Characterization and identification of three novel aldo-keto reductases from Lodderomyces elongisporus for reducing ethyl 4-chloroacetoacetate

Article abstract of DOI:10.1016/j.abb.2014.10.007

Lodderomyces elongisporus LH703 isolated from soil samples contained three novel aldo-keto reductases (AKRs) (LEAKR 48, LEAKR 49, and LEAKR 50). The three enzymes were cloned, expressed, and purified to homogeneity for characterization. These three AKRs shared <40% amino acid identity with each other. LEAKR 50 was identified as a member of AKR3 family, whereas the other two LEAKRs were identified as members of two novel AKR families, respectively. All the three AKRs required nicotinamide adenine dinucleotide phosphate as a cofactor. However, they showed diverse characteristics, including optimum catalyzing conditions, resistance to adverse reaction conditions, and substrate specificity. LEAKR 50 was estimated to be a promising biocatalyst that could reduce ethyl 4-chloroacetoacetate with high enantiomeric excess (98% e. e.) and high activity residue under adverse conditions.

Full text of DOI:10.1016/j.abb.2014.10.007

Archives of Biochemistry and Biophysics 564 (2014) 219–228
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/locate/yabbi
Characterization and identification of three novel aldo–keto reductases
from Lodderomyces elongisporus for reducing ethyl 4-chloroacetoacetate
a,
Chenxi Ning a, Erzheng Su b, , Dongzhi Wei
⇑
⇑
a State Key Laboratory of Bioreactor Engineering, New World Institute of Biotechnology, East China University of Science and Technology, Shanghai 200237, PR China
b Enzyme and Fermentation Technology Laboratory, College of Light Industry Science and Engineering, Nanjing Forestry University, Nanjing 210037, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 17 July 2014
and in revised form 9 October 2014
Available online 22 October 2014
Lodderomyces elongisporus LH703 isolated from soil samples contained three novel aldo–keto reductases
(AKRs) (LEAKR 48, LEAKR 49, and LEAKR 50). The three enzymes were cloned, expressed, and purified to
homogeneity for characterization. These three AKRs shared <40% amino acid identity with each other.
LEAKR 50 was identified as a member of AKR3 family, whereas the other two LEAKRs were identified
as members of two novel AKR families, respectively. All the three AKRs required nicotinamide adenine
dinucleotide phosphate as a cofactor. However, they showed diverse characteristics, including optimum
catalyzing conditions, resistance to adverse reaction conditions, and substrate specificity. LEAKR 50 was
estimated to be a promising biocatalyst that could reduce ethyl 4-chloroacetoacetate with high enantio-
meric excess (98% e. e.) and high activity residue under adverse conditions.
Keywords:
Lodderomyces elongisporus
Aldo–keto reductase
Characterization
Identification
Ethyl 4-chloroacetoacetate
Ó 2014 Elsevier Inc. All rights reserved.
Introduction
Use of biocatalysts for manufacturing chiral chemicals offers
major advantages in terms of enhanced reaction selectivity,
Aldo–keto reductases (AKRs)1 belong to a growing oxidoreduc-
tase superfamily of proteins that bind to cofactor nicotinamide ade-
nine dinucleotide phosphate (NADPH) without a Rossmann-fold
motif [13]. They are present in various mammals, plants, yeast,
and bacteria. Many mammalian AKRs are potential therapeutic tar-
gets. Moreover, structure-based drug designing has produced com-
pounds with desired specificity and clinical efficacy [14]. Microbial
AKRs catalyze the reduction of diverse substrates, including aliphatic
and aromatic aldehydes, monosaccharides, steroids, prostaglandins,
polycyclic aromatic hydrocarbons, and isoflavonoids, with excellent
enantioselectivity and yield [5,20,21]. The resulting alcohols repre-
sent chiral building blocks that act as key intermediates of pharma-
ceutical and agricultural products [4]. Optically pure secondary
alcohols have especially drawn a great deal of attention in introduc-
ing chiral groups into products [9].
reduced cost of raw materials, lowered energy cost, improved
safety, and production sustainability [8]. Several enzymes are
known to be attractive options for the industry. Of these, microbial
AKRs have the potential to serve as versatile tools for producing
various valuable optically pure compounds. Therefore, discovery
of new AKRs is of a significant value.
In this study, we discovered, cloned, heterologously expressed,
and characterized three novel AKR genes from Lodderomyces elon-
gisporus LH703. These three AKR genes (LEAKR 48, LEAKR 49, and
LEAKR 50) shared <50% identity in their gene sequences and <40%
in their amino acid sequences. Although each AKR belonged to dif-
ferent AKR family, they shared the same conserved residues. Two
of these AKRs (LEAKR 48 and LEAKR 49) were identified as members
of novel AKR families. All the three AKRs showed different thermo-
stability, substrate specificity, and enantioselectivity. Detailed
sequence analysis was used to investigate the different character-
istics resulting from sequence diversity and to open the gate for
AKR gene modification.
⇑
Corresponding authors. Tel.: +86 25 64258906; fax: +86 25 85428906 (E. Su).
Tel.: +86 21 64252078; fax: +86 21 64250068 (D. Wei).
Materials and methods
E-mail addresses: ezhsu@njfu.edu.cn (E. Su), dzhwei@ecust.edu.cn (D. Wei).
Abbreviations used: AKRs, aldo–keto reductases; NADPH, nicotinamide adenine
1
dinucleotide phosphate; GC, gas chromatography; HPLC, high performance liquid
chromatography; PCR, polymerase chain reaction; SDS–PAGE, sodium dodecyl sulfate
polyacrylamide gel electrophoresis; MALDI-TOF-MS, Matrix-Assisted Laser Desorp-
tion/Ionization Time of Flight Mass Spectrometry; GDH, D-glucose dehydrogenase;
LEAKRs, L. elongisporus that encoded AKRs.
Chemicals and enzymes
Prochiral ketones and (R)- or (S)-alcohols used in this
study were purchased from Alfa Aesar (Karlsruhe, Germany),
http://dx.doi.org/10.1016/j.abb.2014.10.007
0003-9861/Ó 2014 Elsevier Inc. All rights reserved.

