A new family of D-2-hydroxyacid dehydrogenases that comprises D-mandelate dehydrogenases and 2-ketopantoate reductases

Article abstract of DOI:10.1271/bbb.70827

The gene for the D-mandelate dehydrogenase (DManDH) of Enterococcus faecalis IAM10071 was isolated by means of an activity staining procedure and PCR and expressed in Escherichia coli cells. The recombinant enzyme exhibited high catalytic activity toward various 2-ketoacid substrates with bulky hydrophobic side chains, particularly C3-branched substrates such as benzoylformate and 2-ketoisovalerate, and strict coenzyme specificity for NADH and NAD⁺. It showed marked sequence similarity with known NADP-dependent 2-ketopantoate reductases (KPR). These results indicate that together with KPR, D-ManDH constitutes a new family of D-2-hydroxyacid dehydrogenases that act on C3-branched 2-ketoacid substrates with various specificities for coenzymes and substrates.

Full text of DOI:10.1271/bbb.70827

Biosci. Biotechnol. Biochem., 72 (4), 1087–1094, 2008
A New Family of D-2-Hydroxyacid Dehydrogenases
That Comprises D-Mandelate Dehydrogenases and 2-Ketopantoate Reductases
Yusuke WADA, Saho IWAI, Yusuke TAMURA, Tomonori ANDO, Takeshi SHINODA,
y
Kazuhito ARAI, and Hayao TAGUCHI
Department of Applied Biological Science, Faculty of Science and Technology,
Tokyo University of Science, Noda, Chiba 278-8510, Japan
Received December 26, 2007; Accepted January 12, 2008; Online Publication, April 7, 2008
[doi:10.1271/bbb.70827]
The gene for the D-mandelate dehydrogenase (D-
ManDH) of Enterococcus faecalis IAM10071 was iso-
lated by means of an activity staining procedure and
PCR and expressed in Escherichia coli cells. The
recombinant enzyme exhibited high catalytic activity
toward various 2-ketoacid substrates with bulky hydro-
phobic side chains, particularly C3-branched substrates
such as benzoylformate and 2-ketoisovalerate, and strict
coenzyme specificity for NADH and NADþ. It showed
marked sequence similarity with known NADP-depend-
ent 2-ketopantoate reductases (KPR). These results
indicate that together with KPR, D-ManDH constitutes
a new family of D-2-hydroxyacid dehydrogenases that
act on C3-branched 2-ketoacid substrates with various
specificities for coenzymes and substrates.
changes, such as amino acid replacements, cause drastic
change, in the enzyme function in this family, as for
example, conversions from a D-lactate dehydrogenase
(D-LDH) to a D-hydroxyisocaproate dehydrogenase
(D-HicDH),17,18) and even from a formate dehydrogenase
to a D-2-HydDH.19–21) Although the enzymes in this
family generally exhibit strict specificity toward NAD,
it was reported recently that Holoferax mediterranei
D-2-HydDH, which belongs to the D-2-HydDH family,
utilizes both NAD and NADP as coenzymes, acting on
bulky 2-ketoacid substrates.22) The coenzyme specific-
ities of the enzymes in this family can also be changed
by means of protein engineering.21–24) In spite of the
great variety of their catalytic functions, however, there
is no reported enzyme in the D-2-HydDH family that
exhibits high catalytic activity toward C3-branched
substrates.17,18,22,25)
Enzymes purified from Lactobacillus curvatus,26)
Enterococcus faecalis,27,28) and also the yeast Rhodotor-
ula graminis29) are known to catalyze efficiently the
conversion between benzoylformate (C6H5-CO-COOꢀ)
and D-mandelate (C6H5-CHOH-COOꢀ), which possess
a branched side chain at the C3 position, and are called
D-mandelate dehydrogenases (D-ManDHs). Neverthe-
less, their structural relation to the D-2-HydDH family
remains uncertain, because little is known about
the protein structure of D-ManDHs.28,30) Previously, we
isolated and characterized two forms of D-ManDH, D-
ManDH1 (32 kDa) and D-ManDH2 (40 kDa), from
E. faecalis IAM10071 cells.28) The two enzymes
consistently show broad substrate specificities toward
2-ketoacids that have bulky hydrophobic side chains,
such as D-HicDH25) of the D-2-HydDH family, but
exhibit particularly high activity toward C3-branched
substrates such as 2-ketoisovalerate and benzoylformate,
unlike D-HicDH. To gain information on the primary
structure of D-ManDH, in this study we isolated the gene
for one of the two D-ManDHs (D-ManDH2) from
Key words: D-mandelate dehydrogenase; 2-ketopantoate
reductase; D-2-hydroxyacid dehydrogenase;
Enterococcus faecalis; Enterococcus faeci-
um
D-2-hydroxyacid dehydrogenases (D-2-HydDHs),
which catalyze the oxidation-reduction between 2-
ketoacids and D-2-hydroxyacids using NAD(P) as a
coenzyme, play various physiological roles, with a
distinct preference for substrate side chains. They are
promising for industrial application, because optically
active 2-hydroxyacids are valuable for the synthesis of
useful compounds. It is known that many D-2-HydDHs,
such as the D-phosphoglycerate,1) D-lactate,2–4) D-hy-
droxyisocaproate,5) D-glycerate,6) and D-erythronate-4-
phosphate7) dehydrogenases, vancomycin-resistant pro-
tein H (VanH),8,9) glyoxylate reductase (GR),10) and
transcriptional co-repressor CtBP,11) share a common
protein structure and constitute the D-2-HydDH fam-
ily,12) together with some non-D-2-HydDHs, such as the
formate,13) phosphite,14,15) and L-alanine16) dehydrogen-
ases. It has been reported that only minor structural
y
To whom correspondence should be addressed. Tel: +81-4-7124-1501 ext. 3407; Fax: +81-4-7123-9767; E-mail: htaguchi@rs.noda.tus.ac.jp
Abbreviations: D-LDH, D-lactate dehydrogenase; D-HicDH, D-hydroxyisocaproate dehydrogenase; D-HydDH, D-2-hydroxyacid dehydrogenase;
D-ManDH, D-mandelate dehydrogenase; KPR, 2-ketopantoate reductase; SDS–PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis

