Characterization of a GDP-D-mannose 3″,5″-epimerase from rice

Article abstract of DOI:10.1016/j.phytochem.2005.12.003

The enzymatic characterization of GDP-d-mannose 3″,5″-epimerase (GME), a key enzyme in the biosynthesis of vitamin C in plants is described. The GME gene (Genbank Accession No. AB193582) in rice was cloned, and expressed as a fusion protein in Escherichia coli. Reaction products from GDP-d-mannose, as produced by GME catalysis, were separated by recycling HPLC on an ODS column, and were determined to be GDP-l-galactose and GDP-l-gulose, based on their NMR spectra and sugar analysis. The reaction catalyzed by GME was inhibited by GDP, and was strongly accelerated by NAD⁺ in contrast to the case of GME from Arabidopsis thaliana. This difference in the effect of NAD⁺ on GME activity can be attributed to the NAD binding domain which is conserved in the rice gene, but not in the Arabidopsis thaliana gene. The apparent K _m and k_cat were determined to be 1.20 × 10 ^-5 M and 0.127 s^-1, respectively, in the presence of 20 μM NAD⁺. The fractions of GDP-d-mannose, GDP-l-galactose and GDP-l-gulose, at equilibrium, were approximately 0.75, 0.20 and 0.05, respectively.

Full text of DOI:10.1016/j.phytochem.2005.12.003

PHYTOCHEMISTRY
Phytochemistry 67 (2006) 338–346
www.elsevier.com/locate/phytochem
Characterization of a GDP-D-mannose 300,500-epimerase from rice
a
a
a,b,
*
Kentaroh Watanabe , Kiyoshi Suzuki , Shinichi Kitamura
a
Graduate School of Agriculture and Biological Sciences, Osaka Prefecture University, Gakuen cho 1-1, Sakai, Osaka 599-8531, Japan
Graduate School of Life and Environmental Sciences, Osaka Prefecture University, Gakuen cho 1-1, Sakai, Osaka 599-8531, Japan
b
Received 7 September 2005; received in revised form 24 October 2005
Abstract
The enzymatic characterization of GDP-D-mannose 300,500-epimerase (GME), a key enzyme in the biosynthesis of vitamin C in plants is
described. The GME gene (Genbank Accession No. AB193582) in rice was cloned, and expressed as a fusion protein in Escherichia coli.
Reaction products from GDP-D-mannose, as produced by GME catalysis, were separated by recycling HPLC on an ODS column, and
were determined to be GDP-L-galactose and GDP-L-gulose, based on their NMR spectra and sugar analysis. The reaction catalyzed by
GME was inhibited by GDP, and was strongly accelerated by NAD+ in contrast to the case of GME from Arabidopsis thaliana. This
difference in the effect of NAD+ on GME activity can be attributed to the NAD binding domain which is conserved in the rice gene, but
not in the Arabidopsis thaliana gene. The apparent Km and kcat were determined to be 1.20 · 10ꢀ5 M and 0.127 sꢀ1, respectively, in the
presence of 20 lM NAD+. The fractions of GDP-D-mannose, GDP-L-galactose and GDP-L-gulose, at equilibrium, were approximately
0.75, 0.20 and 0.05, respectively.
ꢀ 2005 Elsevier Ltd. All rights reserved.
Keywords: Oryza sativa; Gramineae; Plant AsA biosynthetic pathway; GDP-D-mannose 300,500-epimerase; GDP-D-mannose; GDP-L-galactose;
GDP-L-gulose
1. Introduction
cell growth and biosynthesis of hormones and cell wall con-
stituents (Davey et al., 2000; Conklin and Barth, 2004). As
shown in Fig. 1, Smirnoff and Wheeler proposed the AsA
biosynthetic pathway in plants, known as the Smirnoff
and Wheeler pathway, in which AsA is synthesized from
D-mannose via GDP-D-mannose (1), GDP-L-galactose (3),
L-galactose and L-galactono-1,4-lactone (Wheeler et al.,
1998; Smirnoff and Wheeler, 2000; Smirnoff et al., 2001,
2004). GDP-D-mannose 300,500-epimerase (GME), a key
enzyme in this pathway, catalyzes the synthesis of GDP-
L-galactose (3) and GDP-L-gulose (2) from GDP-D-man-
nose (1). GME was first isolated and characterized from
a type of green algae, Chlorella pyrenoidosa (Barber,
1975, 1979; Hebda et al., 1979; Hebda and Barber, 1982).
In higher plants, GME activity has been reported in both
pea and Arabidopsis (Wheeler et al., 1998), and was
recently cloned and characterized in Arabidopsis (Wolucka
et al., 2001b; Wolucka and Van, 2003). The reaction
catalyzed by GME is possibly the rate limiting step in the
GDP-L-galactose (3) is a nucleotide-sugar (Fig. 1) that
serves as a biosynthetic precursor of L-galactosyl residues
in polysaccharides and oligosaccharides in plants (Seifert,
2004; Reuhs et al., 2004). The L-galactosyl residue which
is contained in cell wall polysaccharides, glycolipids
and glycoproteins in higher plants is a minor component
and little information is available concerning its
biosynthesis.
