Mechanism of eukaryotic serine racemase-catalyzed serine dehydration

Article abstract of DOI:10.1016/j.bbapap.2020.140460

Eukaryotic serine racemase (SR) is a pyridoxal 5′-phosphate enzyme belonging to the Fold-type II group, which catalyzes serine racemization and is responsible for the synthesis of D-Ser, a co-agonist of the N-methyl-D-aspartate receptor. In addition to racemization, SR catalyzes the dehydration of D- and L-Ser to pyruvate and ammonia. The bifuctionality of SR is thought to be important for D-Ser homeostasis. SR catalyzes the racemization of D- and L-Ser with almost the same efficiency. In contrast, the rate of L-Ser dehydration catalyzed by SR is much higher than that of D-Ser dehydration. This has caused the argument that SR does not catalyze the direct D-Ser dehydration and that D-Ser is first converted to L-Ser, then dehydrated. In this study, we investigated the substrate and solvent isotope effect of dehydration of D- and L-Ser catalyzed by SR from Dictyostelium discoideum (DdSR) and demonstrated that the enzyme catalyzes direct D-Ser dehydration. Kinetic studies of dehydration of four Thr isomers catalyzed by D. discoideum and mouse SRs suggest that SR discriminates the substrate configuration at C3 but not at C2. This is probably the reason for the difference in efficiency between L- and D-Ser dehydration catalyzed by SR.

Full text of DOI:10.1016/j.bbapap.2020.140460

BBA - Proteins and Proteomics 1868 (2020) 140460
Contents lists available at ScienceDirect
BBA - Proteins and Proteomics
journal homepage: www.elsevier.com/locate/bbapap
Mechanism of eukaryotic serine racemase-catalyzed serine dehydration
Tomokazu Itoa,1, Mai Matsuoka , Masaru Gotob,⁎, Soichiro Watanabe , Taichi Mizobuchi ,
a,1
b
b
b
a
a
a,⁎
Kazuma Matsushita , Ryoma Nasu , Hisashi Hemmi , Tohru Yoshimura
Department of Applied Molecular Biosciences, Graduate School of Bioagricultural Sciences, Nagoya University, Furou-chou, Chikusa, Nagoya, Aichi 464-8601, Japan
Department of Biomolecular Science, Faculty of Science, Toho University, Miyama 2-2-1, Funabashi, Chiba 274-8510, Japan
a
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
D-serine
Serine racemase
Pyridoxal 5′-phosphate
Reaction mechanism
Eukaryotic serine racemase (SR) is a pyridoxal 5′-phosphate enzyme belonging to the Fold-type II group, which
catalyzes serine racemization and is responsible for the synthesis of D-Ser, a co-agonist of the N-methyl-D-as-
partate receptor. In addition to racemization, SR catalyzes the dehydration of D- and L-Ser to pyruvate and
ammonia. The bifuctionality of SR is thought to be important for D-Ser homeostasis. SR catalyzes the racemi-
zation of D- and L-Ser with almost the same efficiency. In contrast, the rate of L-Ser dehydration catalyzed by SR
is much higher than that of D-Ser dehydration. This has caused the argument that SR does not catalyze the direct
D-Ser dehydration and that D-Ser is first converted to L-Ser, then dehydrated. In this study, we investigated the
substrate and solvent isotope effect of dehydration of D- and L-Ser catalyzed by SR from Dictyostelium discoideum
(DdSR) and demonstrated that the enzyme catalyzes direct D-Ser dehydration. Kinetic studies of dehydration of
four Thr isomers catalyzed by D. discoideum and mouse SRs suggest that SR discriminates the substrate con-
figuration at C3 but not at C2. This is probably the reason for the difference in efficiency between L- and D-Ser
dehydration catalyzed by SR.
1. Introduction
the dehydration of D- and L-Ser to pyruvate and ammonia in addition to
their racemization as shown in Fig. 1 [14]. The dehydratase activity of
Serine racemase (SR) belongs to the fold-type II group of pyridoxal
′-phosphate (PLP)-dependent enzymes. The enzyme catalyzes the in-
SR is probably important for the regulation of D-Ser concentration in
neuronal cells [15], especially in the forebrain area, where D-amino
acid oxidase bearing the role of D-Ser degradation was reported to be
absent or scarce [16].
5
terconversion between L- and D-Ser, and is responsible for the synthesis
of D-Ser. D-Ser acts as an intrinsic co-agonist of an N-methyl-D-aspartate
glutamate (NMDA) receptor in the mammalian brain and thus plays an
important role in the modulation of brain function [1–4, see references
cited in the recent review 4]. SR is structurally and evolutionarily dif-
ferent from PLP-dependent bacterial amino acid racemases such as
alanine racemase and serine racemase belonging to the fold-type III
group [5–7]. SRs are widely distributed among eukaryotes, including
plants [8], yeast Schizosaccharomyces pombe [5], and the cellular slime
mold Dictyostelium discoideum [9]. In addition to eukaryotes, SR has also
been found in the archaea, Pyrobaculum islandicum [10].
We investigated the reaction mechanism of SR using the D. dis-
coideum SR (DdSR) (Fig. 2), which exhibits 47% and 38% identity with
the primary structures of mouse SR and S. pombe SR, respectively [17].
Upon addition of substrates, PLP forms an internal Schiff base with the
K56 residue of DdSR (Fig. 2, I) which reacts with D- or L-Ser to form an
external Schiff base (II and V). Abstraction of a 2-proton from the serine
moiety of the external Schiff base produces an anionic intermediates
(III) stabilized by an “electron sink” function of the coenzyme. In DdSR,
K56 and S81, situated opposite to each other across PLP abstract the 2-
proton of L- and D-Ser, respectively [17]. Re-protonation at C2 of the
substrate moiety of the planar intermediate on the side opposite to that
of the proton abstraction produces an external Schiff base with the
antipodal Ser (III → V, II), which is then released from the enzyme by
In the mammalian brain, the D-Ser level is considered to be tightly
controlled, and abnormalities in the levels of D-Ser is associated with
various neurological disorders, including schizophrenia, Alzheimer's
disease [4,11,12], and amyotrophic lateral sclerosis [13]. SR catalyzes
Abbreviations: SR, serine racemase; DdSR, serine racemase of Dictyostelium discoideum; SpSR, serine racemase of Schizosaccharomyces pombe; PLP, pyridoxal 5′-
phosphate
⁎
Corresponding authors.
