Structural basis for potent inhibition of D-amino acid oxidase by thiophene carboxylic acids

Article abstract of DOI:10.1016/j.ejmech.2018.09.040

A series of thiophene-2-carboxylic acids and thiophene-3-carboxylic acids were identified as a new class of DAO inhibitors. Structure-activity relationship (SAR) studies revealed that small substituents are well-tolerated on the thiophene ring of both the 2-carboxylic acid and 3-carboxylic acid scaffolds. Crystal structures of human DAO in complex with potent thiophene carboxylic acids revealed that Tyr224 was tightly stacked with the thiophene ring of the inhibitors, resulting in the disappearance of the secondary pocket observed with other DAO inhibitors. Molecular dynamics simulations of the complex revealed that Tyr224 preferred the stacked conformation irrespective of whether Tyr224 was stacked or not in the initial state of the simulations. MM/GBSA indicated a substantial hydrophobic interaction between Tyr244 and the thiophene-based inhibitor. In addition, the active site was tightly closed with an extensive network of hydrogen bonds including those from Tyr224 in the stacked conformation. The introduction of a large branched side chain to the thiophene ring markedly decreased potency. These results are in marked contrast to other DAO inhibitors that can gain potency with a branched side chain extending to the secondary pocket due to Tyr224 repositioning. These insights should be of particular importance in future efforts to optimize DAO inhibitors with novel scaffolds.

Full text of DOI:10.1016/j.ejmech.2018.09.040

Accepted Manuscript
Structural basis for potent inhibition of D-amino acid oxidase by thiophene carboxylic
acids
Yusuke Kato, Niyada Hin, Nobuo Maita, Ajit G. Thomas, Sumire Kurosawa, Camilo
Rojas, Kazuko Yorita, Barbara S. Slusher, Kiyoshi Fukui, Takashi Tsukamoto
PII:
S0223-5234(18)30812-2
10.1016/j.ejmech.2018.09.040
EJMECH 10751
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 27 April 2018
Revised Date: 10 July 2018
Accepted Date: 14 September 2018
Please cite this article as: Y. Kato, N. Hin, N. Maita, A.G. Thomas, S. Kurosawa, C. Rojas, K.
Yorita, B.S. Slusher, K. Fukui, T. Tsukamoto, Structural basis for potent inhibition of D-amino acid
oxidase by thiophene carboxylic acids, European Journal of Medicinal Chemistry (2018), doi: https://
doi.org/10.1016/j.ejmech.2018.09.040.
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.

ACCEPTED MANUSCRIPT

Kato et al.
ACCEPTED MANUSCRIPT
1
2
Structural basis for potent inhibition of D-amino acid
oxidase by thiophene carboxylic acids
3
4
Yusuke Katoa, Niyada Hinb, Nobuo Maitaa, Ajit G. Thomasb, Sumire Kurosawaa,
Camilo Rojasb, Kazuko Yoritaa, Barbara S. Slusherb,c, Kiyoshi Fukuia,* and
Takashi Tsukamotob,c
5
6
7
8
aInstitute for Enzyme Research, Tokushima University, Tokushima 770-8503,
Japan.
9
c
10
11
12
13
14
15
bJohns Hopkins Drug Discovery and Department of Neurology, Johns Hopkins
University, Baltimore, MD 21205, USA
*Corresponding
author.
Tel.:
+81-88-633-7429;
e-mail:
kiyo.fukui@tokushima-u.ac.jp
16 Abstract
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
A series of thiophene-2-carboxylic acids and thiophene-3-carboxylic acids
were identified as a new class of DAO inhibitors. Structure-activity relationship
(SAR) studies revealed that small substituents are well-tolerated on the
thiophene ring of both the 2-carboxylic acid and 3-carboxylic acid scaffolds.
Crystal structures of human DAO in complex with potent thiophene carboxylic
acids revealed that Tyr224 was tightly stacked with the thiophene ring of the
inhibitors, resulting in the disappearance of the secondary pocket observed with
other DAO inhibitors. Molecular dynamics simulations of the complex revealed
that Tyr224 preferred the stacked conformation irrespective of whether Tyr224
was stacked or not in the initial state of the simulations. MM/GBSA indicated a
substantial hydrophobic interaction between Tyr244 and the thiophene-based
inhibitor. In addition, the active site was tightly closed with an extensive network
of hydrogen bonds including those from Tyr224 in the stacked conformation. The
introduction of a large branched side chain to the thiophene ring markedly
decreased potency. These results are in marked contrast to other DAO inhibitors
that can gain potency with a branched side chain extending to the secondary
pocket due to Tyr224 repositioning. These insights should be of particular
1

Kato et al.
ACCEPTED MANUSCRIPT
1
2
importance in future efforts to optimize DAO inhibitors with novel scaffolds.
3
Keywords
4
Flavoenzyme; schizophrenia; drug discovery; X-ray crystallography; molecular
dynamics
5
6
7
Highlights
8
· Therapeutics with DAO/DAAO inhibition is a potential approach to treat
schizophrenia.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
· Thiophene carboxylic acids were identified as a new class of DAO inhibitors.
· Tyr224 of DAO was tightly stacked with the thiophene ring of the inhibitors.
· The hydrophobic interaction and hydrogen bonds between them induced the
stacking.
· The results should be important in future efforts to optimize DAO inhibitors.
Abbreviations
DAO/DAAO, D-amino acid oxidase; D-Ser, D-serine; SAR, Structure-activity
relationship; MD, molecular dynamics; H-bond, hydrogen bond; DOPA,
dihydroxyphenylalanine; CPC, 4-(4-chlorophenethyl)-1H-pyrrole-2-carboxylic
acid; TPC, 4H-thieno[3,2-b]pyrrole-5-carboxylic acid; MM/GBSA, molecular
mechanics energies combined with the generalized Born and surface area
24 1. Introduction
25
26
27
28
29
30
31
32
33
D-Amino acid oxidase (DAO/DAAO) is a flavoenzyme that catalyzes the
oxidation of D-amino acids, producing the corresponding α-keto acids, ammonia,
and hydrogen peroxide [1]. One of the endogenous substrates for DAO in the
brain is the co-agonist of NMDA receptors, D-serine (D-Ser) [2-4]. In the brains
of patients with schizophrenia, the amount of D-Ser decreases presumably due
to increased DAO activity [5-10]. Since NMDA receptor hypofunction is believed
to play a pathophysiological role in the negative symptoms and cognitive
impairment of schizophrenia, inhibition of DAO has been of great interest as a
therapeutic approach distinct from those targeting dopaminergic pathways
2

Kato et al.
ACCEPTED MANUSCRIPT
1
2
[11-13]. Indeed, the past decade has seen a wave of medicinal chemistry efforts
in the search for new DAO inhibitors [14, 15].
3
Nearly all DAO inhibitors possess a carboxylic acid or its bioisostere which
interacts with Tyr228 and Arg283 residues that are responsible for recognizing
the carboxylate group of D-amino acid substrates. The majority of DAO inhibitors
also contain an aromatic ring despite DAO’s ability to oxidize a wide range of
neutral D-amino acids including those with an aliphatic chain [16, 17]. This is at
least partially due to the ability of the aromatic ring to form a π-π interaction with
FAD’s isoalloxazine ring [18]. According to the structural studies, the aromatic
ring of benzoate and its derivative are stacked with the side chain of Tyr224
[18-20]. However, other inhibitors and products including
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
4H-thieno[3,2-b]pyrrole-5-carboxylic acid (TPC),
4-(4-chlorophenethyl)-1H-pyrrole-2-carboxylic acid (CPC) and imino DOPA are
not stacked with Tyr224 because the side chain of Tyr224 moves away from the
active site [21, 22]. The stacked and displaced states of Tyr224 are referred to
as S and D states, respectively, in this paper. The conformational flexibility of
Tyr224 plays a critical role in the structural plasticity of the substrate-binding site
of DAO, which catalyzes a wide range of D-amino acid substrates. In the D state,
an additional pocket (referred to as the secondary pocket) is created as a result
of the movement of Tyr224 to accommodate DAO inhibitors/products with a
branched side chain (see Fig. 1B, D). Although the precise mechanism by which
the conformation of Tyr224 is regulated is poorly understood, the secondary
pocket has been exploited in a number of new DAO inhibitors containing a
branched chain [22-24].
In a search for new scaffolds that inhibit DAO, we screened a variety of
aromatic carboxylic acids. Among them, thiophene-2-carboxylic acid 1a and
thiophene-3-carboxylic acid 2a (Table 1) exhibited low micromolar inhibitory
potency with IC50 values of 7.8 µM and 4.4 µM, respectively. Herein we report
structure-activity relationship (SAR) studies on the two thiophene-based
scaffolds as well as X-ray crystallographic analysis and molecular dynamics
simulations of the complex between thiophene-based compounds and DAO to
elucidate the mechanism underlying their potent interactions.
3

