities and rat PK profiles of isoindolinone derivatives
No.
R
hmGluR1
hmGluR5
Log D7.4b
Rat pharmacokineticsc
IC50 ± SEMa (nM)
IC50 ± SEMa (nM)
F (%)
T1/2 (h)
CLp (mL/min/kg)
1
5.6 ± 1.4
3.3 ± 1.2
>10,000
2.1
18
0.2
0.3
57
19
20
21
22
1100 ± 140
2800 ± 1300
2300 ± 670
5400 ± 1200
1.8
62
100
96
19
3.6 ± 0.94
5.5 ± 1.1
3.5 ± 1.5
2.3
2.2
1.9
1.0
2.2
0.6
5.5
5.9
8.1
100
a
b
c
The IC50 value is the mean of multiple results (at least three independent determinations performed in duplicate) with standard error of the means.
The log D7.4 value was measured using a reported method.
The oral dose was 3 mg/kg and the IV dose 1 mg/kg.
14
log D7.4 value (1–3) for a targeting oral CNS drug.13 In addition, all
isoindolinones listed here had significantly improved bioavailabil-
ity and total clearance compared to lead compound 1 in rats. This
result suggested that the isoindolinone was much suited substruc-
ture for developing orally available mGluR1 antagonists in this
class.
In the course of further characterization of this series, however,
compound 22 generated a GSH adduct at the fluoropyridine part in
rat hepatocytes incubation. In general, a GSH trapping experiment
was used as a surrogate marker assay to help in the structure
metabolism relationship studies aimed at minimizing drug–pro-
tein covalent binding.15,16 The formation of this GSH adduct indi-
cated that this lead class containing the fluoropyridine unit had
the possibility to bind irreversibly to biomolecules such as protein
1
6,17
or DNA, giving rise to drug induced toxicities.
It could be de-
pend on the potential electrophilicity of the fluoropyridine
structure.
We previously reported that oral mGluR1 antagonistic activity
and also metabolic stability of this class are highly dependent on
9
their lipophilicities. The fluoropyridine part of the tetrahydropyr-
idine leads had the great benefit of adjusting their log D7.4 value
without sacrificing potency. As shown in Table 2, most of the dis-
covered isoindolinones had a slight amount of room to increase
lipophilicity for maintaining a log D7.4 value of 1–3, the reported

S. Ito et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5310–5313
5313
Table 3
Profiles of the most advanced isoindolinone analog 23
In vitro profiles
Group 1
In vivo profiles
Brain penetrabilityb
Mouse brain/plasma concd 0.45 nmol/g/0.17 lM
hmGluR1 (IC50 a
)
4.3 ± 0.66 nM
3.6 ± 0.68 nM
1500 ± 170 nM
a
a
hmGluR1 (IC50
hmGluR5 (IC50
)
)
Pharmacokineticsc
Group 2
Rat F: 46%, T1/2 0.7 h, CLp: 20 mL/min/kg
hmGluR2 (IC50)a >10,000 nM
Group 3
Efficacyd
hmGluR8 (IC50)a >10,000 nM
Rat PPI disruption model MED 1 mg/kg, po
Adverse effectd
Quisqualic acid binding site (IC50) >10,000 nM
Rat catalepsy. No effects 30 mg/kg, po
a
b
c
The IC50 value is the mean of multiple results (at least three independent determinations performed in duplicate) with standard error of the means.
At 30 min after oral administration (1 mg/kg).
The oral dose was 1 mg/kg and the IV dose 0.3 mg/kg.
Ref. 8.
d
optimum value for oral CNS drugs,13 and also good metabolic sta-
bility. We promptly replaced the fluoropyridine part of the potent
compounds with promising benzene substructures, which were
well investigated in the tetrahydropyridine prototype lead class.
Among the combined compounds, we finally identified com-
pound 23 as a candidate for further development (Table 3). Com-
pound 23 had a 2,4-difluorobenzene at the triazole N part in the
molecule and its log D7.4 value was 3.1. GSH adducts were not ob-
served at all in rat hepatocytes incubation. In vitro mGluR1 antag-
onistic activities for human and rat were 4.3 and 3.6 nM,
respectively, and there were no species differences. Selectivity for
other mGluRs was sufficient and it did not have any affinity for
the quisqualic binding site, which meant that it is an allosteric
antagonist. In terms of the in vivo DMPK profile, following an oral
dose of 1 mg/kg of compound 23, the mean (n = 3) plasma and
activity, and also an improved PK profile in rats. Moreover, it also
demonstrated an antipsychotic-like effect in an animal model.
The other isoindolinone derivatives 19–22 with a fluoropyridine
part also have quite potent mGluR1 antagonistic activity and in-
creased hydrophilic profiles and would be suitable for develop-
ment of a PET tracer to examine the in vivo pharmacodynamics
of mGluR1 antagonists. These isoindolinone derivatives would
have great potential for the elucidation of the functions of mGluR1
in humans.
1
8
References and notes
1.
Swanson, C. J.; Bures, M.; Johnson, M. P.; Linden, A. M.; Monn, J. A.; Schoepp, D.
D. Nat. Rev. Drug Disc. 2005, 4, 131.
2
3
4
.
.
.
Kew, J. N. C. Pharmacol. Ther. 2004, 104, 233.
Marek, G. Curr. Med. Chem.-Cent. Nerv. Syst. Agents 2002, 2, 29.