1088
Y. WADA et al.
Fig. 1. Nucleotide Sequence of the Enterococcus D-ManDH Gene in the Expression Plasmids.
A, sequence of the Enterococcus D-ManDH gene. The loci used for the PCR primers designed (Table 1) are indicated by bold underlining
(I and III for primers 2 and 3 respectively, and II for primers 1 and 1-anti). The possible promoter and Shine-Dalgarno (SD) sequences are
shadowed. B, the sequence of the C-terminal coding region of the D-ManDH (D-ManDH2ꢂ) gene in pBK-DmanDH.
E. faecalis IAM10071 by means of an activity staining
procedure and subsequent PCR, and overproduced the
recombinant enzyme in Escherichia coli cells.
sham Biotech). After incubation on fresh plates at 37 ꢁC
overnight and then at 4 ꢁC for 4 h, the nylon membranes
were soaked in a thin layer of 50 mM sodium phosphate
buffer containing 1 mM EDTA and 20 mg/ml of hen-egg
lysozyme, and then incubated at 37 ꢁC for 3 h. After
drying on filter papers, the membranes were placed on
1.7% agar plates containing 20 mM sodium D-mandelate,
5 mM NADþ, 0.3 mg/ml of nitroblue tetrazolium,
0.05 mg/ml of phenazinemethosulphate, and 50 mM
sodium phosphate buffer (pH 7.0), and then incubated
at 37 ꢁC. One colony exhibited a clear blue color among
the approximately 1,000 colonies comprising the con-
structed E. faecalis genomic library. The cloned plasmid
(pBK-DmanDH) had a 3.5-kb insert, which contained
only one open reading frame (ORF) that could encode a
protein of molecular size of larger than 30 kDa, but
sequence analysis suggested that the ORF obtained
lacked the C-terminal coding region for the protein (see
below).
On the basis of the ORF sequence (Fig. 1), the ORF
and flanking region were amplified from the genomic
DNA by the inverted PCR procedure to obtain the full
length of D-ManDH gene sequence. EcoRI-digested
genomic DNA fragments of 4.4-kb, which hybridized
with the ORF (data not shown), were extracted from the
agarose gel, self-ligated, and then amplified by PCR
with forward and reverse DNA primers, primer 1 and
Materials and Methods
Construction of an E. faecalis IAM10071 gene li-
brary. Chromosomal DNA was isolated from E. faecalis
IAM10071 cells essentially according to Saito and
Miura.31) After partial digestion with Sau3AI and
electrophoresis on 0.8% agarose, the Sau3AI-fragments
(about 2.0 to 4.0-kb) of DNA were collected from the
agarose gel, ligated with a ZAP Express Predigested
(BamHI/CIAP-treated) Vector (Stratagene, La Jolla,
CA), and then packaged into phage particles with
Gigapack III Gold packaging extracts. The phage library
obtained was converted into a colony library containing
recombinant pBK-CMV phagemids by means of in vivo
excitation32) with a ZAP Express Predigested Vector Kit
(Stratagene).
Isolation of the gene for D-ManDH. E. coli XLOLR
cells with recombinant pBK-CMV phagemids were
plated onto LB agar plates containing kanamycin at
50 mg/ml (200 to 300 colonies/plate). When colonies
had grown to 1 to 2 mm in diameter, they were
transferred to Hybond-Nþ nylon membranes (Amer-

D-Mandelate Dehydrogenase Belongs to a New Enzyme Family
1089
Table 1. Nucleotide Sequences of the Synthetic Oligodeoxynucleo-
tides Used in Cloning and Overexpression of the Enterococcus D-
ManDH Gene
sodium D-mandelate. One unit was defined as the
catalysis of 1 mmol of substrate per min. Protein
concentrations were determined with Bio-Rad Protein
Assay protein reagent according to Bradford34) during
enzyme purification, using bovine serum albumin as the
standard. For enzyme kinetics, the protein concentration
was calculated using an extinction coefficient at 280 nm
(A280) of 0.51 mgꢀ1 cmꢀ1, which was determined from
the deduced amino acid sequence of the enzyme (see
Fig. 3A), and kinetic parameters were determined by
curve fitting of the data with a Kaleidagraph using the
Michaelis-Menten equation.
