2,5-Diketo-gluconic acid reductase from Corynebacterium glutamicum: Characterization of stability, catalytic properties and inhibition mechanism for use in vitamin C synthesis

Article abstract of DOI:10.1016/j.procbio.2012.07.014

2,5-Diketo-d-gluconic acid (2,5-DKG) reductase is an NADPH-dependent, monomeric aldo-keto reductase (AKR) which catalyzes the reduction of 2,5-DKG to 2-keto-l-gulonic acid (2-KLG) - the immediate precursor of vitamin C. The reaction catalyzed by 2,5-DKG reductase is attractive to bypass several chemical steps and produce vitamin C biocatalytically. In a screening of 22 bacterial strains, nine 2,5-DKG reductase producing bacterial strains were found. The gene of Corynebacterium glutamicum 2,5-DKG reductase was cloned and overexpressed in Escherichia coli. By batch fermentation 409 mg L^-1 of 2,5-DKG reductase with a C-terminal His₆-tag were obtained. The purified 2,5-DKG reductase was characterized in detail. The enzyme is most active in a pH range from 5.0 to 8.0 and its stability is high at temperatures below 35 °C. Catalytic constants for 2,5-DKG and NADPH were determined and a weak inhibition by the product 2-KLG was found. 2,5-DKG reductase activity is strongly inhibited by the common process ions Mg²⁺, Ca²⁺, SO₄^3- and Cl^-, which suggests that these should be avoided in the process. The inhibition mechanism for Cl^- was elucidated. It is a competitive inhibitor with respect to NADPH and a noncompetitive inhibitior with respect to 2,5-DKG.

Full text of DOI:10.1016/j.procbio.2012.07.014

Process Biochemistry 47 (2012) 2012–2019
Contents lists available at SciVerse ScienceDirect
Process Biochemistry
journal homepage: www.elsevier.com/locate/procbio
2
,5-Diketo-gluconic acid reductase from Corynebacterium glutamicum:
Characterization of stability, catalytic properties and inhibition mechanism for
use in vitamin C synthesis
∗
Vanja Kaswurm, Claudia Pacher, Klaus Dieter Kulbe, Roland Ludwig
Food Biotechnology Laboratory, Department of Food Science and Technology, BOKU – University of Natural Resources and Life Sciences, Muthgasse 18, 1190 Vienna, Austria
a r t i c l e i n f o
a b s t r a c t
Article history:
2,5-Diketo-d-gluconic acid (2,5-DKG) reductase is an NADPH-dependent, monomeric aldo-keto reduc-
tase (AKR) which catalyzes the reduction of 2,5-DKG to 2-keto-l-gulonic acid (2-KLG) – the immediate
precursor of vitamin C. The reaction catalyzed by 2,5-DKG reductase is attractive to bypass several chem-
ical steps and produce vitamin C biocatalytically. In a screening of 22 bacterial strains, nine 2,5-DKG
reductase producing bacterial strains were found. The gene of Corynebacterium glutamicum 2,5-DKG
Received 14 March 2012
Received in revised form 5 July 2012
Accepted 10 July 2012
Available online 17 July 2012
−
1
reductase was cloned and overexpressed in Escherichia coli. By batch fermentation 409 mg L of 2,5-DKG
reductase with a C-terminal His6-tag were obtained. The purified 2,5-DKG reductase was characterized
in detail. The enzyme is most active in a pH range from 5.0 to 8.0 and its stability is high at tempera-
Keywords:
Ascorbic acid
2
2
,5-Diketo-gluconic acid reductase
-Keto-l-gulonic acid
◦
tures below 35 C. Catalytic constants for 2,5-DKG and NADPH were determined and a weak inhibition
Corynebacterium glutamicum
Inhibition
by the product 2-KLG was found. 2,5-DKG reductase activity is strongly inhibited by the common pro-
2
+
2+
3−
−
cess ions Mg , Ca , SO4 and Cl , which suggests that these should be avoided in the process. The
−
inhibition mechanism for Cl was elucidated. It is a competitive inhibitor with respect to NADPH and a
noncompetitive inhibitior with respect to 2,5-DKG.
©
2012 Elsevier Ltd. All rights reserved.
1
. Introduction
been biochemically characterized. Those include two native DKGRs
from a species of Corynebacterium (2,5-DKG reductase A; AKR5C
and 2,5-DKG reductase B; AKR5D) [7,8], two homologous expressed
2,5-DKG reductase from Escherichia coli (YqhE and YafB) [9,10]
and two heterologous expressed 2,5-DKG reductases from uncul-
tured microbes. Two heterologous expressed 2,5-DKG reductases
from uncultured microbes have been found by screening environ-
mental DNA expression libraries [11]. Structurally, only 2,5-DKG
reductase A from Corynebacterium sp. [12–14] and its quadruple
mutant [15] have been studied. Native 2,5-DKG reductase A from
Corynebacterium sp. is a monomeric enzyme (about 34 kDa) com-
posed of eight ␣-helices and eight parallel ␤-strands (TIM barrel;
The bacterial enzyme 2,5-diketo-d-gluconic acid reductase (2,5-
DKG reductase; 2,5-didehydrogluconate reductase; EC 1.1.1.274)
is one of more than 140 members of the aldo-keto reductases
(
AKRs), an enzyme superfamily of NAD(P)(H)-dependent oxidore-
ductases [1,2]. This enzyme catalyses the stereo specific reduction
of 2,5-diketo-d-gluconic acid (2,5-DKG) at position C-5 to 2-keto-
l-gulonic acid (2-KLG) [3], which is an intermediate that can be
transformed into l-ascorbic acid (vitamin C) by a single chemical
step [4].
The first microorganisms available for conversion of 2,5-DKG to
2
-KLG were isolated from soil and sewage by Sonoyama and col-
(␣/␤) ), similar to most microbial AKRs [1,12]. The reduced form
8
leagues in the 1970s [5,6]. These 2-KLG producing strains belong
to the Brevibacterium, Arthrobacter, Micrococcus, Staphylococcus,
Pseudomonas, Bacillus and Corynebacterium genera. In 1987, the
conversion of 2,5-DKG to 2-KLG by Corynebacterium sp. was identi-
fied as a single catalytic step of 2,5-DKG reductase in the cytosol [7].
To date, only a few enzymes with 2,5-DKG reductase activity have
of pyridine nucleotide NADP(H) is bound to the C-terminal face
of the barrel. The absence of a canonical Rossman fold in active
site set AKRs apart from numerous dehydrogenases [16,17]. Muta-
tion studies of residues in the coenzyme binding site and substrate
binding pocket in the apo and coenzyme-bound form of 2,5-DKG
reductase show that binding of NADPH causes communicated and
coordinated structural changes into these regions [14].