E-mail addresses: goto@biomol.sci.toho-u.ac.jp (M. Goto), yosimura@agr.nagoya-u.ac.jp (T. Yoshimura).
The first two authors contributed equally to this work.
1
https://doi.org/10.1016/j.bbapap.2020.140460
Received 27 February 2020; Received in revised form 14 May 2020; Accepted 26 May 2020
Available online 29 May 2020
1570-9639/ © 2020 Elsevier B.V. All rights reserved.

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
(A) Racemization
type SR.
In this work, we aimed to resolve the controversy regarding the D-
Ser dehydration mechanism through kinetic studies. We analyzed the
substrate and solvent isotope effects on D- and L-Ser dehydration cat-
alyzed by DdSR using 2-deuterated substrates and deuterated water,
respectively. We also obtained the kinetic parameters of the dehydra-
tion reaction of the four stereoisomers of Thr; L-Thr, L-allo-Thr, D-Thr,
and D-allo-Thr when catalyzed by DdSR and mouse SR. The obtained
results suggest that SR directly catalyzes D-Ser dehydration without
converting it to L-Ser, and that the efficiency of dehydration is affected
by the configuration of the carbon atom bearing a hydroxyl group in the
substrates. Therefore, SR probably recognizes the stereochemistry at
serine C3 but not at C2.
COOH
COOH
H N C
H
H C NH2
2
CH OH
CH OH
2
2
L-Ser
D-Ser
(
B) Dehydration
COOH
COOH
H N C
H
H C NH2
CH3 C COOH
+
NH
3
2
CH OH
CH OH
O
2
2
L-Ser
D-Ser
Pyruvic acid
Ammonia
Fig. 1. Reactions catalyzed by SR. SR catalyzes both racemization (A) and de-
hydration (B) of D- and L-Ser.
2. Results and discussion
2
.1. SR directly catalyzes D-Ser dehydration
transaldimination with K56 (V, II → I). Racemization and dehydration
probably share the same mechanism from the external Schiff base for-
mation to 2‑hydrogen abstraction (I → III). Dehydration of serine
proceeds through the elimination of a 3-hydroxyl group from the de-
protonated intermediate to produce a 2-aminoacrylate-PLP Schiff base
SRs, except for the P. islandicum SR, dehydrate D-Ser to pyruvate
and ammonia; however, the rate of this conversion is fairly low com-
pared with that of L-Ser. Two routes can be used for the dehydration of
D-Ser. SR catalyzes D-Ser dehydration directly, or D-Ser is converted to
L-Ser by racemization followed by dehydration. To clarify the route of
D-Ser dehydration, we investigated the substrate and solvent isotope
(
(
III → IV). The subsequent transaldimination releases 2-aminoacrylate
VI), which is hydrolyzed to pyruvate and ammonia. A concerted re-
action of the 2‑hydrogen abstraction and the hydroxyl group elimina-
tion (II, V → IV) is also possible for the mechanism of dehydration.
Racemization and dehydration reactions of SR are modulated with
various factors. ATP and divalent cations such as Mg2 activate the
enzyme (1, 14). In cultured astrocytes, SR activity was reported to be
enhanced by glutamate receptor interacting protein (GRIP1) [18]. SR is
also known to interact with Golgin subfamily A member 3 (Golga3)
protein, which prevents SR from ubiquitylation and subsequent
breakdown by the ubiquitin-proteasomal system [19]. Factors inter-
acting with SR are summarized in the review by Wolosker and Mori [3].
effects on D- and L-Ser dehydration using DdSR. When an amino acid
2
racemase reaction is performed in H
2
O, the 2‑hydrogen of the product
is replaced by deuteron due to exchange of the abstracted 2‑hydrogen
with the solvent deuteron before re-protonation of the anionic inter-
mediate [21]. If the conversion of D-Ser to L-Ser is required upon D-Ser
+
2
2
dehydration, [2− H]L-Ser should be the real substrate in H
case, the kcat value of D-Ser degradation in 2
O should be smaller
than that of D-Ser degradation in H O, provided the 2‑hydrogen ab-
straction is at least a rate-limiting step of the dehydration. We studied
the kinetics of the dehydration of 2-deuterated D- and L-Ser in H O and
2
O. In this
H
2
2
2
2
m
In the racemization reactions, the catalytic efficiencies (kcat/K ) of
D- and L-Ser are similar. However, the dehydration of L-Ser is higher
than that of D-Ser. This tendency is common among fold-type II SRs. For
H
2
O catalyzed by DdSR.