Primer 1
Primer 1-anti 50-CGATGGTTCTTAATCAAGTCTTG-30
50-CAAGACTTGATTAAGAACCATCG-30
Primer 2
Primer 3
50-GGAATTCCATATGATGAAAATAGCAATTGCAGGC-30
50-CGCGGATCCGCGGCTAACTGTGCGATCGTTACC-30
primer 1-anti (Table 1 and Fig. 1). After purification
with a QIAquick PCR Purification Kit (Qiagen,
Valencia, CA), the amplified fragment was ligated to
a pGEMꢀT Easy Vector with pGEMꢀT Easy Vector
System (Promega, Madison, WI), and then sequenced.
On the basis of the DNA sequence obtained, the full
length of the ORF was amplified from the genomic DNA
by PCR with primers 2 and 3, which provide the ORF
with the NdeI and BamHI sites in the 50- and 30-flanking
regions respectively. After purification, the PCR
product was digested with NdeI and BamHI and then
inserted between the NdeI and BamHI sites of pET-3a
(Stratagene). The resulting plasmid (pET-DmanDH)
was transformed into E. coli BL21-Gold(DE3)pLysS
(STRATAGENE) cells to produce the recombinant
D-ManDH. DNA sequence analysis was entrusted to
Custom DNA Sequencing Analysis of Bio Matrx
Research (Chiba, Japan). The oligonucleotide primers
used (Table 1) were purchased from Sigma-Aldrich
and Takara Shuzo.
Results and Discussion
Recombinant D-ManDH was highly produced in
E. coli BL21-Gold(DE3)pLysS cells harboring pET-
DmanDH, reaching approximately 20% of total soluble
proteins. The recombinant enzyme was purified by
essentially the same purification procedure as that used
for native D-ManDH2, and it exhibited an apparent
molecular weight of 40 kDa on SDS–PAGE (Fig. 2A),
in good agreement with that of native D-ManDH2,28)
although somewhat larger than the molecular weight
(34,331 Da) calculated from the deduced primary struc-
ture of the enzyme (see Fig. 3A) for unknown reasons.
Table 2 summarizes the kinetic parameters of the
purified recombinant D-ManDH for various 2-ketoacid
and D-2-hydroxyacid substrates. The recombinant en-
zyme exhibited high catalytic activity toward various
aliphatic and aromatic 2-ketoacid substrates, particularly
C3-branched substrates, such as 2-ketoisovalerate and
benzoylformate and 2-ketoisocaproate, and utilized not
L-mandelate but D-mandelate, as in the case of D-
ManDH1 and D-ManDH2, which exhibit quite similar
catalytic properties to each other, except for the Km
values for substrates. That is, the latter enzyme exhibits
markedly smaller Km values for most substrates than the
former.28) The recombinant enzyme exhibited obviously
smaller Km values for all tested substrates than D-
ManDH1, and values similar to those of D-ManDH2.
Although the Vm values are less different in D-ManDH1
and D-ManDH2 (generally 2 to 3-fold smaller for the
latter enzyme),28) the recombinant enzyme showed Vm
values for most substrates similar to those of D-
ManDH2, rather than D-ManDH1. Together with the
apparent molecular weight (Fig. 1A), these kinetic
parameters also indicate that the recombinant enzyme
obtained here can be considered to be a D-ManDH2. For
simplicity, then, we refer to this recombinant enzyme as
D-ManDH2 in the following. The recombinant D-
ManDH2 exhibited no detectable activity when NADPH
and NADPþ were used in the standard assay instead of
NADH and NADþ respectively, indicating that the
enzyme is a strictly NAD-specific dehydrogenase.
Preparation of recombinant D-ManDH. E. coli XLOR
cells carrying the phagemid pBK-DmanDH were grown
at 37 ꢁC in LB medium containing 50 mg/ml kanamycin,
harvested in the early stationary phase by centrifugation,
and then washed with 50 mM Tris–HCl buffer (pH 7.5).
The E. coli BL21-Gold(DE3)pLysS cells carrying pET-
DmanDH were grown at 30 ꢁC in 2 x YT medium (1.6%
tryptone, 1.0% yeast extract, and 0.5% NaCl) containing
150 mg/ml of ampicillin. When cell growth reached an
optical density of approximately 1.0 at 660 nm, the
cultivation temperature was reduced to 25 ꢁC, and IPTG
(final 1.0 mM) was added to the culture medium. The
cells were harvested after additional cultivation at
25 ꢁC for 2 h. Purification of the recombinant enzymes
expressed by pET-DmanDH, and pBK-DmanDH was
performed according to the procedure for native D-
ManDH2,28) which involves ammonium sulfate precip-
itation, and Butyl Toyopearl, DEAE-Sepharose, and
Resource Q (GE Healthcare Bio-Sciences) or Uno-Q
(Bio-Rad Laboratories) column chromatography. The
purity of the enzyme preparations was examined by
SDS–PAGE following to Laemmli.33)
Assay of recombinant D-ManDH. The enzyme assay
was performed at 30 ꢁC in 50 mM Tris–HCl buffer
(pH 7.5) containing 0.1 mM sodium NADH and various
concentrations of sodium 2-ketoacids (1 mM benzoyl-
formate in the standard assay), or 1 mM NADþ and
E. coli XLOR cells carrying phagemid pBK-
DmanDH, which exhibited a positive color on activity
staining, also produced an active D-ManDH, which

1090
Y. WADA et al.
A
B
C
1
2
3
4
5
1
2
3
4
1
2
120k
91k
67k
43k
30k
67k
43k
67k
43k
30k
20k
48k
38k
30k
30k
20k
20k
Fig. 2. SDS–Polyacrylamide Gel Electrophoresis of Enterococcus D-ManDH.