Kato et al.
ACCEPTED MANUSCRIPT
1
2
2. Results
2.1. Inhibition of D-amino acid oxidase by low molecular weight thiophene
carboxylic acids
3
4
Given that a number of D-amino acid oxidase (DAO/DAAO) inhibitors
reported to date are aryl carboxylic acids, we screened a variety of molecules in
this category in a search for new scaffolds that inhibit DAO. Our screening
efforts identified thiophene-2-carboxylic acid 1a and thiophene-3-carboxylic acid
2a as low micromolar DAO inhibitors with IC50 values of 7.8 µM and 4.4 µM,
respectively. While compound 2a was previously reported as a DAO inhibitor
[25], compound 1a represents a new scaffold for DAO inhibition. Although other
aryl carboxylic acids were previously reported to exhibit substantially higher
inhibitory potency, the low molecular weights of these compounds present
attractive structural features as lead compounds and prompted us to evaluate
their analogs in the DAO assay. In the first phase of SAR studies, we examined
analogs with minimal changes in molecular size. All compounds but one
(compound 1j) were commercially available. As shown in Scheme 1, compound
1j was obtained by fluorination of aldehyde 3 followed by hydrolysis of the
methyl ester group.
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
The results are summarized in Table 1. Among compounds with the
thiophene-2-carboxylic acid scaffold, many of the 5-substituted analogs,
particularly those with a small substituent, potently inhibited DAO. For instance,
5-fluoro (1b), 5-chloro (1c) and 5-bromo (1d) analogs exhibited substantial
improvement compared to the parent compound 1a. While 5-methyl (1e) was
found to be nearly as potent as 1a, a gradual decrease in potency was seen with
the increase in the size of the 5-substituents as shown by 5-difluoromethyl (1f)
and 5-trifluoromethyl (1g) analogs. Inhibitory activity was completely abolished
when a formyl group was incorporated into the 5-position (1h). Small
substituents were also well tolerated in the 4-position as seen for 1i and 1j.
Interestingly, 4,5-disubsituted analogs such as 1k and 1l represented the most
potent DAO inhibitors within the thiophene-2-carboxylic acid series with IC50
values of 0.09 and 0.36 µM, respectively. In contrast, any substitution at the
3-position appears to be detrimental to inhibitory activity as neither 3-fluoro (1n)
nor 3-methyl (1o) analogs inhibited DAO. Incorporation of carboxylic acid
4

Kato et al.
ACCEPTED MANUSCRIPT
1
2
bioisosteres such as tetrazole (1p and 1q) and boronic acid (1r) into the
2-position also resulted in a complete loss of potency. As for the analogs of
thiophene-3-carboxylic acid 2a, 5-chloro (2b) and 5-methyl (2c) derivatives
showed substantially improved potency as compared to the parent compound
2a. Indeed, 5-chlorothiophene-3-carboxylic acid 2b was the most potent
thiophene-carboxylic acid-based DAO inhibitors with an IC50 value of 0.04 µM.
2,5-Dichloro analog 2d, however, exhibited much weaker inhibitory potency.
3
4
5
6
7
8
9
2.2. Preference for the S state in the complexes with 1c and 2b
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
For a better understanding of the good potency of low molecular thiophene
carboxylic acids, we determined the crystal structures of the 1c-DAO and
2b-DAO complexes (Supplementary Table 1, Figs. 1, 2). We found that the
thiophene rings of the thiophene-2-carboxylic 1c and thiophene-3-carboxylic 2b
acid analogs are stacked with the benzene ring of Tyr224. Thus, both of the
complexes are in the S state in which the formation of the secondary pocket is
lost due to the movement of Tyr224 (Fig. 1). The shapes and locations of 1c and
2b in the complexes are almost superimposable, with the exception of the sulfur
atom of the thiophene ring (Fig. 2). Slight differences were observed for the
shapes of the thiophene rings and the orientations of the carboxylate and
chlorine atom. These may cause the difference in the IC50 values between 1c
and 2b (Supplementary Fig. 1).
To assess the stability of the S state conformation, we performed all-atom
molecular dynamics (MD) simulations using the dimer crystal structure of the
2b-DAO complex as an initial structure. We measured the distance between the
centroids of the thiophene ring of 2b and benzene ring of Tyr224 to judge
whether Tyr224 was in the S or D state. The crystal structure of the 2b-DAO
complex shows a distance of ~4 Å between the two centroids. Based on visual
inspection of the trajectories of the 2b-DAO complex, we defined the S state as
having a ~4 Å distance between the two centroids and the D state as having a
distance that was ≥ ~5 Å .
The distance between the centroids of the thiophene ring of 2b and benzene
ring of Tyr224 of Chain A was ~4 Å most of the simulation time, while it
occasionally became ≥ 5 Å (Fig. 3A). Thus the S state was dominant in the
5

Kato et al.
ACCEPTED MANUSCRIPT
1
2
equilibration between the S and D states. We measured the corresponding
distance in Chain B as well in the same MD run; the results were reproducible
(Supplementary Fig. 2). Repeated MD runs again gave similar results.
In contrast, the crystal structure of the TPC-DAO complex (PDB code: 3znn)
shows the D state as seen in the imino DOPA-DAO complex [21, 22]. The
secondary pocket in the TPC-DAO complex is unoccupied, because TPC is a
planar molecule without a branched side chain (Supplementary Table 2). Thus,
it would be possible to form the S state for the TPC-DAO complex without a
steric clash between TPC and DAO’s residues including Tyr224. However, MD
using 3znn as an initial structure indicated that the distance between the
centroids of the pyrrole ring of TPC and benzene ring of Tyr224 remained ~5 Å
most of the time (i.e. the D state) (Fig. 3B). Repeated MD runs gave similar
results reproducibly (Supplementary Fig. 3).
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
To exclude the possibility that the initial structures biased the results above,
we performed additional MD runs with a virtual initial state in which 2b was
substituted for TPC in the D state TPC-DAO complex (Fig. 3C, Supplementary
Fig. 4). The results showed that the S state became dominant within a few
nanoseconds after the simulations were initiated. Conversely, MD runs with an
initial state in which TPC was substituted for 2b in an S state structure showed
that the D state became dominant over time (Fig. 3D, Supplementary Fig. 5).
These simulations suggest that the S state is thermodynamically preferred for
the 2b-DAO complex while DAO adopts the D state when TPC is bound to its
active site.
2.3. Thiophene carboxylic acids containing a branched chain as DAO
inhibitors
In light of the previous findings that some DAO inhibitor scaffolds benefit from
an added branched side chain that occupies the secondary pocket, we
examined whether such a modification can also improve the inhibitory potency of
the thiophene carboxylic acid scaffolds. The previously reported SAR studies
indicate that the incorporation of a side chain to the position across from the
carboxylate attached carbon is most effective in other aryl carboxylic acid-based
DAO inhibitors. This prompted us to evaluate 4- or 5-substituted derivatives of
6