It is also noteworthy that GDP-L-galactose (3) is also
required for the synthesis of L-ascorbic acid (AsA; vitamin
C) in plants. It is well-known that AsA functions as an
antioxidant and an enzymatic cofactor. AsA also plays
important roles in many different processes, including pho-
tosynthesis, photo-protection, stress resistance, control of
*
Corresponding author. Tel.: +81 72 254 9457; fax: +81 72 254 9458.
E-mail address: skita@bioinfo.osakafu-u.ac.jp (S. Kitamura).
0031-9422/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2005.12.003

K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
339
Fig. 1. Schematic diagram of a possible pathway for the biosynthesis of GDP-L-galactose and AsA (L-ascorbic acid). Enzymes catalyzing the numbered
reactions are: (1) mannose-6-phosphate isomerase (EC 5.3.1.8); (2) phosphomannomutase (EC 5.4.2.8); (3) GDP-D-mannose pyrophosphorylase (EC
2.7.7.22); (4) GDP-L-galactose pyrophosphatase (Barber, 1971); (5) L-galactose-L-phospate phosphatase (Laing et al., 2004); (6) L-galactose dehydrogenase
(Gatzek et al., 2002); (7) L-galactono-1,4-lactone dehydrogenase (EC 1.3.2.3); (8) GDP-L-gulose pyrophosphatase; (9) L-gulose-1-phosphate phosphatase;
(10) L-gulose dehydrogenase; (11) L-gulono-1,4-lactone oxidase/dehydrogenase; (12) myo-inositol oxygenase (Lorence et al., 2004); (13) D-glucuronic acid
reductase (EC 1.1.1.19); (14) aldono lactonase; (15) glycosyl transferase. When EC numbers are not shown in parenthesis, they have not yet been identified.
(Baig et al., 1970; Conklin, 2001; Reiter and Vanzin, 2001; Valpuesta and Botella, 2004).
AsA biosynthetic pathway in plants. Experiments with an
Arabidopsis cell suspension culture indicate that the rate
limiting step in AsA biosynthesis occurs between D-man-
nose and L-galactose (Davey et al., 1999). Moreover, a
strong correlation between GME reaction and accumula-
tion of AsA has been reported in the colorless microalga,
Prototheca moriformis (Running et al., 2003).
brane domain nor a signal peptide sequence, from the pre-
diction using the SOSUI program (Mitaku et al., 2002) and
SignalP version 3.0 (Bendtsen et al., 2004). An analysis by
means of iPSORT (Bannai et al., 2002) predicted that the
product of OsGME was localized in the cytoplasm. The
calculated molecular mass and isoelectric point for
the enzyme was 42,778 Da and 5.75, respectively.
In this study, we report on the cloning, expression and
characterization of GME from rice (Oryza sativa L.). There
appears to be a large difference in the activity of this
enzyme between rice and Arabidopsis, which may involve
a difference in amino acid sequence. We also attempted
to establish a method for the efficient separation and recov-
ery of rare nucleotide-sugars, GDP-L-galactose (3) and
GDP-L-gulose (2), which represent key compounds in the
plant AsA biosynthesis pathway (the Smirnoff and Wheeler
pathway).
Using the clustal W program (Thompson et al., 1994),
the amino acid sequence deduced from OsGME indicated
a 91% identity with that from Arabidopsis thaliana
(Fig. 2A), and the sequence of OsGME shared high simi-
larities with putative GME from Zea mays (97%), Sorghum
bicolor (97%), Hordeum vulgare (93%), Triticum aestivum
(92%), Lycopersicon esculentum (92% and 91%), Medicago
truncatula (91%), Lotus corniculatus var. japonicus (91%),
Solanum tuberosum (91%), Mesembryanthemum crystalli-
num (90%), Glycine max (90%), and Nicotiana tabacum
(89%) at the amino acid level. Fig. 2B shows a phylogenetic
tree constructed from the multiple amino acid sequence
alignment obtained for these plants. Such a high similarity
between monocots and dicots is very unique and a charac-
teristic of the GME enzyme. In addition, the GME gene is a
single gene, as no homologous gene has been found in the
same species except for Lycopersicon esculentum.
2. Results and discussion
2.1. Characterization of the OsGME cDNA clone
The cloned cDNA, OsGME (Oryza sativa GDP-D-man-
nose 300,500-epimerase), appeared to encode a polypeptide of
378 amino acids. The peptide sequence, as deduced from
OsGME cDNA is assumed to contain neither a transmem-
The OsGME enzyme contains consensus motifs for the
epimerase/dehydratase family, GxxGxxG and YxxxK
(Fig. 2A and C). The motifs, GxxGxxG and YxxxK have

340
K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
Fig. 2. Amino acid sequence alignment and phylogenetic analysis of GME. (A) Amino acid sequence alignments of GME proteins were aligned with
Oryza sativa AB193582 and Arabidopsis thaliana AX647945 by the pairwise method using the clustalW program (Thompson et al., 1994). Gaps (-) were
introduced to achieve maximum similarity. Identical amino acid residues among the sequences are highlighted by an asterisk (*) under the residue. The
putative nicotinamide-adenine dinucleotide binding motif, GxxGxxG (Bellamacina, 1996), and a putative catalytic domain, Ser and YxxxK motif
(Jornvall et al., 1995) are highlighted in each box. (B) A Neighbor-Joining tree derived from the amino acid sequences of GME from various plants. The
phylogenetic analysis was performed using the TreeView program. The amino acid sequences shown in these diagrams are listed in the DDBJ database
under the following GenBank Accession Nos. Oryza, AB193582; Zea mays, AX647973; Sorghum, AX647983; Hordeum, AX647987; Triticum, AX647985;
Arabidopsis, AX647945; Mesembryanthemum, AX647975; Medicago, AX647981; Lotus, AX647989; Glycine, AX647977; Solanum, AX647979;
Lycopersicon 1, AX647971; Lycopersicon 2, AX647997; Nicotiana, CQ809292. The bar indicates substitutions per site. Plant species which have the
GxxGxxG motif conserved (see Fig. 1C) are highlighted in dashed box. (C) Two motif sequence alignments of GME protein among higher plants was
highlighted. Oryza (rice), Hordeum (barley), Triticum (wheat), Zea mays (maize), and Sorghum (millet) are monocots. Mesembryanthemum (iceplant),
Solanum (potato), Nicotiana (tobacco), Lycopersicon (tomato), Medicago (alfalfa), Lotus, Glycine (soybean), and Arabidopsis are dicots.