Table 1 summarizes the kinetic parameters of Ser dehydration. The
2
kcat value for L-[2– H]Ser was 41.2% of that of non-labeled L-Ser upon
dehydration in H O. The ratio of the kcat value in H O of L-[2-H]Ser to
that of L-[2– H]Ser, (in other words the substrate isotope effect on the
example, the kcat/K
lyzed by DdSR are 0.048 s /mM and 0.034 s /mM, respectively
17]. In contrast, the efficiency of L-Ser dehydration is much higher
than that of D-Ser dehydration. The kcat/K values in the L-Ser and D-
Ser dehydration catalyzed by DdSR obtained in this study were
m
values for L- and D-Ser in the racemization cata-
2
2
−1
−1
2
2
[
2
kcat of L-Ser dehydration) was 2.43. The kcat value in H O for D-[2– H]
m
Ser dehydration was 45.6% of that of non-labeled D-Ser. The substrate
isotope effect with D-Ser was 2.19. Substitution of deuteron for 2-H
gave a similar substrate isotope effect upon both L- and D-Ser dehy-
dration, suggesting that 2‑hydrogen abstraction is a rate-limiting step of
L- and D-Ser dehydration. The kcat values for non-labeled L- and D-Ser
−
1
−1
0
.120 s /mM and 0.025 s /mM, respectively. The difference in the
catalytic efficiency of D- and L-Ser dehydration is more remarkable
with mouse SR. The kcat/K values of racemization of L- and D-Ser
catalyzed by mouse SR were 0.2 s /mM and 0.13 s /mM, respec-
m
−1
−1
in the dehydration reaction in 2
the dehydration performed in H
H
O were 45.6 and 55.8% of those in
2
−1
tively. Those of the dehydration reaction were 0.34 s /mM and
2
O, respectively. Similar solvent isotope
−
1
0
.045 s /mM, respectively [20]. The preference for L-Ser in the de-
effects, 2.19 and 1.79 respectively, were also obtained. According to the
mechanistic studies of D-amino acid oxidase catalyzing the oxidative
deamination of various D-amino acid [22], the isotope effect on 2-H of
substrate and solvent isotope effect showed multiplicative behavior,
hydration reaction is most clearly observed with P. islandicum SR,
which exhibits similar catalytic efficiencies in the D- and L-Ser race-
−
1
m
mization. The kcat/K value of L-Ser dehydration was 28.2 s /mM, but
when the deuterated substrate was reacted in 2
H
O. Such tendencies
were also observed with DdSR. The isotope effects obtained in the de-
hydration of 2-deuterated L- and D-Ser in 2
O were 3.77 and 5.43,
the activity of D-Ser dehydration was below detection limits [10]. The
difference in the dehydration efficiency between D-Ser and L-Ser has
attracted interest because of the physiological importance of the de-
hydration activity of this enzyme.
2
H
2
respectively. These values are close to 4.33 and 4.80, which are ob-
To explain the imbalance between D- and L-Ser dehydration,
Konvalinka's group hypothesized that mouse SR catalyzes only L-Ser
dehydration and that D-Ser is converted to L-Ser by racemization fol-
lowed by dehydration. This hypothesis is supported by the fact that the
mouse SR catalyzes the dehydration of 3-Cl-L-Ala, L-Thr, L-Ser-O-sul-
fate, and L-threo-3-hydroxyaspartate, but their D-isomers were inert as
substrates for dehydration [20]. Meanwhile, Foltyn et al. argued that SR
directly catalyzes the D-Ser dehydration on the basis of studies of D-
and L-Ser metabolism in HEK 293 cells transformed with the wild-type
and mutant mouse SR genes [15], where they showed that most D-Ser is
dehydrated instead of being converted to L-Ser in cells expressing wild-
tained by the product of each substrate and solvent isotope effect with
L- and D-Ser. If D-Ser dehydration proceeds through its conversion to L-
2
Ser by racemization, the real substrate for dehydration of D-Ser in H
2
O
2
would be L-[2– H]Ser. If this is the case, the value of isotope effect upon
D-[2-H]-Ser dehydration in 2
O would be close to 4.33 or 3.77.
However, the isotope effect of 2.19 obtained in the dehydration of L-
H
2
2
2
[
2
2– H]Ser in H O is lower than these values. These results contradict
the hypothesis that D-Ser is converted to L-Ser prior to dehydration.
However, we were concerned about the similarity in kcat values
2
observed for L-[2– H]Ser and non-labeled D-Ser for dehydration in
2
−1
−1
H
2
O; 0.63
±
0.046 s
and 0.47
±
0.013 s , respectively
2

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
Fig. 2. Reaction mechanism of SR. Details are described in the text.
(
Table 1). As described above, if D-Ser dehydration proceeds through its
concluded that DdSR catalyzes D-Ser dehydration directly, without
racemization.
2
conversion to L-Ser, L-[2– H]Ser would be the real substrate upon D-Ser
dehydration in
dehydration in
dration in H
D-Ser dehydration proceeds after its conversion to L-Ser. However, this
would logical only when the concentration of [2– H]L-Ser in the re-
action system of D-Ser dehydration in
the [2– H]L-Ser dehydration system in H
possible if the rate of D- to L-Ser conversion in
than the dehydration of [2– H]L-Ser. To obtain the rate of D- to L-Ser
conversion in
changes in the D- and L-Ser concentrations were analyzed by HPLC. As
a result, no L-Ser formation was observed under the conditions, while
1
obtain the rate of L- to D-Ser conversion in
strate. Similar results were obtained: no D-Ser formation was observed
2
2
H O. If this is the case, the kcat value for [2-H]D-Ser
H O should be similar to that of [2– H]L-Ser dehy-
2
2
2
2
2.2. Dehydration of four threonine isomers catalyzed by DdSR and mouse
SR
2
O. Similar kcat values cannot rule out the possibility that
2
2
2-Hydrogen abstraction from the substrate moiety of the external
Schiff base is a common step in racemization and dehydration. The
small difference in the rate between D- and L-Ser racemization suggests
that another rate-limiting step of Ser dehydration exists besides 2‑hy-
drogen abstraction. We speculated that the imbalance between the D-
and L-Ser dehydration reactions should be derived from the difference
in the efficiency of 3-hydroxyl group elimination. We investigated the
effects of the stereochemistry at C3, that is, the orientation of the 3-
hydroxyl group on the DdSR-catalyzed epimerization and dehydration
with 4 Thr isomers, D-allo-Thr (2R, 3R), L-Thr (2S, 3R), D-Thr (2R, 3S),
and L-allo-Thr (2S, 3S) (Fig. 3). The rates of epimerization of DdSR
catalysis of these 4 Thr isomers were 0.118, 0.105, 0.063, and
2
H O was comparable to that of
2
2
2
O. Such conditions could be
2
H
2
O is notably faster
2
2
2
2 2
H O, D-Ser was incubated with DdSR in H O, and the
0–20% of D-Ser was consumed (data not shown). We also attempted to
2
2
H O with L-Ser as a sub-
even though L-Ser dehydration proceeded. These results suggest that in
2
2
H O, the rates of racemization of D- and L-Ser were lower than those of
−1
0
.043 s , respectively. The differences in the epimerization rates of
their dehydration. In racemization, the solvent isotope effect upon the
re-protonation of the anionic intermediate lowers the reaction rate of
racemization [21]. With the series of isotope effect studies, we
the 4 Thr isomers were less than 3 times, but those in the dehydration
rate were much larger. As shown in Table 2, the kcat values for D-allo-
−1
−1
Thr and L-Thr were 3.81
±
0.129 s
and 1.78
±
0.702 s
,
Table 1
Isotope effects in serine dehydration catalyzed by DdSR.
