Isolation and amino acid sequence of a dehydratase acting on D-erythro-3- hydroxyaspartate from Pseudomonas sp. N99, and its application in the production of optically active 3-hydroxyaspartate

Article abstract of DOI:10.1080/09168451.2017.1295804

An enzyme catalyzing the ammonia-lyase reaction for the conversion of D-erythro-3-hydroxyaspartate to oxaloacetate was purified from the cell-free extract of a soil-isolated bacterium Pseudomonas sp. N99. The enzyme exhibited ammonia-lyase activity toward L-threo-3-hydroxyaspartate and D-erythro-3- hydroxyaspartate, but not toward other 3-hydroxyaspartate isomers. The deduced amino acid sequence of the enzyme, which belongs to the serine/ threonine dehydratase family, shows similarity to the sequence of L-threo-3-hydroxyaspartate ammonia- lyase (EC 4.3.1.16) from Pseudomonas sp. T62 (74%) and Saccharomyces cerevisiae (64%) and serine racemase from Schizosaccharomyces pombe (65%). These results suggest that the enzyme is similar to L-threo-3-hydroxyaspartate ammonia-lyase from Pseudomonas sp. T62, which does not act on D-erythro-3-hydroxyaspartate. We also then used the recombinant enzyme expressed in Escherichia coli to produce optically pure L-erythro-3-hydroxyaspartate and D-threo-3-hydroxyaspartate from the corresponding DL-racemic mixtures. The enzymatic resolution reported here is one of the simplest and the first enzymatic method that can be used for obtaining optically pure L-erythro-3-hydroxyaspartate.

Full text of DOI:10.1080/09168451.2017.1295804

Bioscience, Biotechnology, and Biochemistry
ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: http://www.tandfonline.com/loi/tbbb20
Isolation and amino acid sequence of
a dehydratase acting on d-erythro-3-
hydroxyaspartate from Pseudomonas sp. N99, and
its application in the production of optically active
3-hydroxyaspartate
Hiroyuki Nagano, Kana Shibano, Yu Matsumoto, Atsushi Yokota & Masaru
Wada
To cite this article: Hiroyuki Nagano, Kana Shibano, Yu Matsumoto, Atsushi Yokota &
Masaru Wada (2017): Isolation and amino acid sequence of a dehydratase acting on d-
erythro-3-hydroxyaspartate from Pseudomonas sp. N99, and its application in the production
of optically active 3-hydroxyaspartate, Bioscience, Biotechnology, and Biochemistry, DOI:
10.1080/09168451.2017.1295804
To link to this article: http://dx.doi.org/10.1080/09168451.2017.1295804
Published online: 14 Mar 2017.
Article views: 14
View supplementary material
Submit your article to this journal
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tbbb20
Download by: [University of Newcastle, Australia]
Date: 21 March 2017, At: 03:23

Bioscience, Biotechnology, and Biochemistry, 2017
Isolation and amino acid sequence of a dehydratase acting on D-erythro-3-
hydroxyaspartate from Pseudomonas sp. N99, and its application in the
production of optically active 3-hydroxyaspartate
Hiroyuki Nagano, Kana Shibano, Yu Matsumoto, Atsushi Yokota and Masaru Wada*
Laboratory of Microbial Physiology, Research Faculty of Agriculture, Hokkaido University, Sapporo, Japan
Received December 6, 2016; accepted February 11, 2017
http://dx.doi.org/10.1080/09168451.2017.1295804
An enzyme catalyzing the ammonia-lyase reaction
for the conversion of D-erythro-3-hydroxyaspartate
to oxaloacetate was purified from the cell-free
extract of a soil-isolated bacterium Pseudomonas sp.
N99. The enzyme exhibited ammonia-lyase activity
toward L-threo-3-hydroxyaspartate and D-erythro-3-
hydroxyaspartate, but not toward other 3-hydrox-
yaspartate isomers. The deduced amino acid
sequence of the enzyme, which belongs to the serine/
threonine dehydratase family, shows similarity to
the sequence of L-threo-3-hydroxyaspartate ammo-
nia-lyase (EC 4.3.1.16) from Pseudomonas sp. T62
(74%) and Saccharomyces cerevisiae (64%) and
serine racemase from Schizosaccharomyces pombe
(65%). These results suggest that the enzyme is sim-
ilar to L-threo-3-hydroxyaspartate ammonia-lyase
from Pseudomonas sp. T62, which does not act on
D-erythro-3-hydroxyaspartate. We also then used the
recombinant enzyme expressed in Escherichia coli to
produce optically pure L-erythro-3-hydroxyaspartate
and D-threo-3-hydroxyaspartate from the corre-
sponding DL-racemic mixtures. The enzymatic reso-
lution reported here is one of the simplest and the
first enzymatic method that can be used for obtain-
ing optically pure L-erythro-3-hydroxyaspartate.
mammalian serine racemase7) have been reported.
Thus, 3-hydroxyaspartate is an interesting molecule that
plays an important role in the research of the
mammalian central nervous system. However, 3-hy-
droxyaspartate has two chiral centers, and their four
stereoisomers—D-threo-3-hydroxyaspartate (2R, 3R;
D-THA), L-threo-3-hydroxyaspartate (2S, 3S; L-THA),
D-erythro-3-hydroxyaspartate (2R, 3S; D-EHA), and
L-erythro-3-hydroxyaspartate (2S, 3R; L-EHA)—are dif-
ficult to synthesize in an optically pure form. Various
routes for the synthesis of optically active 3-hydroxyas-
partate isomers have been reported, but most of them
use tartrate derivatives as chiral templates and contain
four or more steps to obtain optically pure isomers.8) In
addition, the synthesis of L-THA from L-aspartate
through a completely stereoselective cyclization to
trans-oxazoline9) and
a
stereoselective synthetic
approach for erythro-hydroxyaspartate through nucle-
ophilic ring opening of racemic trans-aziridine-2,3-di-
carboxylate also has been reported.10) These improved
methods, however, still need chiral starting materials or
special reagents. In contrast, a racemic mixture of
DL-THA and DL-EHA can be easily synthesized from
inexpensive cis- or trans-epoxysuccinate in a one-step
reaction.11,12) Thus, enzymes that can distinguish the
L-form or D-form of 3-hydroxyaspartate isomers are
useful for the enzymatic resolution of DL-racemic 3-hy-
droxyaspartate mixtures.