been reported to be the NAD-binding domain (Bellama-
cina, 1996) and a catalytic domain having short-chain
dehydrogenase activity forms a catalytic triad with Tyr,
Lys and Ser upstream (Jornvall et al., 1995). These two
motifs may contribute to the first step in the dehydratase
and epimerase enzyme mechanisms which occurs via
abstraction of the 4-hydroxyl proton and hydride transfer
from the C4 position of the sugar to the NAD+ bound at
the GxxGxxG motif, i.e. to yield the nucleotide-4-keto
sugar intermediate (Liu et al., 1997). In the GME reaction
mechanism proposed by Barber (1979), using tritium-
labeled water and GME from Chlorella pyrenoidosa, the
mannosyl residue of GDP-D-mannose was also converted
to a 4-keto sugar intermediate, this suggests that these
motifs function as for other epimerase enzymes as well. It
should also be noted that the GxxGxxG motif in Arabidop-
sis is not conserved. In addition, the GxxGxxG motif in
GME from dicot plants is replaced to GxxGxxA, except
for Mesembryanthemum crystallinum (Fig. 2C).
2.2. Identification of products of OsGME
The OsGME expressed in Escherichia coli was success-
fully purified to apparent homogeneity as evidenced by
SDS–PAGE (Fig. 3A). Two products, GDP-L-galactose
(3) and GDP-L-gulose (2), generated from GDP-D-man-
nose (1) by the action of the purified OsGME were also
isolated and characterized. The maximum yield of these
products from the recombinant OsGME was 20–25%
(Fig. 3B). As shown in Fig. 4, these products were
separated by reversed-phase recycle HPLC. The GDP-
D-mannose (1) was removed completely after three recy-
cles. The reaction products were divided into three frac-
tions, products 1, 2 and 3, respectively. These fractions
were collected after the 10th, 12th and 9th cycles,
respectively.
Products 1–3 were identified by NMR spectroscopic
analysis. Published NMR data for GDP-D-rhamnose
(Kneidinger et al., 2001) and GDP-L-fucose (Albermann

K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
341
et al., 2000) were used in the identification of the prod-
ucts. Table 1 indicates the assignments of the NMR
spectra for products 2 and 3; GDP-L-galactose (3) and
GDP-L-gulose (2). When GDP-L-galactose (3) was ana-
1
lyzed by H and 13C NMR spectroscopy, the assignments
of the proton and carbon signals were obtained
from 1H–1H correlation spectroscopy (COSY; Fig. 5B),
13C–1H heteronuclear multiple quantum coherence
(HMQC; Fig. 5A), 13C–1H heteronuclear multiple-bond
correlation (HMBC; not shown), and Homonuclear
Hartmann–Hahn (HOHAHA; not shown). Concerning
GDP-L-gulose (2), the assignments of its proton reso-
nances were determined by 1H NMR spectroscopy and
1H–1H correlation spectroscopy (COSY; Fig. 5C). How-
1
ever, the H NMR spectra of the product 1 showed only
guanine and ribose, suggesting that the product 1 was an
artifact derived from GDP sugars during the preparation
for HPLC. The proton–proton coupling constants
3JH,H+1 for GDP-L-galactose (3) and GDP-L-gulose (2)
are summarized in Table 1.
Data for sugar analysis by gas–liquid chromatography
is shown in Fig. 6. The retention times of the alditol
acetate derivatives for authentic D-mannose, D-altrose,
L-galactose, and L-gulose were 44.5, 45.5, 48.5, and
52.5 min, respectively. D-Altrose is produced from
D-mannose by epimerization at C-3, which suggests that
GDP-D-altrose is one of the products produced by
OsGME. The nucleotide sugars in the reaction mixture
of OsGME were hydrolyzed and the monosaccharides
produced were converted to alditol acetates, with the lat-
ter separated by GLC. As a result, we confirmed the
presence of D-mannose, L-galactose and L-gulose;
however, no derivative of D-altrose was detected. This
suggests that the epimerization step is during the conversion
of GDP-D-mannose (1) to GDP-L-gulose (2) (C5 epimeriza-
tion) and then GDP-L-galactose (3) (C3 epimerization).