Kato et al.
ACCEPTED MANUSCRIPT
1
2
thiophene-2-carboxylic acid (compounds 1s-w) as well as 5-substituted
derivatives of thiophene-3-carboxylic acid (compounds 2e-f). While some of
these compounds were commercially available, compounds 1u, 1w, and 2f were
synthesized using bromothiophenes as starting materials. The key steps
involved in the synthesis include Sonogashira coupling and subsequent catalytic
hydrogenation as illustrated in Schemes 2 and 3.
3
4
5
6
7
As summarized in Table 2, none of these compounds showed substantial
inhibitory activity against DAO. The lack of potency seen with these compounds
represent a sharp contrast to other aryl carboxylic acid scaffolds that benefited
from side chain incorporation. For example, a 4-substituted pyrrole-2-carboxylic
acid discovered by Sunovion, SEP-137, inhibits DAO with a markedly higher
potency [22] than the unsubstituted pyrrole-2-carboxylic acid [26]. Similar
structural modifications to 1a and 2a, however, led to a complete or significant
loss of potency as demonstrated by compounds 1u, 1w, and 2f. It is worth noting
that a much smaller substituent such as ethyl group is sufficient enough to
eliminate the ability to inhibit DAO as seen with compounds 1s, 1v, and 2e even
though the corresponding methyl substituted derivatives 1e, 1i, and 2c showed
potent DAO inhibition (Table 1). These results suggest that DAO is incapable of
accommodating thiophene carboxylic acids with a branched side chain larger
than a methyl group. The S state observed in the co-crystal structures of DAO
with 1c and 2b appears to have little flexibility in responding to the branched side
chain added to the thiophene ring by shifting to the D state and creating the
secondary binding pocket.
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
2.4. Mechanism to stabilize the S state
We calculated the binding free energy (∆Gbind) by molecular mechanics
energies combined with the generalized Born and surface area (MM/GBSA)
using MD trajectories to compare ∆Gbind with ∆Gexp that was derived from
experimental IC50 (Table 3). ∆Gbind and ∆Gexp were in good agreement with each
other for the interactions between 2b and DAO and between TPC and DAO. The
energy decomposition of ∆Gbind on a per residue basis calculated that the
contribution of Tyr224 to ∆Gbind was greater in the 2b-DAO interaction than in
the TPC-DAO interaction. We found a notable difference in ∆Gvdw,Y224 between
7

Kato et al.
ACCEPTED MANUSCRIPT
1
2
the 2b-DAO and TPC-DAO interactions (Table 4). This suggests that Tyr224
contributes to the interaction with low molecular thiophene carboxylic acid-based
inhibitors primarily through hydrophobic interactions.
3
4
To evaluate the contributions of hydrogen bonds (H-bonds) to the 2b-DAO
and TPC-DAO interactions, occupancy of each H-bond between molecules
around the active site was calculated (Fig. 4A). Characteristic differences
between these interactions were observed in the H-bonds between the inhibitors
and Gly313 and between H2O and Tyr224. In the 2b-DAO complex, a H-bond
network composed of Gln53, Pro54, His217, Tyr224, Gly313 and bridging H2O
molecules contributed to stabilization of the S state (Fig. 4B). In contrast, a
H-bond network including Tyr224 was not that extensive in the TPC-DAO
complex and partly explaining the preferred D state for this complex (Fig. 4C).
Taken together the MM/GBSA and H-bond analyses suggest that the
2b-Tyr224 interaction and formation of the H-bond network around Tyr224 were
driving forces to form the S state. This was supported by a finding that Tyr224
underwent the largest conformational change among the inhibitor-interacting
residues in the comparison of the structures of the S and D states (Fig. 5A).
Leu51 and His217 also changed their conformations as a result of direct
interactions with the inhibitors but to a lesser extent. Although the extent of
conformational changes of Tyr55 and Ile223 appeared significant, these
changes were indirectly influenced by the binding of the inhibitors. Consequently,
Loop 216−228 (refereed to as the lid) and Loop 53-62 approached each other in
the 1c-DAO and 2b -DAO complexes. These structural changes were
accompanied by a closing of a cleft between the lid and Loop 53-62 to sequester
the substrate binding pocket from the surrounding environment (Fig. 5B). In
contrast, the cleft is half-open with the lid relaxed in the TPC-DAO complex, in
which water molecules can pass through from the inside of the pocket to the
surrounding environment (Fig. 5C).
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
As a result of the closing of the cleft, the substrate-binding pockets of the
1c-DAO and 2b-DAO complexes appeared to shrink compared with the
complexes in the D state including the TPC-DAO complex (Tables 5 and 6). The
extent of shrinkage was the greatest in the 2b-DAO complex. In addition,
comparison between the averages throughout MD trajectories also indicated a
8

Kato et al.
ACCEPTED MANUSCRIPT
1
2
greater extent of shrinkage in the 2b-DAO complex than in the TPC-DAO
complex.
3
4
3. Discussion
5
The highly potent DAO/DAAO inhibitors 1c and 2b derived from the
6
thiophene-2-carboxylic acid and thiophene-3-carboxylic acid scaffolds,
respectively, allowed us to investigate the structural basis for the potent DAO
inhibition achieved by such small molecules. Crystal structures and MD analysis
of DAO in complex with the thiophene-based inhibitors suggested that the
complexes prefer the S state and that the formation of the S state is driven by
direct interactions of Tyr224 with the inhibitors and H2O. In contrast, the D state
was preferred in the TPC-DAO complex during our MD simulations as shown in
the crystal structure [22]. Given that the D state is preferred by TPC despite the
lack of a side chain, it is conceivable that the core scaffold rather than the
presence or absence of a branched side chain dictates whether the S or D state
is adopted. Further evidence supporting this notion is co-crystal structures of
DAO with 3-hydroxypyridin-2(1H)-one (PDB code: 3w4i) [23]. Even though the
unsubstituted 3-hydroxypyridin-2(1H)-one is small enough to fit into the active
site of the S state structure, the co-crystal structure adopts the D state with a
vacant secondary pocket. One notable structural difference between S and D
state-inducing inhibitors is the presence/absence of a hydrogen bond donor that
can interact with the carbonyl oxygen of Gly313 [23]. In addition, an additional
MD run with a virtual complex in which 2c was substituted for 2b in the 2b-DAO
complex preferred the S state (Supplementary Fig. 6), supporting the notion
that the scaffold dictates whether the S or D state is adopted. The precise
mechanism by which DAO adopts the S or D state needs further investigation.
The energy decomposition of ∆Gbind indicated that the hydrophobic
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
interaction of 2b-Tyr224 was stronger than that of TPC-Tyr224. In addition, the
substitution of a halogen for a methyl moiety in the thiophene ring showed
increased potency. These suggested that the π-π interaction plays a role in the
interaction between Tyr224 and thiophene-based inhibitors. The side chain of
Tyr224 is electron-donating [27]. Thus, it is inferred that the substitution of a
9

Kato et al.
ACCEPTED MANUSCRIPT
1
2
halogen increased potency because the LUMO level of the thiophene ring
decreased due to electron-withdrawing effect of the halogen.
3
A number of scaffolds that promote the formation of the D state have been
published to date. Many of these inhibitors showed potent inhibitory activity by
exploiting the secondary pocket with a large branched chain in combination with
a H-bond with Gly313 [22-24]. In contrast, our thiophene-based inhibitors
including 1c and 2b showed low nanomolar inhibitory potency without relying on
the secondary pocket and H-bond with Gly313, but instead by forming a strong
stacking interaction with Tyr224. Indeed, the cleft between the lid and Loop
53-62 was tightly closed in our S state structures, presenting a sharp contrast to
the D state structures accommodating the half-open cleft. Moreover, a H-bond
network including Tyr224 and H2O was shown to be extensive in the S state
contributing to the closing of the cleft. It appears difficult to accommodate a large
branched chain in the S state in which the cleft is closed due to the loss of the
secondary pocket. This accounts for the substantial loss of potency caused by
the introduction of a large branched chain to the thiophene ring of either the
2-carboxylic acid or the 3-carboxylic acid scaffolds.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 4. Conclusions
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Taken together, the present results suggest that the formation of the S state
is driven by the concerted action of the residues around the cleft including
Tyr224 and thiophene carboxylic acid scaffolds in alliance with solvent
molecules. The mechanism by which the thiophene-based inhibitors achieve
potent DAO inhibition is distinct from that of the D-state promoting inhibitors
which exploit the secondary pocket. These findings collectively highlight two
distinct structural optimization approaches to DAO inhibitors depending on how
a given pharmacophore affects the position of Tyr224. For those inducing the S
state, as seen with 1a and 2a, the addition of a branched chain unlikely results in
improvement of inhibitory potency. The primary focus of the structural
optimization strategy should be to preserve the S state by avoiding sterically
hindered substituents. On the other hand, pharmacophores promoting the D
state can take full advantage of the secondary pocket generated by the
movement of Tyr224 by incorporating a branched chain. These insights should
10

