Improving the acidic stability of Staphylococcus aureus α-acetolactate decarboxylase in Bacillus subtilis by changing basic residues to acidic residues

Article abstract of DOI:10.1007/s00726-014-1898-5

The α-acetolactate decarboxylase (ALDC) can reduce diacetyl fleetly to promote mature beer. A safe strain Bacillus subtilis WB600 for high-yield production of ALDC was constructed with the ALDC gene saald from Staphylococcus aureus L3-15. SDS-PAGE analysis revealed that S. aureus α-acetolactate decarboxylase (SaALDC) was successfully expressed in recombinant B. siutilis strain. The enzyme SaALDC was purified using Ni-affinity chromatography and showed a maximum activity at 45 °C and pH 6.0. The values of K _m and V _max were 17.7 μM and 2.06 mM min^-1, respectively. Due to the unstable property of SaALDC at low pH conditions that needed in brewing process, site-directed mutagenesis was proposed for improving the acidic stability of SaALDC. Homology comparative modeling analysis showed that the mutation (K52D) gave rise to the negative-electrostatic potential on the surface of protein while the numbers of hydrogen bonds between the mutation site (N43D) and the around residues increased. Taken together the effect of mutation N43D-K52D, recombinant SaALDC_N43D-K52D showed dramatically improved acidic stability with prolonged half-life of 3.5 h (compared to the WT of 1.5 h) at pH 4.0. In a 5-L fermenter, the recombinant B. subtilis strain that could over-express SaALDC_N43D-K52D exhibited a high yield of 135.8 U mL^-1 of SaALDC activity, about 320 times higher comparing to 0.42 U mL^-1 of S. aureus L3-15. This work proposed a strategy for improving the acidic stability of SaALDC in the B. subtilis host.

Full text of DOI:10.1007/s00726-014-1898-5

Amino Acids
DOI 10.1007/s00726-014-1898-5
ORIGINAL ARTICLE
Improving the acidic stability of Staphylococcus aureus
α‑acetolactate decarboxylase in Bacillus subtilis by changing basic
residues to acidic residues
Xian Zhang · Zhiming Rao · Jingjing Li ·
Junping Zhou · Taowei Yang · Meijuan Xu ·
Teng Bao · Xiaojing Zhao
Received: 16 July 2014 / Accepted: 9 December 2014
© Springer-Verlag Wien 2014
Abstract The α-acetolactate decarboxylase (ALDC)
can reduce diacetyl fleetly to promote mature beer. A safe
strain Bacillus subtilis WB600 for high-yield production
of ALDC was constructed with the ALDC gene saald
from Staphylococcus aureus L3-15. SDS-PAGE analy-
sis revealed that S. aureus α-acetolactate decarboxylase
(SaALDC) was successfully expressed in recombinant B.
siutilis strain. The enzyme SaALDC was purified using
Ni-affinity chromatography and showed a maximum
activity at 45 °C and pH 6.0. The values of Km and Vmax
were 17.7 μM and 2.06 mM min−1, respectively. Due to
the unstable property of SaALDC at low pH conditions
that needed in brewing process, site-directed mutagen-
esis was proposed for improving the acidic stability of
SaALDC. Homology comparative modeling analysis
showed that the mutation (K52D) gave rise to the nega-
tive-electrostatic potential on the surface of protein while
the numbers of hydrogen bonds between the mutation site
(N43D) and the around residues increased. Taken together
the effect of mutation N43D-K52D, recombinant SaAL-
DCN43D-K52D showed dramatically improved acidic stabil-
ity with prolonged half-life of 3.5 h (compared to the WT
of 1.5 h) at pH 4.0. In a 5-L fermenter, the recombinant
B. subtilis strain that could over-express SaALDCN43D-
activity, about 320 times higher comparing to 0.42 U
mL−1 of S. aureus L3-15. This work proposed a strategy
for improving the acidic stability of SaALDC in the B.
subtilis host.
Keywords α-Acetolactate decarboxylase · Diacetyl ·
Acidic stability · Site-directed mutagenesis · Bacillus
subtilis
Introduction
Diacetyl, an important substance in the beer, can cause
undesirable buttery off-flavor when its content is over fla-
vor threshold (0.02–0.1 mg L−1) (Dulieu et al. 2000). It is
necessary to reduce the content of diacetyl during the tra-
ditional process of beer production. Diacetyl formation
happens under the oxygenolysis of acetolactate, which is
produced by the yeast to be an intermediate of amino acid
leucine and valine synthesis. Diacetyl is then transformed
to acetoin by an enzymatic reaction. Acetoin has a much
higher flavor threshold and does not cause any off-flavor
in the beer. However, the route of diacetyl formation dur-
ing traditional beer fermentation is unavoidable as well as a
rate-limiting step (Dulieu et al. 2000).
exhibited a high yield of 135.8 U mL−1 of SaALDC
K52D
It was reported that α-acetolactate decarboxylase
(ALDC, EC.4.1.1.5) could catalyze α-acetolactate to ace-
toin in one step without the formation of diacetyl (Loken
and Stormer 1970; Suihko et al. 1990), and it was a good
food additive for promoting mature beer fleetly. So ALDC
application in the beer fermentation will bring great eco-
nomic benefit. Besides involving in the catabolic degrada-
tion of acetolactate to acetoin in the 2,3-butanediol path-
way, ALDC is also found to be as a key regulator of valine
and leucine biosynthesis by controlling the acetolactate flux
Electronic supplementary material The online version of this
article (doi:10.1007/s00726-014-1898-5) contains supplementary
material, which is available to authorized users.
X. Zhang · Z. Rao (*) · J. Li · J. Zhou · T. Yang · M. Xu ·
T. Bao · X. Zhao
The Key Laboratory of Industrial Biotechnology of Ministry
of Education, School of Biotechnology, Jiangnan University,
Wuxi 214122, Jiangsu, China
e-mail: raozhm@jiangnan.edu.cn
1 3

X. Zhang et al.
(Curic et al. 1999; GoupilFeuillerat et al. 1997; Monnet
et al. 2000; Monnet and Corrieu 2007; Repizo et al. 2011).