Fig. 3. SDS–PAGE analysis of GME and the HPLC products of GME
reaction. (A) Aliquots of the fractions were loaded onto a 10% SDS–
PAGE gel and subjected to electrophoresis. The gel was stained with
Coomassie brilliant blue R-250. Lane M, size markers; lane 1, MBP-GME
fusion proteins purified by amylose resin; lane 2, GME and MBP proteins
digested by Factor Xa; lane 3, GME protein purified by anion exchange
chromatography. (B) Elution profiles of the products of the GME reaction
with 0.1 mM GDP-D-mannose for 5 min at 37 ꢁC (pH 7.5) detected by
HPLC analysis on the ODS column. Purified protein (1 lM) derived from
E. coli expressing a vector inserted the GME gene and an empty vector are
shown, respectively.
Fig. 4. Recycle HPLC of the products by GME reaction. After partial purification of the GME reaction mixture, the sample was injected into the
recycling HPLC system using a reversed phase column with monitoring and detection by UV absorbance at 254 nm. Six millimolar phosphoric acid and
2 mM potassium chloride were used as the mobile phase (Wolucka et al., 2001a). As the peaks were separated, the purified product was discharged by
operating a valve. As a result, three products were purified and are denoted as products 1, 2 and 3, respectively.

342
K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
Table 1
1H and 13C NMR assignments for GDP-L-galactose (3) and GDP-L-gulose (2)
Sugar residue
1
2
3
4
5
6
b-D-ribose
1H
5.95a
87.8
6.1b
4.77
74.8
5.8
4.54
71.4
3.7
4.37
84.7
2.6
4.23 (2·)
66.4
4.9
13C
3JH,H+1
b-L-galactose
b-L-gulose
1H
4.97
99.5
7.9
3.62
72.2
10
3.69
73.3
3.4
3.93
69.6
3.4
3.75
76.9
3.4
3.75, 3.82
62.2
13C
3JH,H+1
1H
3JH,H+1
5.30
8.1
3.76
3.2
4.11
3.4
3.84
2.7
4.08
4.1
3.74, 3.81
Nucleotide residue
Guanine
2
4
5
6
8
1H
13C
8.13
138.5
155.0
152.7
117.2
160.0
a
Chemical shifts (parts per million) of 1H and 13C NMR signals.
JH,H+1 (Hz) value of 1H NMR signal for nucleotide sugars.
b
2.3. Biochemical characterization of OsGME
limited from 44% to 110%. It is clear that the activity
of the OsGME reaction is stimulated by exogenously
added nicotinamide-adenine dinucleotides, especially the
oxidized forms NAD+ and NADP+. However, it should
be noted that the GME from Arabidopsis (Wolucka and
Van, 2003) is inactivated by the reduced forms, NADH
and NADPH. Moreover, the degrees with which
NAD+ and NADP+ contribute to the increase in activity
differ greatly in rice (805% and 531% of control, respec-
tively) and Arabidopsis (145% and 110% of control,
respectively). These differences in the effect of nicotin-
amide-adenine dinucleotides between the Arabidopsis
GME and OsGME can be attributed to differences
in their affinities for NAD+. As mentioned above, a
NAD-binding motif is conserved in rice, but not in
Arabidopsis.
Equilibrium ratios of GDP-D-mannose (1), GDP-L-
galactose (3) and GDP-L-gulose (2) in the reaction cata-
lyzed by OsGME were determined by monitoring the
amount of these substances produced or consumed in the
reaction (Fig. 7). Using a reversed-phase recycle HPLC,
we were able to separate and collect the enzymatic prod-
ucts, GDP-L-galactose (3) and GDP-L-gulose (2). Thus,
we also performed enzymatic reactions using GDP-L-galac-
tose (3) and GDP-L-gulose (2) as substrates. No differences
in the equilibrium ratios of the reaction catalyzed by
OsGME were observed. The equilibrium ratios of GDP-
D-mannose (1), GDP-L-galactose (3) and GDP-L-gulose
(2) were approximately 75%, 20% and 5%, respectively.
These equilibrium ratios were not changed by the additives
shown in Table 2. Interestingly, only when GDP-L-gulose
(2) was used as a substrate (Fig. 7C), the ratio of GDP-L-
galactose (3) temporarily increased to more than the equi-
librium ratio of GDP-L-galactose (3) before equilibrium.
It should be finally noted that although this enzyme is
considered to be involved in ascorbic acid synthesis via L-
galactose; however, the genetic evidence for this using
transgenic plants will be required to clarify its role in rice.
The enzymatic properties of the purified OsGME were
examined by measuring the amount of products that con-
tained both GDP-L-galactose (3) and GDP-L-gulose (2).
The maximum activity of the enzyme was observed at pH
7.5–8.0 in the absence of NAD+; however, the optimal
pH shifted to 8.0–8.5 when NAD+ was present. Similarly,
the optimal temperature for the OsGME shifted to 25 ꢁC
(addition of NAD+) from 20 ꢁC (no addition of NAD+),
suggesting that NAD+ contributes to the stability of
OsGME.
