Synthesis, SAR study, and biological evaluation of novel 2,3-dihydro-1H-imidazo[1,2-a]benzimidazole derivatives as phosphodiesterase 10A inhibitors

Article abstract of DOI:10.1016/j.bmc.2019.07.010

Phosphodiesterase 10A (PDE10A) inhibitors were designed and synthesized based on the dihydro-imidazobenzimidazole scaffold. Compound 5a showed moderate inhibitory activity and good permeability, but unfavorable high P-glycoprotein (P-gp) liability for bra

Full text of DOI:10.1016/j.bmc.2019.07.010

Accepted Manuscript
Synthesis, SAR study, and biological evaluation of novel 2,3-dihydro-1H-imi-
dazo[1,2-a]benzimidazole derivatives as phosphodiesterase 10A inhibitors
Ayaka Chino, Shugo Honda, Masataka Morita, Koichi Yonezawa, Wataru
Hamaguchi, Yasushi Amano, Hiroyuki Moriguchi, Mayako Yamazaki, Masaki
Aota, Masaki Tomishima, Naoyuki Masuda
PII:
S0968-0896(19)30938-1
https://doi.org/10.1016/j.bmc.2019.07.010
BMC 14994
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry
Received Date:
Revised Date:
Accepted Date:
6 June 2019
3 July 2019
5 July 2019
Please cite this article as: Chino, A., Honda, S., Morita, M., Yonezawa, K., Hamaguchi, W., Amano, Y., Moriguchi,
H., Yamazaki, M., Aota, M., Tomishima, M., Masuda, N., Synthesis, SAR study, and biological evaluation of novel
2
,3-dihydro-1H-imidazo[1,2-a]benzimidazole derivatives as phosphodiesterase 10A inhibitors, Bioorganic &
Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.bmc.2019.07.010
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.

Synthesis, SAR study, and biological evaluation of novel 2,3-dihydro-1H-
imidazo[1,2-a]benzimidazole derivatives as phosphodiesterase 10A inhibitors
*
Ayaka Chino, Shugo Honda, Masataka Morita, Koichi Yonezawa, Wataru Hamaguchi,
Yasushi Amano, Hiroyuki Moriguchi, Mayako Yamazaki, Masaki Aota, Masaki
Tomishima and Naoyuki Masuda
Functional Molecules, Modality Research Laboratories, Drug Discovery Research,
Astellas Pharma Inc., 21, Miyukigaoka, Tsukuba, Ibaraki 305-8585, Japan
*Corresponding author:
Ayaka Chino
Functional Molecules, Modality Research Laboratories
Drug Discovery Research
Astellas Pharma Inc.
2
1 Miyukigaoka, Tsukuba
Ibaraki 305-8585, Japan
Tel: +81-(0)29-829-6261
Fax: +81-(0)29-854-1519
E-mail: ayaka.chino@astellas.com
1

Abstract
Phosphodiesterase 10A (PDE10A) inhibitors were designed and synthesized based on
the dihydro-imidazobenzimidazole scaffold. Compound 5a showed moderate inhibitory
activity and good permeability, but unfavorable high P-glycoprotein (P-gp) liability for
brain penetration. We performed an optimization study to improve both the P-gp efflux
ratio and PDE10A inhibitory activity. As a result, 6d was identified with improved P-gp
liability and high PDE10A inhibitory activity. Compound 6d also showed satisfactory
brain penetration, suppressed phencyclidine-induced hyperlocomotion and improved
MK-801-induced working memory deficit.
2

Keywords
PDE10A inhibitor; schizophrenia; brain penetration; P-gp liability; Y-maze
3