In the cell, synthesis of ALDC at high levels can have a
negative effect on the valine and leucine synthesis. The cell
can often adjust its amount of ALDC in response to dif-
ferent environmental conditions. It is assumed that ALDC
may be required at high levels during carbon limitation or
aerobiosis conditions that induce a shift from lactic homo-
fermentation to mixed acid fermentation, and concomitant
with high fluxes to acetoin (Goupil-Feuillerat et al. 2000).
The utilization of ALDC had also been expanded for the
synthesis of enantiomerically pure diols by Yan et al.
(2009).
L3-15) showed a relatively high enzyme activity compared
with other ALDC. To satisfy its application in the beer
fermentation, the SaALDC gene saald was cloned and
expressed in a safe host strain Bacillus subtilis WB600.
The recombinant SaALDC was purified and characterized.
For further use of this enzyme in beer maturation, its acidic
stability was improved by site-directed mutagenesis. The
recombinant enzyme activity was increased significantly by
scale-up fermentation in the 5-L fermenter, and this study
is of great importance to industrial application of ALDC.
Materials and methods
Recently, the enzyme ALDC has been reported to
be found in several bacteria, such as Lactococcus lactis
(Goupil-Feuillerat et al. 2000), Enterobacter aerogenes
(Sone et al. 1988), Streptococcus lactis (Goelling and Stahl
1988), Bacillus brevis (Diderichsen et al. 1990), Bacillus
licheniformis (Qin et al. 2000), Bacillus subtilis (Guo et al.
2001), Acetobacter aceti (Yamano et al. 1994) and Leucon-
ostoc oenos (Garmyn et al. 1996). ALDC from different
bacteria showed different physical and chemical properties.
The optimum pH and Km value of ALDC from Lactococ-
cus lactis were found to be pH 6.0 and 20 mM, respec-
tively. And this enzyme could be inhibited by Mg2+, Mn2+,
Ba2+, Ca2+, Fe2+, Zn2+ and Sn2+ inordinately (Kisrieva
et al. 2000). Researches from O’Sulivan revealed that the
purified ALDC from Leuconostoc lactis had an optimum
pH of 6.0, pI of 4.2 and Km value of 1.3 mM (O’Sullivan
et al. 2001), while in Rasmussen’s research, the optimum
pH and temperature of the ALDC form Lactobacillus casei
were pH 4.5 and 40 °C, respectively. The enzyme, with the
molecular weight of 48 kDa and Km value of 4.8 mM, could
be inhibited by 1,10-phenanthroline (Phalip et al. 1994).
In this work, we isolated the ALDC from Staphylococcus
aureus L3-15. The SaALDC (ALDC from S. aureus
Strains, plasmids and chemicals
The strains and plasmids used in this study are shown in
Table 1. Staphylococcus aureus was screened and stored in
our laboratory. Escherichia coli JM109, B. subtilis WB600
and the shuttle expression plasmid pMA5 were preserved
in our laboratory. The cloning vector pMD18-T and restric-
tion enzymes were purchased from TaKaRa Co., Ltd.
(Dalian, China). T4 DNA ligase and 4-morpholineethane-
sulfonic acid (MES) monohydrate were purchased from
Sangon Biotech Co., Ltd. (Shanghai, China). Ethyl-2-ace-
toxy-2-methylacetocetate was purchased from Sigma
Aldrich Co. LLC (Shanghai, China). The Fast Mutagenesis
System was purchased from TransGen Biotech Co., Ltd.
(Beijing, China). All other chemicals of high grade were
obtained from commercial source.
Strain cultivation
Staphylococcus aureus and B. subtilis were grown at 37 °C
in Luria–Bertani (LB) medium. Kanamycin with the con-
centration of 50 mg L−1 was added to the culture medium
Table 1 Characteristics, source or reference of the strains and plasmids using in this study
Strains/plasmids
Characteristics
Source or reference
Strainis
Staphylococcus aureus
E. coli JM109
The source of saald
Our lab
Host for gene cloning
TaKaRa
Our lab
B. subtilis WB600
B. subtilis WB600/pMA5
B. Subtilis WB600/pMA5-saald
Plasmids
Host for gene expression
B. subtilis WB600 harboring plasmid pMA5
B. subtilis WB600 harboring plasmid pMA5-saald
This work
This work
pMD18-T
Cloning vector, 2,692 bp, AmpR, lacZ
B. subtilis expression vector, AmpR in E. coli, KmR in B. subtilis, His-Tag coding
TaKaRa
Our lab
pMA5
sequence, HpaII promoter
pMA5-saald
A derivative of pMA5, KmR, harboring the saald gene
This work
AmpR ampicillin-resistant phenotype, KmR kanamycin-resistant phenotype
1 3

Improving the acidic stability
Table 2 Primes used for site-directed mutagenesis
kanamycin at 37 °C. At the end of the exponential growth
phase, cells were harvested by centrifugation of culture for
15 min at 10,000×g and 4 °C. The cells were washed twice
using MES I buffer (50 mM 4-morpholineethanesulfonic
acid, 0.5 g L−1 Brij35, 0.6 M NaCl, pH 6.0), then
resuspended in the same buffer. The harvested cells were
sonicated and centrifuged at 15,000×g for 30 min at 4 °C
to remove the cell debris. The supernatant was used for
SaALDC enzyme activity assays. SDS-PAGE analysis
(12 % separation gel) was basically performed according
to the method described by Laemmli (1970). The protein
concentration was determined by Bradford method using
BSA as standard protein (Bradford 1976).