Substrate
Solvent
Km (mM)
kcat (s−1)
kcat/Km (s−1/mM)
kcat ratio (%)
Isotope effect⁎
L-Ser
H
2
2
O
O
19.8 ± 0.084
18.8 ± 2.15
20.5 ± 2.65
20.6 ± 3.53
41.2 ± 5.16
55.5 ± 9.80
33.5 ± 2.42
16.9 ± 2.27
2.38 ± 0.043
0.98 ± 0.047
1.33 ± 0.074
0.63 ± 0.046
1.03 ± 0.070
0.47 ± 0.051
0.47 ± 0.013
0.19 ± 0.0073
0.120
0.042
0.067
0.031
0.025
0.0083
0.014
0.011
100
1.0
2
L-[α- H]Ser
H
41.2
55.8
26.5
100
45.6
45.6
18.4
2.42
1.79
3.77
1.0
2.19
2.19
5.43
2
L-Ser
L-[α- H]Ser
D-Ser
H
H
2
O
2
2
2
O
H
2
O
O
2
O
2
D-[α- H]Ser
D-Ser
D-[α- H]Ser
H
2
2
H
H
2
2
2
O
Measurements were repeated three times. Other conditions were described in the text.
⁎
kcat value obtained with non-labeled substrate and solvent / kcat value obtained with labeled substrate and solvent.
3

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
L-Thr (2S, 3R)
OH
D-Thr (2R, 3S)
OH
3-hydroxyl group of Ser (Fig. 4).
Analysis of the SpSR active site structure [5] and site-directed mu-
tagenesis studies of DdSR [17] suggest that a loop in SpSR consisting
with the sequence S81-S82-G83-N84-H85 recognizes the serine sub-
strate, and the hydroxyl group of S81 coordinates with the carboxyl
group of the substrate. The residues of the loop are conserved in various
SRs, including DdSR. S80 of DdSR corresponding to S81 of SpSR
probably acts as a binding site for the carboxyl group of L- and D-Ser.
Because the loop is less basic, we hypothesize that the carboxyl group of
L- and D-Ser are protonated in the hydrophobic environment of the
active site. Based on this assumption, we speculate the following re-
action mechanism of DdSR-catalyzed L- and D-Ser dehydrations (Fig. 4
(A)). When L-Ser is a substrate, the 2‑hydrogen of the L-Ser moiety of an
external aldimine is abstracted as a proton by K56. After 2‑hydrogen
abstraction, the proton of the carboxyl group is transferred to the hy-
droxyl group of L-Ser. According to the model structure, the distance
between the hydroxyl group and the carboxyl group oxygen is 2.8 A.
Proton transfer can be possible. The abstracted 2‑hydrogen on K56 is
then transferred to the unprotonated carboxyl group. The following
transaldimination reaction releases 2-aminoacrylate, which is hydro-
lyzed to pyruvate and ammonia. When D-Ser is a substrate, proton
transfer from the carboxyl group to the hydroxyl group of D-Ser is
difficult. The proton comes from other places, for example, the phos-
phate group of PLP is 2.6 A away from the hydroxyl group (Fig. 4 (B)).
The unprotonated phosphate group is possibly protonated with the
O
O
H C
OH
H C
OH
3
3
NH2
D-allo-Thr (2R, 3R)
OH
NH2
L-allo-Thr (2S, 3S)
O
OH
O
H C
OH
H C
OH
3
3
NH2
NH2
Fig. 3. Structure of four Thr isomers.
Table 2
Kinetic parameters of Thr dehydration catalyzed by DdSR.
K
m
(mM)
k
cat (s−1)
k
cat/K
m
(s−1/mM)
−
4
D-allo-Thr (2R, 3R)
L-Thr (2S, 3R)
D-Thr (2R, 3S)
135 ± 7.42
225 ± 121
185 ± 19.5
47 ± 6.1
3.81 ± 0.129
1.78 ± 0.702
0.038 ± 0.0016
0.015 ± 0.00090
282 × 10
−
4
4
4
78.9 × 10
2.04 × 10
3.28 × 10
−
−
L-allo-Thr (2S, 3S)
2
‑hydrogen of D-Ser abstracted by Ser81 [17]. It is quite possible that
Measurements were repeated three times. Other conditions were described in
the text.
the efficiency of hydroxyl group protonation depends on its mechanism.
The difference in the efficiencies is a probable reason for the difference
in the dehydration rate between D- and L-Ser. However, the reaction
mechanism shown in Fig. 4 is constructed on the model structure of the
active-site accommodating L- and D-Ser. The proton transfer to the
hydroxyl group occurs with the anionic intermediate or concertedly
with the 2‑hydrogen abstraction. It is possible that the active-site
structure is different from that of the assumed models. To investigate
the conformation of 3-OH, we are performing X-ray crystallographic
studies of DdSR bound with a phosphopyridoxyl D- and L-Ser, which are
the homologs of the external aldimine intermediate of the reaction.
Table 3
Kinetic parameters of Thr dehydration catalyzed by mouse SR.