Kato et al.
ACCEPTED MANUSCRIPT
1
2
be of particular importance in future efforts to optimize DAO inhibitors with novel
scaffolds.
3
4
5. Experimental section
5
5.1. Chemistry
6
All solvents were reagent grade or HPLC grade. Melting points were obtained
on a Mel-Temp apparatus and are uncorrected. 1H NMR spectra were recorded
at 400 MHz. The HPLC solvent system consisted of distilled water and
acetonitrile, both containing 0.1% formic acid. Preparative HPLC purification was
performed on an Agilent 1200 Series HPLC system equipped with an Agilent
G1315D DAD detector using a Phenomenex Luna 5 ꢀm C18 (2) column (21.2
mm × 250 mm, 5 ꢀm) with a gradient of 40% ACN/60% H2O for 5 minutes
followed by an increase to 100% ACN/0% H2O over 40 minutes and a
continuation of 100% ACN/0% H2O until 50 minutes at a flow rate of 15 mL/min.
Analytical HPLC was performed on an Agilent 1200 Series HPLC system
equipped with an Agilent G1315D DAD detector (detection at 220 nm), and an
Agilent Quadrupole 6120 LC-MS with electrospray ionization (ESI) source. The
analytical HPLC conditions involve a gradient of 20% ACN/80% H2O for 0.25
minutes followed by an increase to 85% ACN/15% H2O over 1.75 minutes and
continuation of 85% ACN/15% H2O until 4 minutes (detection at 220 nm) with a
Luna C18 column (2.1 mm × 50 mm, 3.5 μm) at a flow rate of 0.75 mL/min. All
final compounds tested were confirmed to be of ≥95% purity by the HPLC
methods described above. Compounds 1a, 1c-e, 1i, 1o, and 2d were purchased
from Aldrich. Compounds 1b and 2b were purchased from Ark Pharma, Inc.
Compound 1g was purchased from Enamine. Compounds 1h and 1r were
purchased from TCI America. Compounds 1m and 1q were purchased from Alfa
Aesar. Compounds 1n and 1p were purchased from Parkway Scientific and
ASDI, Inc., respectively. Compounds 2a, 2c, and 1s were purchased from Matrix
Scientific. Compound 1t and 2e were purchased from Oakwood Chemical and
Chembridge, respectively. Synthesis of compounds 1f [28], 1k [29], 1l [30], and
1v [31] were previously reported.
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
5.1.1. Synthesis of 4-(difluoromethyl)thiophene-2-carboxylic acid (1j).
11

Kato et al.
ACCEPTED MANUSCRIPT
1
2
To a solution of methyl 4-formylthiophene-2-carboxylate 3 (0.18 g, 1.06 mmol)
in dichloromethane (7 mL) at rt was added deoxo-fluor (0.27 g, 1.22 mmol) via
syringe. After 4 h of stirring, the reaction was quenched by a careful addition of
sat. NaHCO3. The compound was extracted twice with dichloromethane. The
combined organic layers was dried over sodium sulfate and concentrated to give
an oil which was purified by silica gel column (eluent: 10% EtOAc in hexanes) to
afford 0.075 g (37% yield) of methyl 4-(difluoromethyl)thiophene-2-carboxylate 4
as a colorless oil. To a solution of 4 (0.07 g, 0.36 mmol) in a 1:1 mixture of
dioxane and water (4 mL) was added 1 N aqueous solution of sodium hydroxide
(1 mL). The reaction was heated at 100 °C for 1 h. Upon completion, the reaction
mixture was concentrated in vacuo. The resulting material was acidified with
aqueous 10% KHSO4 solution to pH~4. The product was extracted with EtOAc.
The organic layer was washed with brine, dried over Na2SO4, and concentrated
to give 0.065 g (quantitative yield from 4) of
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
4-(difluoromethyl)thiophene-2-carboxylic acid 1j as a light yellow solid; mp 102
1
°C. H NMR (CDCl3) δ 6.56 (t, J = 56.3 Hz, 1H), 7.85 (m, 1H), 7.97 (m, 1H).
LCMS: retention time 1.58 min, m/z 179.1 [M + H]+.
5.1.2. Synthesis of 5-phenethylthiophene-2-carboxylic acid (1u)
To a solution of methyl 5-bromothiophene-2-carboxylate 5u (0.45 g, 2.04
mmol) in diisopropylamine (10 mL) were added triphenylphosphine (0.21 g, 0.80
mmol), Pd(PhCN)2Cl2 (0.15 g, 0.39 mmol) and copper (I) iodide (0.076 g, 0.40
mmol). Phenylacetylene (0.4 g, 3.92 mmol) was then added under N2 and the
reaction was heated at 70 °C for 72 h. After coolin g to rt, the reaction mixture
was concentrated in vacuo and the residual material was purified by flash
chromatography (eluent: 5% EtOAc/hexanes) to give 0.39 g (79% yield) of
methyl 5-(phenylethynyl)thiophene-2-carboxylate 6u as a tan solid. 1H NMR
(CDCl3) δ 3.92 (s, 3H), 7.24 (d, J = 3.8 Hz, 1H), 7.39-7.40 (m, 3H), 7.54 (m, 2H),
7.71 (d, J = 4.0 Hz, 1H). To a solution of 6u (0.39 g, 1.61 mmol) in EtOAc (50
mL) was added a spatula tip of 10% Pd/C and the mixture was shaken under
hydrogen (50 psi) for 2 h. The catalyst was removed by filtration through a pad of
celite and the filtrate was concentrated to give 0.33 g (83% yield) of methyl
5-(phenylethynyl)thiophene-2-carboxylate 7u as a yellow oil. 1H NMR (CDCl3) δ
3.00 (t, J = 7.3 Hz, 2H), 3.15 (t, J = 8.1 Hz, 2H), 3.88 (s, 3H), 6.76 (dt, J = 0.8, 3.5
12

Kato et al.
ACCEPTED MANUSCRIPT
1
2
Hz, 1H), 7.19 (d, J = 7.3 Hz, 2H), 7.24 (m, 1H), 7.30 (m, 2H), 7.63 (d, J = 3.8 Hz,
1H). To a solution of 7u (0.33 g, 1.34 mmol) in a 1:1 mixture of water and
methanol (15 mL) was added lithium hydroxide (0.080 g, 3.33 mmol) and the
reaction was heated at 40 °C for 19 h. After comple tion of the reaction, the
reaction mixture was concentrated in vacuo. The resulting material was acidified
with 1N HCl to pH~2 and the resulting precipitate was filtered to give 0.29 g
(93% yield) of 5-phenethylthiophene-2-carboxylic acid (1u) as an off white solid;
3
4
5
6
7
1
8
mp 114 °C. H NMR (CDCl3) δ 3.01 (t, J = 7.3 Hz, 2H), 3.18 (t, J = 8.1 Hz, 2H),
9
6.80 (dt, J = 0.8, 3.8 Hz, 1H), 7.19-7.22 (m, 2H), 7.24 (m, 1H), 7.30 (m, 2H), 7.72
(d, J = 3.8 Hz, 1H). LCMS: retention time 2.63 min, m/z 233.1 [M + H]+.
5.1.3. Synthesis of 4-Phenethylthiophene-2-carboxylic acid (1w).
Methyl 4-(phenylethynyl)thiophene-2-carboxylate 6w was prepared as
described for the preparation of methyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
5-(phenylethynyl)thiophene-2-carboxylate 6u except methyl
4-bromothiophene-2-carboxylate 5w was used in place of methyl
5-bromothiophene-2-carboxylate 5u and the reaction was heated at 70 °C for 18
h; brown oil (38% yield). 1H NMR (CDCl3) δ 3.93 (s, 3H), 7.37-7.39 (m, 3H),
7.52-7.54 (m, 2H), 7.70 (d, J = 1.3 Hz, 1H), 7.88 (d, J = 1.3 Hz, 1H). Methyl
4-phenethylthiophene-2-carboxylate (7w) was prepared as described for the
preparation of methyl 5-phenethylthiophene-2-carboxylate (7u) except methyl
4-(phenylethynyl)thiophene-2-carboxylate 6w was used in place of methyl
5-(phenylethynyl)thiophene-2-carboxylate 6u and that the mixture was
hydrogenated for 1 h at 40 psi; yellow oil (93% yield). 1H NMR (CDCl3) δ 2.92 (s,
4H), 3.88 (s, 3H), 7.10 (d, J = 1.5 Hz, 1H), 7.14 (m, 2H), 7.20 (m, 1H), 7.28 (m,
2H), 7.65 (d, J = 1.5 Hz, 1H). To a solution of 7w (0.17 g, 0.69 mmol) in 1:1
mixture of 1,4-dioxane and water (4 mL) was added 1 N aqueous solution of
sodium hydroxide (2.1 mL) and the reaction was heated for 1 h at 100 °C. Upon
completion, the reaction was concentrated in vacuo. The resulting material was
acidified with aqueous 10% KHSO4 solution to pH~4. The product was extracted
with EtOAc, washed with brine, dried over Na2SO4, and concentrated to give 0.15
g (94% yield) of 4-phenethylthiophene-2-carboxylic acid 1w as a yellow solid; mp
1
136 °C. H NMR (CDCl3) δ 2.95 (s, 4H), 7.16-7.24 (m, 4H), 7.28 (m, 2H), 7.72 (d,
J = 1.5 Hz, 1H). LCMS: retention time 2.84 min, m/z 233.1 [M + H]+.
13