Primes
Primer sequences 5′–3′
PN43D
PN43D
PN43E
PN43E
PK52E
PK52E
PK52D
PK52D
F
TACACTAACAGGTTCAGACGGTGAGGTAAT
CTGAACCTGTTAGTGTAGCGATACCTAAGT
TACACTAACAGGTTCAGAAGGTGAGGTAAT
TTCTGAACCTGTTAGTGTAGCGATACCTAA
AATCTTTTTAGATGGAGAAGCTTACCATGC
CTCCATCTAAAAAGATTACCTCACCGTTTG
AATCTTTTTAGATGGAGATGCTTACCATGC
ATCTCCATCTAAAAAGATTACCTCACCGTT
TTAAAAAAATGCATGTAGATATGATGCCGG
TCTACATGCATTTTTTTAAATAAGCCTGAA
TTAAAAAAATGCATGTAGAAATGATGCCGG
TTCTACATGCATTTTTTTAAATAAGCCTGA
R
F
R
F
R
F
R
PR124D
PR124D
PR124E
PR124E
F
R
F
R
Measurement of α-acetolactate decarboxylase activity
Mutation sites are shown in bold and underlined parts
The substrate for SaALDC enzyme assay, α-acetolactate,
was generated by hydrolyzing ethyl-2-acetoxy-2-methy-
lacetocetate according to the reported method (Garmyn
et al. 1996; Loken and Stormer 1970) with minor modifi-
cation. 100 mg ethyl-2-acetoxy-2-methylacetocetate was
hydrolyzed with 6.0 mL 0.5 M NaOH at 20 °C for 30 min.
The pH was adjusted to 6.0 with 0.5 mM HCl. The result-
ing solution was diluted to 50 mL with MES II (50 mM
4-morpholineethanesulfonic acid, pH 6.0), defined as sub-
strate solution and used for α-acetolactate decarboxylase
activity assay immediately.
if necessary. For growing on solid medium, 20 g L−1 agar
was added. Bacillus subtilis was grown at 37 °C in LB in
the phase of engineering bacteria construction and then
grown at 37 °C in fermentation medium (g L−1): soybean
tryptone, 10; glucose, 50; corn steep liquor, 15; urea, 3;
K2HPO4·3H2O, 1.7; KH2PO4, 2.3; MgSO4·7H2O, 0.75;
NaCl, 50; pH 6.8–7.0.
Cloning of the gene and site-directed mutagenesis
Diacetyl could be produced from acetolactate
decarboxylase by oxidation of acetoin under extreme
alkaline conditions. The diacetyl formed in turn reacted with
the guanine group of creatine to produce a strong red color in
the presence of naphthol (Garmyn et al. 1996). The reaction
mixture contained 200 µL substrate solution and 200 µL
enzyme solution. The whole reaction was carried out for
30 min at 20 °C and pH 6.0, The mixture was incubated at
30 °C for 10 min and then 4.6 mL creatine-naphthol reagent
(α-naphthol, 5 g L−1; creatine monohydrate, 0.5 g L−1;
NaOH, 10 g L−1) was added. The absorbance of the resulting
reaction mixture was measured at 520 nm. A control reaction
was run without enzyme. One unit of SaALDC activity was
defined as 1 μmol acetoin produced per minute.
The gene saald was obtained from S. aureus chromosome
and amplified by PCR using the forward (5′-ATCGGATC
CATGACGAATGTCTTGTATC) and reverse (5′-TGACTGC
GGGGCCTTCAGCTTCTCTAATTT) primers (underlined
parts were the restriction sites). The BamH I/Not I site
of saald fragment was cloned onto the plasmid pMA5.
Transformation to the B. subtilis WB600 cell was carried
out by the procedure described by Spizizen (Spizizen 1958).
The recombinant strain B. subtilis WB600/pMA5-saald
was selected by resistance to kanamycin. B. subtilis WB600
harboring the empty plasmid pMA5 was used as control.
Site-directed mutagenesis on the saald was carried out by
using the overlapping PCR-based method (Sambrook and
Russell 2001). All mutants with single substitutions were
constructed with plasmid encoding wild-type saald as the
template. The overlap extension PCR for the generation of
mutants was performed using the primers listed in Table 2
and followed the instructions of Fast Mutagenesis System.
The successful introductions of the desired mutations were
confirmed by sequencing.
SaALDC purification and characterization
The supernatant of cell extraction from B. subtilis strains
were loaded onto a Ni-NTA column. After allowing bind-
ing to proceed for 30 min, the resin was washed with 20
column volumes of buffer McAc-0 (20 mM Tris pH 8.0,
500 mM NaCl, 100 g L−1 glycerol) and eluted with 10 mL
buffer McAc-100 (McAc-0 added with 100 mM imida-
zole). The eluted protein was monitored by SDS-PAGE.
To determine the optimal temperature for SaALDC, the
purified enzyme was assayed from 20 to 80 °C in buffer
Expression of SaALDC in B. subtilis WB600
The recombinant strain B. subtilis WB600/pMA5-
saald was grown in LB medium containing 50 μg mL−1
1 3

X. Zhang et al.
MES I (pH 6.0). The following buffer systems were used to
investigate the pH dependence of SaALDC: 50 mM sodium
acetate buffer (pH 4.0–6.0), 50 mM phosphate buffered
saline buffer (PBS buffer, pH 6.0–8.0) and 50 mM glycine-
NaOH buffer (pH 8.0–11.0). The enzyme stability was
investigated using the standard ALDC assay after it had
been incubated for certain time under different conditions.
shown). Among these bacteria, S. aureus had relatively
high-yield production of ALDC. However, S. aureus was
pathogenic bacterium from which the enzyme product
could not be used for beer fermentation. Because of that,
on the GRAS (generally regarded as safe) list of Food
and Drug Administration (Schallmey et al. 2004), B. sub-
tilis was selected as a safe host bacterium for the produc-
tion of many new and improved products (Kunst et al.
1997). The ALDC encoding gene saald obtained from S.
aureus genome was cloned to B. subtilis WB600 with vec-
tor pMA5, and the recombinant strain B. subtilis WB600/
pMA5-saald was constructed. The B. subtilis WB600 trans-
formed with vector pMA5 (B. subtilis WB600/pMA5) was
used as the control. The gene saald had an open reading
frame of 705 bp that encodes a protein of 234 amino acids.