The effects of the substrate concentration on activity
were examined by measuring enzyme activities with vary-
ing concentrations of GDP-D-mannose. When no NAD+
was added, the apparent Km, kcat, and kcat/Km was
7.12 · 10ꢀ6 M, 3.03 · 10ꢀ2 sꢀ1, and 4.26 · 103 sꢀ1 Mꢀ1
,
respectively. These values were changed to 1.20 · 10ꢀ5 M,
0.127 sꢀ1
,
and 1.06 · 104 sꢀ1 Mꢀ1
,
respectively, when
20 lM NAD+ was added. These kinetics data indicate that
the OsGME enzyme has a high affinity for GDP-D-man-
nose (1) and that the rate of the catalytic reaction from
GDP-D-mannose (1) to GDP-L-gulose (2) and GDP-L-
galactose (3) is low. By the addition of NAD, the affinity
for the substrate decreases slightly and rate of the catalytic
reaction increases, indicating that NAD+ contributes to the
turnover in the OsGME reaction.
The effects of additives on OsGME activity were
examined by measuring enzyme activities with various
additives of nucleotides and nucleotide sugars (Table
2). It is remarkable that NAD(P)+ and NAD(P)H acti-
vated the enzyme reaction by more than twice and that
GDP inhibited it by more than 80%. Inhibition of the
enzyme by GDP was reported in GME from Arabidopsis,
which suggests that the configuration of GDP is signifi-
cant for substrate recognition. In contrast to NAD+,
when almost all other nucleotides or nucleotide pentose
were added, the inhibition of the OsGME reaction was

K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
343
Fig. 6. Sugar analysis by GLC. A GLC of alditol acetates of L-galactose
and L-gulose. The GME reaction mixture was hydrolyzed and the
monosaccharides converted to alditol acetates. The peaks at 15.1 min,
44.1 min, 47.7 min, and 52.4 min correspond to alditol acetates of ribose,
D-mannose, L-galactose, and L-gulose, respectively. Within the framework,
references of alditol acetates of D-mannose, D-altorose, L-galactose, and L-
gulose are shown.
performed, comparing the rice genome sequences with
the GME gene sequence of Arabidopsis thaliana (Genbank
Accession No. AX647945) using the BLAST program
(Altschul et al., 1990). One candidate for the OsGME gene
(Genbank Accession No. AB193582) was found, with a
homology to the Arabidopsis gene sequence of 80%. To
obtain cDNA of the OsGME gene, total RNA was isolated
14 days postanthesis from the japonica rice (Oryza sativa
L.) variety Nipponbare, that had been grown in a green-
house from May to September, 2001 (Suzuki et al.,
2004). First-strand cDNA synthesis was carried out using
5 lg of the total RNA with the OsGME-specific primer
(5-GAGGAGCTGATTGATCTTCATGGTG-3) accord-
ing to the protocol of the SuperScript First-Strand Synthe-
sis System (Invitrogen, Tokyo, Japan). The reaction
mixture was used as a template for PCR in a iCycler (Bio
Rad Laboratories, Hercules, CA). The candidate gene
was amplified with a set of specific primers, S-1 (50-GAT-
CCCTCTCCGACCGACCAAG-3) and A-1 (5-GAG-
GAGCTGATTGATCTTCATGGTG-3) in a volume of
50 ll containing the cDNA template and LA Taq polymer-
ase (Takara-Bio, Shiga, Japan). The PCR program was 35
cycles at 95 ꢁC for 1 min and 62 ꢁC for 2 min, followed by
incubation at 72 ꢁC for 4 min. Five microliters of PCR
product was analyzed by electrophoresis on a 1% agarose
gel with ethidium bromide staining. The nucleotide
sequence was then determined.
Fig. 5. Two-dimensional HMQC and H-H cosy. Panel A, a partial two-
dimensional HMQC spectrum of GDP-L-galactose in D2O shows the
single bond 13C–1H correlations arising from the ribose protons and
the pyranose protons to their respective carbons. Panels B and C show the
connectivities in the ribose and pyranose moieties of GDP-L-galactose and
GDP-L-gulose, respectively. The 1H and 13C spectroscopic assignments are
listed in Table 1.
3. Experimental
3.1. Cloning of the OsGME gene
An annotation of the coding sequences of the Oryza
sativa GDP-D-mannose 3,5-epimerase (OsGME) gene was

344
K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
Table 2
Effects of additives on GME activities
Additives
Reaction rate (%)
Water
NAD
NADH
NADP
NADPH
GMP
100
805[0.11]a
240[14]a
531[0.51]a
297[2.0]a
94
GDP
GTP
19[1.0]b
74
AMP
73
ADP
64
ATP
68
CMP
51
CDP
54
CTP
52
TMP
44
TDP
55
TTP
57
UMP
44
UDP
58
UTP
64
D-mannose
UDP-glucose
UDP-galactose
UDP-xylose
UDP-arabinose
GDP-fucose
a
67
104
110
79
83
73
Concentration (lM) of additive required for 200% activation of the
enzyme reaction.
b
Concentration (lM) of additive required for 50% inhibition of the
enzyme reaction.