220
C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
Sigma-Aldrich (Schnelldorf, Germany), Acros Organics (New Jersey,
USA), and Aladdin (Shanghai, China). Enzymes were obtained from
Takara Biotechnology (Dalian, China) and Fermentas China
(Shenzhen, China). Primers were obtained from Generay Biotech-
nology (Shanghai, China).
Isopropyl-b-D-thiogalactopyranoside (final concentration 0.1 mM)
was added for induction after the cells reached an optical density
of 0.8 at 600 nm. Culturation continued at 20 °C for 20 h
(200 rpm) until cells were collected by centrifugation. The centri-
fuged cells were suspended in 20 mL sodium phosphate buffer
(100 mM, pH 7.0) and were disrupted using an ultrasonic oscillator
(JY92-2D; Scienz Biotechnology Co. Ltd. Ningbo, China). The cell
debris was removed by centrifugation (10 min, 15,000ꢁg, 4 °C).
The supernatant was used as cell-free extract, and was applied to
a 5 mL His-Trap column assembled with Ni-NTA HisBind Resin
(EMD Chemicals, Inc. NY, USA). The protein of interest that con-
tained 6ꢁ His-tag was purified by binding and eluting with NPI
elution buffers (50 mM NaH2PO4, 300 mM NaCl, and varied con-
centrations of imidazole [pH 8.0]). NPI-10, NPI-60, and NPI-200
(corresponding to 10 mM, 60 mM, and 200 mM of imidazole) were
employed. The volume of elution buffers employed in this study
was 10 times volume of Ni-NTA HisBind Resin. Binding procedure
was carried out after balancing the His-Trap column with NPI-10.
NPI-60 was then used to wash the His-Trap column. NPI-200 was
employed to elute the protein of interest. The obtained protein
solution was treated with dialysis overnight in sodium phosphate
buffer (100 mM, pH 7.0), and then concentrated by ultrafiltration
(molecular weight cutoff, 30 kDa). The 6ꢁ His-tag of the protein
were cleaved by incubation with thrombin at 4 °C overnight, fol-
lowed by further purification with His-Trap column again. The
purified protein was examined using sodium dodecyl sulfate poly-
acrylamide gel electrophoresis (SDS–PAGE).
Microorganism cultivation
The medium for screening yeast included 0.6% (wt/vol) KH2PO4,
0.6% (NH4)2SO4, 0.06% MgSO4ꢀ7H2O, 0.006% ZnSO4ꢀ7H2O, 0.006%
FeSO4ꢀ7H2O, 0.001% CuSO4ꢀ5H2O, 0.001% MnSO4ꢀ4H2O, 0.01% NaCl,
0.5% yeast extract, and 3% glucose (pH 7.0). Soil samples were col-
lected from rich fields in Shanghai, China. Each soil sample (1 g)
was added to a test tube containing 10 mL sterile distilled water
and was shaken on a rotary shaker at 180 rpm for approximately
30 min. Next, 1 mL of the soil suspension was added to an Erlen-
meyer flask containing 50 mL yeast screening medium. The flask
was shaken on a rotary shaker at 28 °C, 150 rpm for 48 h. After
48-h incubation, 1 mL of culture was diluted appropriately and
was spread onto a plate containing solid yeast screening medium.
After 48 h of incubation at 28 °C, individual colonies on the plates
were inoculated into a test tube (15 mmØ ꢁ 150 mm) containing
2 mL medium that was constantly shaken for 28 h at 28 °C.
Screening of ethyl 4-chloroacetoacetate-reducing strains
Cells grown in the test tubes were collected by centrifugation,
and were washed. The cells were then suspended in 0.5 mL reac-
tion mixture containing 100 mM sodium phosphate buffer (pH
Protein analysis
7.0), 20 mM ethyl 4-chloroacetoacetate, and 500 mM
D-glucose.
After shaking for 24 h at 30 °C, the reaction mixture was extracted
using 1 mL ethyl acetate, and was centrifuged. The organic layer
was dried over anhydrous sodium sulfate and then analyzed to
determine the conversion and optical purity by gas chromatogra-
phy (GC) and high performance liquid chromatography (HPLC),
respectively. Strains were identified using standard nucleotide
sequence analysis.
Protein concentration was determined using Bradford assay
with bovine serum albumin as the standard [3]. Protein purity
was analyzed using SDS–PAGE. Protein samples for SDS-PAGE were
prepared by heating the protein for 10 min at 100 °C in the pres-
ence of a loading buffer (Takara Biotechnology Co. Ltd). A protein
marker (Fermentas China Co. Ltd) was used to estimate the molec-
ular weight of the obtained proteins. Molecular mass of the puri-
fied enzymes and their subunits were determined using Matrix-
Assisted Laser Desorption/Ionization Time of Flight Mass Spec-
trometry (MALDI-TOF-MS).
Cloning
Genomic DNA was isolated from L. elongisporus LH703 (CCTCC
AY 2014001) using a TaKaRa MiniBEST Bacterial Genomic DNA
Extraction Kit (TaKaRa, China). Genes encoding the three LEAKRs
were amplified using polymerase chain reaction (PCR; 48: forward
primer, 50-CGCCATATGGCATCCAACTT-30 and reverse primer,
50-CGCAAGCTTTTAAGCATCACCATC-30; 49: forward primer, 50-GC
GCCATATGTCGTATAGAT-30 and reverse primer, 50-CAAGCTTTCATG
GAGCATC-30; 50: forward primer, 50- GCGCCATATGTCTTCTCAA-3’
and reverse primer, 50-CGCAAGCTTTTAGGCATT-30 [with NdeI and
HindIII sites being underlined]). The primers were designed using
sequences of conserved hypothetical protein partial mRNA with
GenBank accession numbers: XM_001525808, XM_001526403,
and XM_001526411. Each double-digested PCR product was
inserted into pET-28a vector. Because of the chosen restriction
sites, the N-terminal 6ꢁ His-tag sequence of the plasmid was
retained. The resulting vector was transformed into Escherichia coli
Activity assay
The activity of the LEAKRs was determined using an assay mix-
ture containing sodium phosphate buffer (100 mM, pH 7.0), 6 mM
ethyl 4-chloroacetoacetate as substrate and 0.2 mM cofactor. A
continuous assay using UV absorbance at 340 nm was employed
to monitor the concentration of the cofactor during the reaction
at 30 °C. One unit of activity was defined as the amount of LEAKR
that catalyzed the oxidation of 1
standard conditions [23].