Comparing with those reported proteins in NCBI database
(http://www.ncbi.nlm.nih.gov/protein/312830942), ALDC
of this work shared a 100 % similarity with that from
S. aureus subsp. aureus ECT-R 2 (accession No.FR714927)
in amino acid sequence.
The kinetic parameters of SaALDC
The reaction was performed in MES I (pH 6.0) at 30 °C
for the determination of kinetic parameters. Assays were
performed with active enzyme and α-acetolactate with dif-
ferent concentrations (10–500 μM). The Eadie-Hofstee
plots were used to calculate kinetic parameters Km and Vmax
according to the reactions.
Fermentation of the recombinant B. subtilis
The recombinant B. subtilis strain was cultured in LB
medium supplemented with 50 μg mL−1 kanamycin at
37 °C for 16 h; then it was inoculated into fermentation
medium for 48 h at 37 °C. Furthermore, the fermentation
was scaled up in a 5-L fermenter with a working volume
of 2.5 L for 60 h at 37 °C with optimized fermentation
medium (g L−1): glucose, 50; soybean tryptone, 12; corn
steep liquor, 20; urea, 3; K2HPO4·3H2O, 3.6; KH2PO4,
4.9; MgSO4·7H2O, 0.75; NiSO4·6H2O, 0.3; NaCl, 5; pH
6.8–7.0. The fermentation was carried out with the stir-
ring speed of 300 rpm and the ventilation volume of 1
vvm. Glucose was supplemented when its concentration
was under 10 g L−1. The cell density and SaALDC activity
were measured over the fermentation period.
Functional expression of SaALDC in B. subtilis
SDS-PAGE analysis (Fig. 1) showed that one obvious band
of about 28 kDa was observed in the cell-free extracts of
B. subtilis WB600/pMA5-saald. So, the enzyme SaALDC
was expressed as a soluble protein. The recombinants
were grown in 250 mL shake flasks containing 50 mL LB
medium at 37 °C for 16 h. The SaALDC activity in B. sub-
tilis WB600/pMA5-saald was 36.0 U mL−1, improved by
about 90 times compared to 0.42 U mL−1 of S. aureus and
120 times compared to 0.3 U mL−1 of the control strain B.
Measurement of diacetyl in beer fermentation
Beer fermentation with or without 100 μL L−1 SaALDC
enzyme was conducted in the 5 L fermenter, following by
the description of Lei et al. (Lei et al. 2013). Fermentation
duration was 20 days and the temperature kept at 10 °C.
Determination of diacetyl was conducted during the fer-
mentation period. Diacetyl was evaporated by the distilla-
tion of beer sample. Then it was reacted with o-phenylen-
ediamine to produce 2,3-dimethylchinaxolin which was
absorbed at 335 nm.
Results and discussion
Cloning of ALDC from S. aureus
Fig. 1 SDS-PAGE analysis of SaALDC expression and purification
in B. subtilis WB600. Lane M molecular weight markers. Lane 1 The
purified recombinant SaALDC; Lane 2 B. subtilis/pMA5-saald; Lane
3 B. subtilis/pMA5
More than 20 bacteria that could produce ALDC were
previously isolated from nature samples (data were not
1 3

Improving the acidic stability
subtilis WB600/pMA5. These results suggested that the
enzyme was functionally expressed in B. subtilis W600.
Before we cloned the SaALDC gene to B. Subtilis, we also
expressed the SaALDC in E. coli. But the specific activity
of SaALDC in E. coli was only 12.5 U mL−1. By compari-
son of E. coli, B. Subtilis demonstrated itself a good host
bacteria for SaALDC.
A
100
75
50
25
0
Enzymatic properties of recombinant SaALDC
The purification of SaALDC from the recombinant
strain B. subtilis WB600/pMA5-saald was performed
by a Ni-NTA column (Fig. 1). The effect of pH on
SaALDC activity was measured at different pH ranging
from 4.0 to 11.0. The recombinant SaALDC exhibited
the maximum activity at pH 6.0 (Fig. 2a). SaALDC
could catalyze α-acetolactate to acetoin in acidic pH
conditions. However, when pH was lower than 5.0 or
higher than 7.0, the SaALDC activity declined rapidly
and was completely inactivated at pH 2.0–4.0, and the
recombinant SaALDC in acidic pH conditions was not as
stable as at pH 7.0–9.0 (Fig. 2b). Comparing to more than
75 % retained enzyme activity at pH 7.0–9.0 after pre-
incubated for 10 h, the enzyme activity at pH 6.0 retained
only 41 % after 10 h and decreased much more rapidly
at pH 5.0 after 2.5 h. The optimal temperature for the
recombinant SaALDC activity was between 40 and 50 °C
with a maximum temperature at 45 °C at pH 6.0 (Fig. 3a).
Studies on temperature stability were conducted with the
temperature ranging from −20 to 65 °C. The SaALDC
activity retained nearly 90 % at −20 °C and about 80 %
from 0 to 35 °C after 10 h, while the rapid inactivation
was observed over 45 °C (Fig. 3b).
3
6
9
12
pH
pH 5.0
pH 8.0
pH 6.0
pH 9.0
pH 7.0
pH 10.0
B
100
75
50
25
0
0.0
2.5
5.0
7.5
10.0
Time (h)
Fig. 2 Effect of pH on recombinant SaALDC activity. a Optimum
pH was decided by assaying the recombinant SaALDC activity
at 30 °C with pH ranging from 4.0 to 11.0. b Stability of SaALDC
under different pH conditions. After incubating the enzyme in differ-
ent pH values (5.0–10.0) at 4 °C, the residue activities were deter-
mined at pH 6.0 and 30 °C. (All assays were performed by triplicate
cultures; standard deviations of the biological replicates are repre-
sented by error bars)
The kinetics of the recombinant SaALDC was
analyzed using α-acetolactate as substrate with different
concentrations. The Km and Vmax values were determined
by nonlinear fit analysis based on Eadie-Hofstee plots.
