Discovery of Clinical Candidate N-((1S)-1-(3-Fluoro-4-(trifluoromethoxy)phenyl)-2-methoxyethyl)-7-methoxy-2-oxo-2,3-dihydropyrido[2,3-b]pyrazine-4(1H)-carboxamide (TAK-915): A Highly Potent, Selective, and Brain-Penetrating Phosphodiesterase 2A Inhibitor for the Treatment of Cognitive Disorders

Article abstract of DOI:10.1021/acs.jmedchem.7b00807

Phosphodiesterase (PDE) 2A inhibitors have emerged as a novel mechanism with potential therapeutic option to ameliorate cognitive dysfunction in schizophrenia or Alzheimer's disease through upregulation of cyclic nucleotides in the brain and thereby achieve potentiation of cyclic nucleotide signaling pathways. This article details the expedited optimization of our recently disclosed pyrazolo[1,5-a]pyrimidine lead compound 4b, leading to the discovery of clinical candidate 36 (TAK-915), which demonstrates an appropriate combination of potency, PDE selectivity, and favorable pharmacokinetic (PK) properties, including brain penetration. Successful identification of 36 was realized through application of structure-based drug design (SBDD) to further improve potency and PDE selectivity, coupled with prospective design focused on physicochemical properties to deliver brain penetration. Oral administration of 36 demonstrated significant elevation of 3′,5′-cyclic guanosine monophosphate (cGMP) levels in mouse brains and improved cognitive performance in a novel object recognition task in rats. Consequently, compound 36 was advanced into human clinical trials.

Full text of DOI:10.1021/acs.jmedchem.7b00807

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Article
Discovery of Clinical Candidate N-((1S)-1-(3-Fluoro-4-
trifluoromethoxy)phenyl)-2-methoxyethyl)-7-methoxy-2-
(
oxo-2,3-dihydropyrido[2,3-b]pyrazine-4(1H)-carboxamide
(TAK-915): A Highly Potent, Selective, and Brain-Penetrating
Phosphodiesterase 2A Inhibitor for the Treatment of Cognitive Disorders
Satoshi Mikami, Shinji Nakamura, Tomoko Ashizawa, Izumi Nomura, Masanori Kawasaki, Shigekazu
Sasaki, Hideyuki Oki, Hironori Kokubo, Isaac D. Hoffman, Hua Zou, Noriko Uchiyama, Kosuke
Nakashima, Naomi Kamiguchi, Haruka Imada, Noriko Suzuki, Hiroki Iwashita, and Takahiko Taniguchi
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00807 • Publication Date (Web): 10 Aug 2017
Downloaded from http://pubs.acs.org on August 10, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of Clinical Candidate Nꢀ((1S)ꢀ1ꢀ(3ꢀ
Fluoroꢀ4ꢀ(trifluoromethoxy)phenyl)ꢀ2ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
methoxyethyl)ꢀ7ꢀmethoxyꢀ2ꢀoxoꢀ2,3ꢀ
dihydropyrido[2,3ꢀb]pyrazineꢀ4(1H)ꢀcarboxamide
(TAKꢀ915): A Highly Potent, Selective, and Brainꢀ
Penetrating Phosphodiesterase 2A Inhibitor for the
Treatment of Cognitive Disorders
,†
†
†
†
†
Satoshi Mikami,* Shinji Nakamura, Tomoko Ashizawa, Izumi Nomura, Masanori Kawasaki,
†
†
†
‡
‡
Shigekazu Sasaki, Hideyuki Oki, Hironori Kokubo, Isaac D. Hoffman, Hua Zou, Noriko
†
†
†
†
†
Uchiyama, Kosuke Nakashima, Naomi Kamiguchi, Haruka Imada, Noriko Suzuki, Hiroki
†
†
Iwashita, and Takahiko Taniguchi
AUTHOR AFFILIATIONS:
†
Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 26ꢀ1, Muraokaꢀ
Higashi 2ꢀchome, Fujisawa, Kanagawa 251ꢀ8555, Japan
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
‡
Takeda California, Inc., 10410 Science Center Drive, San Diego, California 92121, United
States
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ABSTRACT
Phosphodiesterase (PDE) 2A inhibitors have emerged as a novel mechanism with
potential therapeutic option to ameliorate cognitive dysfunction in schizophrenia or Alzheimer’s
disease through upregulation of cyclic nucleotides in the brain, and thereby achieve potentiation
of cyclic nucleotide signaling pathways. This article details the expedited optimization of our
recently disclosed pyrazolo[1,5ꢀa]pyrimidine lead compound 4b, leading to the discovery of
clinical candidate 36 (TAKꢀ915), which demonstrates an appropriate combination of potency,
PDE selectivity, and favorable pharmacokinetic (PK) properties, including brain penetration.
Successful identification of 36 was realized through application of structureꢀbased drug design
(SBDD) to further improve potency and PDE selectivity, coupled with prospective design
focused on physicochemical properties to deliver brain penetration. Oral administration of 36
demonstrated significant elevation of 3',5'ꢀcyclic guanosine monophosphate (cGMP) levels in
mouse brains, and improved cognitive performance in a novel object recognition task in rats.
Consequently, compound 36 was advanced into human clinical trials.
INTRODUCTION
ACS Paragon Plus Environment