2
,5-DKG reductase is of high interest for the biocatalytic pro-
duction of the key intermediate 2-KLG by the 2,5-diketo-d-gluconic
acid pathway from d-glucose via d-gluconate, 2-keto-d-gluconate
and 2,5-diketo-d-gluconate. Vitamin C can be obtained through
transformation and refining of 2-KLG [18]. Sonoyama et al. [19]
∗
Corresponding author at: Department of Food Sciences and Technology,
University of Natural Resources and Life Sciences, Muthgasse 18, 1190 Vienna,
Austria. Tel.: +43 147654 6149; fax: +43 147654 6199.
E-mail address: roland.ludwig@boku.ac.at (R. Ludwig).
1
359-5113/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.procbio.2012.07.014

V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
2013
DSM 20301, its expression in E. coli and biochemical characteriza-
tion in regard to process relevant properties of the recombinant
enzyme. A detailed kinetic study of the catalytic mechanism of 2,5-
DKG reductase and its inhibition by cations and anions provides
mechanistic insights for further enzyme and process engineering.
2
. Materials and methods
2
.1. Materials
Salts, acids and bases for enzyme assays and media preparation were purchased
from commercial suppliers at the highest level of purity possible. Media compo-
nents were obtained from Sigma–Aldrich, Roth and Merck. 2,5-Diketo-d-gluconic
acid (2,5-DKG) was produced by fermentation of P. cypripedii as previously described
[
28] and isolated from the culture broth by methanol precipitation. It was further
purified by liquid chromatography, using isocratic elution with ultrapure water on
Amberlite CG120-II (Sigma–Aldrich) in the calcium form. The purity and concentra-
tion of 2,5-DKG was determined by HPLC. Anhydrous 2-KLG was purchased from
Hofmann-La Roche.
2
.2. Microorganisms, plasmids and media
Bacterial strains used for screening experiments are given in Table 1 and were
obtained from DSMZ (Braunschweig, Germany). The freeze-dried cultures were
revived following the suppliers recommendations and periodically subcultured on
−
1
recommended media containing 15 g L agar. The media used and their compo-
sition are given in the supplemental information (Tables S1 and S2). The E. coli
strains TOP10 and BL21/DE3 (Invitrogen, Carlsbad, CA) were used for cloning and
expression. The construction of the pET-21d/dkr vector, which expresses the His-
tagged 2,5-DKG reductase gene from C. glutamicum under the control of the T7
promoter, has been described recently [28]. The integrity of the construct was con-
firmed by DNA sequence analysis (VBC-Biotec, Vienna, Austria). The sequence of dkr
was deposited in the GenBank database; accession number: JQ407590. To optimize
the production of recombinant 2,5-DKG reductase by E. coli BL21 (DE3), the follow-
Fig. 1. Synthetic routes to ascorbic acid. Route A depicts the classical Reichstein
process [26,27], Route B shows the process for utilizing the 2-KLG intermediate
produced by 2,5-DKG-reductase [28].
invented a two-stage fermentation process for 2-KLG production
where glucose is oxidized to 2,5-DKG by a mutated Erwinia sp. and
is then reduced to 2-KLG by a mutant strain of Corynebacterium sp.
The second step reaction is catalyzed by NADPH dependent 2,5-
DKG reductase. A tandem fermentation process to produce 2-KLG
from gluconic acid by using co-immobilized cells of Gluconobac-
ter oxydans and Corynebacterium sp. has also been suggested [20].
Also the genetically engineered Erwinia strains; Erwinia herbicola
−
1
ing media were screened: Luria broth [29], M9-medium (8.5 g L Na2HPO4·2H2O,
− −1 −1 −1
1
3
.0 g L KH2PO4, 0.5 g L NaCl and 1.0 g L NH4Cl supplemented with 20 g L glu-
cose, 10 mL of a 1 M MgSO4 (final concentration = 10 mM) and 0.5 mL of a 1 M CaCl2
solution (final concentration = 0.5 mM)) and MCHGly-medium (M9-medium sup-
−
1
−1
plemented with 10 g L casein and 10 g L glycerol instead of glucose). For M9-
and MCHGly-media the M9-salt solution was prepared in a 10 times concentrated
solution and autoclaved separately.
[
3] and E. citreus [21], which naturally accumulate 2,5-DKG from
2
.3. Screening for 2,5-DKG reductase producing organisms
d-glucose have been employed. The gene encoding for 2,5-DKG
reductase was cloned from Corynebacterium sp. into the above
mentioned Erwinia strains, allowing an elegant one-organism fer-
mentation of 2-KLG directly from d-glucose. According to Powers
For shaking flasks cultivation experiments, 25 mL of the appropriate medium
was inoculated with a single colony from agar-plates and cultivated overnight under
the recommended conditions on an orbital shaker (eccentricity = 2.5 cm, rotational
frequency = 140 rpm). One mL of this starter culture was used to inoculate 100 mL
[
22], the transport of 2,5-DKG into, and the diffusion of 2-KLG
−1
of the production stage medium supplemented with 10 g L 2,5-DKG by using a
out of the 2-KLG synthesizing cells, appear to be the rate-limiting
steps. Based on this, Genencor established an in vitro biocatalytic
four steps method to produce 2-KLG from d-glucose [23,24]. Four
0
.2 ␮m filter for sterilization. Samples were taken aseptically at 24, 48 and 72 h
after inoculation for measuring enzymatic activity and protein concentration. After
reaching the maximum cell density (measured by OD600) the biomass was har-
vested by centrifugation at 10,000 × g for 10 min. The cells were resuspended by
adding 3 mL of 50 mM Bis–Tris buffer, pH 6.4, containing 1 mM phenylmethylsul-
fonylfluoride (PMSF) to the centrifuged cells for homogenization by a French Press.
Using the spectrophotometric assay (see Section 2.7) with 50 mM Bis–Tris buffer,
pH 6.4, 2,5-DKG reductase activity was determined from this lysate (measurements
were performed in triplicates and are reported as mean values and their standard
deviation) and 2,5-DKG reductase activity was verified by small-scale conversion
experiments of 2,5-DKG to 2-KLG. To that purpose, the lysate containing approx.
+
enzymes: NADP dependent glucose dehydrogenase (GDH) from
Thermoplasm acidophilum, NADPH dependent 2,5-DKG reductase
from Corynebacterium sp., gluconate dehydrogenase and 2-keto-
d-gluconate dehydrogenase (both from permeabilized, modified
Pantoea citrea cells with glucose dehydrogenase activity) are
involved in the continuous conversion of d-glucose. The first two,
soluble enzymes are exogenously added and regenerate the coen-
zyme in situ.