The recombinant SaALDC catalyzed α-acetolactate
decarboxylation with Michaelis constant (Km) of 17.7 μM
and maximum reaction rate (Vmax) of 2.06 mM min−1
(Fig. 4).
that was similar to α-acetolactate decarboxylase was
identified as having the highest similarity to SaALDC.
The estimated accuracy model by I-TASSER has the
TM-score of 0.99
0.04 and RMSD (root-mean-square
Modeling and site-directed mutagenesis of SaALDC
deviation) of 1.6 1.4 Å (TM-score and RMSD are known
standards for measuring structural similarity between two
structures) and has the C-score (confidence score of the
model) = 2. In general, the models with C-score > −1.5
have a correct fold, and TM-score >0.5 indicates the
correct topology of the predicted models. With low
C-scoreEC (confidence score of the enzyme commission
number) and no consensus of the EC numbers among the
identified analogs in the function prediction by I-TASSER,
we were advised to consult the GO (gene ontology) term
predictions, wherein the analogs were associated with the
The amino acid sequences of SaALDC were subjected
to the SWISS-MODEL Workspace (http://swissmodel.
expasy.org/workspace/) (Arnold et al. 2006; Biasini et al.
2014; Bordoli et al. 2009) and the I-TASSER server (http://
zhanglab.ccmb.med.umich.edu/I-TASSER/) (Roy et al.
2010, 2012; Zhang 2008) to obtain a relative accurate
tertiary-structure model. After online prediction and
simulation, the crystal structure of the unknown function
protein (PDB ID: 1xv2) with a resolution of 2.00 Å
1 3

X. Zhang et al.
A
100
80
60
40
20
0
20
30
40
50
60
70
80
Temperature (oC)
Fig. 4 Lineweaver-Burk plot for kinetics determinations of purified
SaALDC using α-acetolactate as the substrate
B
100
75
50
25
0
was also proposed. By amino acid sequences alignment of
4tb2, 1xv2 and SaALDC, they are dramatically conserved
in the predicted active center sides (E62, R142 and E251
of BbALDC was involved in the catalytic mechanism,
which were corresponded to E45, R124 and E234 of
SaALDC; the highly conserved H194, H196 and H207 of
BbALDC coordinate a Zn2+ ion, corresponded to H173,
H175 and H186 of SaALDC). Thus, the tertiary structure
of 4tb2 was used as the template for SaALDC molding and
mutagenesis.
-20 oC
25 oC
45 oC
65 oC
0 oC
35 oC
55 oC
0
2
4
6
8
10
The conventional method for down-shifting the pH opti-
mum was to change the basic residues to acidic residues
(Huang et al. 2012). To select the candidate residues, the
amino acid sequences alignment of several α-acetolactate
decarboxylases was conducted (Fig. 5), in which the highly
conserved sequences of K52 have the probability to be sub-
stituted with either aspartic acid or glutamic acid residues.
The recombinant SaALDCK52D and SaALDCK52E were
constructed. The second substitution of N43D was also
constructed to obtain recombinant SaALDCN43D, because
the glutamic acid at this site was conserved in other two
ALDCs. Due to the divergence of predicted active residues
(R124 and E234) between 1xv2 and 4tb2, the basic residue
Time (h)
Fig. 3 Effect of temperature on recombinant SaALDC activity. a
Optimum temperature was examined by assaying the recombinant
SaALDC activity at various temperatures ranging from 20 °C to
80 °C. b Thermal stability of the purified enzymes was determined
by pre-incubation at different temperatures from −20 to 65 °C, then
assaying the residue activity under standard assay conditions. (All
assays were performed by triplicate cultures; standard deviations of
the biological replicates are represented by error bars)
highest level of molecular function, biological process and
cellular location in the GO hierarchy. The model with the
highest C-scoreGO (confidence score of GO term) of 0.80
was also associated with the PDB hit of 1xv2 and was
assigned to the acetolactate decarboxylase activity (GO:
0047605) and polyol metabolic process (GO: 0019751).
The predicted binding site residues were E45, H173, H175
and H186, which should be avoided while the enzyme was
restructured by site-directed mutagenesis.
The amino acid residues involved in the decarboxylation
and rearrangement steps of ALDC with the PDB
deposition of 4bt2 (1.10 Å) from Bacillus brevis had
been experimentally identified (Marlow et al. 2013). The
catalytic mechanism of BbALDC on its molecular-level
of R124 was also substituted to construct SaALDCR124D
,
SaALDCR124E
.
Enzymatic stability at low pH conditions of mutants
The pH sensitivities of the wild-type WT (SaALDC) and
the mutants, with respect to their functional stability, were
studied by incubating the enzymes for 2.5 h at various pH
values ranging from pH 4.0 to 8.0 followed by measurement
of ALDC activity at pH 6.0 (Fig. 6). The wild SaALDC
showed very unstable property at low pH conditions; only
6.0 and 4.0 % activity left after it had been incubated for
1 3

Improving the acidic stability
Fig. 5 Amino acid sequences
alignment of SaALDC, BbA-
LDC (4tb2) and 1xv2. Highly
conserved residues (E45, R124,
H173, H175, H186 and E234)
are highlighted with back-
ground, the catalytic related
residues are indicated by rectan-
gle, and the mutation sites (N43,
K52 and R124) are indicated
by rectangle with filled triangle
below
WT
K52D
K52E
N43D
WT
N43D-K52D
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
pH 4.0
0
1
2
3
4
Time (h)
4
5
6
7
8
4
5
6
7
8
pH
pH
Fig. 6 Effect of pH (ranging from 4.0 to 8.0) on enzyme stability
of mutants K52D, K52E and N43D. (All assays were performed by
triplicate cultures; standard deviations of the biological replicates are
represented by error bars)
Fig. 7 Comparison of effect of pH (ranging from 4.0 to 8.0) on
enzyme stability of mutant N43D-K52D and WT. The half-lives of
N43D-K52D and WT at pH 4.0 were also studied. (All assays were
performed by triplicate cultures; standard deviations of the biological
replicates are represented by error bars)
2.5 h at pH 5.0 and 4.0, respectively. On the contrary,
this enzyme was very stable at neuter and alkalescence
conditions (pH 6.0–8.0). After site-directed mutagenesis,
mutants of N43D, K52D and K52E have almost no
variation on enzyme activity, but exhibited improved
stability over pH 4.0–5.0 than the WT. After pre-incubation
for 2.5 h at pH 5.0, N43D, K52D and K52E retained 56,
53 and 47 % of their enzyme activity, respectively, whereas
their residue activity was 38, 36 and 33 % at pH 4.0,
respectively. The results indicated that the acidic stability
of the enzyme has been improved after rationally changing
basic residues to acidic ones. As might have been expected,
mutants of R124D and R124E retained only about 20 % of
its activity at pH 6.0 and about 10 % at pH 4.0, indicating
a basic residue is necessary for catalysis in this position
(Marlow et al. 2013). Based on the significant improvement
of the acidic stability of N43D and K52D, mutant of
N43D-K52D was constructed, purified and identified by
mass spectrometry compared to the WT (Fig. 7; Fig. S1).