Page 3 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Phosphodiesterases (PDEs) are a superfamily of enzymes that catalyze the hydrolysis of
the second messengers, 3',5'ꢀcyclic adenosine monophosphate (cAMP) and/or 3',5'ꢀcyclic
1
guanosine monophosphate (cGMP) into inactive 5'ꢀAMP and/or 5'ꢀGMP, thereby regulating the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
cellular levels of these cyclic nucleotides, their downstream signaling cascades, and, as a result, a
2
–4
diverse array of biological responses.
PDEs are encoded by 21 separate genes that are
categorized into 11 different PDE families (PDE1–11) on the basis of amino acid sequence
1
,5
similarity, substrate specificity, and mode of regulation.
PDE2A, which is a dualꢀsubstrate enzyme capable of accommodating and degrading both
cAMP and cGMP, has the highest expression in regions of the brain associated with cognitive
6
–11
processes, such as the cortex, hippocampus, and striatum.
In the central nervous system
(CNS), both cAMP and cGMP play important roles in modulating intracellular signal
transduction pathways related to longꢀterm potentiation (LTP), a surrogate measure of synaptic
1
2–16
plasticity,
which is considered a key element of the neurobiological foundations of cognitive
1
7
function.
On the basis of these features, it has been hypothesized that PDE2A inhibition would be
beneficial for the treatment of cognitive dysfunction associated with schizophrenia or dementia
in Alzheimer’s disease, by suppressing the degradation of intracellular cAMP and/or cGMP
levels in brain areas critical for cognitive function and memory. Thus, the search for potent and
1
8–
selective PDE2A inhibitors has attracted considerable attention in the pharmaceutical industry.
36
18
As shown in Figure 1, Bayer reported a highly potent PDE2A inhibitor BAY 60ꢀ7550 (1)
possessing the imidazo[5,1ꢀf][1,2,4]triazinꢀ4(3H)ꢀone motif as a core, which augmented
3
7,38
cognition and memory in animal behavioral models.
Although this compound has been
frequently used preclinically as a key pharmacological tool in this research field, poor CNS
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
39
penetration has limited its clinical use. Meanwhile, Pfizer recently disclosed a highly potent
and selective PDE2A inhibitor PFꢀ05180999 (2) with oral bioavailability and CNSꢀpenetrant
2
6,27
attributes, which was advanced to human clinical trials.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
Me
Me
N
HN
N
N
N
N
N
CF3
N
N
Me
N
OH
MeO
N
N
Me
OMe
1
(BAY 60-7550, Bayer)
2 (PF-05180999, Pfizer)
PDE2A IC : 4.7 nM
PDE2A IC : 1 nM
5
0
50
PDE selectivity: 50x (vs. PDE1C)
PDE selectivity: 2000x (vs. PDE10A)
Figure 1. Structures and profiles of representative PDE2A inhibitors.
4
0
In the preceding publication, we described our leadꢀgeneration approach that began
with highꢀthroughput screening hit 3 from our inꢀhouse compound library and led to the
identification of lead compound 4b, which possessed a reasonable balance of potency, PDE
selectivity, and PK properties including brain penetration (Figure 2). Reflecting these favorable
attributes, oral dosing of 4b significantly elevated cGMP levels in mouse brains, but its high
systemic exposure at effective dosage in vivo and selectivity profile against all other PDE
families (180ꢀfold vs. PDE3A) were insufficient for consideration as a clinical candidate. As
1
0
PDEs are expressed in a variety of tissues, inhibition of offꢀtarget PDEs could induce undesired
side effects. Potential side effects include vasodilation and tachycardia, which are attributed to
41,42
inhibition of PDE1 and PDE3,
or visual disturbances, which are attributed to inhibition of
ACS Paragon Plus Environment

Page 5 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
43–45
PDE6.
Therefore, achieving high selectivity is crucial for the development of a clinically
beneficial PDE inhibitor with minimized safety concerns. Hence, our medicinal chemistry efforts
were focused on moving beyond delivery of an in vivo tool compound into identification of a
clinical candidate with the following attributes: (i) PDE2A IC50 < 5 nM, (ii) >1000ꢀfold
selectivity over all other PDEs, and (iii) reasonable brain penetration. To this end, we applied
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
0
SBDD utilizing a PDE2A coꢀcrystal structure of 4b. Simultaneously, we tracked topological
polar surface area (TPSA), hydrogen bond donor (HBD) count, and multidrug resistance protein
1 (MDR1) efflux ratio to maximize the probability of preserving the excellent brain exposure
characteristics of 4b.
Herein, we disclose a new series of PDE2A inhibitors derived from 4b and the
subsequent optimization efforts employing SBDD, which culminated in the discovery of Nꢀ((1S)ꢀ
1
ꢀ(3ꢀfluoroꢀ4ꢀ(trifluoromethoxy)phenyl)ꢀ2ꢀmethoxyethyl)ꢀ7ꢀmethoxyꢀ2ꢀoxoꢀ2,3ꢀ
2
9
dihydropyrido[2,3ꢀb]pyrazineꢀ4(1H)ꢀcarboxamide 36 (TAKꢀ915, Figure 3), a compound with
fully aligned drug attributes. Furthermore, we describe the binding mode of 36 in the human
PDE2A catalytic domain, which accounts for its potent inhibitory activity and high PDE
selectivity. We also report on the pharmacological profile of 36, which helps demonstrate the
promise of PDE2A inhibition as a potential therapeutic approach in the treatment of cognitive
impairment for a range of neuropsychiatric and neurodegenerative disorders.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0
Figure 2. Transition from highꢀthroughput screening hit 3 to lead compounds 4a and 4b.
O
HN
N
O
O
F
F
F
N
HN
F
MeO
OMe
6 (TAK-915)
3
Figure 3. Structure of 36.
RESULTS AND DISCUSSION
Drug Design Strategies. The program leading to discovery of 36 began with examination of a
4
0
coꢀcrystal structure of 4b (Figure 4). As reported previously, the characteristic interactions are
as follows: (i) the C7 hydrogen of the pyrazolo[1,5ꢀa]pyrimidine core most likely forms a weak
46
C–H hydrogen bond with the carbonyl group of Gln859, a conserved amino acid in all PDEs
that is key to cyclic nucleotide binding, (ii) the core sits within a hydrophobic clamp formed by
Phe862 on the top and Ile826 on the bottom, and is stabilized by π–π and CH–π interactions, (iii)
the amide carbonyl group is engaged in two hydrogen bonding interactions with water molecules,
one of which provides a bridging interaction with Tyr655, and (iv) the rightꢀhand side (RHS)
ACS Paragon Plus Environment