lmol cofactor per minute under
Characterization of recombinant LEAKRs
Optimum temperature and thermal stability of the purified enzymes
The optimum temperature for each LEAKR was determined by
incubating each enzyme in assay mixtures (pH 7.0) for 5 min at
temperatures ranging from 15 to 65 °C. The subsequent reaction
was started by adding ethyl 4-chloroacetoacetate as the substrate.
For thermostability analysis, each enzyme was pre-incubated for 1
and 3 h at temperatures ranging from 15 to 65 °C, and then incu-
bated at 30 °C for 15 min. The residual enzyme activity was deter-
mined at 30 °C under standard conditions.
DH5a competent cells. After isolation from competent cells and
DNA sequencing, the resulting plasmid was finally transformed
into E. coli BL21 (DE3) cells.
Recombinant expression and purification of recombinant LEAKRs
A
single colony was inoculated in
a
test tube
(15 mmØ ꢁ 150 mm) containing 3 mL preculture medium and
was constantly shaken (200 rpm) at 37 °C for 12 h. The cells were
then cultured in 100 mL LB medium with kanamycin (final
concentration, 30 mg lꢂ1) and were shaken (200 rpm) at 37 °C.
Optimum pH and pH stability of the purified enzymes
The effect of pH on enzyme activity was determined using
the following 0.1 M buffer systems: sodium citrate buffer (pH

C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
221
3.0–6.0), sodium phosphate buffer (pH 6.0–8.0) and glycine–NaOH
buffer (pH 8.0–11.0). For pH stability assay, each enzyme was pre-
incubated in the buffer systems described above for 1 and 3 h at
room temperature. The residual enzyme activity was assayed in
sodium phosphate buffer system (pH 7.0) under standard
conditions.
detection was with
volume was 10
S-enantiomer, 12 min; and ethyl 4-chloroacetoacetate, 19 min.
a
DAD set at 220 nm and injection
l
L. Retention times were: R-enantiomer, 11 min;
NCBI accession number for nucleotide sequence
LEAKR 48, GenBank accession number: JQ365664; LEAKR 49,
GenBank accession number: JQ365665; LEAKR 50, GenBank acces-
sion number: JQ365666.
Cofactor and substrate specificity
Cofactor dependence of each LEAKR was analyzed by adding
NADH or NADPH separately as cofactor into the assay mixtures
and by determining LEAKR activity. Substrate specificity of each
LEAKR was determined by employing different substrates (2 mM)
and conducting activity assays under standard conditions.
Results
Strains screening and genes selection
Kinetic parameters
In the screening experiments, the method for detecting ethyl
4-chloroacetoacetate-reducing strains described above was
performed efficiently. The resulting data showed that 41% of the
isolated strains (approximately 400 strains) could catalyze the
reduction of ethyl 4-chloroacetoacetate, however, most of these
strains showed efficiency. One strain, identified by 18S rDNA as
L. elongisporus (LH703) showed the best efficiency and high enanti-
oselectivity in the screening reaction. L. elongisporus LH703 pro-
duced ethyl (R)-4-chloro-3-hydroxybutyrate, with a conversion
rate of 92% and e. e. value of 80%.
The genome of L. elongisporus has already been analyzed for
genes encoding hypothetical conserved proteins with oxidoreduc-
tase activity. Several potential genes encoding such proteins have
been discovered. Of these, three genes were selected for further
identification of their catalyzing activity to ethyl 4-chloroacetoac-
etate. These three genes shared <50% identity with each other.
Kinetic parameters Km and Vmax were obtained from multiple
determinations using Michaelis-Menten equation and program
Origin (Origin Pro, version 8.0). Nonlinear Curve Fit from the anal-
ysis software program was used for least-squares parameter fitting
on the basis of a Levenberg–Marquardt algorithm.
Effect of organic solvents on enzyme activity
The enzymes were pre-incubated for 60 min in a reaction mix-
ture containing various organic solvents with increasing log P val-
ues. LEAKR activity in reaction systems with vol/vol 5% and 10%
concentrations of the organic solvents was assayed under standard
conditions, and reaction rates were compared for reactions with
and without organic solvents.
Effect of metal ions, chelators, and detergents on enzyme activity
Various detergents, metal ions, and chelators were added to the
reaction systems and were pre-incubated for 10 min at 30 °C. The
residual activity of the enzymes was determined under standard
conditions and was presented as percentage of enzyme activity
without any additive.
Purification of recombinant LEAKRs
The purified proteins were obtained using an optimized elution
procedure. The typical yield of pure protein during the entire puri-
fication procedure was 0.12–0.14 mg/g wet cell weight. SDS–PAGE
(Fig. 1) revealed a single band for each LEAKR. MALDI-TOF-MS
assay showed that the molecular mass of each purified enzyme
Enantioselective reduction of ethyl 4-chloroacetoacetate
The enantioselectivity of LEAKRs was determined by employing
NADPH regeneration system composed of D-glucose dehydroge-
nase (GDH) and D-glucose. The reaction mixture included 20 mM
ethyl 4-chloroacetoacetate, 0.2 mM NADPH, 0.03 mg LEAKR
(1.2 mg for LEAKR 49), 100 mM
D-glucose, 0.1 mg GDH, and
100 mM sodium phosphate buffer (pH 7.0), with the final volume
of 0.5 mL. After shaking constantly at 30 °C, the reaction mixture
was extracted with 1 mL ethyl acetate. The organic layer collected
by centrifugation and dried with anhydrous sodium sulfate was
then analyzed to determine the conversion of the substrate and
enantiomeric excess (e. e.) of the products by employing GC and
HPLC as efficient and sensitive tools. A GC station (Agilent
6890N) with a capillary column (DB-5) was used to analyze the
conversion of ethyl 4-chloroacetoacetate while an HPLC station
(Agilent 1100) with a Chiralcel OB-H column was employed for
the optical assay of the products.