We recently reported the identification of a novel
enzyme, D-threo-3-hydroxyaspartate ammonia-lyase
(dehydratase) (EC 4.3.1.27), from Delftia sp. HT23.13)
The production of optically active L-THA from
DL-THA by enzymatic resolution using this enzyme has
also been reported.12) In addition, the enzymatic
syntheses of L-THA from L-aspartate or L-asparagine
have also been reported.14,15) However, no dehydratase
acting on the D-erythro form of 3-hydroxyaspartate has
been reported so far. Therefore, the enzymatic resolu-
tion of DL-EHA to produce L-EHA has not been
achieved to date.
Key words: 3-hydroxyaspartate;
ammonia-lyase;
enzymatic resolution; pyridoxal 5′-phos-
phate; Pseudomonas sp. N99
Hydroxyaspartate is a non-proteinogenic amino acid;
however, it was first found in human cerebrospinal
fluid in the 1960s1) and has been known as an inhibitor
of aspartate aminotransferase2) or the intermediate of
glycolate metabolism in some bacteria3,4) for a long
time. More recently, its biological activities such as
affinity to N-methyl-D-aspartate (NMDA) receptors,5)
competitive blocking of human brain excitatory
amino-acid transporters (EAATs),6) and inhibition of
*Corresponding author. Email: wada@chem.agr.hokudai.ac.jp
Abbreviations: DH, dehydratase; D-EHA, D-erythro-3-hydroxyaspartate; D-THA, D-threo-3-hydroxyaspartate; EAATs, excitatory amino-acid trans-
porters; L-EHA, L-erythro-3-hydroxyaspartate; L-THA, L-threo-3-hydroxyaspartate; MALDI-TOF-MS, matrix-assisted laser desorption/ionization
time-of-flight mass spectrometry; NADH, nicotinamide adenine dinucleotide; NMDA, N-methyl-D-aspartate; PAGE, polyacrylamide gel elec-
trophoresis; PCR, polymerase chain reaction; PLP, pyridoxal 5′-phosphate; SDS, sodium dodecyl sulfate; TLC, thin layer chromatography
© 2017 Japan Society for Bioscience, Biotechnology, and Agrochemistry

2
H. Nagano et al.
Threo-isomers have attracted more attention from
DH) from Delftia sp. HT23.12) Restriction endonucle-
ases were obtained from TaKaRa Bio Inc. (Kusatsu,
Japan). All other chemicals were of analytical grade
and commercially available.
biochemists, because these threo-isomers, especially
L-THA, have more affinity to NMDA receptors5) and
more inhibitory activity toward EAATs in the mam-
malian central nervous system than erythro-isomers.16)
However, L-EHA has been identified as an inhibitor of
mammalian serine racemase, producing D-serine, which
is a co-agonist of NMDA receptors.7) L-EHA also has
affinity to EAATs and NMDA receptors, even
though the affinities are lower than those of threo-
isomers.17) Thus, in addition to L-THA, L-EHA is also
an important molecule for biochemical studies, and an
enzyme capable of degrading D-EHA might be useful
for the enzymatic optical resolution of DL-EHA to pro-
duce optically pure L-EHA.
Therefore, in this study, we screened microorganisms
that can utilize D-EHA as the sole carbon source. A
newly isolated bacterium, Pseudomonas sp. N99, was
found to produce an enzyme that catalyzes the dehy-
dratase reaction for the conversion of D-EHA to
oxaloacetate (Fig. 1). We report here the purification
and amino acid sequence of this novel enzyme, D-EHA
dehydratase (D-EHA DH). An efficient expression
system for this enzyme was also established using
Escherichia coli. Moreover, we also applied this
enzyme for the production of optically pure L-EHA and
D-THA by enzymatic optical resolution. The findings of
this study provide a new approach for obtaining opti-
cally pure L-EHA.
Screening of D-EHA-degrading microorganisms from
soil. Screening of D-EHA-degrading microorganisms
was done as described previously.13) The nitrogen
source for the screening medium used in previous stud-
ies, D-THA, was replaced by D-EHA. Identification of
the isolated bacterium was done by determining the
most variable region of the 16S rDNA sequence (460
bp) using the primer set of V1 (5′-AGAGTTTGATCC-
TGGCTCAG-3′) and V3 (5′-ACCGCGGGKGCTGGC-
3′, K = G/T).
Microorganism and cultivation. Pseudomonas sp.
N99 isolated from soil and identified in our laboratory
was used as the enzyme source. Pseudomonas sp. N99
has been deposited in the AHU Culture Collection of
Hokkaido University under the accession number
AHU2141. Pseudomonas sp. N99 was cultivated as
described previously.19) E. coli DH5α (TaKaRa Bio
Inc.) and E. coli BL21 (DE3) (Merck Millipore,
Darmstadt, Germany) were used as the host cells for
cloning and expression of the gene (dehadh), respec-
tively, and were grown as described previously.13)
Enzyme assays. The activity of 3-hydroxyaspartate
dehydratase was determined as described previously.12)
The enzyme activity was coupled to malate dehydroge-
nase, which reduces oxaloacetate formed by the dehy-
dratase to malate in the presence of NADH. One unit
of the enzyme was defined as the amount capable of
catalyzing the oxidation of 1 μmol of NADH per min.