Page 7 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
phenyl group fits into the hydrophobic pocket newly created by ligandꢀbinding. In addition, the
amide N–H forms an intramolecular hydrogen bond with the N4 nitrogen atom of the core. In
good agreement with our previous findings about the structure–activity relationships (SARs), this
internal hydrogen bond plays a critical role in potency via stabilization of the bound
conformation. Through this analysis of the binding mode of 4b, we envisioned that the
identification of additional interactions with the protein as well as access to pockets unique to the
PDE2A in the PDE family would be key for realizing our potency and selectivity goals. To this
end, we devised four drug design strategies, as shown in Figure 5. Strategy 1 involves
replacement of the pyrazolo[1,5ꢀa]pyrimidine core to enable an unambiguous hydrogen bond
interaction with Gln859. This strategy also expands the structural diversity around the core ring,
and consequently increases the probability of finding a highꢀquality clinical candidate. Strategy 2
involves optimization of the RHS phenyl group to fill the PDE2Aꢀspecific hydrophobic pocket
optimally, which might enhance both potency and selectivity. Strategy 3 involves incorporation
of a polar functionality into the branched ethyl group. As this portion is close to the solvent
accessible region and is surrounded by proteinꢀbound metal ions containing a highly ordered
water network and hydrophilic amino acid residues, such as His656 and Gly826, addition of a
polar functionality may enhance potency while also improving physicochemical properties by
decreasing the overall molecular lipophilicity. Strategy 4 involves introduction of substituents
into the core to fill the surrounding space efficiently. This strategy is supported by the previous
finding that substitution on the pyrazolo[1,5ꢀa]pyrimidine core contributes significantly to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
0
potency and PDE selectivity, and is, therefore, expected to improve both inhibitory activity and
selectivity.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 9 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Xꢀray crystal structure of 4b bound in the PDE2A catalytic site (PDB 5XKM) viewed
from the top (A) and the entrance of the catalytic site (B). The key hydrogen bonding
interactions of 4b with PDE2A and the intramolecular hydrogen bond in 4b are indicated by red
and yellow dotted lines, respectively.
Strategy 1
Replacement of the pyrazolo[1,5-a]pyrimidine core to
enable an unambiguous H-bond interaction with Gln859
N
N
O
O
F
F
Strategy 2
Optimization of the substituent
to fill PDE2A specific pocket
Me
N
N
F
Me
Strategy 4
Introduction of a substituent to fill the
space and/or to target residue differences
Strategy 3
Introduction of a polar group to access
water molecules or His656 around metal ions
Figure 5. Drug design strategies to improve PDE2A inhibitory activity and PDE selectivity.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Strategy 1: Exploration of New Alternative Cores. We focused initially on replacement of the
pyrazolo[1,5ꢀa]pyrimidine core. As the C–H hydrogen bond (C–HꢁꢁꢁO) observed between the
core and Gln859 is generally weaker than a classic hydrogen bond, we hypothesized that
establishing an unambiguous hydrogen bond interaction with either or both the carbonyl and –
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH groups of Gln859 would further increase potency. Notably, one proposed molecular
2
mechanism for the dual cAMP/cGMP specificity of PDEs like PDE2A involves rotation of the
side chain of invariant glutamine (Gln859 in PDE2A) to recognize both cAMP and cGMP
4
7,48
(Figure 6). This is commonly referred to as the ''glutamine switch''.
With this design
hypothesis and glutamine switch mechanism in hand, we designed novel alternative core
structures that could create a classic hydrogen bond with Gln859 in a binding mode similar to
that for cAMP or cGMP (Figure 7). Considering the importance of the intramolecular hydrogen
bond observed in pyrazolo[1,5ꢀa]pyrimidine analogs, a nitrogen atom that could serve as a
hydrogen bond acceptor (HBA) was also incorporated into the newly designed core to ensure a
critical internal hydrogen bond network with the neighboring amide N–H, and thereby adopt a
similar conformation to the RHS benzylamine moiety of 4b. At this design stage,
physicochemical properties, such as TPSA and the number of HBD, were also considered to
optimize the MDR1 efflux ratio, and thus obtain a compound with favorable brain penetration.
2
To minimize Pꢀglycoprotein (Pꢀgp) efflux, TPSA should be maintained below 90 Å , and the
4
9,50
HBD count should be kept below 2.
Within this chemotype series, in which the amide NH
interacts with the nitrogen atom of the core scaffold, the TPSA and HBD count would be
affected by the strength of the intramolecular hydrogen bond; assuming that these polar
functionalities are completely masked, the apparent TPSA and HBD count would decrease by
2
51
approximately 20 Å and 1, respectively. On the basis of these considerations, we targeted
ACS Paragon Plus Environment