Kato et al.
ACCEPTED MANUSCRIPT
1
2
5.1.4. Synthesis of 5-phenethylthiophene-3-carboxylic acid (2f).
To a solution of 5-bromothiophene-3-carboxylic acid 8 (1.0 g, 4.83 mmol) in
acetonitrile (20 mL) was added potassium carbonate (3.3 g, 23.9 mmol) followed
by the addition of benzyl bromide (0.63 mL, 5.30 mmol). The mixture was heated
at 80 °C for 24 h and concentrated in vacuo. The re sulting residue was dissolved
in EtOAc and the resulting solution was subsequently washed with water, dried
over Na2SO4, and concentrated to give a light yellow oil. Purification of the crude
material by flash chromatography (eluent: 5% EtOAc/hexanes) afforded 0.89 g
3
4
5
6
7
8
1
9
(62% yield) of benzyl 5-bromothiophene-3-carboxylate (9) as a colorless oil. H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
NMR (CDCl3) δ 5.31 (s, 2H), 7.37-7.43 (m, 5H), 7.50 (d, J = 1.5 Hz, 1H), 8.03 (d,
J = 1.5 Hz, 1H). Benzyl 5-(phenylethynyl)thiophene-3-carboxylate 10 was
prepared
5-(phenylethynyl)thiophene-2-carboxylate
5-bromothiophene-3-carboxylate
as
described
for
the
preparation
6u except
of
methyl
benzyl
9
was used in place of methyl
5-bromothiophene-2-carboxylate 5u and the reaction was heated at 70 °C for 17
1
h; brown oil (51% yield). H NMR (CDCl3) δ 5.33 (s, 2H), 7.36-7.39 (m, 4H),
7.41-7.46 (m, 4H), 7.69 (d, J = 1.3 Hz, 1H), 8.07 (d, J = 1.3 Hz, 1H). Benzyl
5-phenethylthiophene-3-carboxylate 11 was prepared as described for the
preparation of methyl 5-phenethylthiophene-2-carboxylate 7u except benzyl
5-(phenylethynyl)thiophene-3-carboxylate 10 was used in place of methyl
5-(phenylethynyl)thiophene-2-carboxylate 6u and that the mixture was
hydrogenated for 1 h at 50 psi; yellow oil (quantitative yield). 1H NMR (CDCl3) δ
2.97 (t, J = 7.1 Hz, 2H), 3.10 (t, J = 8.6 Hz, 2H), 5.30 (s, 2H), 7.18-7.24 (m, 3H),
7.28 (m, 3H), 7.37-7.45 (m, 5H), 7.94 (d, J = 1.3 Hz, 1H).
5-Phenethylthiophene-3-carboxylic acid 2f was prepared as described for the
preparation of compound 1u except benzyl 5-phenethylthiophene-3-carboxylate
11 was used in place of methyl 5-phenethylthiophene-2-carboxylate 7u and the
reaction was heated in 2:1 mixture of 1,4-dioxane and water at 50 °C for 17 h and
the crude material was purified by preparative HPLC; white solid (53% yield); mp
1
126 °C. H NMR (CDCl3) δ 2.99 (t, J = 7.1 Hz, 2H), 3.12 (t, J = 8.3 Hz, 2H),
7.19-7.24 (m, 3H), 7.29-7.33 (m, 3H), 8.02 (d, J = 1.5 Hz, 1H). LCMS: retention
time 2.85 min, m/z 233.1 [M + H]+.
14

Kato et al.
ACCEPTED MANUSCRIPT
1
2
5.2. In vitro DAO assay
D-Serine was purchased from Bachem Biosciences Inc, horse radish
peroxidase from Worthington Biochemical Corporation and o-phenylenediamine
from Pierce Biotechnology, Inc. All other chemicals were obtained from
Sigma-Aldrich. A reliable 96-well plate D-amino acid oxidase (DAO/DAAO)
assay was developed based on previously published methods [32]. Briefly,
D-serine (5 mM) was oxidatively deaminated by human DAO in the presence of
molecular oxygen and flavin adenosine dinucleotide (FAD; 10 µM), to yield the
corresponding α–keto acid, ammonia and hydrogen peroxide. The resulting
hydrogen peroxide was quantified using horseradish peroxidase (0.01 mg/mL)
and o-phenylenediamine (180 µg/mL), which turns yellowish-brown upon
oxidation. DAO activity was correlated to the rate formation of the colored
product, i.e., rate of change of absorbance at 411 nm. All reactions were carried
out for 20 min at room temperature in a 100-µL volume in Tris buffer (50 mM, pH
8.5). Additionally, stock solutions and serial dilutions of potential DAO inhibitors
were made in 20:80 DMSO:buffer with a final assay DMSO concentration of 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
5.3. X-ray crystallography
Expression, purification and crystallization of human DAO were previously
described [19]. Briefly, full-length DAO protein was expressed using pET11b
(Novagen) and E. coli BL21 (DE3). After disruption of cells, the extract was
centrifuged, heated and fractionated with ammonium sulfate. The sample was
applied to DEAE (Sigma-Aldrich) and hydroxylapatite columns (Nacalai). DAO
was crystallized with 15% [w/v] PEG 4000, 0.2 M ammonium acetate, 0.1 M Na
citrate at pH 8.0, and 10% [v/v] glycerol. 1c and 2b (1 mM each) were soaked
into the crystals prior to X-ray diffraction experiments. Diffraction data were
collected at Photon Factory AR NW12A and SPring-8 BL44XU. Scaling,
molecular replacement and model building were performed with iMosfilm [33],
XDS [34], MolRep [35] and Coot [36], respectively. Refinement was performed
as previously described using Refmac and Phenix [37-39].
5.4. Molecular dynamics
Dimer structures of the complexes of 2b-DAO (PDB code: 5zj9) and
15

Kato et al.
ACCEPTED MANUSCRIPT
1
2
TPC-DAO (PDB code: 3znn) retaining cofactors and water molecules were used
as initial structures. The initial structure in the D state with 2b was constructed by
superimposing 3znn and 5zj9 and subsequently combining the coordinates of 2b
with the 3znn coordinates without TPC. The construction of the initial structure in
the S state with TPC and 2c followed a similar procedure. Structure optimization
and calculation of electrostatic potentials of ligands were performed with the
HF/6-31G(d) basis set using Gaussian 09 [40]. Atomic charges and atom types
were assigned using Antechamber [41]. Generalized Amber force field (GAFF)
implemented in the LEaP module of the Amber suite was used to parameterize
ligands [42, 43]. Ionization states of charged residues were calculated according
to ProPKA [44]. The Amber ff14SB force field was used for polypeptide chains.
All systems were solvated in TIP3P water boxes with the minimum margin of 10
Å using solvate1.0 and the LEaP module [45]. Cl- was added to neutralize the
systems. Na+ and Cl- were further added to adjust the salt concentration of the
systems to 0.15 M.
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
Energy minimization was performed using Amber 14 with 10-Å cut-off for the
non-bonded interactions [43]. 5000 steps of steepest descent minimization were
performed followed by 5000 steps of conjugate gradient minimization using the
Particle Mesh Ewald method under constant-volume and periodic boundary
conditions. An initial minimization was performed for solvent molecules and
hydrogen atoms followed by minimization for all atoms in the systems.
The systems were gradually heated from 0 to 310 K under constant-volume
conditions. The temperature was controlled using Langevin dynamics with a
collision frequency of 2.0 ps-1. Step size was set to 2 fs with fixed bond lengths
involving all hydrogen atoms using the SHAKE algorithm [46]. Motions of all
solute atoms except hydrogens were restricted during the heating process.
Equilibration and production MD simulations were performed under 1 atm at 310
K without motion restrictions for all atoms as previously described [47, 48].
Equilibration of the systems was monitored as shown in Supplementary Figs.
7-11.
5.5. Binding free energy calculations
Calculations of binding free energy (∆Gbind) of inhibitors to DAO were
16