1 3

X. Zhang et al.
The N43D-K52D mutant showed dramatically improved
acidic stability at pH 4.0 (retained 63 % activity) and pH
5.0 (retained 79 % activity) after incubated for 2.5 h, and
its half-life at pH 4.0 was prolonged from 1.5 to 3.5 h;
meanwhile, the range of the enzyme stability at various pH
conditions was broadened.
fermentation, the total SaALDCN43D-K52D activity reached
the highest value of 66.5 U mL−1. Then the fermentation
was scaled up to a 5-L fermenter with a working volume of
2.5 L. In the growth phase, the maximum cell concentration
in the fermenter was 25.3 (OD600). The SaALDCN43D-K52D
enzyme reached the maximum of 135.8 U mL−1 at 48 h
(Fig. 9), which was 320-fold excess of S. Aureus. To our
knowledge, the recombinant SaALDCN43D-K52D showed
a very high enzyme activity. This could be attributed to
not only the screening of bacterium for high-yield ALDC
production but also the successful expression of saald gene
to the right host bacterium B. subtilis.
Structure analysis of wild SaALDC and mutants
The structure of SaALDC and SaALDCN43D-K52D was mod-
eled using the crystal structure of BbALDC (PDB ID: 4tb2)
from Bacillus brevis as template. The resulting models were
used to explain the differences in acidic stability between
the mutants and WT in terms of structure using the soft-
ware of PyMOL. The electrostatic surfaces of the WT and
mutant enzyme were modeled (shown in Fig. 8). The charge
distribution of the WT enzyme showed a negative-positive–
negative amino acid residue chain (GLU64-LYS52-ASP50-
LYS113-GLU198-LYS196-GLU86) (Fig. 8a). The mutant
enzyme gives rise to a negative potential on the structure
surface around the mutant site (K52D) by disruption of
the negative–positive–negative chain, which significantly
changes the modeled electrostatic potential on the surface
of the protein (Fig. 8b), so the resistance to the acidic con-
dition of the proteins is increased.
Analysis of the hydrogen-bond interaction between
the mutant site of N43D and K52D and their surround-
ing amino acid residues showed that N43D increased the
hydrogen-bond numbers of the surface structure of the cat-
alytic center from 11 to 14 while K52D showed no changed
hydrogen-bond numbers (Fig. 8c, d). Since increasing the
hydrogen-bond numbers can contribute to the enzyme sta-
bility (Joshi et al. 2000; Le Nours et al. 2003; Yang et al.
2013), the mutation also increased the rigidity of this struc-
ture, in which GLU45 (catalysis related amino acid) has
two interaction hydrogen-bonds between N43 and D43,
respectively. However, unexpected, the substitution of N43
by D43 decreased the interaction hydrogen-bonds of the
catalytic center from 8 to 7 by losing the hydrogen-bond
interaction between R124 and E234 (Fig. 8e, f), which
may increase the flexibility for substrate docking and also
improve the acidic stability of this enzyme (Liu et al. 2014;
Marlow et al. 2013). Taken together, the results showed
that the acidic stability of SaALDC had been improved
after changing amino acids on the structure surface into the
acidic ones.
Formation of diacetyl with SaALDCN43D-K52D
The effect of SaALDCN43D-K52D enzyme on beer fermenta-
tion can be seen in Fig. 10. Addition of SaALDCN43D-K52D
accelerated the transformation of α-acetolactate to ace-
toin. The maximum concentration of diacetyl in beer fer-
mentation with SaALDCN43D-K52D was 0.35 mL L−1 and
appeared at the fourth day of fermentation, which indi-
cated that the enzymatic removal of α-acetolactate served
to reduce the content of diacetyl indirectly. Sometimes the
content of diacetyl rises during the storage stage of fin-
ished beer, attributed to the presence of a certain amount
of α-acetolactate. This could also be improved by adding of
ALDC enzyme.
As one approach to prevent the formation of diacetyl,
adding the enzyme ALDC requires an ALDC high produc-
ing microorganism which is also safe for food industry.
Then for the first time, the recombinant SaALDC with a
high activity of 135.8 U mL−1 was expressed in B. Subtillis,
a very safe strain for industrial application. It can be shown
that the enzyme SaALDCN43D-K52D can be used in beer fer-
mentation with improved acidic stability. The results dem-
onstrate that highly active recombinant SaALDCN43D-K52D
can be produced at sufficient levels and can thus be used
as the biocatalyst for the reduction of diacetyl in a cost-
effective way. Moreover, a further investigation with an aim
of improving the enzyme SaALDC stability and activity in
fermentation is in progress.