Serine racemase and serine dehydratase activities were
determined as described previously.13,19)
Materials and methods
Materials. DL-EHA was synthesized by ammonoly-
sis of ( )-trans-epoxysuccinic acid (Tokyo Chemical
Industry, Tokyo, Japan) at 60 °C as described previ-
ously.12) The synthesis was confirmed by silica gel thin
layer chromatography (TLC) using 7:1:2 (v/v/v) of
ethanol/aqueous ammonia (28%)/distilled water as a
developing solvent. L-THA, L-EHA, and DL-THA were
purchased from Tocris Cookson, Ltd. (Bristol, UK),
Wako Pure Chemical Industries (Osaka, Japan), and
Tokyo Chemical Industry, respectively. D-THA was
prepared from DL-THA by enzymatic resolution using
L-THA dehydratase (L-THA DH) from Saccharomyces
cerevisiae.18) D-EHA was prepared from DL-EHA by
enzymatic resolution using D-THA dehydratase (D-THA
Enzyme purification.
Enzyme purification proce-
dures were performed as described previously.13) The
final Mono Q step was performed as follows: The
partially purified enzyme after Superdex-200 chro-
matography was applied to a Mono Q 5/50 GL column
L-threo-3-hydroxyaspartate
D-EHA DH
(L-THA)
Oxaloacetate
D-erythro-3-hydroxyaspartate
(D-EHA)
Fig. 1. Scheme of the ammonia-lyase reactions of D-erythro-3-hydroxyaspartate dehydratase (D-EHA DH).

Novel D-erythro-3-hydroxyaspartate dehydratase
3
(0.5 × 5 cm, GE Healthcare UK Ltd, Little Chalfont,
UK) previously equilibrated with the buffer and was
eluted using the ÄKTA system (GE Healthcare UK
Ltd) with a linear gradient of 0–0.65 M NaCl in 20 mL
of buffer at a flow rate of 2.0 mL/min. The activity-
containing fractions, eluted with approximately 0.3 M
NaCl, were collected, dialyzed against 3 L of the buffer
for 16 h, and used as the purified enzyme.
site of the pET15b expression vector (Merck,
Darmstadt, Germany) to obtain pET15b_dehadh. The
expression vector was then introduced into E. coli
BL21 (DE3) cells, and the nucleotide sequence of the
insert was then confirmed.
Expression and purification of recombinant D-EHA
DH.
Expression and purification of recombinant
D-EHA DH was performed as described previously.13)
Briefly, the E. coli BL21 (DE3) carrying pET15b_de-
hadh were grown at 37 °C in LB medium containing
ampicillin (100 μg/mL). In order to induce gene
expression, 0.1 mM IPTG was added to the culture
medium when the absorbance at 600 nm reached 0.5.
After cultivation for another 8 h, cells were harvested
by centrifugation. For the recombinant enzyme purifica-
tion, HisTrap HP column (1.6 × 2.5 cm; GE Health-
care) connected to the ÄKTA system was used. The
enzyme was eluted with a 20–500 mM liner imidazole
gradient. Active fractions were collected, dialyzed, and
used as the purified enzyme.
Amino acid sequence analysis.
The internal
peptide sequences were determined using the In-Gel
Tryptic Digestion kit (Thermo Fisher Scientific Inc.,
Waltham, MA, USA). After trypsine digestion of the
enzyme, the peptides were separated by HPLC (L-2000
system; Hitachi High-Tech Science Corp., Tokyo,
Japan) on an Inertsil WP300 C8 column (4.6 × 250
mm; GL Science, Tokyo, Japan). Edman degradation
was performed at the Global Facility Center, Hokkaido
University.
Isolation of DNA encoding D-EHA DH (dehadh).
Degenerate PCR was performed as described previ-
ously.19) Oligonucleotide primer pools were designed
based on the amino acid sequences of the N-terminal
(MQIILPT) and internal (EIVHIDT) peptides. The
alignment of the primers was 5′-ATGCARATHATHY-
TICCNAC-3′ (sense strand) 5′-GGNGTRTCDATRT-
GIACDATYTC-3′ (antisense strand) (D = A/G/T,
H = A/C/T, I = inosinic acid, N = A/C/T/G, R = A/G,
Y = C/T). The chromosomal DNA of Pseudomonas sp.
N99 was used as the template. The chromosomal DNA
of Pseudomonas sp. N99 was digested with EcoRI at 37
°C overnight and purified by ethanol precipitation. The
DNA fragments were then circularized with Ligation
Solution I and II in the TaKaRa LA PCR in vitro cloning
kit (Takara Bio Inc.) at 16 °C for 30 min and then used
as a template for inverse PCR. Inverse PCR was per-
formed using the TaKaRa LA PCR in vitro cloning kit
with a sense primer (5′-AGCGAAAGTCTGGAGTTG-
TAGCGTTCT-3′) and an antisense primer (5′-TCATAC-
GATCATCCGGATGTCTTGTCA-3′). The PCR was
performed following with the manufacturer’s protocol.
The inverse PCR product was purified from the agarose
gel and sequenced directly.
Inhibitor
chloride (Kanto Chemical Co. Inc., Tokyo, Japan) and O
(carboxymethyl)hydroxylamine hemihydrochloride
experiments.
Hydroxylammonium
(Sigma-Aldorich Corp., St. Louis, USA) were used as
inhibitors. The inhibitors were added to the standard
enzyme assay mixture minus PLP at the concentration of
5 mM, and then the enzyme activities were measured.
Optical resolutions using recombinant D-EHA DH.
Optical resolutions of DL-EHA and DL-THA were
carried out as described previously by modifying the
reaction temperature (25 °C).12) Final products were
then identified with HPLC and NMR as described pre-
viously.12) Other methods including MALDI-TOF-MS
analysis were done as described previously.19)
Nucleotide sequence accession number.
The
nucleotide sequence of the dehadh gene of
Pseudomonas sp. N99 has been deposited in the
DDBJ/EMBL/GenBank database under the accession
number AB771451.
Construction of the expression plasmid.