K
m
(mM)
k
cat (s−1)
kcat/K
m
(s−1/mM)
−
−
2
1
D-allo-Thr (2R, 3R)
L-Thr (2S, 3R)
D-Thr (2R, 3S)
62.9 ± 18.7
40.8 ± 2.80
ND⁎
4.29 ± 0.816
19.6 ± 0.76
< 0.05
6.82 × 10
4.80 × 10
–
–
L-allo-Thr (2S, 3S)
ND
< 0.05
Measurements were repeated three times. Other conditions were described in
the text.
⁎
3. Conclusion
ND; not determined.
−
1
This study aimed to resolve the controversy regarding the reaction
mechanism of Fold-type II serine racemase (SR)-catalyzed D-Ser dehy-
dration, that is, SR directly dehydrates D-Ser or dehydrates it after
conversion to L-Ser. The substrate and solvent isotope effects on D- and
L-Ser dehydration catalyzed by SR from Dictyostelium discoideum (DdSR)
have demonstrated that DdSR directly catalyzes D-Ser dehydration
without converting it to L-Ser. Kinetic parameters of the dehydration
reaction of the four stereoisomers of Thr; L-Thr, L-allo-Thr, D-Thr, and
D-allo-Thr when catalyzed by DdSR and mouse SR suggested that the
efficiency of dehydration is affected by the configuration of the carbon
atom bearing a hydroxyl group in the substrates. SR probably re-
cognizes the stereochemistry at serine C3 but not at C2.
respectively. Those for D-Thr and L-allo-Thr were 0.038 ± 0.0016 s
−
1
and 0.015 ± 0.0009 s , respectively. These values were approxi-
mately 47–258 times lower than the kcat values for D-allo-Thr and L-
Thr. The rate of Thr dehydration seems to be dependent on the con-
figuration at C3, but not on that at C2. The 3R configuration is pre-
ferred. A similar tendency was observed in Thr dehydration catalyzed
by mouse SR (Table 3). The kcat values for L-Thr and D-allo-Thr were
−1
−1
1
9.54 ± 0.76 s and 4.29 ± 0.816 s , respectively. In contrast, the
rates of D-Thr and L-allo-Thr dehydrations were below the detection
limits. These results suggest that the orientation of the hydroxyl group
of the bound substrate affects the catalytic efficiency. SR distinguishes
the configuration of threonine C3, but not C2. Disregard of the C2
configuration is not surprising for SRs that catalyze the racemization of
Ser and Thr.
4. Experimental section
4.1. Materials
2.3. Plausible mechanisms for serine dehydration
Amino acids, reduced form of nicotinamide adenine dinucleotide
Because a hydroxyl group is an inefficient leaving group, protona-
(NADH), and lactic dehydrogenase from pig heart were obtained from
tion probably occurs prior to the elimination of the hydroxyl group
from C3 as water. We speculated that differences in the efficiency of the
protonation of the hydroxyl group would affect the dehydration effi-
ciency between L- and D-Ser. We carried out X-ray crystallography of
DdSR (entry 6LUT) and constructed models of the active site structure
of the DdSR bound with L-Ser or D-Ser on the basis of the structure of
the modified SpSR, and considered the mechanism of protonation of the
FUJIFILM Wako Pure Chemicals (Osaka, Japan). Deuterium oxide
2
( H
2
O, 99.8%) was purchased from Sigma (St Louis, MO, USA). ATP
was purchased from Sigma-Aldrich Japan (Tokyo, Japan). All other
chemicals were of the highest grade commercially available. Plasmid
pDdSR, an expression vector of DdSR, was prepared as described pre-
viously [23]. pMSR, an expression vector of mouse SR was constructed
by the insertion of the c-DNA of mouse SR into pET28a. pMSR expresses
4

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
(
A) L-Ser dehydration
Lys56
Lys56
Lys56
H2N:
NH3
NH3
2
.8 A
HO
OH
HO
OH
O
O
H
O
O
O
NH
Ser81
HO
NH
Ser81
NH
Ser81
HO
HO
H
H
H
O
O
O
O
O
O
O
O
P
P
P
O
O
O
O
N
N
N
O
O
(
1)
(2)
(3)
(
B) D-Ser dehydration
Lys56
Lys56
Lys56
H2N:
NH3
NH3
O
OH
OH
OH
HO
O
O
HO
H
H
2
.6 A
NH
NH
NH
Ser81
Ser81
Ser81
HO
HO
O
H
H
O
O
O
O
O
O
O
O
O
O
P
P
P
O
O
O
O
N
N
O
N
(4)
(5)
(6)
Fig. 4. Proposed reaction mechanism of L-Ser (A) and D-Ser (B) dehydration catalyzed by SR. Details are described in the text.
the mouse SR with His-tag at C-terminus.
mixed with 1.4 mg of D-serine dehydratase from Saccharomyces cerevi-
siae [25] and 25 mL of water, followed by the incubation for overnight
at 30 °C. The reaction was terminated by boiling for 10 min. After
centrifugation, the supernatant was applied to a DOW 50×-8 column
pre-equilibrated with 0.01 N HCl. The column was washed with water,
4.2. Enzyme preparation
DdSR was overexpressed in E. coli Rosetta II (DE3) cells (Novagen,
3
and the serine was eluted with 1% NH solution. The fractions con-
Madison, WI, USA) harboring the plasmid pDdSR, and purified to
homogeneity according to the method by Ito et al. [23]. Mouse SR was
also purified with the same method, except that pMSR was used instead
of pDdSR.
taining Ser were collected and freeze-dried. The enantiopurity of each
Ser (99%) was determined by HPLC (Fig. S2). Replacement of the
α‑hydrogen by H was confirmed by 1H NMR as described above (Fig.
S1).