The operating conditions for GC were as follows: flow rate was
1.3 mL/min, injector temperature was 250 °C, detector tempera-
ture was 250 °C and column temperature was programmed:
3 min at 90 °C, 10 °C/min to 180 °C and 2 min at 180 °C. Injection
volume was 0.5 lL. Ethyl 4-oxopentanoate was employed as inter-
nal standard. Retention times were: ethyl 4-oxopentanoate,
6.1 min; ethyl 4-chloroacetoacetate, 6.9 min; and ethyl 4-chloro-
3-hydroxybutyrate, 7.3 min.
Chiral detection of product ethyl 4-chloro-3-hydroxybutyrate
was performed with a chiralcel OB-H column (Daicel Chemical
Industries, Ltd.). The mobile phase was 90% hexane/10% isopropa-
nol, flow rate was 0.8 mL/min, column temperature was 25 °C,
Fig. 1. SDS–PAGE analysis of each recombinant Lodderomyces elongisporus AKRs
after purification step. Lane 1, protein marker; lane 2, purified LEAKR 48; lane 3,
purified LEAKR 49; lane 4, purified LEAKR 50.

222
C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
Fig. 2. Amino acid sequence alignment of three Lodderomyces elongisporus AKRs with members of the aldo–keto reductase superfamily. The abbreviations are used: GCY1P,
GCY1 protein from S. cerevisiae; -HSD, -hydroxysteroid dehydrogenase from Rattus norvegicus liver; 2,5-DKGRA, 2,5-diketo- -gluconic acid reductase from
3a
3a
D
Corynebacterium sp.; LEAKR 48, aldo–keto reductase 48 from Lodderomyces elongisporus; LEAKR 49, aldo–keto reductase 49 from Lodderomyces elongisporus; LEAKR 50, aldo–
keto reductase 50 from Lodderomyces elongisporus. The sequences were aligned using the CLUSTAL X program.
and its subunit was 156 kDa and 39 kDa for LEAKR 48, 204 kDa and
34 kDa for LEAKR 49, and 132 kDa and 33 kDa for LEAKR 50,
respectively. LEAKR 48 and LEAKR 50 were found to be homotetra-
mers while LEAKR 49 included six identical subunits. The activity
of each LEAKR was determined under standard conditions. The
activity of LEAKR 48 was 5.2 U/mg protein, LEAKR 49 was 0.14 U/
mg protein, and LEAKR 50 was 7.6 U/mg protein.
48 and LEAKR 49; 29.97% between LEAKR 48 and LEAKR 50; and
27.34% between LEAKR 49 and LEAKR 50) are presented in Fig. 2.
BLAST-P analysis (http://www.ncbi.nlm.nih.gov/BLAST/) was used
to identify these enzymes, results showed that these enzymes
shared high identities with many AKRs or putative members of
AKR superfamily. GCY1 protein from Saccharomyces. cerevisiae
(P14065),
3a-hydroxysteroid dehydrogenase from Rattus
norvegicus liver (P23457), and 2,5-diketo-
D
-gluconic acid reduc-
tase from Corynebacterium sp. (AAA83534) belong to AKR3, AKR1,
and AKR5 families, respectively, and have determined protein
structures. These three AKRs along with the LEAKRs examined in
this study were selected to obtain an alignment of six proteins
for detailed analysis (Fig. 2).
Sequences analysis
Analysis of amino acid sequences of the enzymes encoded by
the three genes showed differences among them. The identities
of amino acid sequences of these enzymes (39.78% between LEAKR

C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
223
is considered as the general acid in the reaction while Lys-82 and
His-113 are thought to be involved in proton relay [10,7]. These
residues are strictly conserved among AKRs. In contrast to short-
chain dehydrogenase superfamily, AKRs bind to the cofactor
NADPH (preferred by most AKRs) in an extended conformation
without a Rossmann-fold motif [14,7,22]. Residues Asp-52, Ser-
154, Asn-155, Gln-178, Leu-207, Lys-249, Arg-255, Asn-259, and
Ser-271 of the three LEAKRs were involved in cofactor binding.
Of these residues, three were a part of the active site pocket, as
mentioned above. Asp-52 was highly conserved and played an
important role in the binding mechanism. This was confirmed by
a structure study of 3a-hydroxysteroid dehydrogenase from R. nor-
vegicus [14]. The resulting sequence alignment indicated that there
were no major gaps in the regions corresponding to the barrel
structure. However, major gaps appeared in the C-terminal of the
alignment, especially in the sequences of A-loop, B-loop, and H1-
helice. Loops of AKRs are known to vary near the active site to fit
different substrates while maintaining the structure of the barrel
scaffold [7]. The rest of conserved residues probably correspond
to other protein characteristics. Based on the nomenclature of
the AKR superfamily [15], LEAKR 50 was identified as a member
of AKR3 family and shared >40% sequence identity with other
members of this family. The other two LEAKRs (LEAKR 48 and
LEAKR 49) were probably the members of two novel AKR families
and shared <40% sequence identity with other AKR families identi-
fied thus far and with each other.
Fig. 3. Effect of reaction temperature on each purified Lodderomyces elongisporus
AKR. Each of their activities at optimal reaction temperatures corresponds to 100%
activity. The experiments were done in triplicate and the error bar represents the
percentage error in each set of reading.
The AKR superfamily proteins share the characteristic
(a/b)8-barrel structure [13] and conserved active site residues.