For
Results
Screening
expression of dehadh in E. coli, a DNA fragment
containing the open reading frame of dehadh was pre-
pared by PCR using Pseudomonas sp. N99 genomic
DNA as the template and the oligonucleotide sense (5′-
ATGCGGATCCGATGCAGATTATTTTGCCGACCT-
ACAGCGAT-3′) and antisense primers (5′-ATGCGGA-
TCCTTAGGCTTGCAGCAAACGACCAAACTTATC-3′)
(the underlined sequence are the BamHI sites). The
PCR was performed using KOD FX Neo (Toyobo Co.,
Ltd., Osaka, Japan). The thermal cycler program was
set at 94°C for 2 min, followed by 35 cycles of 98°C
for 10 s, and 68°C for 1 min. The unique amplified
band corresponding to approximately 1000 bp was
digested with BamHI and then ligated into the BamHI
D-EHA-utilizing microorganisms were screened, and
the strains with high D-EHA-degrading activity were
isolated from soil. Three hundred and ninety strains
were isolated as D-EHA-utilizing microorganisms, and
31 of them were selected by TLC as D-EHA-degrading
microorganisms. Most of these 31 strains showed 0.02–
0.2 unit (mg protein)–1 activity for D-EHA dehydratase,
which produces oxalacetate from D-EHA. Among these
strains, strain N99, which is a rod-shaped gram-nega-
tive bacterium, was selected as the enzyme source,
because it showed the highest D-EHA DH activity
(around 0.2 unit (mg protein)–1).

4
H. Nagano et al.
The most variable region of the 16S rDNA sequence
(a)
(b)
(460 bp) revealed 98–99% identity to that of several
Pseudomonas strains. These results indicated that strain
N99 belongs to the genus Pseudomonas. Thus, we des-
ignated this strain as Pseudomonas sp. N99.
97.4K
66.3
42.4
30.0
Induction of the enzyme activity
When Pseudomonas sp. N99 was cultured in med-
ium containing DL-EHA or DL-THA as the sole nitrogen
source, the cell-free extract showed an enzyme activity
of approximately 0.2 unit (mg protein)–1. However,
when DL-EHA or DL-THA in the medium was replaced
by an equal amount of D-serine, D-threonine, D-aspar-
tate, or peptone, the cell-free extract showed no
enzyme activity or only a trace [<0.02 unit (mg pro-
tein)–1] of enzyme activity. These results suggested that
the enzyme activity was induced by 3-hydroxyaspartate
in the medium. It should be noted that this D-EHA DH
activity was induced not only by DL-EHA but also by
DL-THA. Judging form the substrate specificity of this
enzyme (see below), the actual inducer seems to be
D-EHA and L-THA. However, the detail of enzyme
induction including the actual inducer(s) remains
unclear.
20.1
14.4
Fig. 2. SDS-PAGE analysis of D-EHA DH from Pseudomonas sp.
N99.
Notes: Lane (a), molecular weight markers; Lane (b), purified
D-EHA DH after final Mono Q 5/50 GL chromatography.
putative serine/threonine dehydratases such as those
from Comamonas teststeroni, Pseudomonas fluo-
rescens, and Acinetobacter baumannii AB307-0294.
Purification and molecular weight determination of
the enzyme
The enzyme was purified approximately 21-fold to
almost homogeneity, with approximately 0.4% recovery
(Table 1). The purified enzyme preparation produced
two bands on SDS-PAGE (Fig. 2). Judging from the
enzyme activity of each fraction from final Mono Q
chromatography, the upper band with a relative molec-
ular mass of 39,100 seemed to correspond to the
enzyme having D-EHA DH activity.
Nucleotide sequence of the D-EHA DH gene from
Pseudomonas sp. N99
Based on the partial amino acid sequence, the full-
length open reading frame of the gene encoding the
enzyme was isolated through degenerate and inverse
PCR approaches. The open reading frame was 966 bp
long and encoded a protein of 321 amino acid residues,
with a predicted molecular weight of 34,240. The amino
acid sequence showed high identity with similar
enzymes from various gram-negative bacteria including
putative threo-3-hydroxyaspartate ammonia-lyase from
Pseudomonas chlororaphis subsp. aureofaciens 30–84
(87%) and putative catabolic threonine dehydratase from
Pseudomonas sp. PAMC25886 (75%) and Pseudomonas
sp. Ag1 (74%). In addition, the amino acid sequence
showed high identity with L-THA DH (EC 4.3.1.16)
from Pseudomonas sp. T62 (74%)19) and S. cerevisiae
(64%)18) and with serine racemase from Schizosaccha-
romyces pombe (65%),21) whose functions have already
been analyzed. Figure 3 shows multiple alignment of
these deduced amino acid sequences. These results sug-
gested that enzymes having similar amino acid sequence
N-Terminal and internal amino acid sequence
analysis
Automated Edman degradation of the enzyme was
performed with
a pulsed liquid phase sequencer
and yielded an NH2-terminal amino acid sequence
of MQIILPTYSDVVDASQRLQGVANVTPPL. The
enzyme was then digested by trypsin, and the resultant
peptides were separated by HPLC. One peptide was
isolated, and the amino acid sequence was found to be
SFNAGEIVHIDTPDTIAD. The two sequences were
compared with sequences in a protein sequence data-
base (nr-aa) using the sequence similarity search pro-
gram FASTA.20) Both sequences showed relatively
high identities to the partial amino acid sequences of
Table 1. Purification of D-EHA DH from Pseudomonas sp. N99.
Step
Total Protein (mg)
Total Activity (U)
Specific Activity (U/mg)
Purification (fold)
Yield (%)
Cell extract
903
108
0.12
0.29
0.16
0.77
0.99
2.12
2.54
1.00
2.41
1.33
6.42
8.25
17.7
21.2
100
Hi-Prep Q FF
Ammonium sulfate
Hi-Trap Butyl FF
Resource Q
Superdex-200
Mono Q 5/50 GL
78.6
72.2
9.68
2.21
0.37
0.17
22.8
11.6
7.45
2.19
0.78
0.43
21.1
10.7
6.90
2.03
0.72
0.40

Novel D-erythro-3-hydroxyaspartate dehydratase
5
to D-EHA DH are probably distributed broadly in nature,
not only in gram-negative soil bacteria but also in
eukaryotic microorganisms. According to the PROSITE
database (http://au.expasy.org/prosite/),23) the pyridoxal-
phosphate attachment site of serine/threonine dehy-
dratase, 51-TGSFKFRGA-58, was found. This indicated
that Lys55 of D-EHA DH seems to be a PLP-binding
residue (Fig. 3).
acid sequence. This discrepancy was also observed in
the case of L-THA DH from Pseudomonas sp. T62.19)
The first 25 N-terminal amino acid residues sequenced
using Edman degradation in recombinant His-tagged
enzyme perfectly matched the amino acid sequence
deduced from the nucleotide sequence of pET15b_
dehadh. Moreover, the purified recombinant enzyme
showed high activity, with specific activity of approxi-
mately 20 units (mg protein)-1 toward D-EHA. From
these results, we concluded that this protein is recombi-
nant D-EHA DH.