Kato et al.
ACCEPTED MANUSCRIPT
1
2
performed by the molecular mechanics energies combined with the generalized
Born and surface area (MM/GBSA) method implemented in the AmberTools14
suite [49, 50]. Binding free energy (∆Gbind) was previously defined as below [48,
51, 52].
3
4
5
∆Gbind = ∆GMM + ∆Gsolv = ∆Hbind - T∆SMM
6
where ∆GMM, ∆Gsolv, ∆Hbind and T∆SMM refer to molecular mechanics free energy,
solvation free energy, binding enthalpy and entropy term, respectively.
∆GMM = ∆EMM - T∆SMM
7
8
9
where ∆EMM refers to enthalpy in the gas phase upon complex formation.
∆Hbind = ∆EMM + ∆Gsolv
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
∆EMM =∆Evdw + ∆Eelec
where ∆Evdw and ∆Eelec refer to van der Waals and electrostatic interaction
energies, respectively.
∆Gsolv = ∆Gsolvpolar + ∆Gsolv
nonpolar
where ∆Gsolv
and ∆Gsolv
refer to polar and nonpolar contributions of
nonpolar
polar
solvation free energy, respectively.
∆Gexp was calculated from the IC50 using the following relations [51].
∆Gexp = ~ RT ln IC50
Where R and T are the ideal gas constant and absolute temperature,
respectively.
T∆SMM was calculated using the quasi-harmonic approximation [53]. H-bond
occupancy was calculated using CPPTRAJ implemented in the AmberTools14
suite [54].
5.6. Data availability
The atomic coordinates of inhibitor-DAO complexes were deposited in the
Protein Data Bank. The access codes are 5zja (1c-DAO complex) and 5zj9
(2b-DAO complex).
31 Acknowledgements
32
33
This research was supported in part by a grant for Enzyme Research from
the Japan Foundation for Applied Enzymology, a research grant from the
17

Kato et al.
ACCEPTED MANUSCRIPT
1
2
National Institutes of Health (R01MH091387) and JSPS KAKENHI Grant
Number 18K06580. We thank the beamline staff at the Photon Factory (proposal
No: 2013G075) and the SPring-8 BL44XU (Proposal No. 2015A6537
and 2015B6537) for supporting the data collection. Authors declare no conflict of
interests.
3
4
5
6
7
8
Appendix A. Supplementary data
9
Supplementary data related to this article can be found at
10
11
12 Figure legends
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Fig 1 Structural change of Tyr224 in response to bound inhibitors
A, C: Tyr224 in the S state. In the complex structure of 2b-DAO, Tyr224 (yellow)
is stacked with the thiophene ring of the inhibitor. The secondary pocket is lost in
this state.
B, D: Tyr224 in the D state. In the complex structure between CPC and DAO, the
side chain of Tyr224 is shifted to allow an additional pocket (i.e. the secondary
pocket) to appear, which accommodates the branched side chain of the inhibitor
[22](PDB code: 3zno). A yellow dashed circle indicates the secondary pocket.
The structure and IC50 value of CPC are shown in Supplementary Table 2.
Fig 2 The active sites of the 1c-DAO and 2b-DAO complexes
A: The active site of the crystal structure of the 1c-DAO complex
B: The active site of the crystal structure of the 2b-DAO complex
C: Superimposition of the 1c-DAO (pale cyan) and 2b-DAO (yellow) complexes
Fig 3 Distance between the benzene ring of Tyr224 and the 5-membered
rings of the DAO inhibitors
All panels in this figure indicate the results for Chain A of DAO dimers. The
results for Chain B are shown in Supplementary Figs. 1-4. On top of the panels,
initial and final states under individual simulation conditions are illustrated. An
18

Kato et al.
ACCEPTED MANUSCRIPT
1
2
arrowhead in each panel indicates the distance between the centroids of the
rings of Tyr224 and inhibitors in the initial structure.
3
A: Distance between the centroids of the rings of Tyr224 and 2b during a MD
simulation. The initial structure for this run was a dimer crystal structure of the
2b-DAO complex, which is in the S state. The average distance between the
centroids throughout trajectories was 4.04 Å.
4
5
6
7
B: Distance between the centroids of the benzene ring of Tyr224 and pyrrole ring
of TPC during a MD simulation. The initial structure for this run was a dimer
crystal structure of the TPC-DAO complex, which is in the D state. The structure
and IC50 value of TPC are shown in Supplementary Table 2.
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
C: Distance between the centroids of the rings of Tyr224 and 2b during a MD
simulation. The initial structure for this run was a hypothetical structure in which
a D state structure of DAO derived from the TPC-DAO complex structure was
combined with the coordinates of 2b.
D: Distance between the centroids of the rings of Tyr224 and TPC during a MD
simulation. The initial structure for this run was a hypothetical structure in which
an S state structure of DAO derived from the 2b-DAO complex structure was
combined with the coordinates of TPC.
Fig 4 Analysis of H-bond networks around the active site
A: Occupancy of H-bonds between inhibitors and DAO (left) and between H2O
and DAO (right). The listed H-bonds between H2O and DAO are those bridging
Tyr224 and other DAO residues via H2O.
B: H-bond networks around the active site in the 2b-DAO complex depicted
based on Panel A (Occupancy > 0.4). Cyan lines and red balls indicate H-bonds
and oxygen atoms of H2O molecules, respectively.
C: H-bond networks around the active site in the TPC-DAO complex depicted
based on Panel A (Occupancy > 0.4).
Fig 5 Difference of S and D state structures
A: Loops around the active sites of superimposed DAO structures in complex
with 2b (pale cyan) and TPC (yellow).
B: The 2b-DAO complex viewed from the surface of the protein molecule.
19

Kato et al.
ACCEPTED MANUSCRIPT
1
2
Orange spheres and blue balls indicate 2b and oxygen atoms of H2O,
respectively. Transparent surface in pale cyan corresponds to the lid and Loop
53-62.
3
4
C: The TPC-DAO complex viewed from the surface of the protein molecule.
Orange spheres and blue balls indicate TPC and oxygen atoms of H2O,
respectively. Transparent surface in yellow corresponds to the lid and Loop
53-62.
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
References
[1] Y. Kato, D.H. Tran, H.T.T. Trinh, K. Fukui, D-Amino Acid Oxidase and
D-Aspartate Oxidase., in: T. Yoshimura, T. Nishikawa, H. Homma (Eds.)
D-amino acids : physiology, metabolism, and application, Springer, Japan,
2016, pp. 293-309.
[2] A. Hashimoto, T. Nishikawa, T. Oka, K. Takahashi, Endogenous D-serine
in rat brain: N-methyl-D-aspartate receptor-related distribution and aging,
Journal of neurochemistry, 60 (1993) 783-786.
[3] J.P. Mothet, A.T. Parent, H. Wolosker, R.O. Brady, Jr., D.J. Linden, C.D.
Ferris, M.A. Rogawski, S.H. Snyder, D-serine is an endogenous ligand for the
glycine site of the N-methyl-D-aspartate receptor, Proceedings of the
National Academy of Sciences of the United States of America, 97 (2000)
4926-4931.
[4] M.J. Schell, M.E. Molliver, S.H. Snyder, D-serine, an endogenous synaptic
modulator: localization to astrocytes and glutamate-stimulated release,
Proceedings of the National Academy of Sciences of the United States of
America, 92 (1995) 3948-3952.
[5] P.W. Burnet, S.L. Eastwood, G.C. Bristow, B.R. Godlewska, P. Sikka, M.
Walker, P.J. Harrison, D-amino acid oxidase activity and expression are
increased in schizophrenia, Molecular psychiatry, 13 (2008) 658-660.
[6] K. Ono, Y. Shishido, H.K. Park, T. Kawazoe, S. Iwana, S.P. Chung, R.M.
Abou El-Magd, K. Yorita, M. Okano, T. Watanabe, N. Sano, Y. Bando, K.
Arima, T. Sakai, K. Fukui, Potential pathophysiological role of D-amino acid
oxidase in schizophrenia: immunohistochemical and in situ hybridization
20