A, SDS–PAGE showing the purity of the recombinant D-ManDH produced by E. coli BL21-Gold(DE3)pLysS cells carrying pET-DmanDH
at each purification step. Lane 1, standard calibration marker proteins (pre-stained protein markers (low range) from Nakalai Tesque, Kyoto);
lane 2, crude extract; lane 3, supernatant of the 25% ammonium sulfate pool; lane 4, Butyl Toyopearl M650 pool; lane 5, Uno-Q pool. B, native
D-ManDH from E. faecalis IAM 10071 (28). Lanes 1 and 4, standard caribration marker proteins (LMW electrophoresis calibration kit, GE
Healthcare Bio-Sciences); lane 2, D-ManDH1; lane 3, D-ManDH2. C, recombinant D-ManDH produced by E. coli XLOR cells carrying
phagemid pBK-DmanDH (D-ManDH2ꢂ) (lane 1). Lane 2, standard calibration marker proteins (the same as in lanes 1 and 4 in panel B).
Table 2. Kinetic Parameters of Various Substrates for E. faecalis D-ManDHs
Recombinant D-ManDH2a
Km (mM) for:
Substrate
Km
Vm
Vm=Km
D-ManDH1b
D-ManDH2b
(mM)
(units/mg)
(units/mg/mM)
2-Ketobutyrate
2-Ketovalerate
2-Ketoisovalerate
2-Ketocaproate
2-Ketoisocaproate
Phenylpyruvate
Benzoylformate
Ketopantoate
9.5
0.3
81
160
8.5
530
27
2.8
4.0
0.4
0.18
1.5
500
230
2,800
150
0.5
4.2
0.6
0.15
0.6
0.13
5.7
650
530
5,000
93
0.11
3.4
11
0.25 (0.2)
n.d.
680 (480)
n.d.
2,700 (2,400)
n.d.
2.4
—
0.26
—
D-Mandelate
0.9
3.2
3.6
4.8
0.76
aThe parameters for D-ManDH2ꢂ from pBK-DmanDH are given in parenthesis.
bCited from reference 28.
n.d., Enzyme activity was undetectable.
—, not reported.
reached approximately 5% of the soluble protein (data
not shown). This does not appear to be strange, since the
gene has promoter and Shine-Dalgano (SD) like se-
quences upstream of it (Fig. 1A). It is noteworthy that
this D-ManDH, called D-ManDH2ꢂ below, lacks the 11
amino acids at the C-terminal region of D-ManDH2,
probably because of artificial rearrangement of the
genomic DNA during library construction (Fig. 1B). D-
ManDH2ꢂ was also purified by the same procedure as
native D-ManDH2, and it gave a single protein band on
SDS–PAGE, exhibiting higher mobility than D-ManDH2
(and lower mobility than D-ManDH1) (Fig. 2B and C),
giving an apparent molecular weight of approximately
35 kDa (33,110 Da based on its deduced amino acid
sequence). Like ManDH2, the purified D-ManDH2ꢂ
exhibited a strict coenzyme specificity to NADH and
NADþ, and virtually the same specific activity and Km
value toward benzoylformate as D-ManDH2 (Table 2),
indicating that the C-terminal 11 amino acids are not
essential to the catalytic function of D-ManDH2.
The homology search revealed that the amino acid
sequence of D-ManDH2, which exhibited little if any
significant sequence homology to any enzyme of the
known D-2-HydDH family, is the same as that of a
2-ketopantoate reductase (KPR) homolog (a putative
KPR) of E. faecium DO (Q3Y316). This homolog shows

D-Mandelate Dehydrogenase Belongs to a New Enzyme Family
1091
Fig. 3. Amino Acid Sequence Alignment of Enterococcus D-ManDH with KPRs (and KPR Homologs).
The amino acid residues common to E. coli KPR (AE000148) and D-ManDH (same as E. faecium DO putative KPR; Q3Y316) are shadowed.
The consensus G-X-G-X-X-G motif and residues that might directly interact with the coenzyme or substrate in the KPR 3D structure (37) are
indicated in red. A, the whole amino acid sequences of Enterococcus D-ManDH and E. coli KPR. B, sequences around possible discrimination
sites for coenzymes (Arg31) and substrates (Asn194 and Asn241) in representative KPRs (homologs) of Salmonella typhimurium (P37402),
Serratia proteamaculans (A0IR82), Vibrio vulnificus (Q8DBK6), Psychromonas sp. CNPT3 (Q1ZCM6), Pseudomonas putida (A1FF62),
Pyrocuccus horikoshii (O50098), Bacillus subtillis (O34661), Enterococcus faecalis [1] (Q83Z94), [2] (Q834J5), [3] (Q831Q5), Lactococcus
lactis subsp. cremoris (Q02YL7), Lactobacillus casei (Q039A2), Clostridium perfringens (Q0TT01), Clostridium novyi (A0PXF0),
Staphylococcus aureus (Q1Y7F2), and Salmonella typhi (Q8Z4L0). Amino acids that are common only to E. coli KPR or the D-ManDH are
indicated in green and blue respectively. Amino acid identities are indicated as the ratios of the amino acids that are identical to those in E. coli
KPR (303 amino acids) and E. faecalis IAM10071 D-ManDH2 (312 amino acids).