1. Introduction
Schizophrenia is a chronic and debilitating psychiatric disorder that affects
1
approximately 1% of the world’s population. Schizophrenia is characterized by a
combination of positive symptoms, negative symptoms and cognitive impairments.2
Current anti-psychotics are marginally effective for treating positive symptoms but are
less effective against negative symptoms and cognitive dysfunction. Therefore, novel
drug types that are effective against not only positive symptoms but also negative
symptoms and cognitive impairments would represent a significant advancement in
schizophrenia treatment.
The phosphodiesterases (PDEs) consist of an 11-membered family of enzymes that
catalyze the hydrolysis of secondary signal messengers, such as cyclic adenosine
monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP). PDE10A is a
dual substrate PDE that hydrolyzes both cAMP and cGMP. PDE10A is highly
expressed in the brain, particularly in medium spiny neurons of the striatum, which is
the main input station of the basal ganglia and is strongly associated with motor and
3
cognitive functions. The distribution of PDE10A expression is conserved across
3
mammalian species. Inhibition of PDE10A is thought to be effective for the treatment
of schizophrenia and a range of neurological, psychotic, anxiety, and movement
disorders that benefit from increased levels of cAMP and cGMP in neurons. Therefore,
4

PDE10A inhibitors might prove useful for the treatment of neurological and psychiatric
disorders.4
5
At present, several PDE10A inhibitors are undergoing clinical investigation and
6
there are a large number of studies on PDE10A inhibitors. Preclinical evidence
suggests that these inhibitors may demonstrate anti-psychotic, pro-cognitive, and anti-
negative symptom efficacy.7
We previously reported that the triazolopyrimidine analogue (1) had high inhibitory
8
activity with in vitro efficacy (Fig. 1). From the X-ray co-crystal structure of the
pyrimidoazaindazole derivative (2) with PDE10A enzyme (Fig. 2), it was revealed that
the 1-nitrogen of pyrimidoazaindazole ring of 2 accepted a hydrogen bond from Tyr693
and 4-nitrogen of triazolopyrimidine ring accepted it from Gln726. Unfortunately, 1
displayed a poor brain penetration profile in mice as shown in Table 1 making it
unsuitable as a tool compound for in vivo studies despite its low P-gp liability (Net
Efflux Ratio (NER) = 1.8). Compound 1 showed low blood to brain permeability (Pint
-6
=
3.2×10 cm/sec).
Given that permeability is correlated with lipophilicity or the number of hydrogen
9
bond donors (HBDs), we speculated that the hydroxyl group of 1 was the cause of the
poor permeability and low brain penetration because of the high polarity and the HBD
ability. However, the hydroxyl group was crucial for maintaining high PDE10A
5

8
inhibitory activity. For example, compound 3, in which the hydroxyl group was
removed from 1, showed significantly decreased inhibitory activity (Table 2, entry 1).
Therefore, further research was conducted to find novel scaffolds other than
pyrimidoindazole without a hydroxyl group that showed high inhibitory activity.
N
N
N
N
N
N
N
HO
1
PDE10A IC : 2.0 nM
5
0
logD : 0.3
7
.4
P-gp NER: 1.8
-6
Pint: 3.2×10 cm/sec
Figure 1. Structure of compound 1.
Table 1. Plasma and brain concentration of 1 at 1 h after oral administration to mice
a
Dose (mg/kg)
Plasma (ng/mL)
Brain (ng/g)
K ,brain
p
0
.1
0.59
ND b
ND b
a
Average value of individual Kp (brain/plasma) (n=3).
Not Detected
b
6

N
N
N
N
N
N
N
N
HO
2
PDE10A IC : 2.9 nM
5
0
Figure 2. Crystal structure of compound 2 (purple) bound to PDE10A (PDB code:
KDZ).
6
7