Kato et al.
ACCEPTED MANUSCRIPT
1
2
study of the expression in human and rat brain, Journal of neural
transmission (Vienna, Austria : 1996), 116 (2009) 1335-1347.
3
[7] K. Hashimoto, T. Fukushima, E. Shimizu, N. Komatsu, H. Watanabe, N.
Shinoda, M. Nakazato, C. Kumakiri, S. Okada, H. Hasegawa, K. Imai, M. Iyo,
Decreased serum levels of D-serine in patients with schizophrenia: evidence
in support of the N-methyl-D-aspartate receptor hypofunction hypothesis of
schizophrenia, Archives of general psychiatry, 60 (2003) 572-576.
[8] M.A. Calcia, C. Madeira, F.V. Alheira, T.C. Silva, F.M. Tannos, C.
Vargas-Lopes, N. Goldenstein, M.A. Brasil, S.T. Ferreira, R. Panizzutti,
Plasma levels of D-serine in Brazilian individuals with schizophrenia,
Schizophrenia research, 142 (2012) 83-87.
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
[9] T. Fukushima, H. Iizuka, A. Yokota, T. Suzuki, C. Ohno, Y. Kono, M.
Nishikiori, A. Seki, H. Ichiba, Y. Watanabe, S. Hongo, M. Utsunomiya, M.
Nakatani, K. Sadamoto, T. Yoshio, Quantitative analyses of
schizophrenia-associated metabolites in serum: serum D-lactate levels are
negatively correlated with gamma-glutamylcysteine in medicated
schizophrenia patients, PloS one, 9 (2014) e101652.
[10] K. Hashimoto, G. Engberg, E. Shimizu, C. Nordin, L.H. Lindstrom, M.
Iyo, Reduced D-serine to total serine ratio in the cerebrospinal fluid of drug
naive schizophrenic patients, Progress in neuro-psychopharmacology &
biological psychiatry, 29 (2005) 767-769.
[11] D.C. Goff, M. Hill, D. Barch, The treatment of cognitive impairment in
schizophrenia, Pharmacology, biochemistry, and behavior, 99 (2011) 245-253.
[12] D. Cadinu, B. Grayson, G. Podda, M.K. Harte, N. Doostdar, J.C. Neill,
NMDA receptor antagonist rodent models for cognition in schizophrenia and
identification of novel drug treatments, an update, Neuropharmacology,
(2017).
[13] H.Y. Lane, C.H. Lin, M.F. Green, G. Hellemann, C.C. Huang, P.W. Chen,
R. Tun, Y.C. Chang, G.E. Tsai, Add-on treatment of benzoate for
schizophrenia: a randomized, double-blind, placebo-controlled trial of
D-amino acid oxidase inhibitor, JAMA psychiatry, 70 (2013) 1267-1275.
[14] D.V. Ferraris, T. Tsukamoto, Recent advances in the discovery of
D-amino acid oxidase inhibitors and their therapeutic utility in
21

Kato et al.
ACCEPTED MANUSCRIPT
1
2
schizophrenia, Current pharmaceutical design, 17 (2011) 103-111.
[15] G. Molla, Competitive Inhibitors Unveil Structure/Function
Relationships in Human D-Amino Acid Oxidase, Frontiers in molecular
biosciences, 4 (2017) 80.
3
4
5
[16] R. Konno, S. Uchiyama, Y. Yasumura, Intraspecies and interspecies
variations in the substrate specificity of D-amino acid oxidase, Comparative
biochemistry and physiology. B, Comparative biochemistry, 71 (1982)
735-738.
6
7
8
9
[17] C. Setoyama, Y. Nishina, H. Mizutani, I. Miyahara, K. Hirotsu, N.
Kamiya, K. Shiga, R. Miura, Engineering the substrate specificity of porcine
kidney D-amino acid oxidase by mutagenesis of the "active-site lid", Journal
of biochemistry, 139 (2006) 873-879.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
[18] R. Miura, C. Setoyama, Y. Nishina, K. Shiga, H. Mizutani, I. Miyahara,
K. Hirotsu, Structural and mechanistic studies on D-amino acid oxidase x
substrate complex: implications of the crystal structure of enzyme x
substrate analog complex, Journal of biochemistry, 122 (1997) 825-833.
[19] T. Kawazoe, H. Tsuge, M.S. Pilone, K. Fukui, Crystal structure of
human D-amino acid oxidase: context-dependent variability of the backbone
conformation of the VAAGL hydrophobic stretch located at the si-face of the
flavin ring, Protein science : a publication of the Protein Society, 15 (2006)
2708-2717.
[20] A. Mattevi, M.A. Vanoni, F. Todone, M. Rizzi, A. Teplyakov, A. Coda, M.
Bolognesi, B. Curti, Crystal structure of D-amino acid oxidase: a case of
active site mirror-image convergent evolution with flavocytochrome b2,
Proceedings of the National Academy of Sciences of the United States of
America, 93 (1996) 7496-7501.
[21] T. Kawazoe, H. Tsuge, T. Imagawa, K. Aki, S. Kuramitsu, K. Fukui,
Structural basis of D-DOPA oxidation by D-amino acid oxidase: alternative
pathway for dopamine biosynthesis, Biochemical and biophysical research
communications, 355 (2007) 385-391.
[22] S.C. Hopkins, M.L. Heffernan, L.D. Saraswat, C.A. Bowen, L. Melnick,
L.W. Hardy, M.A. Orsini, M.S. Allen, P. Koch, K.L. Spear, R.J. Foglesong, M.
Soukri, M. Chytil, Q.K. Fang, S.W. Jones, M.A. Varney, A. Panatier, S.H.
22

Kato et al.
ACCEPTED MANUSCRIPT
1
2
Oliet, L. Pollegioni, L. Piubelli, G. Molla, M. Nardini, T.H. Large, Structural,
kinetic, and pharmacodynamic mechanisms of D-amino acid oxidase
inhibition by small molecules, Journal of medicinal chemistry, 56 (2013)
3710-3724.
3
4
5
[23] T. Hondo, M. Warizaya, T. Niimi, I. Namatame, T. Yamaguchi, K.
Nakanishi, T. Hamajima, K. Harada, H. Sakashita, Y. Matsumoto, M. Orita,
M. Takeuchi, 4-Hydroxypyridazin-3(2H)-one derivatives as novel D-amino
acid oxidase inhibitors, Journal of medicinal chemistry, 56 (2013) 3582-3592.
[24] N. Hin, B. Duvall, D. Ferraris, J. Alt, A.G. Thomas, R. Rais, C. Rojas, Y.
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
Wu,
K.M.
Wozniak,
B.S.
Slusher,
T.
Tsukamoto,
6-Hydroxy-1,2,4-triazine-3,5(2H,4H)-dione Derivatives as Novel D-Amino
Acid Oxidase Inhibitors, Journal of medicinal chemistry, 58 (2015)
7258-7272.
[25] M. Katane, N. Osaka, S. Matsuda, K. Maeda, T. Kawata, Y. Saitoh, M.
Sekine, T. Furuchi, I. Doi, S. Hirono, H. Homma, Identification of novel
D-amino acid oxidase inhibitors by in silico screening and their functional
characterization in vitro, Journal of medicinal chemistry, 56 (2013)
1894-1907.
[26] T. Sparey, P. Abeywickrema, S. Almond, N. Brandon, N. Byrne, A.
Campbell, P.H. Hutson, M. Jacobson, B. Jones, S. Munshi, D. Pascarella, A.
Pike, G.S. Prasad, N. Sachs, M. Sakatis, V. Sardana, S. Venkatraman, M.B.
Young, The discovery of fused pyrrole carboxylic acids as novel, potent
D-amino acid oxidase (DAO) inhibitors, Bioorganic & medicinal chemistry
letters, 18 (2008) 3386-3391.
[27] M.L. Fonda, B.M. Anderson, D-amino acid oxidase. II. Studies of
substrate-competitive inhibitors, The Journal of biological chemistry, 243
(1968) 1931-1935.
[28] S. Kasai, H. Igawa, M. Takahashi, T. Maekawa, K. Kakegawa, T.
Yasuma, A. Kina, J. Aida, U. Khamrai, M. Kundu, Preparation of
benzimidazole and imidazopyridine derivatives and analogs as MCH
receptor antagonists for treating obesity, in, Takeda Pharmaceutical
Company Limited, Japan . 2013.
[29] P. Wang, M. Ji, F. Sha, An efficient and facile process for synthesis of
23