Page 11 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
TPSA values below 110 Å and HBD counts of less than 3. Additionally, an MDR1 efflux ratio
of less than 2.0 was considered as our target range.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Glutamine switch mechanism proposed to explain the accommodation of both cyclic
47,48
nucleotide substrates in dualꢀsubstrate PDEs.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Drug design concepts (A)–(D) to identify new alternatives to the pyrazolo[1,5ꢀ
a]pyrimidine core.
Representative results from our exploration of new core scaffolds are summarized in
Table 1. All compounds synthesized possessed TPSA values within our target range. For
comparison, pyrazolo[1,5ꢀa]pyrimidine derivatives 4a and 5 are also included in the table.
Notably, 4a is 9ꢀfold more potent than 5, which highlights the significant contribution of the
4
0
methyl group at the 6ꢀposition of 4a to potency enhancement. Compound 6, which contains a
lactam moiety seemingly capable of achieving a more effective hydrogen bond with Gln859,
exhibited greater PDE2A inhibitory activity when compared with 5, revealing the superiority of
ACS Paragon Plus Environment

Page 13 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the 3,4ꢀdihydropyrido[2,3ꢀb]pyrazinꢀ2(1H)ꢀone core over the pyrazolo[1,5ꢀa]pyrimidine core.
However, the interaction energy gained between the lactam moiety and Gln859 may be partially
offset by attenuation of the π–π and CH–π interactions with Phe862 and Ile826, respectively,
owing to the decreased πꢀelectron density of the core. To confirm this speculation, the PDE2A
inhibitory activity of 7, which possesses a lactam moiety and a similar aromaticity to 5, was
characterized. Compound 7 demonstrated improved potency (PDE2A IC50 = 21 nM) in
comparison with 6, which suggests that interactions with both Gln859 and Phe862/Ile826 may
contribute to the potency enhancement. On the other hand, 8, designed to form a hydrogen bond
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
between the core and the –NH group of Gln859 in a cAMPꢀlike binding mode, exhibited
2
significantly decreased potency compared with 4a. Moreover, a similar trend was observed with
9, a reversed lactam counterpart of 7. Although we cannot exclude the possibility of this
representing suboptimal placement of the HBA and/or HBD, we postulated that the loss in
potency was associated with the side chain of Gln859 being unlikely to freely rotate to
accommodate the cores of 8 and 9. From the Xꢀray crystal structure of 4b, the side chain of
Gln859 forms a hydrogen bond with neighboring Tyr827, and the energetic penalty of disrupting
this preexisting hydrogen bonding network was likely not overcome by the interaction energy
gained between each core and Gln859. Meanwhile, to confirm the role of the pyridine nitrogen
atom in the core in 6, this atom was replaced by a CH group to afford 10. Compound 10
exhibited a pronounced loss of inhibitory activity, reinforcing the importance of the
intramolecular hydrogen bond on the potency in this core series, as was the case with
pyrazolo[1,5ꢀa]pyrimidine core derivatives. Thus, the examination of various cores targeting the
interaction with Gln859 identified two promising novel cores in 6 and 7. Unfortunately, further
profiling revealed that 7 had significant in vitro phototoxicity, presumably inherent in the 1,5ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
naphthyridinꢀ2(1H)ꢀone core, and also exhibited decreased CNS penetration, likely associated
with the more acidic nature of the N–H proton relative to that of the core in 6 (data not shown).
Consequently, the 3,4ꢀdihydropyrido[2,3ꢀb]pyrazinꢀ2(1H)ꢀone motif in 6 was chosen as a core
for further investigation.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 1. In Vitro Activities of Derivatives Possessing Various Fused Bicyclic Cores
a
b
2
compd
core
anticipated binding mode
PDE2A IC (nM)
TPSA (Å )
5
0
4a
53 (47–59)
69
69
(
A)
A)
5
6
480 (350–650)
(
78 (61–98)
21 (18–26)
84
84
(A) or (B)
7
(
A) or (B)
8
9
6500 (4200–9800)
81
84
(
C)
19000 (15000–24000)
(
C) or (D)
ACS Paragon Plus Environment