significant amino acid identity to an actual KPR, such as
the E. coli enzyme (PanE) (AE000148), which has a
protein structure distinct from those of the enzymes of
the known D-HydDH family.35) Therefore, it can be
concluded that D-ManDH and KPR diverged from a
common ancestral protein, constituting a new D-HydDH
family, one which is evolutionally separate from the
known D-HydDH family. The nucleotide sequence
of the D-ManDH2 gene also agreed with that of the
E. faecium DO enzyme (EMBL AAAK03000003), with

1092
Y. WADA et al.
exception of one nucleotide (GAG for Glu41 in the
E. faecium protein is replaced by GAA in the D-
ManDH2 gene). In addition, the gene of E. faecium
DO is followed by another structural gene, which
encodes a putative membrane protein (Q3Y317), possi-
bly in the same transcription unit, and this gene was also
found downstream of the D-ManDH2 gene (Fig. 1A).
These facts suggest that E. faecalis IAM10071 is closely
related to E. faecium DO, and also that the two gene
products coordinately function in Enterococcus cells,
although the physiological roles of these proteins remain
uncertain.
KPR catalyzes the conversion of 2-ketopantoate to
D-pantoate concomitantly with the oxidation of NADPH
to NADPþ, and is involved in the pantothenate synthesis
pathway in bacteria, yeast, and plants. It is thus notable
that KPR exhibits high catalytic activity toward C3-
branched 2-ketoacid and D-2-hydroxyacid substrates,
such as D-ManDH. Panel A in Fig. 3 shows amino acid
sequence alignments of E. coli KPR and Enterococcus
D-ManDH2, in which 23% of the amino acids are
identical. The structure-function relationship of E. coli
KPR has been extensively studied,35–40) and several
amino acid residues that are directly involved in
coenzyme or substrate binding or in hydrogen transfer
during the chemical reaction step have been identified.
Among these residues, Lys176 and Glu256 have been
found to be particularly important both in ligand binding
and in the subsequent proton and hydride transfer steps.
D-ManDH2 appears to be equipped with essentially the
same catalytic machinery as E. coli KPR, because the
amino acids essential to the catalytic function of E. coli
KPR, such as Lys176 and Glu256, are highly conserved
in the D-ManDH2 sequence.
On the other hand, D-ManDH2 differs markedly from
E. coli KPR in catalytic properties. First, the former
enzyme exhibits strict specificity to NADH and NADþ,
while the latter greatly prefers NADPH and NADPþ.
Like many NAD(P)-dependent dehydrogenases, E. coli
KPR possesses a domain comprising the Rossmann-
fold-like ꢀ/ꢁ structure (residues 1 to 161) and G-X-G-
X-X-G (where X can be any amino acid) sequence,
which is a consensus motif in the nucleotide binding
site,41) in the tight ꢁ1/ꢀA turn (residues 7 to 12) of this
domain.35–37) It is known that the ꢁ2-ꢀB loop of this
domain is responsible for NADH-NADPH discrimina-
tion by usual NAD(P)-dependent dehydrogenases.23,42,43)
In the case of E. coli KPR, the guanidino group of
Arg31 in the corresponding loop (ꢁ2-ꢁ3 loop in the case
of the KPR due to an inserted anti-parallel ꢁ-structure)
forms a hydrogen bond with the 20-phosphate of the
NADP adenine ribose.37) In contrast, D-ManDH2 lacks
Arg31, and instead has an Asp and Glu at positions 30
and 34 (the numbering of amino acid residues of the
enzyme follows that for E. coli KPR) respectively, while
the enzyme has the consensus G-X-G-X-X-G motif at
the corresponding positions (Fig. 3A). This is in good
agreement with the fact that NAD-specific enzymes
commonly possess an acidic amino acid, Asp or Glu,
in the ꢁ2-ꢀB loop, which forms a hydrogen bond with
the 20-hydroxyl group of NAD, but not with the 20-
phosphate of NADP, due to static repulsion. Secondly,
D-ManDH2 and E. coli KPR exhibit greatly different
specificities for 2-ketoacid substrates. D-ManDH2
exhibited no detectable activity toward 2-ketopantoate,
but high activity toward 2-ketoacids with bulky hydro-
phobic side chains, such as 2-ketoisovalerate, benzoyl-
formate, and 2-ketoisocaproate (Table 1). In contrast,
E. coli KPR greatly prefers 2-ketopantoate to 2-keto-
isovalerate (about 360-fold more as to kcat=Km value)
although no activity of the enzyme toward benzoylfor-
mate or 2-ketoisocaproate has been reported.40) In the
ternary complex of E. coli KPR with NADPþ and D-
pantoate,37) the C4 hydroxyl group of the D-pantoate
forms a hydrogen bond with the side chain of Asn194 on
the second domain (residues 162 to 303) of the enzyme,
and possibly with that of Asn241, which might also
interact with the amide of the coenzyme nicotinamide
ring. On the other hand, D-ManDH2 lacks both Asn194
and Asn241 at the corresponding positions (Fig. 3A).