2
2
2
4
.3. Synthesis of D-[α- H]serine and L-[α- H]serine
4
.4. Assay of enzymatic activity
2
1
2
D-[α- H]Ser was obtained by the incubation of D-[α- H]Ser in H
2
O
2
2
with D-amino acid aminotransferase as described previously [21] ex-
cept that D-Ala was replaced by D-Ser. Replacement of the α‑hydrogen
by H was confirmed by AVANCE600 H NMR (Brucker, Billerica, MA,
The rates of dehydration of D-Ser, L-Ser, D-[α- H]Ser, L-[α- H]Ser,
D-Thr, L-Thr, D-allo-Thr, and L-allo-Thr catalyzed by DdSR and mouse
SR were assayed by a lactate dehydrogenase coupling method as pre-
viously described [9]. The reaction mixture for the rate assay mixture
(1 mL) consisted of 50 mM Tris–HCl buffer (pH 8.5), each substrate
amino acid, 20 μM PLP, 1 mM ATP, 1 mM MgCl , 0.3 mM NADH, 10 U
2
of lactic dehydrogenase, and appropriate amount of DdSR or mouse SR.
Substrate amino acids and their concentrations used were as follows: L-
Ser, 1–50 mM; D-Ser, 5–50 mM; L-[α- H]Ser, 5–50 mM; D-[α- H]Ser,
5–50 mM; L-Thr, 2–100 mM; D-Thr, 50–500 mM; L-allo-Thr,
3.75–112.5 mM; D-allo-Thr, 2–100 mM. The reaction was started by the
2
1
2
1
USA) (Fig. S1). L-[α- H]Ser was prepared by the conversion of D-[α- H]
2
2
1
Ser to L-[α- H]Ser with the mutant alanine racemase in H
2
O. D-[α- H]
Ser (5 mmol) was incubated with 0.5 mg of Y354N mutant of G.
stearothermophilus alanine racemase catalyzing serine racemization [24]
2
in a buffered
H
2
O containing 100 mM potassium phosphate buffer
O (5 mL). After incubation at 37 °C for 24 h, the re-
action mixture was boiled for 10 min and centrifuged at 20,000 xg. To
2
2
2
(
p H 8.0) in 2H
2
1
2
degrade D-[α- H]Ser and D-[α- H]Ser, the supernatant solution was
5

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
addition of SR and the decrease in absorbance at 340 nm was recorded
Table 4
with UV2450 spectrophotometer (Shimadzu, Japan). The K
m
and kcat
Data collection and processing.
values were determined by fitting data points to the Michaelis–Menten
equation using the Kaleida Graph program.
Racemization activities were assayed by HPLC (LC20 system,
Shimadzu, Japan) by measuring the amount of each antipodal Ser
formed from the corresponding enantiomer, respectively. The reactions
were performed in a 200-μL mixture containing 50 mM Tris–HCl
Diffraction source
KEK, PF, BL-17A
Wavelength (Å)
Temperature (K)
Detector
Crystal-detector distance (mm)
Rotation range per image (°)
Total rotation range (°)
Exposure time per image (s)
Space group
0.98000
100
ADSC Quantum 315r
181.3
0.5
200
2.0
P1
(
pH 8.5), 20 μM PLP, 1 mM ATP, 1 mM MgCl
2
, 50 mM substrate, and an
appropriate amount of DdSR or mouse SR at 30 °C for 30 min. The
reaction was terminated by the addition of 200 μL of 10% tri-
chloroacetic acid to the reaction mixture. After incubation at 4 °C for
a, b, c (Å)
α, β, γ (°)
Mosaicity (°)
40.64, 58.07, 64.53
87.40, 72.75, 69.60
0.22
10 min, the reaction mixture was centrifuged at 20,000 xg for 15 min.
After 3 times extraction of the supernatant with water-saturated diethyl
ether, the aqueous layer was collected. The amino acids were deriva-
tized to fluorescent diastereomers with N-acetyl-L-cysteine (NAC) and
o-phthalaldehyde (OPA) as follows. OPA-NAC reagent was prepared by
mixing of 20 mg OPA in 1 mL methanol and 10 mg NAC. The amino
acids were mixed with the same volume of 1:9 mixture of OPA-NAC
reagent and 0.4 M borate-NaOH buffer (pH 9.4), and then applied to the
Cosmosil 5C18-MSII column (4.6 × 250 mm, Nacalai tesque, Kyoto,
Resolution range (Å)
Total No. of reflections
No. of unique reflections
Completeness (%)
Redundancy
36.75–1.35 (1.42–1.35)
237,696
109,316
94.4 (90.1)
2.2 (2.2)
8.7 (2.0)
〈 I/σ(I)〉
R
r.i.m.
0.066 (0.476)
11.292
2
Overall B factor from Wilson plot (Å )
Values for the outer shell are given in parentheses.
2 4
Japan). Mobile phase buffer A was 50 mM Na HPO (pH 6.5), and
mobile phase buffer B was a 4:6 mixture of mobile phase buffer A and
methanol. The separations of serine and threonine enantiomers were
achieved by the linear gradient of mobile phase buffer B from 15% to
Table 5
Structure solution and refinement.
Resolution range (Å)
36.753–1.350
41.2% in mobile phase buffer A in 30 min. Total flow rate was 0.8 mL/
min and the column was kept at 28 °C.
Completeness (%)
σ cutoff
94.42
0
103,890
5411
0.2478
0.2681
2
The reactions performed in
method those performed in H O as described above. The buffer and
chemicals used for the experiment were dissolved in
2
H O were assayed with the same
No. of reflections, working set
No. of reflections, test set
Final Rcryst
2
2
2
H O.
Epimerization rates with D-allo-Thr, L-Thr, D-Thr, and L-allo-Thr
were assayed as follows. The reactions were performed in a 100 μL
mixture containing 50 mM Tris–HCl (pH 8.5), 20 μM PLP, 1 mM ATP,
a
Final Rfree
No. of non-H atoms
Protein
Ligand (PLP)
Water
R.m.s. deviations
Bonds (Å)
Angles (°)
Average B factors (Å )
Protein
Ligand (PLP)
Water
4529
30
534
1
3
2
mM MgCl , 50 mM each substrate, and 0.04 μM DdSR at 30 °C for
0 min for D-allo-Thr and L-Thr epimerization, for overnight for D-Thr
and L-allo-Thr epimerization. The reaction was terminated by the ad-
dition of 100 μL of 10% trichloroacetic acid. Epimerization activities
were assayed by HPLC as described above.