2
. Chemistry
The synthesis of target compounds is shown in Schemes 1–6. The synthetic route of
,7-dimethyl[1,2,4]triazolo[1,5-a]pyrimidine derivatives 4a–e is outlined in Scheme 1.
5
A Curtius rearrangement of (5,7-dimethyl[1,2,4]triazolo[1,5-a]pyrimidin-2-yl)acetic
acid 7 with benzyl alcohol gave the corresponding carbamate 8. Catalytic hydrogenation
of 8 yielded amine 9. Buchwald-Hartwig cross coupling reaction or ipso substitution of
aryl chlorides with 9 gave 4a–e. Compound 4f was prepared by alkylation of 4d with
1
,2-dibromoethane as shown in Scheme 2.
O
N
N
N
N
N
N
N
N
a
O
b
N
N
c or d
N
N
Ar
N
N
HO
N
H
N
N
H
N
H2N
O
4
a: Ar = 2-quinoline
b: Ar = 2-quinoxalin
c: Ar = 2-quinazolin
d: Ar = 2-benzimidazol
e: Ar = 1-methyl-2-benzimidazole
7
8
9
4
4
4
4
Scheme 1. Reagents and conditions: (a) BnOH, diphenylphosphorylazide, triethylamine,
toluene, 69%; (b) H , Pd-C, MeOH, quant.; (c) Ar-Cl, Pd(OAc) , (S)-1-[(R)-2-
2
2
(dicyclohexylphosphino)ferrocenyl]ethyldi-tert-butylphosphine ((S)-(R)-JOSIPHOS), t-
BuONa, DME, 13%; (d) Ar-Cl, NMP, 14–96%.
N
N
N
N
N
N
N
N
a
N
H
N
N
N
N
N
H
4d
4f
Scheme 2. Reagents and conditions: (a) 1,2-dibromoethane, tetrabuthylammonium
bromide, NaOH aq., DCE, 31%.
8

The synthesis of [1,2,4]triazolo[1,5-a]pyrimidine analogues 5a, b, g and h is illustrated
in Scheme 3. Ipso substitution of 2-chloro-1H-benzimidazole 10 with amino alcohols
1
1a, b, g and h gave alcohols 12a, b, g and h, which were chlorinated with SOCl2.
Intramolecular alkylation gave 13a, b, g and h, and subsequent condensation with (5,7-
dimethyl-[1,2,4]triazolo[1,5-a]pyrimidin-2-yl)-acetic acid 14 gave 5a, b, g and h.
a
b, c
N
N
L
Cl
L
N
N
H
+
H2N
O H
N
NH
N
H
L
N
H
O H
1
1
1
1
2
3a: L = C H4
1
0
11a: L = C H4
2
12a: L = C H4
2
3b: L = C H6
1
1
1
1b: L = C H6
3
3
12b: L = C H6
3
3g: L = CH CHMe
1g: L = CHMeCH2
2
12g: L = CHMeCH2
3h: L = CHMeCH2
1h: L = CH CHMe
2
12h: L = CH CHMe
2
N
N
d, e or f
N
N
L
N
N
N
N
+
HO
N
N
N
N
O
NH
N
L
O
1
1
1
1
3a: L = C H4
14
5a: L = C H4
2
2
3b: L = C H6
5b: L = C H6
3
3
3g: L = CH CHMe
5g: L = CH CHMe
2
2
3h: L = CHMeCH2
5h: L = CHMeCH2
Scheme 3. Reagents and conditions: (a) neat, 24–97%; (b) SOCl , DCE; (c) xylene, 14–
2
9
7% (2 steps); (d) WSC•HCl, HOBt, triethylamine, DMF, 54–58%; (e) WSC•HCl,
HOBt, DMF, 44%; (f) HATU, DIPEA, CH Cl , 59%.
2
2
Imidazo[1,2-a]benzimidazolyl derivatives 5c–f and 5i were synthesized in accordance
with the procedures outlined in Scheme 4. Buchwald-Hartwig cross coupling reaction of
imidazolidinones 16a–b with 2-nitrohalobenzene 15c–f and 15i, reduction of nitro
substitution with hydrazine using an iron-based catalyst, and chlorination and
9