Characterization of the recombinant enzyme
We purified recombinant His-tagged D-EHA DH
from E. coli BL21 (DE3) cells and characterized its
enzymatic properties. The molecular weight determined
using SDS-PAGE (Fig. S1, 41,800) was not in agree-
ment with that calculated from the deduced amino acid
sequence of the recombinant enzyme including His-tag
(36,900). However, the molecular weight determined
using MALDI-TOF-MS analysis (36,847) showed good
agreement with that calculated from the deduced amino
Substrate specificity of the recombinant enzyme
Next, we attempted to determine the substrate
specificity of D-EHA DH. Table 2 shows the substrate
specificity and kinetic parameters of D-EHA DH. In
addition to D-EHA, L-THA also acts as a good substrate
of the enzyme (around 190% kcat/Km value compared
to that of D-EHA). Other 3-hydroxyaspartate isomers
Fig. 3. Multiple alignment of the amino acid sequences of D-EHA DH from Pseudomonas sp. N99 (D-EHA DH N99), L-THA DH from
Pseudomonas sp. T62 (L-THA DH T62), L-THA DH from S. cerevisiae (L-THA DH Yeast), and serine racemase from Schizosaccharomyces
pombe (SR Pombe).
Notes: The alignment was drawn using Clustal W and BOXSHADE.22) The numbers on the left side indicate the residues of each amino acid
sequence. White letters in the black box indicate the identical residues and white letters in the gray box indicate similar residues. The asterisk
indicates a possible PLP-binding residue.

6
H. Nagano et al.
Table 2. Kinetic parameters of D-EHA DH.
(A)
100
80
60
40
20
0
Substrate
kcat (s−1
)
Km (mM)
kcat/Km
D-THA
L-THA
D-EHA
L-EHA
L-serine
N.D.
8.61
2.19
N.D.
0.84
–
–
15.2
8.01
–
0.57
0.27
–
144
0.00584
Note: N.D., below the detection limit, i.e. <0.01 μmol/min/(mg protein).
0
2
4
6
8
10
12
14
-20
Reaction time (h)
such as D-THA, and L-EHA could not be used as sub-
strates for the enzyme. In addition, the enzyme showed
very low (around 0.073% kcat/Km value compared to
that of D-EHA) dehydratase activity toward L-serine,
but no activity toward D-serine, L-threonine, and D-thre-
onine was observed. No serine racemase activity was
detected using the sensitive fluorescence detection
HPLC, as in the case of L-THA DH.19) Normal hyper-
bolic kinetics were observed with D-EHA, L-THA, and
L-serine. The Km, kcat, and kcat/Km values calculated
from the Lineweaver–Burk plots are shown in Table 2.
100
80
60
40
20
0
(B)
0
2
4
6
8
10
12
14
Absorption spectrum and effect of inhibitors
The recombinant enzyme exhibited absorption max-
ima at 280 and 412 nm (Fig. S2). The solutions of the
pure enzyme were distinctly yellow. In addition, both
the inhibitors tested almost completely inhibited the
enzyme activity at 5 mM (data not shown). These
results suggested that the enzyme contains PLP as its
prosthetic group.
Reaction time (h)
Fig. 4 Optical resolution of DL-EHA and DL-THA.
Notes: (A) Biocatalytic preparation of L-EHA from DL-EHA using
purified D-EHA DH in a reaction mixture containing 100 mM Tris-
HCl (pH 8.5), 0.01 mM PLP, 0.1 mM MnCl2, 200 mM DL-EHA, and
20.8 unit (for D-EHA) of the enzyme (6.05 mg) in a total volume of
100 mL. During incubation, the D-form and L-form of 3-hydroxyas-
partate in the reaction mixture were monitored using HPLC.
The enantiomeric excess (e.e.) was calculated from the peak area of
the stereoisomers. (B) Biocatalytic preparation of D-THA from the
DL-THA using purified D-EHA DH. The methods were the same as
mentioned above, except for an enzyme amount (20.8 unit for
L-THA, 1.72 mg). The curves shown in the figure are representative
value of several experiments, which gave similar results.
Properties of the recombinant enzyme
The optimal pH and temperature of the recombinant
enzyme were 8.5 and 45 °C, respectively, under the
standard assay conditions. The thermal stability of the
recombinant enzyme was also determined. After 15
min of incubation at various temperatures, the residual
enzyme activity toward D-EHA was measured. The
enzyme was stable below 25 °C and lost its activity
above 45 °C.
and optically pure L-EHA was obtained from DL-EHA
with 35.1% yield in the isolated form.
The optical resolution of DL-THA to obtain optically
pure D-THA was also conducted in the same manner.
Optically pure D-THA (>99% e.e.) was obtained with
37.4% yield in the isolated form (Figs. 4(B), S6).
NMR analysis also showed that the chemical purity of
the isolated D-THA was >99% (Figs. S7 and S8). DL-
THA standard (purchased from Tokyo Chemical Indus-
L-EHA and D-THA production by enzymatic optical
resolution
The enzyme showed remarkable activity toward
D-EHA, while no dehydratase activity was observed
toward L-EHA (Table 2). Thus, we conducted the
optical resolution of DL-EHA to obtain optically pure
L-EHA using the recombinant enzyme produced by the
E. coli cells. After 14 h of reaction, no D-EHA was
detected on the HPLC, and L-EHA remained with an
optical purity of >99% e.e. (Figs. 4(A), S3).
