Stabilization of an amine transaminase for biocatalysis

Article abstract of DOI:10.1016/j.molcatb.2015.11.022

The amine transaminase from Chromobacterium violaceum (Cv-ATA) is a well-known enzyme to achieve chiral amines of high enantiomeric excess in laboratory scales. However, the low operational stability of Cv-ATA limits the enzyme applicability on larger scales. In order to improve the operational stability of Cv-ATA, and thereby extending its applicability, factors (additives, co-solvents, organic solvents and different temperatures) targeting enzyme stability and activity were explored in order to find out how to store and apply the enzyme. The present investigation shows that the melting point of Cv-ATA is improved by adding sucrose or glycerol, separately. Further, by storing the enzyme at higher concentrations and in co-solvents, such as; 50% glycerol, 20% methanol or 10% DMSO, the active dimeric structure of Cv-ATA is retained. Enzyme stored in 50% glycerol at -20 °C was e.g., still fully active after 6 months. Finally, the enzyme performance was improved 5-fold by a co-lyophilization with surfactants prior to usage in isooctane.

Full text of DOI:10.1016/j.molcatb.2015.11.022

Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Stabilization of an amine transaminase for biocatalysis
Shan Chen1, Henrik Land1, Per Berglund∗, Maria Svedendahl Humble∗
KTH Royal Institute of Technology, School of Biotechnology, Division of Industrial Biotechnology, AlbaNova University Center, SE-106 91 Stockholm, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
The amine transaminase from Chromobacterium violaceum (Cv-ATA) is a well-known enzyme to achieve
chiral amines of high enantiomeric excess in laboratory scales. However, the low operational stability
of Cv-ATA limits the enzyme applicability on larger scales. In order to improve the operational stability
of Cv-ATA, and thereby extending its applicability, factors (additives, co-solvents, organic solvents and
different temperatures) targeting enzyme stability and activity were explored in order to find out how to
store and apply the enzyme. The present investigation shows that the melting point of Cv-ATA is improved
by adding sucrose or glycerol, separately. Further, by storing the enzyme at higher concentrations and in
co-solvents, such as; 50% glycerol, 20% methanol or 10% DMSO, the active dimeric structure of Cv-ATA
is retained. Enzyme stored in 50% glycerol at −20 ◦C was e.g., still fully active after 6 months. Finally,
the enzyme performance was improved 5-fold by a co-lyophilization with surfactants prior to usage in
Received 11 September 2015
Received in revised form
25 November 2015
Accepted 26 November 2015
Available online 2 December 2015
Keywords:
Amine transaminase
Amine
Pyridoxal-5ꢀ-phosphate (PLP)
Biocatalysis
isooctane.
© 2015 Elsevier B.V. All rights reserved.
Enzyme stability
1. Introduction
of reactions such as transamination, racemization, decarboxyla-
tion, substitution and elimination [6,7]. One largely represented
Enzymes have acquired extensive attention as catalysts in
applied chemistry due to their wide range of applications in
the synthesis of pharmaceuticals, agro-chemicals, biofuels or fine
chemical building blocks. Although biocatalysts have numerous
advantages, such as high chemo-, regio- and enantioselectivity,
high turnover and ability to operate in mild reaction conditions, the
poor operational stability of enzymes still restricts their application
in industrial process conditions [1,2]. Generally, enzyme stability
can be defined as: thermodynamic (conformational) stability and
kinetic (long-term) stability [1,2]. Thermodynamic stability con-
cerns the reversible unfolding of the protein conformation, while
kinetic stability determines how long enzyme activity is main-
tained before irreversible denaturation occurs. The operational
stability, which involves thermal stability, stability in organic sol-
vents and stability over time, is related to both thermodynamic-
and kinetic stability [3–5]. In order to improve the operational sta-
bility of an enzyme both of these factors need to be considered.
activity among the PLP-dependent enzymes is transamination. The
largest substrate group within that activity is amino acids but
amine transaminases (ATAs) have recently gained a lot of inter-
est as catalysts for the production of chiral amines. ATAs can be
used to perform asymmetric synthesis of chiral amines from the
corresponding prochiral ketones with high yields and excellent
enantioselectivities in mild reaction conditions [8,9]. The opera-
tional stability of ATAs is however poor and attempts to improve
it using various strategies have been reported. Immobilization is a
common method to improve enzyme stability and several methods
for ATA immobilization have been reported over the last decades
[10–23]. Also, directed evolution has been successfully applied to
improve both the stability of Arthrobacter citreus ATA and of an (R)-
selective ATA for the production of substituted 2-aminotetralin [24]
and the anti-diabetic compound Sitagliptin [25]. Both these cases
show that improvement of operational stability is crucial for the
application of ATAs in industrial processes. Another recently pub-
lished method to improve ATA stability is the global incorporation
of fluorotyrosine, which enhanced the stability of Vibrio fluvialis
ATA [26].
Enzymes dependent on the cofactor pyridoxal-5-phosphate
´
(PLP) are a large class of enzymes consisting of several protein fold
types and activities. The versatility of PLP as a cofactor leads to
the PLP-dependent enzymes being able to catalyze a wide range
The ATA from Chromobacterium violaceum (Cv-ATA, CV2025)
is an extensively studied enzyme with a broad substrate scope
toward aromatic and aliphatic amines [27–30]. Recently, the Cv-
ATA activity toward serine was improved by different single point
mutants [31]. We have previously reported on important char-
acteristics of the Cv-ATA enzyme, such as thermostability [32]
and enantioselectivity [33], as well as how to perform active-site
∗
Corresponding authors.
E-mail addresses: per.berglund@biotech.kth.se (P. Berglund),
maria.humble@biotech.kth.se (M.S. Humble).
1
Shared first authorship and contributed equally.
http://dx.doi.org/10.1016/j.molcatb.2015.11.022
1381-1177/© 2015 Elsevier B.V. All rights reserved.