cyclization gave imidazo[1,2-a]benzimidazols 17c–f and 17i. Amine 17c–f and 17i were
condensed with carboxylic acid 14 to give 5c–f and 5i.
4
3
4
R
Y
3
a, b, c
R
d or e
3
Y
5
N
N
N
N
4
5
NO2
5
N
HN
NH
6
N
R
+
N
6
N
N
X
NH
Y
O
O
6
Y
Y
Y
1
1
1
1
1
5c: R = 3- Me, X = Br 16a: Y = H
17c: R = 3- Me, Y = H
17d: R = 4- Me, Y = H
17e: R = 5- Me, Y = H
17f: R = 6- Me, Y = H
17i: R = H, Y = Me
5c: R = 3- Me, Y = H
5d: R = 4- Me, Y = H
5e: R = 5- Me, Y = H
5f: R = 6- Me, Y = H
5i: R = H, Y = Me
5d: R = 4- Me, X = I
5e: R = 5- Me, X = I
5f: R = 6- Me, X = Br
5i: R = H, X = I
16b: Y = Me
Scheme 4. Reagents and conditions: (a) Pd (dba) , 4,5-bis(diphenylphosphino)-9,9-
2
3
dimethylxanthene (Xantphos), Cs CO , 1,4-dioxane; (b) FeCl •6H O, N H •H O,
2
3
3
2
2
4
2
activated carbon, EtOH, H O; (c) POCl , 2.0–63% (3 steps); (d) 14, WSC•HCl, HOBt,
2
3
DMF, 41–84%; (e) 14, HATU, DIPEA, CH Cl , 44%.
2
2
Imidazo[1,2-a]benzimidazolyl derivatives 5j–l were synthesized as depicted in
Scheme 5. An ipso substitution of 2-nitrophenyl isocyanate 18 with alkyl amine 19j–l
gave the urea analogues 20j–l. Subsequently, imidazo[1,2-a]benzimidazolyl derivatives
5
j–l were prepared in a manner similar to that described for 5c–f and 5i.
NO2
f
N
N
N
N
NH2
R
b or c, d, e
N
a
NH
NO2
N
N
N
NH
R
N
R
HN
O
+
NCO
O H
O
R
O H
18
19j: R = (2S)- Ph
20j: R = (2S)- Ph
20k: R = (2R)- Ph
20l: R = (2S)- Me
21j: R = (2S)- Ph
21k: R = (2R)- Ph
21l: R = (2S)- Me
5j: R = (2S)- Ph
5k: R = (2R)- Ph
5l: R = (2S)- Me
19k: R = (2R)- Ph
9l: R = (2S)- Me
1
Scheme 5. Reagents and conditions: (a) THF, quant.; (b) FeCl •6H O, N H •H O,
3
2
2
4
2
activated carbon, EtOH, H O; (c) H , Pd-C, EtOH, THF; (d) POCl ; (e) xylene, 13–24%
2
2
3
(3 steps); (f) 14, WSC•HCl, HOBt, DMF, 46–73%.
1
0

As shown in Scheme 6, synthesis of imidazo[1,2-a]benzimidazolyl derivatives 6a–d
started with cyanation of methyl chlorides 22a–c. Subsequently, hydrolysis with NaOH
aqueous solution yielded carboxylic acids 23a–c. Amidation of 23a–c with amines 13a
or 21l gave 6a–d.
Y
N
N
Z
Y
Y
c, d or e
N
X
N
X
Z
a, b
Z
N
N
N
N
N
X
Cl
HO
O
O
R
6a: R = H, X = CH, Y = N, Z = CMe
6b: R = H, X = N, Y = CH, Z = CMe
6c: R = H, X = N, Y = CMe, Z = N
6d: R = Me, X = N, Y = CMe, Z = N
22a: X = CH, Y = N, Z = CMe
22b: X = N, Y = CH, Z = CMe
22c: X = N, Y = CMe, Z = N
23a: X = CH, Y = N, Z = CMe
23b: X = N, Y = CH, Z = CMe
23c: X = N, Y = CMe, Z = N
Scheme 6. Reagents and conditions: (a) NaCN, DMF, H O; (b) NaOH aq., 1,4-dioxane,
2
4
0–81% (2 steps).; (c) 13a, WSC•HCl, HOBt, DIPEA, DMF 53%; (d) 13a, HATU,
DIPEA, DMF, 71%; (e) 13a or 21l, WSC•HCl, HOBt, DMF, 55–68%.
1
1