0.0071
1.2656
2
19.329
15.558
30.131
4.5. Crystallization
Ramachandran plot
Most favoured (%)
Allowed (%)
The crystals of DdSR were grown over a period of 7–14 days by
hanging-drop vaper diffusion at 5 °C. The reservoirs consisted of
0–26% PEG 1540, 0.1 M NaCl, and 5% 2-methyl-2,4-pentandiol buf-
98.2
1.8
2
Values for the outer shell are given in parentheses.
fered at pH 6.5 with 0.1 M PIPES adjusted with NaOH. The drops were
a
R free was monitored with 5% of the reflection data excluded
from the refinement.
composed of 3 μL of the protein solution and 3 μL of reservoir solution.
The crystals had unit-cell parameters
a = 40.64, b = 58.07,
c = 64.53 Å, α = 87.40, β = 72.75, γ = 69.60° with two DdSR mo-
value of 26.8%. Graphical representations were designed using the
CCP4mg package [31]. The refined coordinates and experimental
structure factors have been deposited in PDB (entry 6LUT). Refinement
statistics are summarized in Table 5.
lecules in the asymmetric unit of the P1 unit cell.
Data collection and processing.
Diffraction-quality DdSR crystals were soaked in a cryoprotectant
consisting of mother liquor supplemented with 10% (v/v) glycerol and
were flash-cooled in a nitrogen stream at 100 K. Diffraction data were
recorded on an ADSC Quantum 315r detector on beamline BL17A at the
KEK Photon Factory at Tsukuba, Japan. The diffraction data were
processed with XDS [26] and were merged and scaled with POINTLESS
4.7. Molecular modeling
The models of the active site structures of the DdSR bound with L-
Ser or D-Ser was constructed on the bases of the coordinate of SpSR,
whose active-site lysine is modified with D-alanyl PLP (PDB ID; 2ZR8).
The structure is suitable for the modeling of SR bound with substrate.
We first calculated the most stable configurations of L- and D-serine
with using Spartan ‘14 software (Wavefunction Inc.). They were ob-
tained by molecular mechanics calculation with conformational search,
rotating each bond of L- and D-Ser at 10 degree intervals for C2 - C3
bond and 120 degree intervals for other bonds. Density functional
theory calculations at the B3LYP/6-31G* level with water solvation
were conducted to the most stable conformer obtained above to
[
27] and SCALA [27]. Data-collection and processing statistics are
summarized in Table 4.
4.6. Structure solution and refinement
The crystal structure of DdSR was solved by molecular replacement
with AmoRe [28]. The structure of SpSR (PDB ID; 1V71; 5) was used as
a search model. Refinement was carried out using REFMAC5 [29] with
iterative cycles of model building in Coot [30]. The Rfαctor value of
DdSR was estimated to be 24.8% at a resolution of 1.35 Å, with an Rfree
6

T. Ito, et al.
B
B
A- Proteins and Proteomics 1868 (2020) 140460
optimize structure. The obtained configurations were used for the in-
itial conditions for calculation of the most stable configuration of both
amino acids in water. The resultant configuration of each L- and D-Ser
was manually replaced by D-alanyl moiety of PLP-D-Ala-Lys of SpSR.
[11] T. Nishikawa, Analysis of free D-serine in mammals and its biological relevance, J.
Chromatogr. B Analyt. Technol. Biomed. Life Sci. 879 (2011) 3169–3183.
[
[
[
[
12] J.T. Coyle, D.T. Balu, The role of serine Racemase in the pathophysiology of brain
disorders, Adv. Pharmacol. 82 (2018) 35–56, https://doi.org/10.1016/bs.apha.
2017.10.002.
13] J. Sasabe, T. Chiba, M. Yamada, K. Okamoto, I. Nishimoto, M. Matsuoka, S. Aiso, D-
serine is a key determinant of glutamate toxicity in amyotrophic lateral sclerosis,
EMBO J. 26 (2007) 4149–4159.
14] J. De Miranda, R. Panizzutti, V.N. Foltyn, H. Wolosker, Cofactors of serine racemase
that physiologically stimulate the synthesis of the N-methyl-D-aspartate (NMDA)
receptor coagonist D-serine, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 14542–14547.
15] V.N. Foltyn, I. Bendikov, J. De Miranda, R. Panizzutti, E. Dumin, M. Shleper, P. Li,
M.D. Toney, E. Kartvelishvily, H. Wolosker, Serine racemase modulates in-
tracellular D-serine levels through an α, β-elimination activity, J. Biol. Chem. 280
Declaration of Competing Interest
There is no conflict of interest to declare.
Acknowledgments
(2005) 1754–1763.
This work was supported in part by the Grant-in-Aid for Scientific
Research JSPS KAKENHI Grant Number 19H0288222380058 (to T. Y.)
from Japan Society for the Promotion of Science(JSPS).
[
[
[
16] K. Horiike, H. Tojo, R. Arai, M. Nozaki, T. Maeda, D-amino-acid oxidase is confined
to the lower brain stem and cerebellum in rat brain: regional differentiation of
astrocytes, Brain Res. 652 (1994) 297–303.
17] T. Ito, T.M. Maekawa, S. Hayashi, M. Goto, H. Hemmi, T. Yoshimura, Catalytic
mechanism of serine racemase from Dictyostelium discoideum, Amino Acids 44
(2013) 1073–1084.