3.2. Expression and purification of recombinant enzyme
Fig. 7. Time courses of the GME reaction. Transition of GDP-D-mannose
(1) (d), GDP-L-galactose (3) (h) and GDP-L-gulose (2) (m) in the GME
enzyme reaction. The enzyme reaction was performed in 40 mM Tris–HCl
buffer (pH 8) at 20 ꢁC using 0.1 mM substrate. Panels A–C show moment-
to-moment change using GDP-D-mannose (1) (d), GDP-L-galactose (3)
(h) and GDP-L-gulose (2) (m) as a substrate, respectively.
The coding region of the cloned cDNA of GME was
amplified with the following set of specific primers,
compS-1 (50-AACAACCTCGGGATCGAGGGAAGGA-
TGGGGAGCTCGGAGAAGAAC-30) and compA-1
(50-CGCGGATCCTTACTCCTTGCCATCGGCAGC-30),
which contain restriction sites for BsoBI and BamHI,
respectively. This product was inserted between the BsoBI
and BamHI sites of pMAL-C2X (New England Biolabs,
Beverly, MA). The construct that subcloned OsGME gene
in pMAL-C2X and insert-free pMAL-C2X (negative con-
trol) was transformed to E. coli JM109. The E. coli cells
were grown at 37 ꢁC, and the production of recombinant
protein was induced by treatment with 1 mM isopropyl
b-D-thiogalactopyranoside for 6 h. The E. coli cells were
harvested and sonicated in a buffer containing 50 mM
Tris–HCl buffer (pH 8.0), 50 mM NaCl, 1 mM EDTA,
and 1 mM DTT. The crude soluble extract was placed on
an amylose resin column (New England Biolabs), with
the latter washed with 10 mM Tris–HCl buffer, pH 7.4,
containing 0.8% NaCl and 0.02% KCl; the bound protein
was then eluted with the same buffer containing 10 mM
maltose. The purified recombinant enzyme (1 mg) was
digested with 10 lg of Factor Xa (Haematologic Technol-
ogies, Vermont, USA) at 4 ꢁC for 24 h in order to split
off the fused maltose binding protein. The sample was
applied onto a TOYOPEARL SuperQ-650M (Tosoh,
Tokyo, Japan) column with 10 mM Tris–HCl buffer (pH
7.5), and eluted with a linear gradient of 10–500 mM
Tris–HCl. Active fractions were collected, with the purified
recombinant enzyme examined for purity, specific activity,
and substrate specificity. The concentration of GME
obtained was determined by the Bradford method (Brad-
ford, 1976).
3.3. Enzymatic reaction and NMR spectroscopic analysis of
products
In order to identify products formed by the action of the
recombinant enzyme, a relatively large scale reaction was
employed. A reaction mixture consisting of 40 mM
Tris–HCl buffer, pH 7.5, 40 mg of GDP-D-mannose (Cal-
biochem, Darmstadt, Germany), and 31 mg of the recom-
binant GME in a final volume of 15 ml was incubated at

K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
345
37 ꢁC for 30 min. To terminate the enzyme reaction, the
enzymes were removed using Centriprep filter devices with
a molecular weight cut-off of 10,000 (Millipore, Bedford,
MA), and the synthesized products were confirmed by
HPLC analysis on an ODS column (UK-C18; 20 mm
i.d. · 250 mm; Imtakt, Kyoto, Japan). Concerning the
purification of the synthesized products, they were roughly
fractionated by HPLC and the absorbance of the eluate
was monitored at 254 nm. The roughly purified products
were separated again by a recycle HPLC system using a
UK-C18 column. A solution of 6 mM phosphoric acid
and 2 mM potassium chloride was used for the mobile
phase (Wolucka et al., 2001a) in both the standard HPLC
system and the recycling HPLC system. Each of the three
types of products were absorbed on a SuperQ Toyopearl
650M column in H2O and eluted with 0.3 M NaCl. Finally,
each product was desalted on a Sephadex G-10 gel filtra-
tion chromatography column (Pharmacia, Freiburg, Ger-
many) and lyophilized.
20 ꢁC, and the reaction was then started by the addition
of GDP-D-mannose. After incubation at 20 ꢁC for one to
15 min, the reaction was terminated by dipping the mixture
into a heated block at 100 ꢁC for 1 min. The reaction prod-
ucts were subjected to reversed phase HPLC on an ODS
column (4.6 mm i.d. · 250 mm; Imtakt). The mobile phase
solution was a mixture of H2O–Et3N–AcOH (100:0.2:0.1,
v:v:v) that had been filtered and degassed under reduced
pressure, prior to use. Separation was carried out isocrati-
cally at a flow rate of 0.8 ml/min at 37 ꢁC, and the respec-
tive samples were detected by their absorption at a
wavelength of 254 nm. In the kinetic studies, the trials were
performed in triplicate, at a minimum and the standard
errors were less than 10%. Primary initial velocity data
were fitted to the Michaelis–Menten equation by non-
linear regression (Origin, version 5.0).
References
Lyophilized samples were dissolved in D2O and re-
lyophilized, and then dissolved in D2O. Spectra were
recorded at 25 ꢁC at 400 MHz for 1H and at 100 MHz
for 13C NMR using JNM-AL400 and JNM-A500 spec-
trometers (JEOL, Tokyo, Japan). The analysis of the spec-
tra obtained by measurement of 1H–1H correlated
spectroscopy (COSY), 13C–1H heteronuclear multiple
quantum coherence (HMQC), 13C–1H heteronuclear multi-
ple-bond correlation (HMBC), and Homonuclear Hart-
mann–Hahn (HOHAHA) were performed using the
ALICE software program (JEOL, Tokyo, Japan).