3. Results and Discussion
We first changed the pyrimidoindazole scaffold to contain a small number of other
hetero aromatics that could interact with Tyr693, and assayed the resulting compounds.
Some of the results are shown in Table 2. The 2-amino quinoline derivative 4a showed
moderate inhibitory activity, and the binding mode of 4a with the PDE10A enzyme was
similar to that of 1 (Fig. 3). The quinoxaline derivative 4b and quinazoline derivative 4c
had dramatically lower potency, whereas the benzimidazole derivatives 4d and 4e had
higher potency than 4a.
In the view of the brain penetration, it would be better that the NH of benzimidazole is
capped with substituent to reduce the number of HBDs. We synthesized 2,3-dihydro-
1
H-imidazo[1,2-a]benzimidazole derivative 4f and found that it had a similar PDE10A
inhibitory to 4d. Furthermore, 4f showed much better permeability (Pint = 19.2 x 10-6
cm/sec) than 1, but had high P-glycoprotein (P-gp) liability (NER = 3.6). Our target
NER value was less than 2.0. While compound 5a, in which a carbonyl substituent was
introduced into the linker for the immobilization of the structure, showed more potent
inhibitory activity while maintaining similar permeability to 4f (Pint = 16.8 x 10-6
cm/sec). Compound 5a showed lower P-gp liability than 4f, though the value was still
high (NER = 2.3). We therefore moved to further optimize 5a to improve potency and
reduce P-gp liability.
1
2

Table 2. PDE10A inhibitory activity and P-gp net efflux ratio of heterocycles
N
R
N
N
N
Compds
R
PDE10A IC50
nM)
P-gp NER
(
3
78
NT a
N
N
N
4
4
4
4
4
a
b
c
294
1219
2083
105
88
NT a
NT a
NT a
NT a
NT a
N
N
N
N
H
N
H
N
N
N
H
N
H
N
d
e
N
H
N
H
N
N
1
3

N
N
4
f
138
3.6
2.3
N
N
N
O
5
a
52
N
a
Not tested
N
N
N
N
N
N
H
Figure 3. Crystal structure of compound 4a (green) bound to PDE10A (PDB code:
KDX).
6
1
4

Table 3 describes the SAR of substituents on the imidazo[1,2-a]benzimidazole ring for
improving the inhibitory activity. First, replacement of 2,3-dihydro-1H-imidazo[1,2-
a]benzimidazole with 1,2,3,4-tetrahydropyrimido[1,2-a]benzimidazole (5b) resulted in
a 20-fold loss of affinity, indicating that the 2,3-dihydro-1H-imidazo[1,2-
a]benzimidazole core was suitable for further optimization.
A methyl group was introduced into the phenylene ring of the 2,3-dihydro-1H-
imidazo[1,2-a]benzimidazole core to increase inhibitory activity. Unfortunately, it
resulted in a decrease in inhibitory activity. This result indicates that any substituent
could not be tolerated on the phenylene ring. (5c, 5d, 5e and 5f) In particular, a methyl
substituted next to the 8-nitrogen atom (5c) resulted in a loss of inhibitory activity,
suggesting that the 8-nitrogen atom is important for interaction with the PDE10A
enzyme. We then shifted our efforts to introducing a methyl group onto the imidazoline
moiety. Racemic 2-methyl derivative (5g) had the same potent activity as 5a, while 3-
methyl derivative (5h) showed weaker activity. As additional substituents to the 2-
position seemed tolerable, we introduced other substituents to this position. Changing
the mono methyl substituent to dimethyl substituents (5i) did not substantially change
the activity. We subsequently synthesized chiral phenyl substituent compounds 5j and
5
k. (S)-phenyl analogue 5j showed significantly higher inhibitory activity than (R)-
1
5