It is reasonable to suppose that Asn194 is replaced with
a hydrophobic amino acid (Met) in D-ManDH2, since
the enzyme prefers the hydrophobic side chains of
2-ketoacid substrates, although analysis of the 3D
structure and mutant enzymes is required to explain
the substrate recognition by D-ManDH2. It is likely that
the two enzymes diverged for their distinct coenzyme
and substrate specificities through minor structural
changes during evolution.
There are many amino acid sequences of KPR and
D-ManDH homologs, most of which are designated as
putative KPRs, in the present protein databases. Panel B
in Fig. 3 lists representative homologs, in all of which
residues essential to catalytic function, such as Lys176
and Glu256, are highly conserved, for sequence
alignment around the possible coenzyme and substrate
discrimination sites. The proteins listed in group 1
consistently possess Arg31, Asn194, and 241, indicating
that they are NADP-dependent KPRs in spite of their
highly divergent primary structures. On the other hand,
all of those in group 2 lack Arg31 and instead have
Asp30, and many of them Glu34 or Asp34, strongly
suggesting that all of these proteins are NAD-specific
enzymes. Furthermore, these proteins consistently lack
both Asn194 and 241, but have homologous amino
acid sequences at positions corresponding to those in
D-ManDH2. This strongly suggests that the group 2
proteins exhibit similar coenzyme and substrate specif-
icities to those of D-ManDH2 rather than KPR.
The sequence alignment (Fig. 3B) also indicates
that D-ManDH2-like proteins are widely distributed in
microorganisms, in particular Enterococci and related
bacteria, with highly divergent primary structures. It is
in addition notable that vancomycin-resistant E. faecalis
V58344) possesses several (at least three) D-ManDH2-
like proteins (E. faecalis [1], [2], and [3] in Fig. 3B).

D-Mandelate Dehydrogenase Belongs to a New Enzyme Family
1093
precursor by vancomycin resistance proteins VanH and
VanA. Biochemistry, 30, 10408–10415 (1991).
10) Ohshima, T., Nunoura-Kominato, N., Kudome, T., and
This suggests that it is not rare that one bacterial strain
contains plural D-ManDHs within the cell, like E. fae-
calis IAM10071, particularly in the case of Enter-
ococcus species, although the physiological roles and
relation to the antibiotic-resistance of the bacteria
remain uncertain in this stage.
This study identifies a novel D-2-HydDH family that
comprises enzymes that act highly on C3-branched 2-
ketoacid and 2-hydroxyacid substrates, such as E. coli
KPR and Enterococcus D-ManDH. The enzymes of this
family are widely distributed in organisms and shows
great variety in their primary structures and specificities
to coenzymes and substrates, and are promising for
practical application, particularly in stereospecific syn-
thesis of C3-branched D-hydroxyacids.
Sakurabe, H.,
A novel hyperthermophilic archaeal
glyoxylate reductase from Thermococcus litoralis:
characterization, gene cloning, nucleotide sequence and
expression in Escherichia coli. Eur. J. Biochem., 268,
4740–4747 (2001).
11) Schaeper, U., Boyd, J. M., Verma, S., Uhlmann, E.,
Subramanian, T., and Chinnadurai, G., Molecular clon-
ing and characterization of a cellular phosphoprotein
that interacts with a conserved C-terminal domain of
adenovirus E1A involved in negative modulation of
oncogenic transformation. Proc. Natl. Acad. Sci. USA,
92, 10467–10471 (1995).
12) Grant, G. A., A new family of 2-hydroxyacid dehydro-
genases. Biochem. Biophys. Res. Commun., 165, 1371–
1374 (1989).
13) Popov, V. O., and Lamzin, V. S., NADþ-dependent
formate dehydrogenase. Biochem. J., 301, 625–643
(1994).
14) Costas, A. M. G., White, A. K., and Metcalf, W. W.,
Purification and characterization of a novel phosphorus-
oxidizing enzyme from Pseudomonas stutzeri WM88.
J. Biol. Chem., 276, 17429–17436 (2001).
15) Woodyer, R., Wheatley, J. L., Relyea, H. A., Rimkus, S.,
and van der Donk, W. A., Site-directed mutagenesis
of active site residues of phosphite dehydrogenase.
Biochemistry, 44, 4765–4774 (2005).
16) Baker, P. J., Sawa, Y., Shibata, H., Sedelnikova, S. E.,
and Rice, D. W., Analysis of the structure and substrate
binding of Phormidium lapideum alanine dehydrogen-
ase. Nat. Struct. Biol., 5, 561–567 (1998).
17) Tokuda, C., Ishikura, Y., Shigematsu, M., Mutoh, H.,
Tsuzuki, S., Nakahira, Y., Tamura, Y., Shinoda, T., and
Taguchi, H., Conversion of Lactobacillus pentosus D-
lactate dehydrogenase through a single amino acid
replacement. J. Bacteriol., 185, 5023–5026 (2003).