18] P.M. Kim, H. Aizawa, P.S. Kim, A.S. Huang, S.R. Wickramasinghe, A.H. Kashani,
R.K. Barrow, R.L. Huganir, A. Ghosh, S.H. Snyder, Serine racemase: activation by
glutamate neurotransmission via glutamate receptor interacting protein and med-
iation of neuronal migration, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 2105–2110.
19] E. Dumin, I. Bendikov, V.N. Foltyn, Y. Misumi, Y. Ikehara, E. Kartvelishvily,
H. Wolosker, Modulation of D-serine levels via ubiquitin-dependent proteasomal
degradation of serine racemase, J. Biol. Chem. 281 (2006) 20291–20302.
20] K. Strísovský, J. Jirásková, A. Mikulová, L. Rulísek, J. Konvalinka, Dual substrate
and reaction specificity in mouse serine Racemase: identification of high-affinity
Dicarboxylate substrate and inhibitors and analysis of the β-Eliminase activity,
Biochemistry 44 (2005) 13091–13100.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bbapap.2020.140460.
[
[
References
[
1] H. Wolosker, K.N. Sheth, M. Takahashi, J.P. Mothet, R.O. Brady Jr., C.D. Ferris,
S.H. Snyder, Purification of serine racemase: biosynthesis of the neuromodulator D-
serine, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 721–725.
[
21] A. Watanabe, Y. Kurokawa, T. Yoshimura, T. Kurihara, K. Soda, N. Esaki, Role of
[
2] 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, Proc. Natl. Acad. Sci. U. S. A. 97 (2000)
lysine 39 of alanine racemase from Bacillus stearothermophilus that binds pyridoxal
′-phosphate. Chemical rescue studies of Lys39 – > Ala mutant, J. Biol. Chem. 274
5
4
926–4931.
(1999) 4189–4194.
[
[
[
3] H. Wolosker, H. Mori, Serine racemase: an unconventional enzyme for an un-
conventional transmitter, Amino Acids 43 (2012) 1895–1904.
4] H. Wolosker, The neurobiology of D-serine Signaling, Adv. Pharmacol. 82 (2018)
[22] L. Pollegioni, W. Blodig, S. Ghisla, On the mechanism of D-amino acid oxidase.
Structure/linear free energy correlations and deuterium kinetic isotope effects using
substituted phenylglycines, J. Biol. Chem. 272 (1997) 4924–4934.
23] T. Ito, H. Murase, M. Maekawa, M. Goto, S. Hayashi, H. Saito, M. Maki, H. Hemmi,
T. Yoshimura, Metal ion dependency of serine racemase from Dictyostelium dis-
coideum, Amino Acids 43 (2012) 1567–1576.
[24] W.M. Patrick, J. Weisner, J.M. Blackburn, Site-directed mutagenesis of Tyr354 in
Geobacillus stearothermophilus alanine racemase identifies a role in controlling
substrate specificity and a possible role in the evolution of antibiotic resistance,
Chembiochem. 3 (2002) 789–792.
[
3
25–348, https://doi.org/10.1016/bs.apha.2017.08.010.
5] M. Goto, T. Yamauchi, N. Kamiya, I. Miyahara, T. Yoshimura, H. Mihara,
T. Kurihara, K. Hirotsu, N. Esaki, Crystal structure of a homolog of mammalian
serine racemase from Schizosaccharomyces pombe, J. Biol. Chem. 284 (2009)
2
5944–25952.
[
6] M.A. Smith, V. Mack, A. Ebneth, I. Moraes, B. Felicetti, M. Wood, D. Schonfeld,
O. Mather, A. Cesura, J. Barker, The structure of mammalian serine racemase:
evidence for conformational changes upon inhibitor binding, J. Biol. Chem. 285
[
25] T. Ito, H. Hemmi, K. Kataoka, Y. Mukai, T. Yoshimura, A novel zinc-dependent D-
serine dehydratase from Saccharomyces cerevisiae, Biochem. J. 409 (2008) 399–406.
[26] W. Kabsch, XDS, Acta Cryst D66, (2010), pp. 125–132.
(2010) 12873–12881.
[
7] T. Yoshimura, M. Goto, D-amino acids in the brain: structure and function of pyr-
idoxal phosphate-dependent amino acid racemases, FEBS J. 275 (2008) 3527–3537.
8] Y. Fujitani, N. Nakajima, K. Ishihara, T. Oikawa, K. Ito, M. Sugimoto, Molecular and
biochemical characterization of a serine racemase from Arabidopsis thaliana,
Phytochemistry 67 (2006) 668–674.
[27] P.R. Evans, Scaling and assessment of data quality, Acta Cryst (2006) 72–82 D62.
[28] J. Navaza, AMoRe: an automated package for molecular replacement, Acta
Crystallogr. (1994) 157–163. A47.
[29] G.N. Murshudov, P. Skuba, K.A.A. Lebedev, N.S. Pannu, R.A. Steiner, R.A. Nicholls,
M.D. Winn, F. Long, A.A. Vagin, REFMAC5 for the refinement of macromolecular
crystal structures, Acta Cryst D67 (2011) 355–367.
[30] P. Emsley, B. Lohkamp, W.G. Scott, K. Cowtan, Features and development of coot,
Acta Cryst (2010) 486–501 D66.
[31] S. McNicholas, E. Potterton, K.S. Wilson, M.E.M. Noble, Presenting your structures:
the CCP4mg molecular-graphics software, Acta Cryst. (2011) 386–394 D67.
[
[
9] T. Ito, H. Murase, M. Maekawa, S. Hayashi, H. Saito, M. Maki, H. Hemmi,
T. Yoshimura, Metal ion dependency of serine racemase from Dictyostelium dis-
coideum, Amino Acids 43 (2012) 1567–1576.
[
10] M. Ohnishi, M. Saito, S. Wakabayashi, M. Ishizuka, K. Nishimura, Y. Nagata,
S. Kasai, Purification and characterization of serine racemase from a hy-
perthermophilic archaeon, Pyrobaculum islandicum, J. Bacteriol. 190 (2008)
1
359–1365.
7