Nowadays a remarkable part of vitamin C industrial production
is performed in a two-step fermentation process [4,25,26], but the
traditional Reichstein process [27] which involves several environ-
mentally hazardous chemical and energy consuming steps is still
utilized for vitamin C synthesis (Fig. 1, process route A). In our pre-
vious published work the synthesis of 2-KLG from d-glucose was
established (process route B) [28]. First, d-glucose is converted by
Pectobacter cypripedii strain HEPO1 (DSMZ 12393) into 2,5-DKG,
which is then enzymatically converted with NADPH-dependent
1
U of 2,5-DKG reductase activity was supplemented with 25 mM 2,5-DKG and an
equimolar amount of NADPH as coenzyme in 1 mL 50 mM Bis–Tris buffer, pH 6.4,
for 24 h at 25 C. These samples were analyzed with HPLC and 2-KLG production
was determined by comparison to standards. To further confirm the identity of
the accumulated product, the formed 2-KLG was converted to the final product l-
ascorbic acid. Therefore, a reaction mixture containing 0.5 mL of 2-KLG solution and
0.5 mL H SO (98%) was incubated for 30 min at 100 C. The reaction was stopped
by adding 10 mL of cold, distilled water and the amount of l-ascorbic acid formed by
chemical conversion was determined using a colorimetric test kit from R-Biopharm
◦
◦
2
4
(
Darmstadt, Germany) following the manufacturer’s recommendations.
2
.4. Expression of 2,5-DKG reductase in E. coli
2
,5-DKG reductase from Corynebacterium glutamicum to 2-KLG.
Media optimization experiments were carried out in baffled shaking flasks
−
1
with Luria broth, M9-medium and MCHGly-medium containing 0.1 mL L
NADPH is regenerated in situ by GDH from Bacillus sp. and d-glucose
in the second biocatalytic step. Using this bi-enzymatic system,
−
1
antifoam and 0.1 g L ampicillin. To induce dkr gene expression, isopropyl-␤-d-
thiogalactopyranoside (IPTG) or lactose in different concentrations were added:
2
,5-DKG is completely reduced to 2-KLG. Here, we describe the
−1
−1
−1
IPTG: 0.01 mM, 0.1 mM, 1 mM; lactose: 1 g L , 5 g L , 10 g L . Induction was started
◦
screening and cloning of a 2,5-DKG reductase from C. glutamicum
at an OD600 of 0.5–0.7. Cultures were grown at temperatures ranging from 25
C

2
014
V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
Table 1
Bacteria screened for 2,5-DKG reductase activity.
Activitya (U L−1
)
2-KLG (mg mL )b
−1
l-Ascorbic acidc
Strain
Strain identifier
Cultivation
time (hours)
Maximum
OD600
Acetobacter hansenii
Acetobacter lovaniensis
Arthrobacter agilis
Arthrobacter roseus
Bacillus alvei
DSMZ 5602
DSMZ 4491
DSMZ 20550
DSMZ 14508
DSMZ 29
42
47
47
47
42
42
42
24
24
24
24
47
24
24
24
42
42
42
42
42
24
24
9.65
4.62
1.54
1.02
3.76
7.22
1.1
4.88
4.81
8.65
8.84
2.03
7.11
12.88
13.45
6.12
7.99
5.24
8.6868
8.85
8.43
5.74
32 ± 2
750 ± 311
31 ± 18
19 ± 1
1.5
ND
ND
ND
ND
0.9
0.4
0.8
1.0
2.0
ND
0.5
ND
0.2
ND
ND
ND
ND
0.7
ND
ND
ND
+
8 ± 1
Bacillus funiculus
DSMZ 15141
DSMZ 20530
DSMZ 20166
DSMZ 20145
DSMZ 20301
DSMZ 20138
DSMZ 7148
DSMZ 20032
DSMZ 20033
DSMZ 20315
DSMZ 30176
DSMZ 30177
DSMZ 20641
DSMZ 50259
DSMZ 1241
DSMZ 7068
DSMZ 9930
64 ± 8
+
+
+
+
+
Brevibacterium imperiale
Brevibacterium testaceum
Corynebacterium barkeri
Corynebacterium glutamicum
Corynebacterium ilicis
Gluconobacter asaii
Kocuria kristinae
Kocuria varians
Micrococcus lylae
Pantoea stewarti subsp. stewartii
Pectobacterium chrysanthemi
Pimelobacter jenseii
Pseudomonas cichorii
Pseudomonas syringae
Staphylococcus muscae
Staphylococcus pulvereri
19 ± 1
30 ± 17
48 ± 31
186 ± 34
360 ± 85
830 ± 127
53 ± 4
+
+
26 ± 2
12 ± 4
22 ± 3
84 ± 25
10 ± 6
235 ± 7
220 ± 14
14 ± 4
+
17 ± 4
a
Activity was measured with spectrophotometric assay.
ND. . .2-KLG peak not detected by HPLC.
Detection of l-ascorbic acid colorimetric test kit from R-Biopharm (Darmstadt; Germany).
b
c
◦
to 37 C. After harvest and homogenization of the biomass, the 2,5-DKG reductase
activity and protein concentration were measured in the supernatant. A large-scale
batch cultivation of E. coli BL21 Star (DE3) transformants bearing the dkr gene was
carried out in a 42-L stirred-tank bioreactor (Applikon, Schiedam, Netherlands) filled
2.7. Measurement of enzymatic activity
2,5-DKG reductase activity was determined spectrophotometrically by follow-
−
1
−1
◦
ing the decrease of NADPH absorbance at 340 nm (ε340 = 6.22 mM cm ) and 25 C
−1
with 30 L of MCHGly medium containing 0.1 g L ampicillin. The cultivation was
for 3 min. The assay mixture contained buffer (50 mM ammonium acetate buffer
pH 6.4, unless otherwise stated), 8.256 mM 2,5-DKG, 0.2 mM NADPH and a suitable
activity of 2,5-DKG reductase in a final volume of 1 mL in cuvettes of 1 cm optical
path length. The substrate and co-substrate stock solutions were freshly prepared
started by inoculation with shaking flask cultures (5% v/v, OD600 ∼0.5). For induction
−1
lactose was added to a final concentration of 5 g L after an OD600 of 0.8 was reached.
The pH was maintained at pH 6.5 by automatic titration with sterile NaOH and the
concentration of dissolved oxygen (DO) was kept at 20% air saturation by sparging
−
1
each day. A volumetric activity of 6.4 U L was determined to be the lower limit
of reliable quantification. One unit of 2,5-DKG reductase activity was defined as the
−1
with filtered air (0–20 L min ) and adjusting the stirrer velocity (200–800 rpm).