Page 15 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
70000 (40000–120000)
71
(A) or (B)
a
b
Refer to Figure 7. IC values (95% confidence intervals given in parentheses) were calculated
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
50
from percent inhibition data (duplicate, n = 1). All values are rounded to two significant digits.
As 6 showed attractive in vitro activity, chiral separation was performed. Upon HPLC
separation of the two enantiomers, we observed that the activity of the racemic 6 resided
predominantly in the 6a, which displayed a PDE2A IC50 of 66 nM (Table 2). Furthermore,
eutomer 6a exhibited moderate PDE selectivity, with a minimum 30ꢀfold selectivity over
PDE1A. Notably, the PDE selectivity profile of eutomer 6a almost tracked that of racemic 6,
suggesting that distomer 6b is also inactive toward other PDEs.
Table 2. In Vitro Activity Profiles of 6 and Its Enantiomers 6a and 6b
a
compd
stereo
rac
eutomer
distomer
PDE2A IC (nM)
50
6
78 (61–98)
66 (46–93)
>100000
6
a
6
b
6
6a
a
b
a
b
PDE subtypes IC (nM)
Selectivity ratio
IC (nM)
Selectivity ratio
5
0
50
PDE1A
PDE2A3
PDE3A
3200
78
8100
41
–
100
2000
66
4200
30
–
64
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
PDE4D2
PDE5A1
PDE6AB
PDE7B
PDE8A1
PDE9A2
PDE10A2
PDE11A4
>10000
9800
>130
130
>10000
>150
>150
>150
>150
>150
>150
>150
>150
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>130
>130
>130
>130
>130
>130
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
IC values were calculated from percent inhibition data (duplicate, n = 1). All values are
5
0
b
rounded to two significant digits. Selectivity ratio (rounded to two significant digits) = PDE''X''
IC /PDE2A IC .
5
0
50
Strategy 2: Optimization of the RHS Phenyl Group. Given the hydrophobic nature and
limited space of the binding pocket around the RHS phenyl group (Figure 4), we replaced the
trifluoromethoxy group with less lipophilic and compact substituents (11–14) (Table 3).
Consistent with our previous SARs for the pyrazolo[1,5ꢀa]pyrimidine analogs represented by 4a,
replacing the trifluoromethoxy group with a methoxy group (11) resulted in an approximately 8ꢀ
fold decrease in potency. The potency decrease may be attributed to the following two factors: (i)
differences in conformational preferences, as the O–CF bond prefers an orthogonal orientation
3
5
2
to the phenyl plane, whereas the O–CH bond favors an inꢀplane conformation; and (ii)
3
decreased hydrophobic contacts between the methoxy group and the surrounding hydrophobic
residues in the binding pocket, as evidenced by the significant change in the LogD values
(
∆LogD = 0.83). Trifluoromethyl and cyclopropyl analogs 12 and 13, with different threeꢀ
dimensional conformations of the trifluoromethoxy group, exhibited comparable inhibitory
activity to 6, but reduced PDE selectivity. These results indicate that the geometry of the
substituent at this position affects PDE selectivity rather than potency. In particular,
trifluoromethyl analog 12 had a slightly reduced LogD value relative to 6. Furthermore, larger
groups, such as an azetidine ring (14), significantly decreased the PDE2A inhibitory activity,
indicative of the limited space around this position. To examine the effects of other substitution
ACS Paragon Plus Environment

Page 17 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
patterns, metaꢀ and orthoꢀtrifluoromethoxy derivatives 15 and 16 were synthesized; however,
both compounds showed considerable losses in potency. This is likely due to the more restricted
pocket size around positions other than the paraꢀposition and/or an unfavorable conformational
change brought about by introduction of an orthoꢀsubstituent. We next expanded the scope of
our investigations to explore disubstituted analogs by keeping the trifluoromethoxy group at the
paraꢀposition and, simultaneously, introducing minimally sized hydrophobic groups, such as a
fluorine atom into the metaꢀ or orthoꢀpositions. Incorporation of a fluorine atom into the metaꢀ
position (17) led to a 2ꢀfold gain in inhibitory activity, whereas substitution of a fluorine atom at
the orthoꢀposition (18) was detrimental to the potency relative to 6. Taken together, the
exploration of various substituents on the RHS phenyl group yielded some promising substituted
phenyl groups, as typified by 6, 12, and 17.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 3. SAR of the RHS Phenyl Group
a
b
c
compd
Ar
PDE2A IC (nM)
PDE selectivity
LogD
3.61
5
0
6
78 (61–98)
41ꢀfold (PDE1A)
NT
11
610 (480–780)
2.78
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
12
72 (46–110)
80 (64–100)
4.2ꢀfold (PDE1A)
3.45
3.75
3.13
NT
13
14
15
16
4.4ꢀfold (PDE1A)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
500 (410–590)
4100 (3400–4900)
8100 (5500–18000)
42 (33–52)
NT
NT
NT
3.52
3.67
17
15ꢀfold (PDE1A)
18
210 (160–290)
28ꢀfold (PDE1A)
3.83
a
IC50 values (95% confidence intervals given in parentheses) were calculated from percent
inhibition data (duplicate, n = 1). All values are rounded to two significant digits. Minimum
selectivity (rounded to two significant digits) over other PDEs. The least selective PDE was
given in parenthesis. LogD values at pH 7.4.
b
c
Strategy 3: Conversion of the Branched Ethyl Group. We next focused on varying the
branched ethyl group in 6 (Table 4). On the basis of our aboveꢀdescribed hypothesis, polar
functionalities containing HBA or HBD were introduced to exploit new interactions with
adjacent water molecules or polar amino acid residues such as His656 and Gly826, around metal
ions (Figure 4). Compound 19, possessing a methoxymethyl group, exhibited a 4ꢀfold increase in
potency and significant improvement in PDE selectivity. This modification also significantly
decreased the LogD value by nearly 1 unit. Elongation of the side chain by one carbon atom (20)
ACS Paragon Plus Environment