Asp-52, Ala-54, Tyr-57, Lys-82, Leu-111, Ser-154, Asn-155, and
Tyr-204 (the number of amino acid residues in this article corre-
sponds to that of LEAKR 50) were found to be conserved in the
three LEAKRs based on the alignment. Together with residue 28,
which varied in different LEAKRs, they formed the bottom portion
of the active site pocket. Residues comprising the rim of the pocket
at 56, 84, 114, and 286, except His-113, were partially different
among the three LEAKRs [18]. Of all the residues, Asp-52, Tyr-57,
Lys-82 and His-113 were crucial for the catalytic activity. Tyr-57
Optimum temperature and thermostability
Enzyme activity and thermostability were assayed for the three
LEAKRs at various temperatures (Figs. 3 and 4). LEAKR 49 was
Fig. 4. Thermostability of each purified Lodderomyces elongisporus AKR: LEAKR 48 (a); LEAKR 49 (b); LEAKR 50 (c). Each of their activities at optimal temperatures corresponds
to 100% activity. The experiments were done in triplicate and the error bar represents the percentage error in each set of reading.

224
C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
Fig. 5. Effect of pH on each purified Lodderomyces elongisporus AKR: LEAKR 48 (a); LEAKR 49 (b); LEAKR 50 (c). The activity was measured in the following 0.1 M buffers:
sodium citrate buffer (pH 3.0–6.0), sodium phosphate buffer (pH 6.0–8.0) and glycine–NaOH buffer (pH 8.0–11.0). The experiments were done in triplicate and the error bar
represents the percentage error in each set of reading.
found to be heat sensitive with an optimum temperature of 30 °C,
and showed a sharp decline in activity when it was pre-incubated
at higher temperatures. In contrast, LEAKR 48 and LEAKR 50
showed stronger heat resistance with maximal activity at 55 °C
(10 U/mg) and 45 °C (4.2 U/mg), respectively. LEAKR 48 was stable
at temperature <40 °C for 3 h and showed an activity of 92% after
pre-incubation at 45 °C for 1 h. LEAKR 50 showed a high activity
of >86% at a wide temperature range from 30 to 50 °C.
Several substrates of keto esters and aromatic ketones were
chosen to determine the substrate specificity of the LEAKRs. By
considering the catalyzing activity of LEAKRs to propiophenone
as 100, the activities to other substrates were determined rela-
tively (Table 1). Ethyl 4-chloroacetoacetate was found to be the
preferred substrate by LEAKR 48 (39,870% compared to the activity
against propiophenone) and LEAKR 50 (36,735% compared to the
activity against propiophenone) while 3-chloropropiophenone
was found to be the preferred substrate for LEAKR 49 (2,143% com-
pared to the activity against propiophenone). The size of substitu-
ent group and the length of carbon chain were found to affect the
activity considerably. Among the keto esters tested, ethyl 4-chloro-
acetoacetate was considered to possess the most suitable structure
as the substrate for LEAKR 48 and LEAKR 50. We also observed that
all the three LEAKRs showed poor activities against heterocyclic
ketones. As for aromatic ketones with phenyl group, various elec-
tron-donating and electron-withdrawing substituents were con-
firmed to have a great influence on the catalyzing activity of the
LEAKRs.
Optimum pH and stability
The effect of pH on enzyme activity was measured using buffer
systems with various pH. Optimum pH was determined for three
LEAKRs separately: pH 6.0 for LEAKR 48, pH 7.0 for LEAKR 49
and LEAKR 50 (Fig. 5). As for pH stability, LEAKR 49 was stable at
pH from 6.0 to 7.0, however, it lost most of its activity at pH 5.0.
LEAKR 48 and LEAKR 50 retained most of their activity at extreme
pH ranging from 3.0 to 10.0 for 3 h. In addition, LEAKR 48 retained
75% of its maximal activity at pH 11.0 (Fig. 6).
Enzyme kinetics
Cofactor and substrate specificity
The kinetic parameters of the three LEAKRs were determined
for substrates in Table 1 and cofactor NADPH. Curves of activity
versus substrate concentration were obtained. The three enzymes
showed large differences in kinetic properties. LEAKR 49 showed
a distinctly low affinity for ethyl 4-chloroacetoacetate and NADPH,
To determine the cofactor specificity of each LEAKR, 0.2 mM
NADPH or NADH was used as the cofactor in a standard reaction
mixture. The ability to catalyze various substrates was negligible
in the absence of NADPH. Therefore, NADPH was considered as
the only available cofactor for reactions catalyzed by these three
LEAKRs.
with Km values of 18.79 mM and 270
lM, respectively, compared
with the other two enzymes. LEAKR 48 and LEAKR 50 showed

C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
225
Fig. 6. 1 h (blank) and 3 h (solid) pH stability of each purified Lodderomyces elongisporus AKR: LEAKR 48 (a); LEAKR 49 (b); LEAKR 50 (c). The activity was measured in the
following 0.1 M buffers: sodium citrate buffer (pH 3.0–6.0), sodium phosphate buffer (pH 6.0–8.0) and glycine–NaOH buffer (pH 8.0–11.0). Each of their activities at optimal
pH corresponds to 100% activity. The experiments were done in triplicate and the error bar represents the percentage error in each set of reading.
Table 1
over other substrates obtained from kinetic study confirmed the
substrate specificities of LEAKR 48 and LEAKR 50 (Table 1).
Substrate specificity of the three purified Lodderomyces elongisporus AKRs.
Substratea
Relative activityb (%)
LEAKR 48 LEAKR 49
Effect of organic solvents on enzyme activity
LEAKR 50
Aromatic ketones
40-Methoxyacetophenone
40-Chloroacetophenone
20-Bromoacetophenone
40-Bromoacetophenone
40- Fluoroacetophenone
Propiophenone
54 0.2
0.0
96 0.1
1103 5.0
96 0.2
73 0.3
1498 5.0
632 2.0
724 4.0
105 0.2
100 0.0
2143 0.0
128 0.3
89 0.3
23 0.0
A series of organic solvents were used in this study to test their
effect on the catalyzing activity of the three LEAKRs. Hydrophobic-
ity of the organic solvents was indicated using log P values
(Table 3). Different amounts of organic solvents were used in the
assay. All the three LEAKRs showed low sensitivity to: dimethyl
sulfoxide and methanol (low log P values of ꢂ1.3 and ꢂ0.76), ethyl
acetate (medium log P value of 0.68), and n-hexane and isooctane
(high log P values of 3.5 and 4.5). Benzene, which is toxic toward
many proteins, strongly impaired the activity of LEAKR 48. Octanol
also had a large effect on the activities of LEAKR 48 and LEAKR 49.