The cells were cultivated at 37 C. Shortly before induction of dkr gene expression
the temperature was reduced to 25 C. All setpoint parameters were controlled
◦
−1
amount of enzyme necessary for reduction of 1 ␮mol min of 2,5-DKG, which corre-
◦
+
sponds to the production of 1 ␮mol of NADP per min under assay conditions. At high
by the Bio Controller ADI 1030 (Applikon). Samples were taken at 1-h interval to
measure OD600, dry cell weight (DCW), glycerol and lactose concentration, protein
concentration and 2,5-DKG reductase activity.
substrate concentrations the measurement of the enzymatic activity by absorbance
was unreliable by spectral interference and was therefore verified by a fluoromet-
ric assay. This assay, which is not interfered by high concentrations of 2,5-DKG, is
based on the oxidation of the fluorescent coenzyme species NADPH (Ex = 380 nm,
Em = 480 nm). The optimal excitation and emission wavelengths were determined
in ammonium acetate buffer pH 6.4 and in the presence of 2,5-DKG. The kinetics of
purified enzyme was assayed in 50 mM ammonium acetate buffer at pH 6.4 and at
2.5. Protein purification
An ÄKTA Explorer 100 (GE Healthcare) was used for 2,5-DKG reductase purifica-
◦
tion from E. coli cell-free lysate. An XK 26/20 column packed with 60 mL Ni-charged
Chelating Sepharose Fast Flow (particle size 45–165 ␮m, GE Healthcare) was pre-
equilibrated with 10 column volumes of loading buffer (0.02 M sodium phosphate,
pH 7.4 containing 0.25 M NaCl and 0.01 M imidazole). Then the crude extract was
25 C in a final volume of 200 ␮L. Assays were carried out in a VARIAN Cary Eclipse
instrument and 3 mm path length quartz cuvettes. To determine steady-state kinetic
constants the 2,5-DKG concentration and the NADPH concentration were varied
within 2–100 mM and 0.002–0.3 mM, respectively. The data obtained for the for-
ward reaction in the absence of product were fitted to the kinetic rate equation for
an irreversible, ordered bi-bi mechanism [31]. Product inhibition by 2-KLG and the
−1
applied to the column at a linear flow rate of 1.06 cm min . Unbound proteins were
washed out with loading buffer until a constant UV (280 nm) and conductivity base-
line were obtained. The enzyme was eluted by a linear gradient from 0% to 100% of
elution buffer containing 0.5 M NaCl and 0.5 M imidazole over 10 column volumes
and the eluate was collected in fractions of 2.5 mL. Fractions with the highest specific
activity were pooled, concentrated and diafiltrated with Amicon Ultra Centrifugal
Filter Device with a 10 kDa cut-off membrane (Millipore; Billerica, MA, USA). The
concentrated, imidazole free enzyme solution in 50 mM Bis–Tris buffer, pH 6.4 was
+
oxidized coenzyme NADP were also determined. For these experiments one sub-
strate concentration was varied at several fixed concentrations of a single product,
the other substrate concentration was kept constant at a non-saturated concentra-
tion according to Segel [31]. All measurements were performed in triplicates and
nonlinear least squares regression of unweighted averages of data was calculated
with SigmaPlot 12 (Systat Software, San Jose, and CF, USA). Experimental data are
shown as double-reciprocal Linewaever–Burk plots. Global fits of the data to various
single-substrate single-inhibitor models were performed using the Enzyme Kinetics
Wizard of SigmaPlot 12. All available inhibition models, including full competitive,
full noncompetitive and hyperbolic mixed-type inhibition [31] were evaluated and
−1
◦
diluted to a protein concentration of 20 mg mL and stored at 4 C until further use.
2.6. Protein analysis
2
The protein concentration was determined by the dye-binding method of Brad-
ranked according to R values and the Akaike Information Criteria (AIC).
ford [30], using the Bio-Rad Protein Assay Kit (Bio-Rad, Hercules, CA) with bovine
serum albumin as a standard. SDS-PAGE as well as isoelectric focusing was carried
out with a PhastSystem unit (Pharmacia LKB/GE Healthcare Life Sciences, Uppsala,
Sweden) according to the manufacturer’s instructions. SDS-PAGE was performed
with precast gels (PhastGel Gradient) with continuous acrylamide gradients in
the ranges 8–25%, using the High Precision Dual Color (10–250 kDa, Bio-Rad) as
molecular weight marker. Isoelectric focusing (IEF) was performed with precast
homogeneous (5% T, 3% C) polyacrylamide PhastGel IEF gels with immobilized pH
gradient (pH 3.0–9.0). Marker-proteins used to determine pI values were in a range
of 3.5–10.7 (Serva Electrophoresis, Heidelberg, Germany). Bands on the SDS-PAGE
and IEF Gels were visualized by silver staining according to the manufacturer’s
instructions.
2.8. pH dependence of 2,5-DKG reductase activity
The effect of pH on 2,5-DKG reductase activity was determined in the range of
pH 4.5–9.5, using the following 50 mM buffers: sodium acetate (pH 4.5–6.5), ammo-
nium acetate (pH 4.5–6.5), sodium citrate (pH 4.5–6.5), Bis–Tris (pH 6.0–7.5), MES
(5.7–6.7), MOPS (6.7–7.9), tricine (7.7–8.7) and borate (pH 7.5–9.5). Measurements
were performed in triplicates using the spectrophotometric assay.

V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
2015
hansenii (1.5 mg mL 1) and Bacillus funiculus (0.9 mg mL ). In addi-
tion to HPLC, the identity of accumulated 2-KLG was confirmed by
conversion to l-ascorbic acid.
−
−1
2
.9. Effect of ions on enzymatic activity
To detect inhibitors or activators of 2,5-DKG reductase activity, enzyme samples
were assayed in the presence of various salts. The enzymatic activity was measured
by the photometric assay at 25 C and varying amounts of monovalent and diva-
◦
lent cations in the chloride, sulphate or acetate form. Fitting curves were calculated
by nonlinear least square regression to a single exponential function (SigmaPlot
3.2. Heterologous overexpression of dkr from C. glutamicum
1
2). The inhibition kinetics of NaCl was examined in detail. Therefore, to assess
the mode of inhibition with respect to NADPH, NaCl concentrations were varied
0–100 mM) in the presence of sub-saturating concentration of 2,5-DKG (23 mM)
The 2,5-DKG reductase coding gene from C. glutamicum was
(
amplified by a single PCR, cloned and expressed in E. coli BL21 (DE3),
the full procedure is described in a patent [28]. The nucleotide
sequence of full length dkr (GenBank accession number: JQ407590)
contains an open reading frame of 846 bp, encoding 281 amino
acids residues and a stop codon, and has 51.4% content of G + C
base pairs. The amino acid sequence of C. glutamicum 2,5-DKG
reductase shares a sequence identity of 74% with putative 2,5-
DKG reductase from C. efficiens YS-314 (GenBank accession number
NP 738878). 62% with 2,5-DKG reductase A from Corynebacterium
sp. ATCC 31090 (GenBank accession number P06632), 40% with
and increasing concentration of NADPH (0.002–0.12 mM). The effect of NaCl on 2,5-
DKG turnover was measured in the presence of 0.2 mM NADPH and varying amounts
of salts (0–300 mM) and 2,5-DKG (1.5–60 mM), respectively. As explained above, by
fitting the initial reaction rates to the to the corresponding equations of single-
substrate single-inhibitor models using the SigmaPlot software the observed data
were analyzed using nonlinear least-square regression.