Albermann, C., Distler, J., Piepersberg, W., 2000. Preparative synthesis of
GDP-b-L-fucose by recombinant enzymes from enterobacterial
sources. Glycobiology 10, 875–881.
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990.
Basic local alignment search tool. J. Mol. Biol. 215, 403–410.
Baig, M.M., Kelly, S., Loewus, F., 1970. L-Ascorbic acid biosynthesis in
higher plants from 4-lactone and 4-lactone. Plant Physiol. 46, 277–280.
Bannai, H., Tamada, Y., Maruyama, O., Nakai, K., Miyano, S., 2002.
Extensive feature detection of N-terminal protein sorting signals.
Bioinformatics 18, 298–305.
Barber, G.A., 1971. The synthesis of L-glucose by plant enzyme systems.
Arch. Biochem. Biophys. 147, 619–623.
Barber, G.A., 1975. The synthesis of guanosine 50-diphosphate-l-galactose
by extracts of Chlorella pyrenoidosa. Arch. Biochem. Biophys. 167,
718–722.
3.4. Gas–liquid chromatography for sugar analysis
Barber, G.A., 1979. Observations on the mechanism of the reversible
epimerization of GDP-D-mannose to GDP-L-galactose by an enzyme
from Chlorella pyrenoidosa. J. Biol. Chem. 254, 7600–7603.
Bellamacina, C.R., 1996. The nicotinamide dinucleotide binding motif: a
comparison of nucleotide binding proteins. FASEB J. 10, 1257–1269.
Bendtsen, J.D., Nielsen, H., Heijne, G., Brunak, S., 2004. Improved
prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340, 783–795.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of protein-
dye binding. Anal. Biochem. 72, 248–254.
Conklin, P.L., 2001. Recent advances in the role and biosynthesis of
ascorbic acid in plants. Plant Cell Environ. 24, 383–394.
Conklin, P.L., Barth, C., 2004. Ascorbic acid, a familiar small molecule
intertwined in the response of plants to ozone, pathogens, and the
onset of senescence. Plant Cell Environ. 27, 959–970.
Nucleotide sugars were hydrolyzed by treatment with
trifluoroacetic acid at 100 ꢁC for 6 h, and the monosaccha-
rides were converted to alditol acetates by reduction, fol-
lowed by treatment with acetic anhydride in an equal
volume of pyridine at 120 ꢁC for 2 h. Separation of the
alditol acetates was carried out on a chromatographic
device GC-17A (Shimadzu, Kyoto, Japan) on a DB-225
column (J&W scientific, Folsom, CA). One microliter of
the sample was injected under split conditions at an oven
temperature of 180 ꢁC. After 45 min at 180 ꢁC, the oven
temperature was raised to 200 ꢁC at 2 ꢁC/min, and held
at 200 ꢁC for 15 min. The injector and detector tempera-
tures were 230 ꢁC.
Davey, M.W., Gilot, C., Persiau, G., Ostergaard, J., Han, Y., Bauw, G.C.,
Van, M.M.C., 1999. Ascorbate biosynthesis in Arabidopsis cell
suspension culture. Plant Physiol. 121, 535–543.
3.5. Enzyme assay
Davey, M.W., Van, M.M., Inze, D., Sanmartin, M., Kanellis, A.,
Smirnoff, N., Benzie, I.J.J., Strain, J.J., Favell, D., Fletcher, 2000.
Plant L-ascorbic acid: chemistry, function, metabolism, bioavailability
and effects of processing. J. Sci. Food Agric. 80, 825–860.
Gatzek, S., Wheeler, G.L., Smirnoff, N., 2002. Antisense suppression of L-
galactose dehydrogenase in Arabidopsis thaliana provides evidence for
its role in ascorbate synthesis and reveals light modulated L-galactose
synthesis. Plant J. 30, 541–553.
The activity of the GME enzyme was determined by
monitoring the formation of both GDP-L-galactose (3)
and GDP-L-gulose (2) from GDP-D-mannose (1). In a typ-
ical run, the reaction mixture contained 40 mM Tris–HCl
buffer (pH 8.0), 10 lM GDP-D-mannose, 10 lM NAD+,
and enzyme. In a study of the effect of nucleotide and
nucleotide sugars on GME activity, the GME protein
was separately mixed with each additive for 10 min at
Hebda, P.A., Behrman, E.J., Barber, G.A., 1979. The guanosine 50-
diphosphate D-mannose: guanosine 50-diphosphate L-galactose
epimerase of Chlorella pyrenoidosa. Chemical synthesis of guanosine

346
K. Watanabe et al. / Phytochemistry 67 (2006) 338–346
50-diphosphate L-galactose and further studies of the enzyme and the
reaction it catalyzes. Arch. Biochem. Biophys. 194, 496–502.
Hebda, P.A., Barber, G.A., 1982. GDP-D-mannose: GDP-L-galactose
epimerase from Chlorella pyrenoidosa. Meth. Enzymol. 83, 522–
525.