18) Ishikura, Y., Tsuzuki, S., Takahashi, O., Tokuda, C.,
Nakanishi, R., Shinoda, T., and Taguchi, H., Recognition
site for the side chain of 2-ketoacid substrate in D-lactate
dehydrogenase. J. Biochem., 138, 741–749 (2005).
19) Shinoda, T., Arai, K., Shigematsu-Iida, M., Ishikura, Y.,
Tanaka, S., Yamada, T., Kimber, M. S., Pai, E. F.,
Fushinobu, S., and Taguchi, H., Distinct conformation-
mediated functions of an active site loop in the catalytic
reactions of NAD-dependent D-lactate dehydrogenase
and formate dehydrogenase. J. Biol. Chem., 280, 17068–
17075 (2005).
References
1) Tobey, K. L., and Grant, G. A., The nucleotide sequence
of the serA gene of Escherichia coli and the amino acid
sequence of the encoded protein, D-3-phosphoglycerate
dehydrogenase. J. Biol. Chem., 261, 12179–12183
(1986).
2) Taguchi, H., and Ohta, T., D-Lactate dehydrogenase is a
member of the D-isomer-specific 2-hydroxyacid dehy-
drogenase family: cloning, sequencing, and expression
in Escherichia coli of the D-lactate dehydrogenase gene
of Lactobacillus plantarum. J. Biol. Chem., 266, 12588–
12594 (1991).
3) Bernard, N., Ferain, T., Garmyn, D., Hols, P., and
Delcour, J., Cloning of the D-lactate dehydrogenase gene
from Lactobacillus delbrueckii subsp. bulgaricus by
complementation in Escherichia coli. FEBS Lett., 290,
61–64 (1991).
4) Kochhar, S., Hunziker, P. E., Leong-Morgenthaler, P.,
and Hottinger, H., Primary structure, physicochemical
properties, and chemical modification of NADþ-depend-
ent D-lactate dehydrogenase: evidence for the presence
of Arg-235, His-303, Tyr-101, and Trp-19 at or near the
active site. J. Biol. Chem., 267, 8499–8513 (1992).
5) Lerch, H. P., Blocker, H., Kallwass, H. K. W., Hoppe, J.,
Tsai, H., and Collins, J., Cloning, sequencing and
expression in Escherichia coli of the D-2-hydroxyiocap-
roate dehydrogenase gene of Lactobacillus casei. Gene,
78, 47–57 (1989).
6) Goldberg, J. D., Yoshida, T., and Brick, P., Crystal
structure of an NAD-dependent D-glycerate dehydrogen-
˚
ase at 2.4 A resolution. J. Mol. Biol., 236, 1123–1140
(1994).
20) Shinoda, T., Arai, K., and Taguchi, H., A highly specific
glyoxylate reductase derived from a formate dehydro-
genase. Biochem. Biophys. Res. Commun., 355, 782–787
(2007).
21) Tishkov, V. I., and Popov, V. O., Protein engineering
of formate dehydrogenase. Biomol. Eng., 23, 89–110
(2006).
22) Domenech, J., and Ferrer, J. A., A new D-2-hydroxyacid
dehydrogenase with dual coenzyme-specificity from
Haloferax mediterranei: sequence analysis and heterol-
ogous overexpression. Biochim. Biophys. Acta, 1760,
1667–1674 (2006).
7) Ha, J. Y., Lee, J. H., Kim, K. H., Kim, D. J., Lee, H. H.,
Kim, H.-K., Yoon, H.-J., and Suh, S. W., Crystal
structure of D-erythronate-4-phosphate dehydrogenase
complexed with NAD. J. Mol. Biol., 366, 1294–1304
(2007).
8) Arthur, M., Molinas, C., Dutka-Malen, S., and
Courvalin, P., Structural relationship between the van-
comycin resistance protein VanH and 2-hydroxycarbox-
ylic acid dehydrogenases. Gene, 103, 133–134 (1991).
9) Bugg, T. D., Wright, G. D., Dutka-Malen, S., Arthur, M.,
Courvalin, P., and Walsh, C. T., Molecular basis for
vancomycin resistance in Enterococcus faecium
BM4147: biosynthesis of a depsipeptide peptidoglycan
23) Bernard, N., Johnsen, K., Holbrook, J. J., and Delcour,
J., D175 discriminates between NADH and NADPH in

1094
Y. WADA et al.
the coenzyme binding site of Lactobacillus delbruekii
subsp. bulgaricus D-lactate dehydrogenase. Biochem.
Biophys. Res. Commun., 208, 895–900 (1995).
structure of Escherichia coli ketopantoate reductase at
1.7 A resolution and insight into the enzyme mechanism.
Biochemistry, 40, 14493–14500 (2001).
24) Relyea, H. A., and van der Donk, W. A., Mechanism and
applications of phosphite dehydrogenase. Bioorg. Chem.,
33, 171–189 (2005).
36) Lobley, C. M., Ciulli, A., Whitney, H. M., Williams, G.,
Smith, A. G., Abell, C., and Blundell, T. L., The crystal
structure of Escherichia coli ketopantoate reductase with
NADPþ bound. Biochemistry, 44, 8930–8939 (2005).