2.10. Temperature dependence of activity and stability
The temperature dependence of enzyme activity was determined by incubat-
ing the reaction mixture (1 mL) that consisted of 50 mM ammonium acetate buffer,
◦
◦
◦
pH 6.4 and 8.256 mM 2,5-DKG at temperatures from 20 C to 70 C (5 C intervals)
by starting the reaction with an aliquot of the enzyme and 20 ␮L of NADPH stock
solution. For the determination of the energy of activation (Ea) Arrhenius plots was
prepared from the initial rates obtained in the temperature experiments. For ther-
2
,5-DKG reductase B from the mutant strain Corynebacterium sp.
SHS752001 (P15339), 51% with 2,5-DKG reductase A from E. coli
(YghZ; AKR14A1; Q46857) and 42% with 2,5-DKG reductase B from
E. coli (YafB; P30863). The low sequence identities even within one
genus suggest that 2,5-DKG reductases from Corynebacterium spp.
might differ in their physical and catalytic properties.
−1
mal stability experiments, enzyme solution (2 U mL ) in 50 mM ammonium acetate
◦
◦
buffer, pH 6.4 was incubated at various temperatures from 4 C to 50 C. The enzyme
activity was measured at specified intervals using the standard spectrophotomet-
ric assay. The inactivation of enzyme was analyzed assuming first-order kinetics,
kt
For overexpression in E. coli an expression vector containing
dkr under the control of a lactose- or IPTG-inducible T7 promoter
A = A0 × e , where A and A0 stand for enzyme activity at various time points (t) and
initial enzyme activity, respectively. The inactivation is represented by kin. The half-
life of the enzyme activity t1/2, defined as the time needed for the activity of enzyme
to drop to one-half of the original value, was determined as t1/2 = ln 2/kin.
was fused in frame with a C-terminal His -tag. The resulting vec-
6
tor was transformed into the production strain BL21 Star (DE3), and
ampicillin-resistant clones were tested for the presence of 2,5-DKG
reductase activity by spectrophotometric assays and batch conver-
sion of 2,5-DKG using the raw extract after small-scale cultivation.
The best clone was selected for enzyme production. To maximize
the heterologous expression, different media, temperatures and
inducer concentrations were evaluated in shaking flasks. The max-
2.11. High performance liquid chromatography
2
,5-DKG and 2-KLG were separated and quantified by HPLC with UV detection
at a wavelength of 210 nm (PDA-100 Photodiode Array Detector), using a Dionex DX
00 system (Dionex; Sunnyvale, CA, USA) and a Luna Amino column (Phenomenex,
5
◦
−1
Torrance, California) at 25 C with 15 mM NH4H2PO4 as eluent (1 mL min ).
◦
imum volumetric activity for all three media was obtained at 25 C
3
. Results and discussion
−
1
and the highest 2,5-DKG reductase activity (0.1 U mg ) produced
by E. coli BL21 Star (DE3) carrying pET-21d/dkr was obtained after
3.1. Screening
−
1
◦
2
6 h of induction with 1 g L lactose in MCHGly medium at 25 C.
The production of 2,5-DKG reductase under strictly controlled
growth conditions, was performed in a 30-L stirred bioreactor
Twenty-two bacterial strains were screened in shaking flask
cultures for the ability to reduce 2,5-DKG to 2-KLG (Table 1). Dif-
(
Supplementary Fig. S1). The recombinant enzyme production was
◦
ferent optimal growth conditions (22–37 C, 6 different media,
◦
performed in MCHGly medium at 25 C and at an oxygen concen-
tration of 20% air saturation. Induction with 5 g L
pH 6.0–7.3, for detailed information see: Supplementary Tables
S1 and S2) resulted in different specific growth rates. Therefore,
the cultivation time to reach maximum cell densities was deter-
mined in preliminary experiments (20–50 h) and three classes of
organisms were defined, which were harvested after 24, 42 or 47 h
of cultivation. Still, the obtained biomass differed vastly between
OD600 = 13.5 after 24 h and OD600 = 1.0 after 47 h. The enzymatic
activity was measured in the lysates of the harvested cells and
no correlation between biomass and activity was found. All tested
strains showed at least minor activity, but overlapping substrate
specificities of reductases and dehydrogenases in microbial cells
−
1
lactose was
started at the beginning of the exponential growth phase of the
−
1
inoculum (OD600 = 0.5). The specific growth rate was 0.2 h dur-
ing the exponential growth phase and it is noteworthy that the
enzyme production stopped within the growth phase with 70% of
−
1
the glycerol left. Also, the lactose concentration after 21 h (∼2 g L
)
was still sufficient for induction. The time-course of the enzymatic
activity showed a plateau after 16 h of cultivation with a maximal
volumetric activity of 180 U L after 21 h. By using the specific
activity of the homogeneous enzyme a calculated concentration
−
1
−
1
of 409 mg L
2,5-DKG reductase was as active, soluble protein
[
1] rendered these low measured activities ambiguous. To rescreen
(
no inclusion bodies were observed). Compared to other reported
the obtained results, unambiguous conversion experiments of 2,5-
DKG to 2-KLG and further to l-ascorbic acid were performed with
the crude extracts. HPLC analysis shows that only 9 of the 22
strains actually formed the desired product (2-KLG). It was found
that even some strains with high enzymatic activities did not pro-
duce 2-KLG in the small-scale batch conversions, which verifies
the assumption of interfering reductase/dehydrogenase activities.
The presence of a 2,5-DKG reductase was confirmed for the genera
Acetobacter, Bacillus, Brevibacterium, Corynebacterium, Gluconobac-
ter, Kocuria and Pseudomonas. The highest amount of 2-KLG was
−
1
results, e.g. 50 U L for 2,5-DKG reductase from Corynebacterium
sp. in E. coli (pQE-82L-dkr) on modified Luria broth (with 3% glu-
cose) with 0.1 mM IPTG as inducer [32], the obtained expression
yield is quite high.