Seifert, G.J., 2004. Nucleotide sugar interconversions and cell wall
biosynthesis: how to bring the inside to the outside. Curr. Opin. Plant
Biol. 7, 277–284.
Smirnoff, N., Wheeler, G.L., 2000. Ascorbic acid in plants: biosynthesis
and function. Crit. Rev. Biochem. Mol. 35, 291–314.
Jornvall, H., Persson, B., Krook, M., Atrian, S., Gonzalez-Duarte, R.,
Jeffery, J., Ghosh, D., 1995. Short-chain dehydrogenases/reductases
(SDR). Biochemistry 34, 6003–6013.
Smirnoff, N., Conklin, P.L., Loewus, F.A., 2001. Biosynthesis of ascorbic
acid in plants: A renaissance. Annu. Rev. Plant Physiol. Mol. Biol. 52,
437–467.
Kneidinger, B., Graninger, M., Adam, G., Puchberger, M., Kosma, P.,
Zayni, S., Messner, P., 2001. Identification of two GDP-6-deoxy-D-
lyxo-4-hexulose reductases synthesizing GDP-D-rhamnose in Aneu-
rinibacillus thermoaerophilus L420-91T. J. Biol. Chem. 276, 5577–
5583.
Laing, W.A., Bulley, S., Wright, M., Cooney, J., Jensen, D., Barraclough,
D., MacRae, E., 2004. A highly specific L-galactose-1-phosphate
phosphatase on the path to ascorbate biosynthesis. Proc. Natl. Acad.
Sci. USA 101, 16976–16981.
Liu, Y., Thoden, J.B., Kim, J., Berger, E., Gulick, A.M., Ruzicka, F.J.,
Holden, H.M., Frey, P.A., 1997. Mechanistic roles of tyrosine 149 and
serine 124 in UDP-galactose 4-epimerase from Escherichia coli.
Biochemistry 36, 10675–10684.
Lorence, A., Chevone, B.I., Mendes, P., Nessler, C.L., 2004. myo-Inositol
oxygenase offers a possible entry point into plant ascorbate biosyn-
thesis. Plant Physiol. 134, 1200–1205.
Mitaku, S., Hirokawa, T., Tsuji, T., 2002. Amphiphilicity index of polar
amino acids as an aid in the characterization of amino acid preference
at membrane-water interfaces. Bioinformatics 18, 608–616.
Reiter, W.D., Vanzin, G.F., 2001. Molecular genetics of nucleotide sugar
interconversion pathways in plants. Plant Mol. Biol. 47, 95–113.
Reuhs, B.L., Glenn, J., Stephens, S.B., Kim, J.S., Christie, D.B., Glushka,
J.G., Zablackis, E., Albersheim, P., Darvill, A.G., O’Neill, M.A., 2004.
L-Galactose replaces L-fucose in the pectic polysaccharide rhamnoga-
lacturonan II synthesized by the L-fucose-deficient mur1 Arabidopsis
mutant. Planta 219, 147–157.
Smirnoff, N., Running, J.A., Gatzak, S., 2004. Ascorbate biosynthesis: a
diversity of pathways. In: Asard, H., May, J.M., Smirnoff, N. (Eds.),
Vitamin C: Functions and Biochemistry in Animals and Plants. BIOS
Scientific Publishers, London & New York, pp. 7–29.
Suzuki, K., Watanabe, K., Masumura, T., Kitamura, S., 2004. Charac-
terization of soluble and putative membrane-bound UDP-glucuronic
acid decarboxylase (OsUXS) isoforms in rice. Arch. Biochem.
Biophys. 431, 169–177.
Thompson, J.D., Higgins, D.G., Gibson, T.J., 1994. CLUSTAL W:
improving the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties and weight
matrix choice. Nucl. Acids. Res. 22, 4673–4680.
Valpuesta, V., Botella, M.A., 2004. Biosynthesis of L-ascorbic acid in
plants: new pathways for an old antioxidant. Trends Plant Sci. 9, 573–
577.
Wheeler, G.L., Jones, M.A., Smirnoff, N., 1998. The biosynthetic pathway
of vitamin C in higher plants. Nature 393, 365–369.
Wolucka, B.A., Davey, M.W., Boerjan, W., 2001a. A high-performance
liquid chromatography radio method for determination of L-ascorbic
acid and guanosine 50-diphosphate-L-galactose, key metabolites of the
plant vitamin C pathway. Anal. Biochem. 294, 161–168.
Wolucka, B.A., Persiau, G., Van, D.J., Davey, M.W., Demol, H.,
Vandekerckhove, J., Van, M.M., Zabeau, M., Boerjan, W., 2001b.
Partial purification and identification of GDP-mannose 300,500-epimer-
ase of Arabidopsis thaliana, a key enzyme of the plant vitamin C
pathway. Proc. Natl. Acad. Sci. USA 98, 14843–14848.
Running, J.A., Burlingame, R.P., Berry, A., 2003. The pathway of L-
ascorbic acid biosynthesis in the colourless microalga Prototheca
moriformis. J. Exp. Bot. 54, 1841–1849.
Wolucka, B.A., Van, M.M., 2003. GDP-mannose 30,50-epimerase forms
GDP-L-gulose, a putative intermediate for the de novo biosynthesis of
vitamin C in plants. J. Biol. Chem. 278, 47483–47490.