Page 19 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
did not affect the potency in comparison with 6. Moreover, replacement of the oxygen atom of
19 with a methylene group (21) decreased the potency by an order of magnitude. These results
indicate that the ether oxygen atom in 19 is likely engaged in an additional interaction.
Encouraged by this result, we pursued other polar functionalities. The primary alcohol in 22 was
tolerated, but had no significant impact on either potency or PDE selectivity compared with 6,
whereas the corresponding tertiary alcohol in 23 increased potency and PDE selectivity,
suggesting that the gemꢀdimethyl group may assist in suitably positioning the alcohol moiety by
restricting bond rotation. Moreover, 24, possessing a sulfone group, displayed a similar potency
enhancement to 23, demonstrating that a range of HBAs or HBDs could contribute to potency
enhancement. Thus, the installation of polar functionalities not only enhanced potency and PDE
selectivity, but also of reduced LogD values; however, these advantageous changes were
accompanied by increased TPSA values and, in some cases, HBD counts outside the range of our
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
aforementioned criteria (i.e., TPSA < 110 Å , HBD count < 3), resulting in unacceptable MDR1
efflux ratios, as evidenced by 22–24. On the basis of improved PDE2A inhibitory activity and
PDE selectivity in conjunction with a significant reduction in lipophilicity and no sign of Pꢀgp
efflux, the methoxymethyl branched group in 19 was selected for further optimization.
Table 4. SAR of the Branched Group at the Benzylic Position of the RHS Amine Moiety
a
b
c
2
d
compd
R
PDE2A IC (nM)
PDE selectivity
41ꢀfold (PDE1A)
LogD
3.61
TPSA (Å ) MDR1
84 0.53
50
6
–Et
78 (61–98)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
2
2
9
0
1
–CH
–CH
–nꢀPr
–CH
–C(Me)
–CH SO
2
2
OMe
CH OMe
19 (15–24)
310ꢀfold (PDE1A) 2.95
93
93
84
104
104
118
0.68
62 (37–100)
250 (190–330)
65 (35–120)
24 (18–33)
14ꢀfold (PDE5)
NT
54ꢀfold (PDE1A)
3.09
4.05
2.16
0.77
0.59
2.8
2.8
26
2
22
2
OH
OH
Me
2
3
4
2
150ꢀfold (PDE1A) 2.54
220ꢀfold (PDE1A) 1.91
2
2
2
29 (22–38)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
IC50 values (95% confidence intervals given in parentheses) were calculated from percent
inhibition data (duplicate, n = 1). All values are rounded to two significant digits. Minimum
selectivity (rounded to two significant digits) over other PDEs. The least selective PDE was
given in parenthesis. LogD values at pH 7.4. MDR1 efflux ratios in Pꢀgp overexpressing cells.
b
c
d
Strategy 4: Introduction of Substituents into the Core. A comparison of the aligned
sequences of all PDEs, particularly around the core (i.e., Tyr827, Leu858, and Met847 for
5
3,54
PDE2A), revealed some variation of amino acid residues across the PDE family.
Such
residue variation provides each member of the PDE family with various degrees of differences in
the shape, size, and polarity of the binding pocket. In support of this finding, introduction of a
substituent into the pyrazolo[1,5ꢀa]pyrimidine core significantly affected both potency and PDE
40
selectivity. Therefore, targeting such residue differences around the core could be expected to
improve both inhibitory activity and selectivity. Thus, we explored substitutions on the 3,4ꢀ
dihydropyrido[2,3ꢀb]pyrazinꢀ2(1H)ꢀone core of 19, while maintaining the methoxymethyl
branched linker, as summarized in Table 5. Introduction of a methyl group into the 6ꢀposition
(25) or 8ꢀposition (27) decreased inhibitory activity, whereas methyl substitution at the 7ꢀ
position in 26 resulted in a 2ꢀfold boost in potency with the requisite selectivity of 1100ꢀfold
over other PDEs. Armed with these promising results, our focus was directed toward varying the
substituent at the 7ꢀposition. The more lipophilic and bulkier cyclopropyl group in 28 further
increased the potency concomitant with an improvement of PDE selectivity. On the other hand,
the electronꢀwithdrawing chloro group in 29 was detrimental to both potency and PDE
selectivity. Based on these results, we presumed that incorporation of an electronꢀdonating
ACS Paragon Plus Environment

Page 21 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
substituent, such as an alkyl group, at this position may positively affect the potency in two
ways: (i) space filling around the 7ꢀposition increases interactions with the surrounding amino
acid residues and (ii) increased electron density of the core results in enhanced π–π and CH–π
interactions with Phe862 and Ile826. Although 7ꢀalkyl analogs provided notable advantages with
respect to PDE2A inhibitory activity and PDE selectivity, they suffered from high lipophilicity
as evidenced by higher LogD values. Therefore, we introduced other less lipophilic substituents
at the 7ꢀposition. The methoxy group in 30 produced a minimal increase in lipophilicity with
comparable potency to and improved selectivity over cyclopropyl analog 28. On the other hand,
larger substituents, such as an isopropoxy group (31), were not well tolerated, suggesting limited
space in the binding pocket around the 7ꢀposition. It is also worth noting that 30 possessed a
TPSA value within our target range, and consequently, an acceptable MDR1 efflux ratio, despite
the incorporation of the additional methoxy group, thus validating our initial criteria for TPSA
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
values (TPSA < 110 Å ) of taking into account the expected effect of an intramolecular hydrogen
bond.
Table 5. SAR of Substitutions at the 3,4ꢀDihydropyrido[2,3ꢀb]pyrazinꢀ2(1H)ꢀone Core
a
b
c
TPSA
d
compd
R
PDE2A IC (nM)
PDE selectivity
LogD
2
MDR1
5
0
(
Å )
93
93
93
1
2
2
9
5
6
H
6ꢀMe
7ꢀMe
19 (15–24)
40 (31–52)
8.7 (7.6–10)
310ꢀfold (PDE1A)
50ꢀfold (PDE1A)
1100ꢀfold (PDE5)
2.95
3.38
3.35
0.68
0.63
1.0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
2
7
8
9
8ꢀMe
7ꢀcꢀPr
7ꢀCl
7ꢀMeO
7ꢀiPrO
85 (67–110)
NT
3.16
93
93
93
102
102
0.74
1.0
0.58
1.3
3.5 (3.0–4.1)
33 (30–36)
2.8 (2.6–3.1)
77 (66–90)
1300ꢀfold (PDE3A) 3.85
210ꢀfold (PDE5)
2700ꢀfold (PDE5)
NT
3.74
3.20
3.94
30
3
1
1.1
a
IC50 values (95% confidence intervals given in parentheses) were calculated from percent
inhibition data (duplicate, n = 1). All values are rounded to two significant digits. Minimum
selectivity (rounded to two significant digits) over other PDEs. The least selective PDE was
given in parenthesis. LogD values at pH 7.4. MDR1 efflux ratios in Pꢀgp overexpressing cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
b
c
d
Profiling Enantiomers of 30 and Its Close Analogs. Having identified 30 as a promising
racemic compound that met our initial objectives of potency and PDE selectivity, the compound
was subjected to chiral separation and in vitro profiles were obtained for both enantiomers. In
addition, the RHS phenyl portion of 30 was replaced with other promising counterparts,
exemplified by 12 and 17 (Table 4), to afford eutomers 34 and 36, respectively, after chiral
separation (Table 6). Compound 32 displayed potent inhibitory activity and excellent selectivity
against other PDEs, whereas 34 was less potent and selective than expected. Of the synthesized
analogs, 36 was the best with respect to potency (IC = 0.61 nM) and PDE selectivity (4100ꢀ
5
0
fold vs. PDE1A; for a more detailed selectivity profile, refer to Table 7). In addition, 36 showed
a favorable MDR1 efflux ratio with a TPSA value in the range of our criteria. On the basis of it
possessing the best combination of potency, PDE selectivity and MDR1 efflux ratio, 36 was
selected for further in vitro and in vivo profiling.
a
Table 6. In Vitro Profiles and Physicochemical Properties of Analogs 32–37
ACS Paragon Plus Environment