Their activities sharply declined with the increasing amount of oct-
anol added. LEAKR 50 generally showed a strong resistance to all
the organic solvents, except isopropyl alcohol which reduced the
activity by 42% with 10% added.
0
595 2.0
472 4.0
722 4.0
54 0.1
100 0.0
177 0.1
98 0.4
106 0.2
102 0.3
79 0.1
100 0.0c
200 3.0
1168 3.0
393 1.0
1382 4.0
64 0.6
3-Chloropropiophenone
2-Acetylpyridine
3-Acetylpyridine
4-Acetylpyridine
2-Acetylthiophene
244 0.0
110 0.3
Keto esters
Ethyl acetoacetate
Ethyl 3-oxohexanoate
Ethyl 4-methyl-3-
oxopentanoate
Ethyl benzoylacetate
Ethyl 4-oxopentanoate
Ethyl 4-chloroacetoacetate
1193 3.0
150 0.4
50 0.4
133 0.1
87 0.2
132 0.2
2213 3.0
56 0.0
65 0.0
276 2.0
87 0.1
473 3.0
95 0.3
210 0.0 36,735 7.0
184 0.3
39 0.0
Effect of metal ions, chelators, and detergents on enzyme activity
39,870
9
a
b
c
Concentration of each substrate was 2.0 mM.
The activity of each LEAKR with propiophenone was considered as 100%.
The experiments were done in triplicate and the error bar represents the
The effect of various compounds on the activity of LEAKR was
assayed by adding each of these to the reaction system. Metal ions
highly affected the enzyme activity (Table 4). The effects that were
mostly negative differed thoroughly among the three LEAKRs. Che-
lators reduced the activity of LEAKR 49, but, slightly accelerated
the reaction catalyzed by LEAKR 48 and LEAKR 50. Similarly, deter-
gents, except SDS, reduced the activity of LEAKR 49 but increased
the activity of LEAKR 48 and LEAKR 50. This indicated that LEAKR
standard error of the mean.
Vmax/Km values of 8.632 and 2.028 U mgꢂ1 mMꢂ1 against ethyl 4-
chloroacetoacetate, much higher than those against other com-
pounds (Table 2). The preference to ethyl 4-chloroacetoacetate

226
C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
Table 2
Steady-state kinetic constants of three Lodderomyces elongisporus AKRs.
Substrate
Km (mM)a
Vmax (U mg-1
)
Vmax/Km (U mgꢂ1 mMꢂ1
)
LEAKR
48
LEAKR
49
LEAKR
50
LEAKR
48
LEARK
49
LEAKR
50
LEAKR
48
LEAKR
49
LEAKR
50
Aromatic ketones
40-Methoxyacetophenone
40-Chloroacetophenone
20-Bromoacetophenone
40-Bromoacetophenone
40- Fluoroacetophenone
Propiophenone
3-Chloropropiophenone
2-Acetylpyridine
3-Acetylpyridine
1.26
–
10.32
7.64
0.49
1.98
7.14
4.38
1.78
1.15
18.9
5.04
6.6
8.47
14
0.01
–
0.129
1.87
0.01
0.41
0.08
0.345
0.118
0.033
0.025
0.167
0.011
0.63
0.008
–
0.012
0.245
0.426
0.193
0.018
0.02
0.607
0.046
0.013
0.046
0.019
0.001
0.029
0.042
0.039
0.003
0.006
0.022
0.004
0.024
0.003
0.01
b
6.6
20.9
3.4
89.3
1.05
4.54
15.3
30.4
11.68
1.92
8.89
45.1
5.28
1.12
39.8
0.45
183.2
1.09
0.051
1.56
0.032
0.545
0.038
0.473
0.42
2.78
0.054
0.209
0.383
0.128
0.086
1.08
0.053
0.25
0.23
0.008
0.075
0.009
0.006
0.036
0.104
0.027
0.091
0.005
4-Acetylpyridine
2-Acetylthiophene
0.124
0.011
Keto esters
Ethyl acetoacetate
6.8
6.81
2.14
6.54
158.8
28.66
18.79
0.49
13.7
5.92
0.26
18.1
10.7
0.65
0.154
0.047
0.149
10.1
0.381
0.58
0.24
0.039
0.024
0.018
0.03
0.096
0.03
0.009
0.045
0.015
8.632
0.023
0.022
0.023
0.064
0.013
0.031
0.49
Ethyl 3-oxohexanoate
Ethyl 4-methyl-3-oxopentanoate
Ethyl benzoylacetate
Ethyl 4-oxopentanoate
Ethyl 4-chloroacetoacetate
0.93
0.97
1.21
1.1
0.028
0.009
0.054
0.017
6.56
0.003
0.004
0.069
0.002
2.028
0.76
21.7
Cofactor
NADPH
0.08
0.27
0.07
a
The activities were measured at pH 7.0 and 30 °C in all reactions. The experiments were done in triplicate and the average values were given.
b
LEAKR 48 showed no activity on 40-chloroacetophenone.
Table 3
Furthermore, all the LEAKRs showed different optical configuration
preferences, as indicated in Table 5.
Effect of organic solvents on the activity of each purified Lodderomyces elongisporus
AKR.