1
try): H-NMR (500 MHz D2O): δ = 4.06 (d1H J = 2.0
Hz); 4.56 (d1H J = 1.5 Hz); 13C-NMR (125 MHz
D2O): 60.68; 74.08; 176.46; 178.91. The resolution
product: 1H-NMR (500 MHz D2O): δ = 3.82 (d1H
J = 2.0 Hz); 4.46 (d1H J = 2.0 Hz); 13C-NMR (125
MHz D2O): 60.76; 74.32; 174.89; 179.04.
As shown in Figs. S3 and S6 (HPLC chro-
matograms), e.e. values of the substrate racemic mix-
ture (DL-EHA and DL-THA) were not zero. The reason
is unclear, however, e.e. value of laboratory-synthesized
DL-EHA was around –18% for L-form, while that of
commercially available DL-THA (Tokyo Chemical
Industries) was around 3% for D-form. The reaction
conditions of ring-opening ammonolysis may affect
optical purity of the product. As shown in Fig. 4(A),
The chemical purity of the isolated product was then
verified by NMR analysis (Figs. S4 and S5). L-EHA
standard (purchased Wako Pure Chemical Industries):
1H-NMR (500 MHz D2O): δ = 4.03 (d1H J = 3.5Hz);
4.34 (d1H J = 3.5Hz); 13C-NMR (125 MHz D2O):
59.81; 73.58; 174.90; 179.04. The resolution product:
1H-NMR (500 MHz D2O): δ = 3.99 (d1H J = 3.5Hz);
4.32 (d1H J = 3.5Hz); 13C-NMR (125 MHz D2O):
60.55; 75.39; 179.48; 180.68. Finally,>99% chemically

Novel D-erythro-3-hydroxyaspartate dehydratase
7
all the D-EHA in the reaction mixture were degraded
within 12 h, even though e.e. value of the starting
material were minus for desired L-EHA.
deduced amino acid sequence, which indicated that the
enzyme belongs to the serine/threonine dehydratase
family, showed similarity to L-THA DH (EC 4.3.1.16)
from Pseudomonas sp. T62 (68%)19) and S. cerevisiae
(64%) (Fig. 3).18) In addition, D-EHA DH also showed
significant activity toward L-THA (Table 2). Together,
these results suggested that the enzyme is similar to L-
THA DH from Pseudomonas sp. T62, but with broader
substrate specificity, because the T62 enzyme does not
act on D-EHA.19) The primary structure of D-EHA DH
also showed significant similarity to the structure of
eukaryotic serine racemases (EC 5.1.1.18) from
Schizosaccharomyces pombe (64%),21) mouse (35%),30)
human (34%),31) and Caenorhabditis elegans (28%).32)
Among these serine racemases, the enzymes from
mouse30) and C. elegans32) have been shown to have
activity toward L-THA. These results indicated that the
enzymes having 3-hydroxyaspartate dehydratase activ-
ity are very broadly distributed in nature from bacteria
to mammals. However, like L-THA DH from Pseu-
Discussion
In this study, we isolated an enzyme acting on D-
EHA and purified it to almost homogeneity from the
newly isolated soil bacterium, Pseudomonas sp. N99.
Although 3-hydroxyaspartate dehydratases acting on L-
THA or D-THA have been identified and characterized
previously by our research group,13,18,19,24) an enzyme
acting on D-EHA has not yet been reported. To the best
of our knowledge, the present study is the first to report
an enzyme that catalyzes the deamination of D-EHA.
Therefore, we designated this enzyme D-erythro-3-hy-
droxyaspartate dehydratase (D-EHA DH), although it
also showed considerable activity toward L-THA.
The reason why substrate specificity of D-EHA DH
in this study is different from L-THA DH of Pseu-
domonas sp. T6219) in spite of the high primary struc-
ture identity (72%) is still unknown. However,
according to the 3D structure of D-THA DH,12) which
catalyzes the very similar reaction (only the stereoselec-
tivity is different) to this enzyme, very small change
around the substrate-binding pocket may affect stereos-
electivity of the enzyme reaction. The local environ-
ment around the substrate-binding pocket of these two
enzymes, i.e. D-EHA DH and L-THA DH, may be dif-
ferent from each other. The detailed substrate recogni-
tion mechanism, however, should be investigated in
further study such as 3D structure analysis.
Based on this substrate specificity, the enzyme was
found to be clearly distinct from the erythro-3-hydrox-
yaspartate dehydratase (EC 4.3.1.20) reported by Gibbs
and Morris many years ago,25) which acts only on L-
EHA but not on D-EHA. The D-EHA DH isolated and
purified in this study also showed weak activity toward
L-serine in addition to D-EHA and L-THA. The L-serine
dehydratase activity of this enzyme, however, was less
than 1% of the D-EHA DH activity (Table 2). The pro-
duction of D-EHA DH by Pseudomonas sp. N99 cells
was not induced by L-serine, but by the DL-EHA or DL-
THA in the culture medium. These results strongly
suggested that the physiological substrates of this
enzyme are 3-hydroxyaspartate isomers such as D-EHA
and L-THA, but not L-serine.
Although the physiological function of D-EHA DH in
Pseudomonas sp. N99 remains unknown, one explanation
is possible. 3-Hydroxyaspartate is a rare but naturally
occurring amino acid. It has been found in human cere-
brospinal fluid1), mammalian urine hydrolysate,26) and
some peptide antibiotics such as cinnamycin,27) and cor-
mycin A.28) Most 3-hydroxyaspartate found in peptides
are L-threo form. Thus, natural habitat of Pseudomonas
sp. N99, i.e. soil, may contain free L-THA, which is toxic
to many bacteria.29) Thus, D-EHA DH may play a role in
detoxification of free 3-hydroxyaspartate in Pseudomonas
sp. N99 cells. The facts that D-EHA DH is inducible in
Pseudomonas sp. N99 and D-EHA DH acts both on
D-EHA and L-THA also support this hypothesis.
domonas sp. T6219) and S. cerevisiae18)
, D-EHA DH
did not show any detectable serine racemase activity.