Conclusions
Heterologous over-expression of ALDC from S. aureus
in B. subtilis WB600 was successfully achieved in this
work. The recombinant enzyme SaALDC was expected
to promote mature beer by reducing of diacetyl. However,
SaALDC showed very low enzyme stability at low
pH conditions (pH 4.0), which is needed during beer
maturation process. Based on protein molecular structure
modeling and site-directed mutagenesis, the mutant protein
Fermentation for the recombinant variant
SaALDCN43D-K52D production in 5 L fermenter
The recombinant strain B. subtilis WB600/pMA5-saaldN43D-
K52D was cultivated for 48 h at 37 °C in 250 mL shaker flasks
containing 50 mL fermentation medium at first. During the
1 3

Improving the acidic stability
Fig. 8 Structure overview of SaALDC and SaALDCN43D-K52D. a
Charge distribution on the surface of SaALDC; b charge distribution
on the surface of SaALDCN43D-K52D, in which K52D showed signifi-
cantly changed charge distribution; c interaction hydrogen bonds of
N43 and its nearby residues; c interaction hydrogen bonds of D43
and its nearby residues; e interaction hydrogen bonds of catalytic-
related residues, Zn2+ and N43; f interaction hydrogen bonds of cat-
alytic-related residues, Zn2+ and D43, in which N43D resulted in the
loss of hydrogen bond between E45 and E234. The hydrogen bonds
are indicated by broken pink lines. This figure was generated using
PyMOL software (color figure online)
1 3

X. Zhang et al.
30
25
20
15
10
5
150
120
90
60
30
0
70
60
50
40
30
20
10
0
Development of Jiangsu Higher Education Institutions, the 111 Project
(No. 111-2-06), and the Jiangsu province “Collaborative Innovation
Center for Advanced Industrial Fermentation” industry development
program. We thank for Rongxia Zhang and Lab of Brewing Science
and Technology of Jiangnan University for performing the experiments
Glucose concentration
SaALDC activity
OD600
of beer fermentation with SaALDCN43D-K52D
.
Conflict of interest The authors declare no conflict of interest.
References
0
0
12
24
36
48
60
Time (h)
Arnold K, Bordoli L, Kopp J, Schwede T (2006) The SWISS-
MODEL workspace: a web-based environment for protein struc-
ture homology modelling. Bioinformatics 22:195–201
Biasini M et al (2014) SWISS-MODEL: modelling protein tertiary
and quaternary structure using evolutionary information. Nucleic
Acids Res 42:W252–W258
Bordoli L, Kiefer F, Arnold K, Benkert P, Battey J, Schwede T (2009)
Protein structure homology modeling using SWISS-MODEL
workspace. Nat Protoc 4:1–13
Fig. 9 Fermentation curves of B. subtilis WB600/pMA5-saaldN43D-K52D
in a 5-L fermenter. (All assays were performed by triplicate cultures;
standard deviations of the biological replicates are represented by error
bars)
1.5
Bradford MM (1976) A rapid and sensitive method for the quantita-
tion of microgram quantities of protein utilizing the principle of
protein-dye binding. Anal Biochem 72:248–254
Without enzyme
With SaALDCN43D/K52D
Curic M, Stuer-Lauridsen B, Renault P, Nilsson D (1999) A general
method for selection of alpha-acetolactate decarboxylase-defi-
cient Lactococcus lactis mutants to improve diacetyl formation.
Appl Environ Microbiol 65:1202–1206
Diderichsen B, Wedsted U, Hedegaard L, Jensen BR, Sjoholm C
(1990) Cloning of aldB, which encodes alpha-acetolactate
decarboxylase, an exoenzyme from Bacillus brevis. J Bacteriol
172:4315–4321
Dulieu C, Moll M, Boudrant J, Poncelet D (2000) Improved perfor-
mances and control of beer fermentation using encapsulated
alpha-acetolactate decarboxylase and modeling. Biotechnol Prog
16:958–965
Garmyn D, Monnet C, Martineau B, Guzzo J, Cavin JF, Divies C
(1996) Cloning and sequencing of the gene encoding alpha-ace-
tolactate decarboxylase from Leuconostoc oenos. FEMS Micro-
biol Lett 145:445–450
1.0
0.5
0.0
0
4
8
12
16
20
Fermentation time (d)
Fig. 10 Formation of diacetyl in beer without (filled square) and
with 100 μL L−1 recombinant SaALDC enzyme (filled circle) (All
assays were performed by triplicate cultures; standard deviations of
the biological replicates are represented by error bars)
Goelling D, Stahl U (1988) Cloning and expression of an alpha-
acetolactate decarboxylase gene from Streptococcus lactis
subsp. diacetylactis in Escherichia coli. Appl Environ Microbiol
54:1889–1891
GoupilFeuillerat N, CocaignBousquet M, Godon JJ, Ehrlich SD,
Renault P (1997) Dual role of alpha-acetolactate decarboxylase
in Lactococcus lactis subsp. lactis. J Bacteriol 179:6285–6293
Goupil-Feuillerat N, Corthier G, Godon JJ, Ehrlich SD, Renault P
(2000) Transcriptional and translational regulation of alpha-ace-
tolactate decarboxylase of Lactococcus lactis subsp. lactis. J Bac-
teriol 182:5399–5408
Guo W, He X, Tie C, Zhang B (2001) Expression of alpha-acetolac-
tate decarboxylase gene from Bacillus subtilis in Brewer’s yeast.
Wei Sheng Wu Xue Bao 41:105–108
of SaALDCN43DK52D showed improved acidic stability
and prolonged half-live at pH 4.0. Structure analysis of
the mutant enzyme exhibited that mutation of K52D gave
rise to the negative potential on the structure surface while
mutation of N43D increased the hydrogen-bond interaction
of the enzyme. This work provides reference for improving
acidic stability of ALDC. The successful expression of
saald gene in B. subtilis indicates it a suitable host for
scale-up ALDC production.
Huang L et al (2012) Improving the acidic stability of a methyl par-
athion hydrolase by changing basic residues to acidic residues.
Biotechnol Lett 34:1115–1121
Joshi MD, Sidhu G, Pot I, Brayer GD, Withers SG, McIntosh LP
(2000) Hydrogen bonding and catalysis: a novel explanation for
how a single amino acid substitution can change the pH optimum
of a glycosidase. J Mol Biol 299:255–279
Kisrieva YS, Serebrennikov VM, Zagustina NA, Bezborodov AM
(2000) Isolation and purification of acetolactate synthase and ace-
tolactate decarboxylase from the culture of Lactococcus lactis.