Oragnic solvents
Concentration
(v/v, %)
Relative activity (%)
Discussion
LEAKR 48
LEAKR 49
LEAKR 50
None
100 0.0a
95 0.1
93 0.2
100 0.4
92 0.3
100 0.6
76 0.2
92 0.0
60 0.1
100 0.0
82 0.4
83 0.3
57 0.6
96 0.3
92 0.1
92 0.3
55 0.2
100 0.0
98 0.2
85 0.4
99 0.8
92 0.5
98 0.5
66 0.0
90 0.3
72 0.6
95 0.1
93 0.1
80 0.7
69 0.2
85 0.0
81 0.1
47 0.0
34 0.0
97 0.5
96 0.2
100 0.1
99 0.1
100 0.0
90 0.1
95 0.1
100 0.4
90 0.2
95 0.2
83 0.2
94 0.2
58 0.1
100 0.0
95 0.4
94 0.2
83 0.0
96 0.5
83 0.8
85 0.9
79 0.2
100 0.1
95 0.1
92 0.5
90 0.4
To the best of our knowledge, no enzyme from L. elongisporus
has been reported to have explicit AKR activity. These gaps were
filled in this study. Three genes from L. elongisporus that encoded
AKRs (LEAKRs) were characterized, cloned, and heterologously
overexpressed. The three LEAKRs showed <40% amino acid identity
with each other but shared highly conserved residues with other
members of the AKR superfamily. According to sequence analysis,
LEAKR 50 was identified as a member of the AKR3 family while the
other two LEAKRs were identified as the members of two novel
AKR families based on the latest AKR family list. Considering their
novelty, a detailed sequence analysis was performed in this study
by using sequence comparison and structure relativity. LEAKRs
shared the same crucial part of the active site and similar NADPH
binding site. This might explain why they all catalyzed the same
ketone substrates and preferred NADPH as cofactor. Furthermore,
major gaps appeared in loops of C-terminal which were considered
to have a large effect on the substrate specificity of AKRs. This was
confirmed from the data acquired in this study.
DMSOb
5
10
5
10
5
10
5
10
5
10
5
10
5
10
5
10
5
Methanol
Acetonitrile
Isopropanol
Ethyl acetate
Benzene
Toluene
1-Octanol
n-Hexane
Isooctane
100
0
10
5
10
91 0.5
100 0.6
94 0.6
a
The experiments were done in triplicate and the error bar represents the
The members of the AKR superfamily, which are found in
diverse sources, catalyze numerous metabolic pathways involving
a wide range of substrates. Some of the identified AKRs are excel-
lent industrial catalysts with outstanding characteristics
[11,17,12,16]. For example, an alcohol dehydrogenase belonging
to the AKR superfamily identified from a hyperthermophilic archa-
eon Pyrococcus furiosus showed increasing catalyzing activity at up
to 100 °C [19]. Over 150 AKRs have been identified thus far, how-
ever, these only represent a small part of the entire AKR superfam-
ily. In this study, we identified novel members of the AKR family
from a new source. It is notable that the three enzymes showed dif-
ferent characteristics in this study. LEAKR 48 and LEAKR 50 showed
particularly high activities toward ethyl 4-chloroacetoacetate,
which probably resulted from the similarity of ethyl 4-chloroace-
toacetate to their natural substrates. Compared to LEAKR 49, the
standard error of the mean.
b
DMSO is short for dimethyl sulfoxide.
49, unlike LEAKR 48 and LEAKR 50, might need metal ions for its
catalyzing activity.
Enantioselective reduction of ethyl 4-chloroacetoacetate
The enantioselectivity of each LEAKR was tested using ethyl
4-chloroacetoacetate as substrate. Absolute configurations of
product ethyl 4-chloro-3-hydroxybutyrate were measured. All
the LEAKRs showed 100% conversion with different e. e. values
(70% for LEAKR 48, 58% for LEAKR 49, and 98% for LEAKR 50).

C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
227
Table 4
Effect of metal ions, chelators and detergents on the activity of each purified Lodderomyces elongisporus AKR.
Compound
Concentration
Relative activity (%)
LEAKR 48
LEAKR 49
100
LEAKR 50
100
None
100 0a
0
0
Metal ions
Ca2+
1.0(mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
1.0 (mM)
87 0.2
82 0.7
10 0.1
84 0.1
58 0.3
41 0.0
94 0.3
80 0.6
97 0.1
62 0.5
69 0.1
45 0.1
10 0.0
24 0.2
119 0.4
98 0.5
81 0.0
95 0.1
86 0.4
67 0.5
Mg2+
Zn2+
5
0.0
Cu2+
73 0.0
54 0.1
13 0.1
34 0.2
43 0.1
96 0.1
Fe2+
Mn2+
Ni2+
Co2+
Li1+
Chelators and detergents
EDTAꢀ2Na
0.005% (wt/v)
0.005% (wt/v)
0.005% (wt/v)
0.005% (wt/v)
0.005% (wt/v)
0.005% (wt/v)
113 2.0
69 0.5
105 2.0
18 0.2
132 0.0
165 0.5
156 0.8
160 1.0
SDS
0
0.0
0
0.0
Triton X-100
Tween-20
Tween-60
136 0.2
182 1.0
166 0.3
154 0.7
91 0.1
93 0.3
82 0.6
89 0.2
Tween-80
a
The experiments were done in triplicate and the error bar represents the standard error of the mean.
Table 5
also created confusion about the overlap in activity between sev-
eral enzymes within a single cell. Microbial genome data have pre-
dicted that a common prokaryote such as E. coli may possess six
AKRs [1]. Although it is believed that each enzyme has evolved
for a particular metabolic purpose, its ability to act on a range of
substrates stays apparent. For example, LEAKR 48 and LEAKR 50
were found to prefer the same substrate but catalyzed its reaction
at different optimum reaction conditions. One prediction was that
the metabolic functions performed by these enzymes were similar
in reaction type but differed with the changing conditions. Deter-
mining and observing different AKRs from the same microbial
source may also reveal their physiological roles.
Optical preference of each purified Lodderomyces elongisporus AKR.