phenyl group 5k, although 5j showed similar potency to 5g. This indicates that the (S)-
isomer is crucial for inhibitory activity, while larger 2-substituents may not have an
appreciable contribution to potency because the 2-substituent is exposed to the solvent.
As a result, (S)-methyl analogue 5l displayed higher potency than its racemate 5g. The
X-ray co-crystal structure of 5l and PDE10A enzyme revealed that the nitrogen atom at
the 9-position of the imidazo[1,2-a]benzimidazol unit binds to Tyr693 (Fig. 4). The (S)-
methyl substituent of 5l underwent van der Waals interactions with Gly725, which
might underlie the enhanced potency. While we were able to enhance the PDE10A
inhibitory activity, 5l was still suffered from high P-gp liability (NER = 2.4). To address
this issue, we turned our attention to optimizing the triazolopyrimidine ring.
Table 3. PDE10A inhibitory activity and P-gp net efflux ratio of heterocycles
O
N
R
N
N
N
Compds
R
PDE10A IC50
nM)
P-gp NER
(
N
5
5
a
52
2.3
N
N
N
N
b
1476
NT a
N
1
6

N
N
5
5
c
>10000
NT a
N
N
N
N
d
441
NT a
N
5
5
e
f
416
279
NT a
N
N
N
NT a
N
N
N
N
5
5
5
g
h
i
69
2.4
N
N
N
292
NT a
N
N
74
2.1
N
N
1
7

N
N
N
5
j
75
3.8
N
N
N
5
k
1920
NT a
N
5
l
28
2.4
N
N
a
Not tested
1
8

9
N
N
N
N
N
N
N
2
O
Figure 4. Crystal structure of compound 5l (magenta) bound to PDE10A (PDB code:
6
KE0).
Previous reports suggest that the number and position of heteroatoms are correlated
1
0
with P-gp liability. We therefore focused on changing the number and position of
nitrogen atoms on the triazolopyrimidine ring to reduce the P-gp liability. (Table 4)
Although the P-gp NER was improved in imidazopyrimidine (6a) and triazolopyridine
(6b), they showed a loss of PDE10A inhibitory activity, indicating that the 1- and 4-
nitrogen atoms on the triazolopyrimidine ring were important for the inhibitory activity
and were crucial for P-gp liability. Therefore, triazolopyrazine (6c), which contained a
nitrogen atom at the 5-position instead of the 4-position, was evaluated. Changing the
scaffold of 5a to that of 6c improved the P-gp NER while maintaining the potency
observed in 5a. We then introduced 2-methyl substituent on the imidazobenzimidazol
moiety of compound 6c to enhance the potency. As a result, 6d exhibited single-digit
nanomolar inhibitory activity while maintaining low P-gp liability.
1
9

Table 4. PDE10A inhibitory activity and P-gp NER of heterocycles
N
R2
N
N
O
R1
Compds
R1
R2
PDE10A IC50
nM)
P-gp NER
(
N
N
N
N
5
6
6
a
a
b
H
H
H
N
N
N
52
2.3
N
275
135
1.0
1.2
N
N
N
N
N
6
6
c
H
44
1.3
1.2
N
N
N
N
N
d
Me
8.4
Additional profiles of 6d are shown in Fig. 5. Compound 6d exhibited good solubility,
2
0

metabolic stability and permeability, and low CYP1A2, 2C8, 2C9, 2C19, 2D6 and
CYP3A4 inhibition.
N
N
N
N
N
N
N
O
PDE10A IC : 8.4 nM
5
0
logD : 2.0
7
.4
b
Solubility :  100 μM
CYP inhibition (1A2, 2C8, 2C9, 2C19, 2D6, 3A4); residual activity : 98%, 99%, 96%,
0%, 102%, 90%
a
9
CLint c (mL/min/kg) (human, mouse): No Dep. , 175.3
d
P-gp NER: 1.2
-6
Pint: 27.5 x 10 cm/sec
Figure 5. Profiles of compound 6d.
a
Residual activities of HLM for metabolism of each substrate in the presence of test
compounds were determined following pre-incubation for 30 min.
b
Aqueous solubility in the Japanese Pharmacopoeia 2nd fluid for disintegration test
(JP2; pH = 6.8).
c
intrinsic clearance
d
No depletion in this condition
Table 5. Plasma and brain concentration of 6d at 1 h after oral administration to mice
a
Dose (mg/kg)
Plasma (ng/mL)
Brain (ng/g)
K ,brain
p
0
.1
0.91
2.2
2.3
a
Average value of individual Kp (brain/plasma) (n=3).
2
1