3.3. Enzyme purification and physical properties
The purification of His -tagged 2,5-DKG reductase consisted of
6
a Ni-immobilized metal affinity chromatographic step followed
by diafiltration, which resulted in an apparently homogeneous
preparation by means of SDS-PAGE (Supplementary Fig. S2A). The
enzyme was purified 5.5-fold from the lysate with a yield of 68.2%
−
1
detected in the crude extract of C. glutamicum (2 mg mL ), which
was the reason for its overexpression and characterization. Good
−
1
results were also obtained with C. barkeri (1 mg mL ), Acetobacter

2
016
V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
Fig. 2. Effect of pH on recombinant 2,5-DKG reductase activity. The buffers used
were 50 mM sodium acetate (circles), 50 mM ammonium acetate (squares), 50 mM
sodium citrate (diamonds), 50 mM Bis–Tris (triangles up), 50 mM MES (crosses),
5
0 mM MOPS (stars), 50 mM tricine (crosshairs) and 50 mM borate buffer (horizontal
marks).
and a specific activity of 0.44 U mg−1 was obtained. SDS-PAGE of
recombinant 2,5-DKG reductase resulted in a single band at 30 kDa,
which compares well to the calculated molecular mass of the
enzyme (31.033 kDa). Native-PAGE (data not shown) gave also a
molecular mass of approx. 30 kDa, which indicates a monomeric
structure of the active enzyme. This is in agreement with other
reported 2,5-DKG reductases from various microorganisms from
Corynebacterium spp. [7,8], E. coli [9] and an uncultured organ-
ism [11]. The isoelectric point of recombinant 2,5-DKG reductase
was determined by isoelectric focusing under native conditions
and one compact band was observed at pH 5.3 (Supplementary
Fig. S2B), which is noticeably higher than for 2,5-DKG reductase A
from Corynebacterium sp. (4.4) [7]. The pI of C. glutamicum 2,5-DKG
reductase calculated from its amino acid sequence including the
Fig. 3. Temperature-dependent deactivation of recombinant 2,5-DKG reductase. (A)
◦
◦
◦
◦
In absence of NADPH at 4 C (circles), 20 C (diamonds), 25 C (stars), 30 C (triangles
His -tag is 5.36, which matches the experimental result very well.
6
◦
up) and 35 C (triangles down). The inset shows the deactivation in presence (grey
◦
symbols) and absence (black symbols) of NADPH at 25 (stars) and 30 C (triangles
3.4. Effects of pH and temperature on 2,5-DKG reductase activity
up). (B) In absence of NADPH at 40 ◦C (crosshairs) and 50 ◦C (crosses). The inset
shows the deactivation in presence (grey symbols) and absence (black symbols) of
◦
NADPH at 40 C (crosshairs).
Recombinant 2,5-DKG reductase exhibited a bell shaped pH
profile with pH optima between 6.2 and 7.0 depending on the
buffer used (Fig. 2). The enzymatic activity is little influenced by
different buffers. The enzyme exerts more than 50% of its max-
imum activity over a pH range from 5.75 to 8.0, which favours
the coenzyme regeneration in a biocatalytic process. The obtained
values compare well with reported data for 2,5-DKG reductases
from Corynebacterium sp. [7,8] whereas for the 2,5-DKG reductases
from uncultured organisms [11] a preference for more acidic pH
values was found. For the biocatalytic synthesis of 2-KLG ammo-
nium acetate is considered to be a convenient buffer system, which
was further used to determine the catalytic properties at pH 6.4.
Its pH optimum favours coenzyme regeneration in a biocatalytic
process.
deactivation of the enzyme was observed in the presence of NADPH.
2
3
3
,5-DKG reductase showed first-order inactivation kinetics below
◦
◦
◦
◦
5 C. The enzyme is very stable at 4 C and 25 C. At 30 C and
◦
5 C, 40% of enzymatic activity was lost after 15 and 7 days,
respectively. The half-life of the enzymatic activity (t ) was cal-
1/2
◦
−1
◦
culated to be 462 h at 30 C (k = 0.0015 h ), and 217 h at 35 C
kin = 0.0032 h ). The inactivation at 40 C and 50 C showed two
in
−
1
◦
◦
(
distinct phases (both of first-order): a first phase of relatively rapid
−
1
◦
−1
◦
inactivation (kin = 3.44 h at 40 C and k = 23.56 h at 50 C) and
in
a slower second phase for the remaining 20% of enzymatic activ-
−
1
◦
−1
◦
ity (kin = 1.65 h at 40 C and k = 4.91 h at 50 C). These data
in
◦
are in agreement with thermal denaturation studies by differen-
tial scanning calorimetry (DSC, Fig. S3A), which found a thermal
The temperature optimum was only 35 C in 50 mM ammo-
nium acetate buffer, pH 6.4. The linear region of the Arrhenius
plot was assumed between 293 K and 308 K from the initial rates
of activity measurements at various temperatures. The activation
◦
transition midpoint temperature (Tm) of 39 C, at which both the
folded and unfolded states are equally populated. This compares
◦
−
1
well to 2,5-DKG reductase A from Corynebacterium sp. (Tm = 38 C)
energy (Ea) for the substrate was calculated to be 34 ± 1.1 kJ mol
.
◦
and 2,5-DKG reductase B (Tm = 32 C) from Corynebacterium sp. [22].
To investigate the thermal stability in more detail, the enzymatic
In contrast to the above described time-dependent inactivation
activity was measured over time at 4, 20, 25, 30, 35 (Fig. 3A),
◦
◦
studies, a higher Tm (44 C, Fig. S3B) was observed for C. glutam-
4
6
0 and 50 C (Fig. 3B) in 50 mM ammonium acetate buffer, pH
.4 without NADPH. Additionally, stability measurements in the
icum 2,5-DKG reductase in the presence of NADPH. This shows that
the enzyme is stabilized against thermal unfolding by the bound
coenzyme.
presence of 200 ␮M NADPH were conducted at 25, 30 (Fig 3A,
◦
inset) and 40 C (Fig 3B, inset). No difference in the time-dependent

V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
2017
3.5. Effect of ions on 2,5-DKG reductase activity
Very few studies with respect to the effect of various ions
2+
have been done. Two reports indicate that metal ions Cu and
2+
Ni inhibit 2,5-DKG reductases strongly from Corynebacterium sp.,
2
+
2+
2+
whereas Ca , Mg and Mn have little or no noticeable effect.
However, the tested concentrations were below 1 mM [7,8]. We
measured the effects of various ions on 2,5-DKG reductase activ-
ity to avoid inhibition in enzymatic or bioconversion experiments.
The enzymatic activity was not influenced by the monovalent
cations Na+ and K , while bivalent Mg and Ca cations inhibit
+
2+
2+
2+
2+
the enzyme severely (I50 Mg = 97 mM, I50 Ca = 141 mM). The
acetate anion used in these experiments (Fig. 4A) exerts no inhibit-
ing effect. Contrary to the acetates, for all sulphate- (Fig. 4B) and
chloride salts (Fig. 4C) a strong inhibition was found. By inspection
of the curves and I50 values (Na SO = 215 mM, K SO = 124.5 mM,
2
4
2
4
MgSO = 46 mM, NaCl = 119 mM, KCl = 119 mM, MgCl = 20 mM,
4
2
CaCl = 32.5 mM (all data fit a very well to a single exponential equa-
2
2
tion, R > 0.98)) the following conclusions can be drawn: (1) the
+
+
acetate anion and monovalent Na and K cations have no effect on
the activity of 2,5-DKG reductase, (2) bivalent Mg2 and Ca cations
show a strong inhibition on 2,5-DKG reductase activity, (3) chlo-
ride is a more potent inhibitor than sulphate, the first has the same
inhibitory effect as bivalent cations, the second a slightly lower, (4)
the inhibitory effect of cations and anions is additive.