Kato et al.
ACCEPTED MANUSCRIPT
1
2
4,5-dichlorothiophene-2-carboxylic acid using N-chlorosuccinimide, J. Chem.
Res., 38 (2014) 622-624.
3
[30] D.R. Parry, I.R. Matthews, G. Mitchell, A.G. Williams, N.J. Barnes, J.M.
Cox, K.J. Gillen, M.P. Ensminger, K. Khodayari, H. Nakayama, Preparation
4
5
of
cyclopropylcarbonylaminopyrrolidinones,
-thiazolidinones,
or
6
-oxazolidinones as herbicides, in, Zeneca Limited, UK . 2000.
7
[31] S. Gronowitz, T. Klingstedt, L. Svensson, U. Hansson, On the syntheses
of branched saturated fatty acids, Lipids, 28 (1993) 889-897.
[32] S.P. Cook, I. Galve-Roperh, A. Martinez del Pozo, I. Rodriguez-Crespo,
Direct calcium binding results in activation of brain serine racemase, The
Journal of biological chemistry, 277 (2002) 27782-27792.
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
[33] T.G. Battye, L. Kontogiannis, O. Johnson, H.R. Powell, A.G. Leslie,
iMOSFLM: a new graphical interface for diffraction-image processing with
MOSFLM, Acta crystallographica. Section D, Biological crystallography, 67
(2011) 271-281.
[34] W. Kabsch, XDS, Acta crystallographica. Section D, Biological
crystallography, 66 (2010) 125-132.
[35] A. Vagin, A. Teplyakov, MOLREP: an automated program for molecular
replacement, J. Appl. Crystallogr., 30 (1997) 1022-1025.
[36] P. Emsley, B. Lohkamp, W.G. Scott, K. Cowtan, Features and
development of Coot, Acta Crystallogr., Sect. D: Biol. Crystallogr., 66 (2010)
486-501.
[37] G.N. Murshudov, A.A. Vagin, E.J. Dodson, Refinement of
macromolecular structures by the maximum-likelihood method, Acta
Crystallogr., Sect. D: Biol. Crystallogr., D53 (1997) 240-255.
[38] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, I.W. Davis, N. Echols,
J.J. Headd, L.W. Hung, G.J. Kapral, R.W. Grosse-Kunstleve, A.J. McCoy,
N.W. Moriarty, R. Oeffner, R.J. Read, D.C. Richardson, J.S. Richardson, T.C.
Terwilliger, P.H. Zwart, PHENIX: a comprehensive Python-based system for
macromolecular structure solution, Acta Crystallogr., Sect. D: Biol.
Crystallogr., 66 (2010) 213-221.
[39] Y.S. Kato, F. Yumoto, H. Tanaka, T. Miyakawa, Y. Miyauchi, D.
Takeshita, Y. Sawano, T. Ojima, I. Ohtsuki, M. Tanokura, Structure of the
24

Kato et al.
ACCEPTED MANUSCRIPT
1
2
Ca(2+)-saturated C-terminal domain of scallop troponin C in complex with a
troponin I fragment, Biological chemistry, 394 (2013) 55-68.
3
[40] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R.
Cheeseman, G. Scalmani, V. Barone, G.A. Petersson, H. Nakatsuji, X. Li, M.
Caricato, A. Marenich, J. Bloino, B.G. Janesko, R. Gomperts, B. Mennucci,
H.P. Hratchian, J.V. Ortiz, A.F. Izmaylov, J.L. Sonnenberg, D.
Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A.
Petrone, T. Henderson, D. Ranasinghe, V.G. Zakrzewski, J. Gao, N. Rega, G.
Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M.
Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J.
Montgomery, J. A., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E.
Brothers, K.N. Kudin, V.N. Staroverov, T. Keith, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi,
J.M. Millam, M. Klene, C. Adamo, R. Cammi, J.W. Ochterski, R.L. Martin, K.
Morokuma, O. Farkas, J.B. Foresman, D.J. Fox, Gaussian 09, Revision A.02,
in, Gaussian, Inc., Wallingford CT, 2016.
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
[41] J. Wang, W. Wang, P.A. Kollman, D.A. Case, Automatic atom type and
bond type perception in molecular mechanical calculations, J. Mol. Graphics
Modell., 25 (2006) 247-260.
[42] J. Wang, R.M. Wolf, J.W. Caldwell, P.A. Kollman, D.A. Case,
Development and testing of a general Amber force field, J. Comput. Chem.,
25 (2004) 1157-1174.
[43] D.A. Case, V. Babin, J.T. Berryman, R.M. Betz, Q. Cai, D.S. Cerutti, I.
Cheatham, T.E. , T.A. Darden, R.E. Duke, H. Gohlke, A.W. Goetz, S. Gusarov,
N. Homeyer, P. Janowski, J. Kaus, I. Kolossváry, A. Kovalenko, T.S. Lee, S.
LeGrand, T. Luchko, R. Luo, B. Madej, K.M. Merz, F. Paesani, D.R. Roe, A.
Roitberg, C. Sagui, R. Salomon-Ferrer, G. Seabra, C.L. Simmerling, W.
Smith, J. Swails, R.C. Walker, J. Wang, R.M. Wolf, W. X., K. P.A., Amber 14,
in, University of California, San Francisco, 2014.
[44] M.H.M. Olsson, C.R. Sondergaard, M. Rostkowski, J.H. Jensen,
PROPKA3: Consistent Treatment of Internal and Surface Residues in
Empirical pKa Predictions, J. Chem. Theory Comput., 7 (2011) 525-537.
[45] H. Grubmüller, V. Goll, SOLVATE1.0, in, University of Munich, 1996.
25

Kato et al.
ACCEPTED MANUSCRIPT
1
2
[46] J.-P. Ryckaert, G. Ciccotti, H.J.C. Berendsen, Numerical integration of
the cartesian equations of motion of a system with constraints: molecular
dynamics of n-alkanes, Journal of Computational Physics, 23 (1977) 327-341.
[47] Y.S. Kato, T. Yagi, S.A. Harris, S.Y. Ohki, K. Yura, Y. Shimizu, S. Honda,
R. Kamiya, S.A. Burgess, M. Tanokura, Structure of the microtubule-binding
domain of flagellar dynein, Structure (London, England : 1993), 22 (2014)
1628-1638.
3
4
5
6
7
8
[48] Y. Kato, H. Kihara, K. Fukui, M. Kojima, A ternary complex model of
Sirtuin4-NAD(+)-Glutamate dehydrogenase, Comput Biol Chem, 74 (2018)
94-104.
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
[49] A. Onufriev, D. Bashford, D.A. Case, Exploring protein native states and
large-scale conformational changes with a modified generalized born model,
Proteins, 55 (2004) 383-394.
[50] B.R. Miller, 3rd, T.D. McGee, Jr., J.M. Swails, N. Homeyer, H. Gohlke,
A.E. Roitberg, MMPBSA.py: An Efficient Program for End-State Free
Energy Calculations, Journal of chemical theory and computation, 8 (2012)
3314-3321.
[51] M. Malaisree, T. Rungrotmongkol, N. Nunthaboot, O. Aruksakunwong, P.
Intharathep, P. Decha, P. Sompornpisut, S. Hannongbua, Source of
oseltamivir resistance in avian influenza H5N1 virus with the H274Y
mutation, Amino acids, 37 (2009) 725-732.
[52] Y. Zou, F. Wang, Y. Wang, W. Guo, Y. Zhang, Q. Xu, Y. Lai, Systematic
study of imidazoles inhibiting IDO1 via the integration of molecular
mechanics and quantum mechanics calculations, European journal of
medicinal chemistry, 131 (2017) 152-170.
[53] J. Schlitter, Estimation of absolute and relative entropies of
macromolecules using the covariance matrix, Chemical Physics Letters, 215
(1993) 617-621.
[54] D.R. Roe, T.E. Cheatham, 3rd, PTRAJ and CPPTRAJ: Software for
Processing and Analysis of Molecular Dynamics Trajectory Data, Journal of
chemical theory and computation, 9 (2013) 3084-3095.
26

ACCEPTED MANUSCRIPT
Tables and Schemes
Table 1 Inhibition of human DAO by low molecular weight thiophene
carboxylic acids
Cmpd
Structure
IC50 (µM)
7.8 ± 1.7
1a
1.4 ± 0.1
0.72 ± 0.07
1.3 ± 0.2
4.6 ± 0.5
8.8 ± 1.3
21 ± 2
1b
1c
1d
1e
1f
1g
1h
1i
>100
1.3 ± 0.2
1.3 ± 0.1
1j
0.090 ± 0.016
0.36 ± 0.04
1k
1l
1

ACCEPTED MANUSCRIPT
7.3 ± 0.7
>100
1m
1n
>100
>100
>100
1o
1p
1q
>100
1r
2a
2b
2c
4.4 ± 0.6
0.036 ± 0.007
0.22 ± 0.02
27 ± 1
2d
2