Page 23 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b
c
d
2
e
compd
stereo
PDE2A IC (nM)
PDE selectivity
LogD TPSA (Å ) MDR1
5
0
32
33
34
R
S
1.6 (1.4–1.9)
3300ꢀfold (PDE1A)
3.19
–
102
–
1.5
–
39000 (28000–54000)
7.2 (6.4–8.0)
21000 (10000–42000)
0.61 (0.53–0.70)
910 (720–1200)
–
f
NDf
83ꢀfold (PDE1A)
3.03
–
93
–
0.79
–
35
ND
R
–
3
a3
6
7
4100ꢀfold (PDE1A)
–
3.32
–
102
–
0.87
–
S
Racemic compounds were chirally separated and each enantiomer was profiled. The absolute
configurations of 33 and 36 were determined as S and R, respectively, via single crystal Xꢀray
analysis of the corresponding RHS benzylamine precursors (see the Supporting Information).
IC50 values (95% confidence intervals given in parentheses) were calculated from percent
inhibition data (duplicate, n = 1). All values are rounded to two significant digits. Minimum
selectivity (rounded to two significant digits) over other PDEs. The least selective PDE was
given in parenthesis. LogD values at pH 7.4. MDR1 efflux ratios in Pꢀgp overexpressing cells.
b
c
d
e
f
Not determined.
Table 7. PDE Selectivity Profile of 36
a
b
PDE subtypes
IC (nM)
selectivity ratio
50
PDE1A
PDE2A3
PDE3A
2497
0.61
>30000
14882
>30000
>30000
>30000
>30000
>30000
>30000
>30000
4100
–
>49000
24000
>49000
>49000
>49000
>49000
>49000
>49000
>49000
PDE4D2
PDE5A1
PDE6AB
PDE7B
PDE8A1
PDE9A2
PDE10A2
PDE11A4
a
IC50 values were calculated from percent inhibition data (duplicate, n = 1). All values are
rounded to two significant digits. Selectivity ratio (rounded to two significant digits) = PDE''X''
IC /PDE2A IC .
b
5
0
50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As shown in Table 8, 36 exhibited relatively high, but acceptable, rat and mouse liver
microsomal clearance, and demonstrated increased metabolic stability in human liver
microsomes. PK evaluations of 36 in rats and mice were favorable with excellent oral
bioavailability in both species, and moderate (rat) to low (mouse) systemic clearances. Moreover,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
36 demonstrated excellent brain permeability with a brainꢀtoꢀplasma ratio of 1.01 in rats and 0.91
in mice. These PK results, together with improved human in vitro microsomal stability relative to
rat and mouse counterparts, indicated the potential of 36 for good oral absorption in humans.
a
Table 8. PK Parameters and K Values of 36 in Rat and Mouse
p
rat
mouse
human
b
Metabolic stability (ꢂL/min/mg)
84
39.3
3429
64.2
1.7
242.5
56.7
1.01
73
8.6
1602
201.0
1.7
1076.8
55.6
0.91
33
–
–
–
–
–
–
–
c
CL
(mL/min/kg)
total
d
Vd (mL/kg)
ss
e
Cmax (ng/mL)
Tmax (h)
f
g
AUC
(ng·h/mL)
0
–8h
h
F (%)
i
Kp
a
b
d
Cassette dosing at 0.1 mg/kg, iv and 1 mg/kg, po (nonꢀfasted). Average of three rats or mice.
c
Metabolic stability was determined by incubation with liver microsomes. Total clearance.
Volume of distribution at steady state. Maximum plasma concentration. Time of maximum
e
f
g
h
concentration. Area under the plasma concentration vs time curve (0–8 h). Oral bioavailability.
i
Brainꢀtoꢀplasma ratio at 2 h after oral administration of 36 at a dose of 10 mg/kg.
X-ray Crystal Structure of 36 Bound in the PDE2A Catalytic Domain. To rationalize the
structural basis for the potent PDE2A inhibitory activity and excellent PDE selectivity of 36, its
Xꢀray coꢀcrystal structure in complex with PDE2A was determined (Figure 8A and 8B). Close
inspection of the obtained crystal structure revealed that 36 binds to PDE2A in a manner similar
to the binding mode of 4b, and that the R configuration is the active form. As expected, the
ACS Paragon Plus Environment