+
2+
The I50 values show that extremely negative effects can be
expected in a bioconversion process with the wrong buffer. To
the best of our knowledge, inhibition of 2,5-DKG reductase activ-
ity by bivalent cations and monovalent and trivalent anions is a
new fact has not been reported so far, but is of high importance
for its biocatalytic application in the production of l-ascorbic acid
via 2-KLG. To investigate the inhibition mechanism in more detail,
chloride, the most severe inhibitor was chosen for further stud-
ies. The inhibition pattern (by using NaCl) was obtained for NADPH
and 2,5-DKG turnover. The analysis of the Lineweaver–Burk plot for
NADPH kinetics shows an increased apparent Km, while Vmax is not
altered. A series of plots with the same y intercept (1/Vmax) and the x
intercepts (−1/Km) closer and closer to the origin as NaCl increases,
indicates that NaCl acts as full competitive inhibitor, which affects
the binding of NADPH (Fig. 5A). The dissociation constant for NaCl
(
K ) estimated from this and the Dixon plot is 7.4 ± 0.7 mM. With
i
respect to 2,5-DKG a full noncompetitive inhibition was observed
(
Fig. 5B), which implies that the chloride ion directly interferes with
the binding of the coenzyme but not with the substrate binding.
The inhibitor dissociation constant (K ) and the inhibitor dissocia-
ic
tion constant of the enzyme–substrate–inhibitor complex (Kiu) are
equal (263 ± 23 mM). These inhibition patterns are indicative of a
bi–bi ordered mechanism with NADPH binding first [31,33]. The
inhibitory effect of other ions is likely to follow a similar pattern.
By sequence alignment and homology modeling with the deposited
structure of 2,5-DKGR-A (accession number 1A80) it was found that
Asp45, Lys232, Arg238 and Glu241 (numbering according to the
template 1A80) are involved in the binding of the adenine moiety,
ꢀ
the ribose, the phosphate group on the 2 position of the ribose ring
Fig. 4. Effect of different ions on recombinant 2,5-DKG reductase activity. The effect
of the salts of acetic acid (A), sulfuric acid (B) and hydrochloric acid (C) were eval-
uated. A dotted line connects non-fitted data sets, short dashes indicate a single
exponential fit. Sodium (circles), potassium (squares), calcium (triangles up) and
magnesium (diamonds).
and the pyrophosphate. We propose that these amino acids will be
targeted by the investigated ions.
3.6. Catalytic properties
To evaluate the enzyme’s capacity for 2,5-DKG conversion
under suitable for industrial process conditions, the catalytic
constants were measured in 50 mM ammonium acetate buffer at
pH 6.4. Neither ammonium nor acetate ions inhibit the enzymatic
activity. The kinetic measurements had to be performed with a flu-
orometric enzyme assay (in triplicates) because of the significant
absorption of the substrate 2,5-DKG at 340 nm. At lower 2,5-DKG
concentrations these results were verified with the spectropho-
tometric assay, both methods gave similar results. The derived
Michaelis constants for NADPH (Ka) is 0.014 ± 0.003 mM, which
is very low and allows to use a low concentration of the coen-
zyme in the process (50–70 ␮M). The Michaelis constant for the
substrate 2,5-DKG (K ) is 23.1 ± 3.8 mM and the turnover number
b

2
018
V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
Fig. 5. (A) Lineweaver–Burk plot for the inhibition of recombinant 2,5-DKG reductase by NaCl with respect to NADPH. Without inhibitor (circles), 10 mM NaCl (squares),
0 mM NaCl (diamonds), 50 mM NaCl (triangles up), and 100 mM NaCl (triangles down) at a 23 mM 2,5-DKG concentration. (B) Lineweaver–Burk plot for the inhibition of
3
recombinant 2,5-DKG reductase by NaCl with respect to 2,5-DKG. Without inhibitor (circles), 10 mM NaCl (squares), 30 mM NaCl (diamonds), 100 mM NaCl (triangles up),
and 300 mM NaCl (triangles down) at a 0.2 mM NADPH concentration. (C) Lineweaver–Burk plot for the inhibition of recombinant 2,5-DKG reductase by 2-KLG with respect
to NADPH. Without inhibitor (circles), 10 mM 2-KLG (squares), 30 mM 2-KLG (diamonds), 100 mM 2-KLG (triangles up), and 300 mM 2-KLG (triangles down) at a 23 mM
2
2
,5-DKG concentration. (D) Lineweaver–Burk plot for the inhibition of recombinant 2,5-DKG reductase by 2-KLG with respect to 2,5-DKG. Without inhibitor (circles), 10 mM
-KLG (squares), 30 mM 2-KLG (diamonds), 100 mM 2-KLG (triangles up), 300 mM 2-KLG (triangles down) at a 0.08 mM NADPH concentration.
(
kcat) is 1.87 ± 0.18 s−1. When assuming a 2,5-DKG concentration
chromatography. Its physical and catalytic properties make the
enzyme ideally suited for the application in the synthesis of 2-KLG,
a precursor of l-ascorbic acid production. The investigated inhibi-
tion mechanism of 2,5-DKG reductase by anions and cations is of
high relevancy for further process engineering. Having the recom-
binant enzyme in hands, powerful protein engineering tools can
be used to further optimize the enzyme for its application in the
production of ascorbic acid.
of 300 mM, most of the substrate (80%) can be converted before
reaching the rate limiting concentration below 60 mM (3 × K ). The
b
product 2-KLG was found to be an inhibitor of the forward reaction.
Analysis of the results by double-reciprocal plots (Fig. 5C and D)
indicates that 2-KLG is a hyperbolic mixed inhibitor [31] with
respect to NADPH (Ki2-KLG;NADPH = 110 ± 28 mM; ˛ = 3.02 ± 0.82;
ˇ = 0.44 ± 0.12) and 2,5-DKG (K
= 375 ± 26 mM;
i2-KLG;2,5-DKG
˛
= 0.20 ± 0.085; ˇ = 0.66 ± 0.038). The factor
˛
modifies the
Michaelis–Menten constant, whereas the factor ˇ modifies the
maximal velocity for a given substrate in the presence of an
inhibitor (2-KLG). Ki2-KLG is the equilibrium dissociation constant
of 2-KLG. The determined constants show that inhibition under
process conditions is low (concentrations of 2,5-DKG above 10 mM
and NADPH above 20 ␮M) and affects the conversion rate only
Acknowledgements
The authors thank Mr. Dominik Jeschek for superb DSC mea-
surements and MSc. Shima Khazaneh for careful reading and
discussions.
+
minimally. Inhibition by the oxidized coenzyme NADP was not
observed.
Appendix A. Supplementary data
4
. Conclusions
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.procbio.
2012.07.014.