37) Ciulli, A., Chirgadze, D. Y., Smith, A. G., Blundell,
T. L., and Abell, C., Crystal structure of Escherichia coli
ketopantoate reductase in a ternary complex with
NADPþ and pantoate bound: substrate recognition,
conformational change, and cooperativity. J. Biol.
Chem., 282, 8487–8497 (2007).
38) Zheng, R., and Blanchard, J. S., Kinetic and mechanistic
analysis of the E. coli panE-encoded ketopantoate
reductase. Biochemistry, 39, 3708–3717 (2007).
39) Zheng, R., and Blanchard, J. S., Identification of active
site residues in E. coli ketopantoate reductase by
mutagenesis and chemical rescue. Biochemistry, 39,
16244–16251 (2000).
25) Hummel, W., Schutte, H., and Kula, M.-R., D-2-
¨
Hydroxyisocaproate dehydrogenase from Lactobacillus
casei: a new enzyme suitable for stereospecific reduction
of 2-ketocarboxylic acids. Appl. Microbiol. Biotechnol.,
21, 7–15 (1985).
26) Hummel, W., Schutte, H., and Kula, M.-R., D-(ꢀ)-
¨
Mandelic acid dehydrogenase from Lactobacillus cur-
vatus. Appl. Microbiol. Biotechnol., 28, 433–439 (1988).
27) Yamazaki, Y., and Maeda, H., Enzymatic synthesis of
optically pure (R)-(ꢀ)-mandelic acid and other 2-hy-
droxycarbonic acids: screening for the enzyme, and its
purification, characterization and use. Agric. Biol.
Chem., 50, 2621–2631 (1986).
28) Tamura, Y., Ohkubo, A., Iwai, S., Wada, Y., Shinoda,
T., Arai, K., Mineki, S., Iida, M., and Taguchi, H., Two
forms of NAD-dependent D-mandelate dehydrogenase
in Enterococcus faecalis IAM10071. Appl. Environ.
Microbiol., 68, 947–951 (2002).
40) Zheng, R., and Blanchard, J. S., Substrate specificity and
kinetic isotope effect analysis of the Eschericia coli
ketopantoate reductase. Biochemistry, 42, 11289–11296
(2003).
29) Baker, D. P., and Fewson, C. A., Purification and
characterization of D(ꢀ)-mandelate dehydrogenase from
Rhodotorula graminis. J. Gen. Microbiol., 135, 2035–
2044 (1989).
41) Wierenga, R. K., De Maeyer, M. C. H., and Hol,
W. G. J., Interaction of pyrophosphate moiety with
ꢀ-helixes in dinucleotide-binding proteins. Biochemistry,
24, 1346–1357 (1985).
30) Basak, A. K., Baker, D. P., Fewson, C. A., and Stuart,
D. I., Preliminary crystallographic study of D(ꢀ)-
mandelate dehydrogenase from Rhodotorula graminis.
J. Mol. Biol., 233, 781–783 (1993).
31) Saito, H., and Miura, K., Preparation of transforming
deoxyribonucleic acid by phenol treatment. Biochim.
Biophys. Acta, 72, 619–629 (1963).
32) Short, J. M., Fernandez, J. M., Sorge, J. A., and Huse,
W. D., Lambda ZAP: a bacteriophage lambda expression
vector with in vivo excision properties. Nucleic Acids
Res., 16, 7583–7600 (1988).
33) Laemmli, U. K., Cleavage of structural proteins during
the assembly of the head of bacteriophage T4. Nature,
227, 680–685 (1970).
34) Bradford, M. M., A rapid and sensitive method for the
quantitation of microbram quantities of protein utilizing
the principle of protein-dye binding. Anal. Biochem., 72,
248–254 (1976).
35) Matak-Vinkovic, D., Vinkovic, M., Saldanha, S. A.,
Ashurst, J. L., von Delft, F., Inoue, T., Miguel, R. N.,
Smith, A. G., Blundell, T. L., and Abell, C., Crystal
42) Feeney, R., Clarke, A. R., and Holbrook, J. J., A single
amino acid substitution in lactate dehydrogenase im-
proves the catalytic efficiency with an alternative
coenzyme. Biochem. Biophys. Res. Commun., 166,
667–672 (1990).
43) Fan, F., Lorenzen, J. A., and Plapp, B. V., An aspartate
residue in yeast alcohol dehydrogenase I determines
the specificity for coenzyme. Biochemistry, 30, 6397–
6401 (1991).
44) Paulsen, I. T., Banerjei, L., Myers, G. S., Nelson, K. N.,
Seshadri, R., Read, T. D., Fouts, D. E., Eisen, J. A., Gill,
S. R., Heidelberg, J. F., Tettelin, H., Dodson, R. J.,
Umayam, L., Brinkac, L., Beanan, M., Daugherty, S.,
DeBoy, R. T., Durkin, S., Kolonay, J., Madupu, R.,
Nelson, W., Vamathevan, J., Tran, B., Upton, J., Hansen,
T., Shetty, J., Khouri, H., Utterback, T., Radune, D.,
Ketchum, K. A., Dougherty, B. A., and Fraser, C. M.,
Role of mobile DNA in the evolution of vancomycin-
resistant Enterococcus faecalis. Science, 299, 2071–
2074 (2003).