Appl Biochem Microbiol 36:109–114. doi:10.1007/Bf02737903
Acknowledgments This work was supported by the National
Basic Research Program of China (973 Program) (12CB720806), the
National Natural Science Foundation of China (21276110), the High-
tech Research and Development Programs of China (2014AA021304),
the Research Project of Chinese Ministry of Education (113033A),
the Fundamental Research Funds for the Central Universities
(JUSRP51306A), the Project Funded by the Priority Academic Program
1 3

Improving the acidic stability
Repizo GD, Mortera P, Magni C (2011) Disruption of the alsSD
operon of Enterococcus faecalis impairs growth on pyruvate at
low pH. Microbiology 157:2708–2719
Roy A, Kucukural A, Zhang Y (2010) I-TASSER: a unified platform
for automated protein structure and function prediction. Nat Pro-
toc 5:725–738
Roy A, Yang JY, Zhang Y (2012) COFACTOR: an accurate compara-
tive algorithm for structure-based protein function annotation.
Nucleic Acids Res 40:W471–W477
Sambrook J, Russell DW (2001) Molecular cloning: a laboratory
manual. Cold spring harbor laboratory press
Schallmey M, Singh A, Ward OP (2004) Developments in the use
of Bacillus species for industrial production. Can J Microbiol
50:1–17
Sone H, Fujii T, Kondo K, Shimizu F, Tanaka J, Inoue T (1988)
Nucleotide sequence and expression of the Enterobacter aero-
genes alpha-acetolactate decarboxylase gene in brewer’s yeast.
Appl Environ Microbiol 54:38–42
Spizizen J (1958) Transformation of biochemically deficient strains of
Bacillus subtilis by deoxyribonucleate. Proc Natl Acad Sci USA
44:1072–1078
Kunst F et al (1997) The complete genome sequence of the gram-pos-
itive bacterium Bacillus subtilis. Nature 390:249–256
Laemmli UK (1970) Cleavage of structural proteins during the assem-
bly of the head of bacteriophage T4. Nature 227:680–685
Le Nours J, Ryttersgaard C, Lo Leggio L, Ostergaard PR, Borchert
TV, Christensen LL, Larsen S (2003) Structure of two fungal
beta-1,4-galactanases: searching for the basis for temperature and
pH optimum. Protein Sci 12:1195–1204
Lei H, Zheng L, Wang C, Zhao H, Zhao M (2013) Effects of worts
treated with proteases on the assimilation of free amino acids and
fermentation performance of lager yeast. Int J Food Microbiol
161:76–83. doi:10.1016/j.ijfoodmicro.2012.11.024
LiuY, Cui W, Liu Z, CuiY, XiaY, Kobayashi M, Zhou Z (2014) Effect
of flexibility and positive charge of the C-terminal domain on the
activator P14 K function for nitrile hydratase in Pseudomonas
putida. FEMS Microbiol Lett 352:38–44
Loken JP, Stormer FC (1970) Acetolactate decarboxylase from Aero-
bacter aerogenes. Purif Prop Eur J Biochem 14:133–137
Marlow VA, Rea D, Najmudin S, Wills M, Fulop V (2013) Structure
and mechanism of acetolactate decarboxylase Acs. Chem Biol
8:2339–2344
Suihko ML, Blomqvist K, Penttila M, Gisler R, Knowles J (1990)
Recombinant brewer’s yeast strains suitable for accelerated brew-
ing. J Biotechnol 14:285–300
Monnet C, Corrieu G (2007) Selection and properties of alpha-ace-
tolactate decarboxylase-deficient spontaneous mutants of Strepto-
coccus thermophilus. Food Microbiol 24:601–606
Yamano S, Tanaka J, Inoue T (1994) Cloning and expression of the
gene encoding alpha-acetolactate decarboxylase from Acetobacter
aceti ssp. xylinum in brewer’s yeast. J Biotechnol 32:165–171
Yan Y, Lee CC, Liao JC (2009) Enantioselective synthesis of pure (R,
R)-2,3-butanediol in Escherichia coli with stereospecific secondary
alcohol dehydrogenases. Org Biomol Chem 7:3914–3917
Yang H, Liu L, Shin HD, Chen RR, Li J, Du G, Chen J (2013)
Structure-based engineering of histidine residues in the catalytic
domain of alpha-amylase from Bacillus subtilis for improved
protein stability and catalytic efficiency under acidic conditions.
J Biotechnol 164:59–66
Monnet C, Aymes F, Corrieu G (2000) Diacetyl and alpha-acetolac-
tate overproduction by Lactococcus lactis subsp lactis biovar dia-
cetylactis mutants that are deficient in alpha-acetolactate decar-
boxylase and have a low lactate dehydrogenase activity. Appl
Environ Microbiol 66:5518–5520
O’Sullivan SM, Condon S, Cogan TM, Sheehan D (2001) Purification
and characterisation of acetolactate decarboxylase from Leucon-
ostoc lactis NCW1. FEMS Microbiol Lett 194:245–249
Phalip V, Monnet C, Schmitt P, Renault P, Godon JJ, Divies C (1994)
Purification and Properties of the Alpha-Acetolactate Decarboxy-
lase from Lactococcus-Lactis Subsp Lactis Ncdo-2118. FEBS
Lett 351:95–99. doi:10.1016/0014-5793(94)00820-5
Zhang Y (2008) I-TASSER server for protein 3D structure prediction.
BMC Bioinform 9:14. doi:10.1186/1471-2105-9-40
Qin YJ, Gao D, Wang ZN (2000) Alpha-acetolactate decarboxylase
from B. licheniformis AS10106: cloning and expression gene in
E. coli and S. cerevisiaeYi chuan xue bao =. Acta genetica Sinica
27:165–169
1 3