2
,5-DKG reductase from C. glutamicum can be recombinantly
produced by E. coli in high yields and easily purified by affinity

V. Kaswurm et al. / Process Biochemistry 47 (2012) 2012–2019
2019
References
[17] Barski OA, Tipparaju SM, Bhatnagar A. The aldo-keto reductase superfam-
ily and its role in drug metabolism and detoxification. Drug Metab Rev
2
008;40:553–624.
[
[
[
1] Ellis EM. Microbial aldo-keto reductases. FEMS Microbiol Lett 2002;
16:123–31.
2] Jez JM, Penning TM. The aldo-keto reductase (AKR) superfamily: an update.
Chem Biol Interact 2001;130–132:499–525.
3] Anderson S, Marks CB, Lazarus R. Production of 2-keto-l-gulonate, an inter-
mediate in l-ascorbate synthesis, by a genetically modified Erwinia herbicola.
Science 1985;230:144–9.
[
18] Hancock RD, Viola R. Biotechnological approaches for l-ascorbic acid produc-
tion. Trends Biotechnol 2002;20:299–305.
2
[
19] Sonoyama T, Tani H, Matsuda K, Kageyama B, Tanimoto M, Kobayashi K, Yagi S,
Kyotani H, Mitsushima K. Production of 2-keto-l-gulonic acid from d-glucose
by two-stage fermentation. Appl Environ Microbiol 1982;43:1064–9.
20] Aiguo J, Peiji G. Synthesis of 2-Keto-l-gulonic acid from gluconic acid by co-
immobilized Gluconobacter oxydans and Corynebacterium sp. Biotechnol Lett
[
[
[
[
[
[
[
4] Bremus C, Herrmann U, Bringer-Meyer S, Sahm H. The use of microorganisms
in l-ascorbic acid production. J Biotechnol 2006;124:196–205.
5] Sonoyama T, Kageyama B, Honjo T. Process for producing 2-keto-l-gulonic acid.
US Patent US 3922194 (75.11.25).
6] Sonoyama T, Tani H, Kageyama B, Kobayashi K, Honjo T, Yagi S. Process for
producing 2-keto-l-gulonic acid. US Patent. US 3963574 (76.06.15).
7] Miller J, Estell D, Lazarus R. Purification and characterization of 2,5-diketo-
1
998;20:939–42.
21] Grindley J, Payton M, Van de Pol H, Hardy K. Conversion of glucose to 2-keto-
l-gulonate, an intermediate in l-ascorbate synthesis, by a recombinant strain
of Erwinia citreus. Appl Environ Microbiol 1988;54:1770–5.
22] Powers DP. Structure/function studies of 2,5-diketo-d-gluconic acid reduc-
tases. USA: University of Medicine and Dentistry of New Jersey; Ph.D. thesis;
1
996.
d-gluconate reductase from Corynebacterium sp.
016–20.
J Biol Chem 1987;262:
[
23] Boston GM, Swanson BA. Method for producing ascorbic acid intermediates.
US Patent. US 6599722B2 (03.07.29).
24] Banta S, Boston M, Jarnagin A, Anderson S. Mathematical modeling of in vitro
enzymatic production of 2-keto-l-gulonic acid using NAD(H) or NADP(H) as
cofactors. Metab Eng 2002;4:273–84.
25] Zhang J, Liu J, Shi Z, Liu L, Chen J. Manipulation of B. megaterium growth for
efficient 2-KLG production by K. vulgare. Process Biochem 2010;45:602–6.
26] Hancock RD. Recent patents on vitamin C: opportunities for crop improve-
ment and single-step biological manufacture. Recent Pat Food Nutr Agric
9
[
8] Sonoyama T, Kobayashi K. Purification and properties of two 2,5-diketo-d-
gluconate reductases from a mutant strain derived from Corynebacterium sp. J
Ferment Technol 1987;65:311–7.
[
[9] Yum D, Lee B, Pan J. Identification of the yqhE and yafB genes encoding two
[
2
,5-diketo-d-gluconate reductases in Escherichia coli. Appl Environ Microbiol
1
999;65:3341–6.
[
[
[
10] Habrych M, Rodriguez S, Stewart J. Purification and identification of an
Escherichia coli ␤-keto ester reductase as 2,5-diketo-d-gluconate reductase
YqhE. Biotechnol Prog 2002;18:257–61.
11] Eschenfeldt WH, Stols L, Rosenbaum H, Khambatta ZS, Quaite-Randall E, Wu
S, Kilgore DC, Trent JD, Donnelly MI. DNA from uncultured organisms as
a source of 2,5-diketo-d-gluconic acid reductases. Appl Environ Microbiol
2
010;1:39–49.
[
[
[
[
27] Reichstein T. Eine ergiebige Synthese der l-Ascorbinsäure (C Vitamin). Helv
Chim Acta 1934;17:311–28.
28] Pacher C, Kulbe KD, Steiner E, Rembart G. Method for producing ascorbic acid
using Pectobacter cypripedii. WIPO Patent. WO/2008/144792 (08.12.04).
29] Sambrook J, Russell DW. Molecular cloning: a laboratory manual. New York:
Cold Spring Harbor Laboratory Press; 2001. p. A2.2.
30] Bradford MM. A rapid and sensitive method for the quantitation of micro-
gram quantities of protein utilizing the principle of protein-dye binding. Anal
Biochem 1976;72:248–54.
31] Segel IH. Enzyme kinetics: behavior and analysis of rapid equilibrium and
steady-state enzyme systems. New York: John Wiley & Sons. Inc; 1975.
32] Osterath B. Prozessentwicklung zur Produktion von 2-Keto-l-Gulonsäure, einer
Vitamin C-Vorstufe. Germany: University of Bonn; Ph.D. thesis; 2009.
33] Cleleand WW. The kinetics of enzyme-catalyzed reactions with two or more
substrates or products. I. Nomenclature and rate equations. Biochim Biophys
Acta 1963;67:104–37.
2
001;67:4206–14.
[
12] Khurana S, Powers DB, Anderson S, Blaber M. Crystal structure of 2,5-diketo-
d-gluconic acid reductase A complexed with NADPH at 2.1-Å resolution. Proc
Natl Acad Sci USA 1998;95:6768–73.
13] Khurana S, Sanli G, Powers D, Anderson S, Blaber M. Molecular modeling of sub-
strate binding in wild-type and mutant Corynebacteria 2,5-diketo-d-gluconate
reductases. Proteins Struct Funct Genet 2000;39:68–75.
14] Sanli G, Blaber M. Structural assembly of the active site in an aldo-keto reduc-
tase by NADPH cofactor. J Mol Biol 2001;309:1209–18.
15] Sanli G. Structural alteration of cofactor specificity in Corynebacterium 2,5-
diketo-d-gluconic acid reductase. Protein Sci 2004;13:504–12.
[
[
[
[
[
[
[
16] Jez J, Bennett M, Schlegel B, Lewis M, Penning T. Comparative anatomy of the
aldo-keto reductase superfamily. Biochem J 1997;326:625–36.