We next examined the pharmacokinetic profile of 6d in mice at a single dose of 0.1
mg/kg. As shown in Table 5, compound 6d showed good intracerebral transferability in
mice (Kp, brain=2.3). The good brain permeability of 6d prompted us to assess its in
vivo effect on hyperlocomotion induced by phencyclidine (PCP) in mice, an animal
model for positive symptoms of schizophrenia. As shown in Fig. 6, compound 6d dose-
dependently attenuated the PCP-induced hyperlocomotion after oral administration.
Figure 6. Effect of oral administration of 6d on PCP-induced hyperlocomotion in
2
2

mice. PCP was administered subcutaneously (sc). The data represent the mean ± SEM:
## p <0.001 versus normal group (Student’s t test); **p < 0.01 vs control, *p < 0.05 vs
#
control (Dunnett’s test).
We examined the effect of 6d on working memory disruption in mice by observing
spontaneous alternation behavior in a Y-maze following treatment with MK-801. This
model is generally used to screen for agents that improve cognitive deficits because
working memory deficit is a common symptom of disorders involving cognitive
dysfunction, such as dementia and schizophrenia. MK-801 is a noncompetitive
antagonist of the NMDA receptor, and animals treated with MK-801 show various
memory and learning deficits caused by inhibition of signaling through NMDA
receptors. 11
Alternation was defined as consecutive entries into each of the three arms. In this study,
the alternation rate in the normal group was approximately 70%, which significantly
decreased to approximately 50% after administration of MK-801 (Fig. 7). Doses of
0
.1mg/kg and 0.3mg/kg of 6d significantly attenuated the MK-801-induced decrease in
alternation rate (p < 0.01 and p < 0.05, respectively). It was reported that PDE10A plays
a pivotal role in regulating the tone of dopamine D1 receptor signaling, 12 and an
“
inverted-U” relationship between prefrontal dopamine transmission and cortical
2
3

1
3
efficacy that varies as a function of working memory load was also suggested, which
may explain the inverted “U-shaped” dose–response curve of 6d in MK-801-induced
working memory deficit in mice during Y-maze test. The mean number of arm entries
was not affected by 6d administration except at a dose of 1 mg/kg compared with the
control group (data not shown). Compound 6d ameliorated MK-801-induced decreases
in alternation behavior, suggesting that 6d would be effective for treating symptoms of
cognitive disorders.
Figure 7. Effect of 6d on MK-801-induced disruption of spontaneous alternation
behavior in mice. Data represent the mean  S.E.M. alternation rate in mice. ##p < 0.01
2
4

vs normal group (Student’s t-test); **p < 0.01 vs control, *p < 0.05 vs control
Dunnett’s test). N: normal (vehicle/saline-treated group), C: control (vehicle/MK801-
treated group).
(
1
6
Existing antipsychotic drugs are known to induce catalepsy in rodents , which is an
index of extrapyramidal side effects (EPS).14 We therefore evaluated whether 6d
induces catalepsy in mice. Compound 6d did not significantly induce catalepsy at up to
1
00 mg/kg, p.o. (Fig. 8).
30
20
10
0
0
3
10
30
100
6d (mg/kg, p.o.)
Figure 8. Effects of 6d on catalepsy in mice Catalepsy time (sec) (max = 120 sec) of 6d.
2
5