S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
21
titration [34]. However, our previous experiments in harsh condi-
tions have shown that the operational stability of Cv-ATA is low
and the wild type enzyme can therefore not be applied in an indus-
trial process with satisfying results. In addition, we have previously
observed that this enzyme activates over time upon storage in room
temperature, which sometimes results in reproducibility problems.
Previously, we explored the crystal structure of Cv-ATA [32].
To promote protein stabilization prior to crystallization, a buffer
screen was made to explore individual stabilization effects of
various buffers supplemented by additives and/or co-solvents by
melting point (Tm) measurements using differential scanning flu-
orimetry. In that study, the enzyme displayed a high Tm of 78 ◦C
in HEPES (100 mM, pH 7.4, 100 mM NaCl) buffer. The enzyme was
considered to be thermostable due to this high melting point. How-
ever, it turned out that the enzyme is not active and cannot perform
catalysis at this high temperature. Therefore, we decided to further
explore parameters that may affect the stability of Cv-ATA. The use
of additives and co-solvents showed positive stabilization effects in
our previous study and was therefore used as a starting point in this
investigation. The reasons for Cv-ATA activation and inactivation
in buffer solutions have been explored through various stability
investigations combining enzyme activity measurements and the
amount of active dimer present.
(50 mM, pH 8.2) was performed on a PD10 column (GE Healthcare),
according to the manufacturer’s protocol. After adding 3 mM of PLP,
the enzyme was stored in fridge overnight. The following day, a
second buffer change to HEPES buffer (50 mM, pH 8.2) was made
to remove the excess of PLP, after incubation at 37 ◦C for 1 h. The
protein concentration of the enzyme preparation was measured on
NanoDropTM 1000 Spectrophotometer at 280 nm. The enzyme was
stored in fridge prior to use according to the procedures below.
2.3. Melting point measurement
Melting points were measured using differential scanning fluo-
rimetry [45]. Samples were prepared at the desired concentrations
(Fig. 1) in a 96-well PCR plate (BIO-RAD). Enzyme was added to a
final concentration of 1 mg/mL. Finally, 3.75× SYPRO® Orange pro-
tein gel stain (Sigma–Aldrich, S5692) was added and the melting
points were analyzed on a CFX96TM Real-Time PCR Detection Sys-
tem and C1000TM Thermal Cycler. Temperature was set to 25 ◦C
and increased at a rate of 1 ◦C/min until it reached 95 ◦C. Data was
extracted and MS Excel was used to perform a Boltzmann fit of the
data to determine melting points.
2.4. Enzyme activity assay
There are only a few reports of cell-free ATA-catalyzed reactions
in hydrophobic organic solvents [35–39]. Generally, enzymes are
structurally considered to be more rigid in organic solvents. Also,
the enzyme activity is commonly reduced if the enzyme prepa-
ration is not lyophilized [40]. Lyophilization is a well-established
technique to prolong enzyme shelf-life [3–5]. The lyophilization
process involves both freezing and dehydration, which can alter
both enzyme activity and enzyme conformation by removing
waters surrounding the enzyme surface or important structural
waters. In the worst case, the dehydration process can result in pro-
tein denaturation [41,42]. Surfactants have shown a great potential
to shield enzyme structure from denaturation during lyophiliza-
tion [43,44]. Here, we show the first example of the use of an ATA
enzyme co-lyophilized with surfactants which enables cell-free Cv-
ATA-catalyzed transamination reactions in organic solvents to yield
high conversions.
To measure enzyme activity, a Cv-ATA-catalyzed reaction con-
taining 2 ␮g (or 150 ␮g) enzyme, 2.5 mM (S)-1-phenylethylamine
((S)-1-PEA), 2.5 mM pyruvate and HEPES buffer (50 mM, pH 8.2)
with a final volume of 1 mL was performed. The initial rate of prod-
uct (acetophenone) formation was measured at 245 (or 290 nm
when interference occured) [46] on a Cary®50 UV–vis spectropho-
tometer.
2.5. Enzyme storage stability at room temperature or 65 ◦C
A purified enzyme preparation of Cv-ATA (0.1 mg/mL) was
stored at room temperature or at 65 ◦C in a range of selected addi-
tives or water miscible co-solvents. At selected time points, enzyme
samples (2 ␮g) were withdrawn for remaining enzyme activity
measurement according to the enzyme activity assay.
2.6. Enzyme activity in co-solvents at 4 ◦C
2. Experimental
An enzyme preparation (1.5 mg/mL) was stored in aliquots
of 1 mL in different concentrations of co-solvents (glycerol and
methanol) at 4 ◦C. Enzyme samples were taken directly from the
fridge and were measured for residual activity using the enzyme
activity assay at 290 nm. Thereafter, all samples were incubated in
37 ◦C. Samples were taken every hour to measure its re-activation.
2.1. Chemicals
All chemicals were purchased from Sigma–Aldrich and used
without further purification.
2.2. Protein expression and purification
2.7. Enzyme initial activity in co-solvents
The gene encoding the Cv-ATA was previously inserted in a
pET28a(+) vector, including an N-terminal His6-tag [34]. For protein
expression in Escherichia coli BL21(DE3), a 10 mL overnight culture
was added to 1 L TB medium (Terrific Broth) medium (1.2% pep-
tone, 2.4% yeast extract, 72 mM K2HPO4, 17 mM KH2PO4 and 0.4%
glycerol) supplemented with 50 mg/L kanamycin. Firstly, the cul-
ture was incubated at 37 ◦C and 220 rpm for 2–3 h until the OD600
reached 0.7–0.9. Then, the culture was induced by addition of IPTG
(1 mM). After 24 h of cultivation at 20 ◦C and 220 rpm, the cells were
harvested by centrifugation (30 min, 8000 rpm) and re-suspened in
IMAC binding buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imi-
dazole, pH 8.0). Cells were disrupted by sonication and the cell
debris was removed by centrifugation (30 min, 20,000 rpm) fol-
lowed by a final filtration (0.45 ␮m). The IMAC purification process
was performed on a Ni-NTA Sepharose column from IBA according
to the manufacturerı´s protocol. A buffer change to HEPES buffer
Initial activity of Cv-ATA in increasing concentrations of co-
solvent (5–50% of methanol, glycerol or DMSO) was determined
using the enzyme activity assay. When DMSO was applied initial
activity was determined at 290 nm (30 ␮g Cv-ATA) due to interfer-
ence from DMSO at 245 nm. Both incubations and measurements
are made in room temperature.
2.8. Enzyme stability and remaining activity after storage in
co-solvents at room temperature
Aliquots of an enzyme preparation were stored in three different
co-solvents (DMSO, glycerol and methanol) at room temperature
with a final concentration of 1.5 mg/mL. Samples from each aliquot
were taken and analyzed every day during the first week and
with reduced frequency later. The samples were analyzed by the