Page 25 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
lactam moiety of the 3,4ꢀdihydropyrido[2,3ꢀb]pyrazinꢀ2(1H)ꢀone core forms a hydrogen bonding
interaction with the conserved Gln859 residue as in cGMPꢀlike binding pose 1 (Figure 6A),
whereas the core is sandwiched between Phe862 and Phe830/Ile826 and stabilized by π–π and
CH–π interactions. The methoxy group at the 7ꢀposition of the core serves two purposes with
respect to potency and PDE selectivity. First, its electronꢀdonating properties enhance the π–π
and CH–π interactions of the bicyclic core with Phe862 and Phe830/Ile826, which accounts for
the potency increase over the nonꢀsubstituted analog. Second, it efficiently occupies the small
hydrophobic cavity consisting of the side chains of Tyr827, Leu858, and Met847, which is likely
one of the key factor for enhanced potency. Utilization of this pocket also improved selectivity
over PDE1A, as PDE2A possesses a more hydrophobic binding environment in this pocket
owing to the presence of Leu858 in contrast to the more polar Ser416 in PDE1A. We, therefore,
reasoned that such distinctions in interactions with PDE2A vs PDE1A may improve selectivity
over PDE1A. Further, the carbonyl group of the urea portion interacts with the OH group of
Tyr655 via a bridging water molecule. The branched methoxymethyl group is oriented toward
the solvent accessible region, most likely leading to a partial gain in solvation energy, although
no explicit interactions with water molecules were observed. In addition to the expected
intramolecular hydrogen bond between the urea NH and the N5 nitrogen atom of the core
scaffold, the oxygen atom in the branched methoxymethyl group forms another key
intramolecular hydrogen bond with the urea NH. These two separate internal hydrogen bonds
would assist additively in locking the RHS phenyl moiety into an optimal conformation for
effectively occupying the bindingꢀinduced hydrophobic pocket.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 27 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Xꢀray crystal structure of 36 bound in the PDE2A catalytic site (PDB 5VP0) viewed
from the top (A) and the entrance of the catalytic site (B). The key hydrogen bonding
interactions of 36 with PDE2A and the intramolecular hydrogen bond in 36 are indicated by red
and yellow dotted lines, respectively.
In Vivo Pharmacology of 36. Compound 36 was assessed for its ability to increase the levels of
cyclic nucleotides in the hippocampus, one of the brain regions where PDE2A is particularly
6
–11
localized (Figure 9).
Oral dosing of 36 (3 or 10 mg/kg) in mice produced a doseꢀdependent
increase in cGMP levels, with significant cGMP increases observed at a dose of 10 mg/kg.
However, 36 did not have an appreciable impact on cAMP (data not shown). These results are in
good agreement with and further validate previous observations using our lead compound 4b or
23,37,40,55
other reported compounds,
in which PDE2A inhibitors predominantly contribute to the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 97
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
elevation of cGMP under physiological conditions. We next examined the effect of 36 on
5
6
cognitive performance in a novel object recognition (NOR) task in rats, predictive of potential
proꢀcognitive activity of drugs. The test was designed to take advantage of the instinctive
preference of rats to explore novel objects rather than familiar objects. In this study, 36 was
orally administered 2 h prior to the acquisition trial and the exploration time for two identical
objects was recorded (Figure 10A). After 48 h of the acquisition trial, one of the familiar objects
was replaced with a novel object, the time spent investigating each of the objects was recorded
(Figure 10B), and the novelty discrimination index (NDI) was calculated as the percentage of
novel object interaction time relative to total interaction time in the retention trial (Figure 10C).
As shown in Figure 10A, oral administration of 36 (0.01–1 mg/kg) did not significantly affect
the total exploration time in the acquisition trial. On the other hand, during the retention trial
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Figure 10B), rats treated with 0.1 and 1 mg/kg of 36 explored the novel object for a greater time,
indicative of preserved memory for the familiar object presented during the acquisition trial,
whereas rats in vehicle conditions or dosed with 0.01 mg/kg of 36 did not exhibit statistically
meaningful differences between exploration times for familiar and novel objects, indicating
deterioration or loss of memory for the familiar object. In addition, a 1 mg/kg oral dose of 36
significantly increased the NDI, as shown in Figure 10C. These results suggest that 36 enhanced
recognition memory in a NOR task in rats. Inꢀdepth studies of the efficacy of 36 in various
behavioral animal models relevant to schizophrenia will be reported in due course.
ACS Paragon Plus Environment

Page 29 of 97
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 9. Effect of 36 on cGMP contents in mouse hippocampi. Male C57BL/6J mice were
sacrificed by focused microwave irradiation of the brain 60 min after administration of 36.
Values are expressed as pmol/mg tissue weight (mean ± S.E.M., n = 12 per each group). The
statistical significance was determined by a oneꢀtailed Williams’ test with significance set at #p ≤
0.025 (vs vehicle group).
ACS Paragon Plus Environment