The absorption spectrum of the purified recombinant
enzyme and the inhibitor experiment revealed that
D-EHA DH contains PLP, as do other bacterial
dehydratases (Fig. S2).33) Although it has not yet been
identified, Lys55 is most probably the PLP-binding
residue based on the sequence similarity to other
related enzymes (Fig. 3).
There have previously been a few reports on the
enzymatic production of L-THA. Mutated asparagine
oxygenase (AsnO-D241N) from Streptomyces coeli-
color A3(2) has been used for the direct hydroxylation
of L -aspartate to produce L-THA.14) This route of
enzymatic synthesis is very concise and economical,
but the mutant enzyme, AsnO-D241N, seems to be
unstable, and the reaction can only be performed at
16°C. Hara et al. also reported whole cell catalysis
from asparagine to L-THA using asparaginase-deficient-
and wild-type AsnO-expressing E. coli cells.15) How-
ever, the enzymatic production of L-EHA has not yet
been reported. In this study, we successfully achieved
the enzymatic resolution of DL-EHA and DL-THA to
obtain optically pure L-EHA and D-THA, respectively,
using the recombinant D-EHA DH produced by E. coli
cells. This synthetic route is one of the simplest meth-
ods that can be used to obtain optically pure L-EHA,
because no hydroxylase yielding L-EHA as a reaction
product has been reported so far. To the best of our
knowledge, this is the first example of the enzymatic
production or resolution of L-EHA. In addition, the
concentration of L-EHA in the reaction mixture
achieved in this study (100 mM) was much higher than
previously reported values achieved using enzymatic
L-THA production (about 10 mM using AsnO-D241N
system,14) and about 48 mM using wild-type AsnO-ex-
pressing E. coli system15)). We also analyzed the final
products using TLC, HPLC, and NMR spectroscopy
and confirmed them to be chemically and optically
pure L-EHA/D-THA. Pure compounds could be isolated
by direct recrystallization from the reaction mixtures
by taking advantage of the lower solubility of
3-hydroxyaspartate in acidic water (pH 2.0) compared
to that of the other catalytic products of the enzyme.12)
We also successfully isolated the gene encoding
D-EHA DH using the inverse PCR approach. The

8
H. Nagano et al.
Therefore, this enzymatic approach has an advantage
for obtaining the pure products, as it does not require
any chromatographic purification. It is also efficient
and friendly to the environment because of minimal or
no use of organic solvents.
References
[1] Perry TL, Jones RT. The amino acid content of human cere-
brospinal fluid in normal individuals and in mental defectives. J
Clin Invest. 1961;40:1363–1372.
[2] Jenkins WT. Glutamic-aspartic transaminase VI. The reaction
with certain β-substituted aspartic acid analogues. J Biol Chem.
1961;236:1121–1125.
[3] Gibbs RG, Morris JG. Assay and properties of β-hydroxyaspar-
tate aldolase from Micrococcus denitrificans. Biochim Biophys
Acta. 1964;85:501–503.
[4] Kornberg HL, Morris JG. The utilization of glycollate by Micro-
coccus denitrificans: the β-hydroxyaspartate pathway. Biochem J.
1965;95:577–586.
[5] Grimwood S, Foster AC, Kemp JA. The pharmacological speci-
ficity of N-methyl-D-aspartate receptors in rat cerebral cortex:
correspondence between radioligand binding and electrophysio-
logical measurements. Br J Pharmacol. 1991;103:1385–1392.
[6] Shimamoto K. Glutamate transporter blockers for elucidation of
the function of excitatory neurotransmission systems. Chem Rec.
2008;8:182–199.
[7] Střísovšký K, Jirásková J, Mikulová A, et al. 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. 2005;44:13091–13100.
[8] Saito S, Komada K, Moriwake T. Diethyl (2S,3R)-2-(N-tert-butoxy-
carbonyl)amino-3-hydroxysuccinate. Org Synth. 1996;73:184–191.
[9] Cardillo G, Gentilucci L, Tolomelli A, et al. A practical method
for the synthesis of β-amino α-hydroxy acids. synthesis of enan-
tiomerically pure hydroxyaspartic acid and isoserine. Synlett.
1999;11:1727–1730.
In conclusion, in this study, we have discovered
D-EHA DH for the first time, and established an
enzymatic resolution process for obtaining optically
pure L-EHA, which is an important compound for
neurophysiological research. This process is also one of
the simplest methods that can be used to obtain pure
L-EHA, though the maximum yield of this process was
theoretically 50% as a DL-racemic mixture of hydrox-
yaspartate used as substrates. Using this enzyme
together with the already reported D-THA DH from
Delftia sp. HT23 acting on D-THA and L-EHA, four
isomers of 3-hydroxyaspartate can now be prepared
from the inexpensive racemic mixture of DL-THA and
DL-EHA. L-THA and D-EHA can be prepared from
DL-THA and DL-EHA, respectively, using D-THA DH
reported previously,12) and D-THA and L-EHA can be
prepared from DL-THA and DL-EHA, respectively, using
the D-EHA DH reported here. These enzymatic resolu-
tions may provide efficient and inexpensive ways for
obtaining optically pure 3-hydroxyaspartate isomers
and thus contribute to the functional analysis of 3-hy-
droxyaspartate isomers in the future.
[10] Hanessian S, Vanasse B. Novel access to (3R)- and (3S)-3-
hydroxy-L-aspartic acids, (4S)-4-hydroxy-L-glutamic acid, and
related amino acids. Can J Chem. 1993;71:1401–1406.
[11] Kaneko T, Katsura H. The synthesis of four optical isomers of
β-hydroxyaspartic acid. Bull Chem Soc Jpn. 1963;36:899–903.