Data represent the mean ± S.E.M. (n = 6–8 for each group).
2
6

4
Conclusion
In this study, we investigated the SAR of various dihydro-imidazobenzimidazole
derivatives and their PDE10A inhibitory activity and P-gp liability to obtain potent
PDE10A inhibitors with good CNS penetration. The 2,3-dihydro-1H-imidazo[1,2-
a]benzimidazole derivative 5a, which was chosen as a new scaffold, showed moderate
inhibitory activity and good permeability but unfavorable high P-gp liability.
Substitution of a 2-methyl group into imidazo[1,2-a]benzimidazole resulted in
improved PDE10A inhibitory activity. Moreover, replacement of the triazolopyrimidine
core of 6c with a triazolopyrazine core (6d) resulted in potent PDE10A inhibitory
activity and acceptable P-gp liability. PK study indicated that 6d exhibited good brain
penetration. Further, 6d attenuated the hyperlocomotor activity induced by PCP after
oral administration at 1 mg/kg to 30 mg/kg, and significantly improved MK-801-
induced working memory deficit in the Y-maze test following oral administration at 0.1
and 0.3 mg/kg. These effects suggest that 6d may satisfy current unmet medical needs
for the treatment of schizophrenia.
2
7

5. Experimental
5
.1. Chemistry
1H NMR spectra were recorded on a Varian 400-MR and BRUKER AV-III HD500, and
chemical shifts were expressed as δ (ppm) values with tetramethylsilane as an internal
reference (s=singlet, d=doublet, t=triplet, q=quartet, quint=quintet, m=multiplet,
dd=double doublet, dt=double triplet, ddd=double double doublet, and br=broad peak).
Mass spectra (MS) were recorded on a Waters UPLC/SQD and Waters Acquity
UPLC/ZQ. Elemental analyses were performed using a Yanaco JM10 (C, H, N) and
were within ±0.4% of theoretical values. Electrospray ionization positive high-
resolution mass spectra (HRMS) were obtained using a Thermo EXACTIVE-Plus
Waters LCT Premier. Specific rotation was obtained using Horiba SEPA-500 and OHM
Electric OCE-TCR12075WL. Unless otherwise noted, all reagents and solvents
obtained from commercial suppliers were used without further purification.
2
8

5
.1.1.
2-[2-(5,7-Dimethyl[1,2,4]triazolo[1,5-a]pyrimidin-2-
yl)ethyl]pyrido[2',3',3,4]pyrazolo[1,5-a]pyrimidin-4-ol (2)
To a mixture of 3-(5,7-dimethyl[1,2,4]triazolo[1,5-a]pyrimidin-2-yl)propanoic acid
2.20 g, 10.0 mmol) in THF (33 mL) was added 1,1’-carbonyldiimidazole (1.95 g, 12.0
(
mmol), and the mixture was stirred at 50 °C for 1 h. To the reaction mixture were added
ethyl potassium malonate (3.40 g, 20.0 mmol), magnesium chloride (1.90 g, 20.0 mmol),
triethylamine (3.48 mL, 25.0 mmol), and the mixture was stirred at 50 °C for 12 h. After
cooling to room temperature, to the reaction mixture was added 1 M HCl aqueous
solution (50 mL), and the mixture was stirred at room temperature for 1 h. The mixture
was extracted with EtOAc. The organic layer was washed with brine, dried over
anhydrous Na SO , filtered and concentrated in vacuo. To a half of the residue were
2
4
added 1,4-dioxane (15 mL), 1H-pyrazolo[3,4-b]pyridin-3-amine (615 mg, 4.58 mmol)
and AcOH (4 mL), and stirred at 90 °C for 36 h. After cooling to room temperature, the
precipitate was collected by filtration. The solid was washed with EtOAc and saturated
1
NaHCO aqueous solution to give 2 (201 mg, 11%) as a pale brown solid. H NMR
3
(
DMSO-d ) δ 2.56 (s, 3H), 2.70 (d, 3H, J = 0.88 Hz), 2.99–3.17 (m, 2H), 3.21–3.28 (m,
6
2
H), 5.86 (s, 1H), 6.84 (dd, 1H, J = 7.9, 4.4 Hz), 7.10 (d, 1H, J = 0.66 Hz), 8.31 (dd, 1H,
+
J = 7.9, 1.8 Hz), 8.54 (dd, 1H, J = 4.2, 1.8 Hz); MS (ESI) m/z 361 [M+H] ; HRMS
2
9