Marino, M. J.; Conn, P. J. Curr. Opin. Pharmacol. 2006, 6, 98.
brain concentrations in mice were 0.17
lM and 0.45 nmol/g at
3
0 min post-dose, respectively. The brain level of the compound
5. Kohara, A.; Takahashi, M.; Yatsugi, S. I.; Tamura, S.; Shitaka, Y.; Hayashibe, S.;
Kawabata, S.; Okada, M. Brain Res. 2008, 1191, 168.
after oral administration seemed to be good enough to elicit
in vivo efficacy given its great in vitro intrinsic potency. In addition,
it maintained an acceptable oral bioavailability in rats (46%).
To reiterate, we previously demonstrated that blockage of
mGluR1 mimics some effects of antipsychotics. Compound 1 inhib-
ited psychostimulant methamphetamine (MAP)-induced behav-
ioral alterations such as hyperlocomotion and disruption of
6
7
.
.
Schkeryantz, J. M.; Kingston, A. E.; Johnson, M. P. J. Med. Chem. 2007, 50, 2563.
Suzuki, G.; Kimura, T.; Satow, A.; Kaneko, N.; Fukuda, J.; Hikichi, H.; Sakai, N.;
Maehara, S.; Kawagoe-Takaki, H.; Hata, M.; Azuma, T.; Ito, S.; Kawamoto, H.;
Ohta, H. J. Pharmacol. Exp. Ther. 2007, 321, 1144.
8.
Satow, A.; Maehara, S.; Ise, S.; Hikichi, H.; Fukushima, M.; Suzuki, G.; Kimura,
T.; Tanaka, T.; Ito, S.; Kawamoto, H.; Ohta, H. J. Pharmacol. Exp. Ther. 2008, 326,
577.
9. Ito, S.; Satoh, A.; Nagatomi, Y.; Hirata, Y.; Suzuki, G.; Kimura, T.; Satow, A.;
Maehara, S.; Hikichi, H.; Hata, M.; Kawamoto, H.; Ohta, H. Bioorg. Med. Chem.
prepulse inhibition (PPI) at intraperitoneal doses of 10 and
2008, 16, 9817.
8
3
0 mg/kg. We next tested the developed compound 23 in PPI as-
10. Sakamoto, T.; Uchiyama, D.; Kondo, Y.; Yamanaka, H. Heterocycles 1993, 35,
1273.
say system to confirm its potential for the development of an orally
available antipsychotic. As a result, an antipsychotic-like effect was
observed from an oral dose of 1 mg/kg in rats. On the other hand,
we previously examined haloperidol, a typical marketed antipsy-
chotic, in this PPI efficacy model and a catalepsy adverse effect
model in rats. The minimum effective dose of the efficacy is
1
1. Grey, J.; Dyck, B.; Rowbottom, M. W.; Tamiya, J.; Vickers, T. D.; Zhang, M.; Zhao,
L.; Heise, C. E.; Schwarz, D.; Saunders, J.; Goodfellow, V. S. Bioorg. Med. Chem.
Lett. 2005, 15, 999.
12. Almansa, R. C.; Virgili, B. M. WO 2007000337, 2006; Chem. Abstr. 2006, 146,
21696.
1
1
3. John, E. A. C. In Drug Bioavailability: Estimation of Solubility, Permeability,
Absorption and Bioavailability; van de Waterbeemd, H., Lehhemas, H.,
Artursson, P., Eds.; Wiley-VCH, Weinheim, Germany; p 23.
0
.03 mg/kg and the catalepsy was observed at a dose of 0.3 mg/
14. Dohta, Y.; Yamashita, T.; Horiike, S.; Nakamura, T.; Fukami, T. Anal. Chem. 2007,
kg by subcutaneous injections of haloperidol. Compound 23 at up
to 30 mg/kg did not cause any catalepsy and there was an over
7
9, 8312.
1
5. Lévesque, J. F.; Day, S. H.; Chauret, N.; Seto, C.; Trimble, L.; Bateman, K. P.; Silva,
J. M.; Berthelette, C.; Lachance, N.; Boyd, M.; Li, L.; Sturino, C. F.; Wang, Z.;
Zamboni, R.; Young, R. N.; Nicoll-Griffith, D. A. Bioorg. Med. Chem. Lett. 2007, 17,
3
0-fold window. These results suggest that potent mGluR1 alloste-
ric antagonist 23 is an attractive lead to go into pre-clinical phase
as a novel antipsychotic.
3038.
1
6. Kumar, S.; Kassahun, K.; Tschirret-Guth, R. A.; Mitra, K.; Baillie, T. A. Curr. Opin.
In summary, a series of 5-(1-aryl-3-methyl-1,2,3-triazole-4-yl)
isoindolin-1-one derivatives were rapidly screened for a developed
allosteric mGluR1 antagonist by using both library and conven-
tional medicinal chemistry techniques. Representative compound
Drug Disc. Dev. 2008, 11, 43.
17. Untrecht, J. Annu. Rev. Pharmacol. Toxicol. 2007, 47, 513.
1
1
8. Analytical data of 23: H NMR (400 MHz, CDCl3) d: 1.33 (6H, d, J = 6.8 Hz), 2.45
(
(
3H, d, J = 1.7 Hz), 4.43 (2H, s), 4.72 (1H, sept.), 7.03–7.18 (2H, m), 7.52–7.62
1H, m), 7.80 (1H, dd, J = 7.8, 1.2 Hz), 7.95 (1H, d, J = 7.8 Hz), 8.00 (1H, s). MS
+
+
2
3 has quite potent mGluR1 antagonistic activity, low off-target
(ESI ): m/z 369.1 [M+H] .