[12] Matsumoto Y, Yasutake Y, Takeda Y, et al. Structural insights
into the substrate stereospecificity of D-threo-3-hydroxyaspartate
dehydratase from Delftia sp. HT23: a useful enzyme for the syn-
thesis of optically pure L-threo- and D-erythro-3-hydroxyaspar-
tate. Appl Microbiol Biotechnol. 2015;99:7137–7150.
Author contribution
HN conducted the screening, enzyme purification,
and gene cloning. KS characterized the recombinant
enzyme and carried out the enzymatic resolution. YM
prepared the substrates and carried out the enzymatic
resolution. YA and MW conceived and designed the
study and wrote the article.
[13] Maeda T, Takeda Y, Murakami T, et al. Purification,
characterization and amino acid sequence of a novel enzyme,
D-threo-3-hydroxyaspartate dehydratase, from Delftia sp. HT23.
J Biochem. 2010;148:705–712.
Acknowledgments
[14] Strieker M, Essen LO, Walsh CT, et al. Non-heme hydroxylase
engineering for simple enzymatic synthesis of L-threo-hydrox-
yaspartic acid. ChemBioChem. 2008;9:374–376.
We are grateful to Dr. Eri Fukushi and Mr. Yusuke
Takata at the GC-MS &NMR Laboratory, Faculty of
Agriculture, Hokkaido University for the NMR and
MALDI-TOF-MS analysis. We also thank the Global
Facility Center, Hokkaido University for the Edman
degradation analysis.
[15] Hara R, Nakano M, Kino K. One-pot production of L-threo-3-
hydroxyaspartic acid using asparaginase-deficient Escherichia
coli expressing asparagine hydroxylase of Streptomyces coeli-
color A3(2). Appl Environ Microbiol. 2015;81:3648–3654.
[16] Balcar VJ, Johnston GAR, Twitchin B. Stereospecificity of the
inhibition of L-glutamate and L-aspartate high affinity uptake in
rat brain slices by threo-3-hydroxyaspartate. J Neurochem. 1977;
28:1145–1146.
[17] Foster AC, Li YX, Runyan S, et al. Activity of the enantiomers
of erythro-3-hydroxyaspartate at glutamate transporters and
NMDA receptors. J Neurochem. 2016;136:692–697.
Disclosure statement
No potential conflict of interest was reported by the authors.
[18] Wada M, Nakamori S, Takagi H. Serine racemase homologue of
Saccharomyces cerevisiae has L-threo-3-hydroxyaspartate dehy-
dratase activity. FEMS Microbiol Lett. 2003;225:189–193.
[19] Murakami T, Maeda T, Yokota A, et al. Gene cloning and
expression of pyridoxal 5′-phosphate-dependent L-threo-3-hy-
droxyaspartate dehydratase from Pseudomonas sp. T62, and
characterization of the recombinant enzyme. J Biochem. 2009;
145:661–668.
Funding
This work was supported in part by a Grant-in-Aid from Institute
for Fermentation, Osaka (IFO), awarded to MW.
[20] Pearson WR, Lipman DJ. Improved tools for biological
sequence comparison. Proc Nat Acad Sci USA. 1988;85:
2444–2448.
Supplemental material
[21] Goto M, Yamauchi T, Kamiya N, et al. Crystal structure of a
homolog of mammalian serine racemase from Schizosaccha-
romyces pombe. J Biol Chem. 2009;284:25944–25952.
The supplemental material for this paper is available at
http://dx.doi.org/10.1080/09168451.2017.1295804.

Novel D-erythro-3-hydroxyaspartate dehydratase
9
[22] Larkin MA, Blackshields G, Brown NP, et al. Clustal W and
Clustal X version 2.0. Bioinformatics. 2007;23:2947–2948.
[23] Sigrist CJA, Cerutti L, de Castro E, et al. PROSITE, a protein
domain database for functional characterization and annotation.
Nucleic Acids Res. 2010;38:D161–D166.
from Pseudomonas corrugata: cormycin A. Biochem J. 2004;
384:25–36.
[29] Ishiyama T, Furuta T, Takai M, et al. L-threo-β-hydroxyaspartic
acid as an antibiotic amino acid. J Antibiot. 1975;28:821–823.
[30] Wolosker H, Blackshaw S, Snyder SH. Serine racemase: a glial
enzyme synthesizing D-serine to regulate glutamate-N-methyl-D-
aspartate neurotransmission. Proc Nat Acad Sci USA. 1999;96:
13409–13414.
[31] De Miranda J, Santoro A, Engelender S, et al. Human serine
racemase: moleular cloning, genomic organization and functional
analysis. Gene. 2000;256:183–188.
[32] Katane M, Saitoh Y, Uchiyama K, et al. Characterization of a
homologue of mammalian serine racemase from Caenorhabditis
elegans: the enzyme is not critical for the metabolism of serine
in vivo. Genes Cells. 2016;21:966–977.
[33] Eliot AC, Kirsch JF. Pyridoxal phosphate enzymes: mechanistic,
structural, and evolutionary considerations. Annu Rev Biochem.
2004;73:383–415.
[24] Wada M, Matsumoto T, Nakamori S, et al. Purification and char-
acterization of
a novel enzyme, L-threo-3-hydroxyaspartate
dehydratase, from Pseudomonas sp. T62. FEMS Microbiol Lett.
1999;179:147–151.
[25] Gibbs RG, Morris JG. Purification and properties of erythro-β-
hydroxyaspartate dehydratase from Micrococcus denitrificans.
Biochem J. 1965;97:547–554.
[26] Ikegami T. Studies on the metabolism of β-hydroxyaspartic acid.
Acta Med Okayama. 1975;29:241–247.
[27] Kaletta C, Entian KD, Jung G. Prepeptide sequence of
cinnamycin (Ro 09-0198): the first structural gene of a duramy-
cin-type lantibiotic. Eur J Biochem. 1991;199:411–415.
[28] Scaloni A, Dalla Serra M, Amodeo P, et al. Structure,
conformation and biological activity of a novel lipodepsipeptide