LEAKR 48
LEAKR 49
LEAKR 50
R
Optical configurationa
e. e. (%)
R
S
70 0.5b
58 0.5
98
0
a
Ethyl 4-chloroacetoacetate was used as substrate and the absolute configura-
tions of the ethyl 4-chloro-3-hydroxybutyrate were measured.
b
The experiments were done in triplicate and the error bar represents the
standard error of the mean.
other two LEAKRs showed superior characteristics of resistance to
high temperatures and extreme pH conditions. Use of ethyl 4-chlo-
roacetoacetate as the substrate showed great disparity in the opti-
cal purities of products. It is also worth mentioning that LEAKR 50
was found to be outstanding among the three LEAKRs, with high
activity of >86% at a wide range of reaction temperatures (30 to
50 °C), strong resistance to adverse pH environments (pH
3.0–10.0), and high (R)-enantioselectivity (e. e., 98%), thus making
it a potential catalyst for industry.
Acknowledgments
This work was supported by the Natural Science Foundation of
China (No. 21276084) and the National Major Science and Technol-
ogy Projects of China (No. 2012ZX09304009).
References
The diverse characteristics of the three LEAKRs provided us a
pathway to partially identify the roles of some key amino acids
that are important for protein modifications. In the past, processes
have been designed and improved around the limitations of an
enzyme. Nowadays, an enzyme is engineered to fit process specifi-
cations [2]. Directed evolution of proteins has arisen as an efficient
way to modify proteins and obtain enzymes with superior stability,
substrate specificity, and enantioselectivity. Numerous successful
attempts have drawn more attention of researchers. For instance,
site-saturation mutagenesis methods were applied in a semi-
[1] F.R. Blattner, G. Plunkett III, C.A. Bloch, N.T. Perna, V. Burland, M. Riley, J.C.
Vides, J.D. Glasner, C.K. Rode, G.F. Mayhew, J. Gregor, N.W. Davis, H.A.
Kirkpatrick, M.A. Goeden, D.J. Rose, B. Mau, Y. Shao, Science 277 (1997) 1453–
1462.
[2] U.T. Bornscheuer, G.W. Huisman, R.J. Kazlauskas, S. Lutz, J.C. Moore, K. Robins,
Nature 485 (2012) 185–194.
[3] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[4] M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keßeler, R. Stürmer, T. Zelinski,
Angew. Chem. Int. Ed. 43 (2004) 788–824.
[5] N.C. Bruce, D.L. Willey, A.F. Coulson, J. Jeffery, Biochem. J. 299 (1994) 805–811.
[6] E. Champion, F. Guérin, C. Moulis, S. Barbe, T.H. Tran, S. Morel, K. Descroix, P.
Monsan, L. Mourey, L.A. Mulard, S. Tranier, M.R. Siméon, I. André, J. Am. Chem.
Soc. 134 (2012) 18677–18688.
rational approach, to alter an amylosucrase (a-transglucosidase)
from Neisseria polysaccharea. The best mutant showed 400-fold
enhanced catalytic efficiency toward the donor substrate sucrose
and a non-natural acceptor substrate allyl 2-acetamido-2-deoxy-
[7] E.M. Ellis, FEMS Microbiol. Lett. 216 (2002) 123–131.
[8] W.D. Fessner, N.J. Turner, M.X. Wang, Adv. Synth. Catal. 353 (2011) 2189–
2190.
[9] K. Goldberg, K. Schroer, S. Lütz, A. Liese, Appl. Microbiol. Biotechnol. 76 (2007)
237–248.
alpha-D-glucopyranoside [6]. A structure study of the three LEAKRs
[10] E. Hur, D.K. Wilson, Chem-Biol. Interact. 130 (2001) 527–536.
[11] K. Ishihara, H. Yamaguchi, H. Hamada, K. Nakamura, N. Nakajima, J. Mol. Catal.
B-Enzym. 10 (2000) 419–428.
[12] N. Itoh, H. Asako, K. Banno, Y. Makino, M. Shinohara, T. Dairi, R. Wakita, M.
Shimizu, Appl. Microbiol. Biotechnol. 66 (2004) 53–62.
[13] J.M. Jez, M.J. Bennett, B.P. Schlegel, M. Lewis, T.M. Penning, Biochem. J. 326
(1997) 625–636.
identified in this study is underway for modification purposes.
Identification of diverse characteristics and comparison of
sequences are expected to identify some valuable residues, thus
contributing to structure-guided rational engineering of AKRs.
It is possible that more than one enzyme that can catalyze the
reaction of one particular substrate may exist in a single cell. Char-
acterization of the three AKRs from the same microorganism has
[14] J.M. Jez, T.G. Flynn, T.M. Penning, Biochem. Pharmacol. 54 (1997) 639–647.
[15] J.M. Jez, T.M. Penning, Chem-Biol. Interact. 130 (2001) 499–525.

228
C. Ning et al. / Archives of Biochemistry and Biophysics 564 (2014) 219–228
[16] K.J. Jing, Z.N. Xu, Y. Liu, X.X. Jiang, L. Peng, P.L. Cen, Prep. Biochem. Biotechnol.
35 (2005) 203–215.
[20] T.M. Penning, J.E. Pawlowski, B.P. Schlegel, J.M. Jez, H.K. Lin, S.S. Hoog, M.J.
Bennett, M. Lewis, Steroids 61 (1996) 508–523.
[17] M. Kataoka, K. Yamamoto, H. Kawabata, M. Wada, K. Kita, H. Yanase, S.
Shimizu, Appl. Microbiol. Biotechnol. 51 (1999) 486–490.
[18] S. Khurana, D.B. Powers, S. Anderson, M. Blaber, Proc. Natl. Acad. Sci. USA 95
(1998) 6768–6773.
[21] J.M. Petrash, I. Tarle, D.K. Wilson, F.A. Quiocho, Diabetes 43 (1994) 955–959.
[22] G. Sanli, M. Blaber, J. Mol. Biol. 309 (2001) 1209–1218.
[23] S. Shimizu, M. Kataoka, M. Katoh, T. Morikawa, T. Miyoshi, H. Yamada, Appl.
Environ. Microbiol. 56 (1990) 2374–2377.
[19] R. Machielsen, A.R. Uria, S.W.M. Kengen, J.V.D. Oost, Appl. Environ. Microbiol.
72 (2006) 233–238.