22
S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
Scheme 1. Cv-ATA-catalyzed transamination of methoxyacetone by (S)-1-PEA in
dry organic solvents (MTBE, isooctane or toluene) without addition of buffer.
enzyme activity assay using 0.15–0.3 mg Cv-ATA in 1 mL HEPES
buffer (50 mM, pH 8.2) containing different concentrations of co-
solvent. The initial rate of product formation (acetophenone) was
measured at 290 nm, to avoid absorption of co-solvents at 245 nm.
2.9. Surfactant treatment, lyophilization and ATA-catalyzed
reactions in organic solvent
The surfactants, Brij® C10 (15 mg) and octyl ␤-d-
glycopyranoside (15 mg) [47], were dissolved in 3 mL HEPES
buffer (50 mM, pH 8.2) by a 1 h incubation at 37 ◦C at 220 rpm.
Then, the enzyme was added (3 mL, 6 mg/mL). After mixing, the
solution was rapidly frozen in liquid N2 and lyophilized for 16 h.
To compare, half of the enzyme batch was lyophilized without the
addition of surfactants. Reactions (1 mL) were performed using
20 mg lyophilized or surfactant treated lyophilized Cv-ATA, (S)-
1-PEA (50 mM), methoxyacetone (150 mM) and decane (20 mM)
in dry organic solvents (Scheme 1). The solvents were dried over
molecular sieves (3 Å) prior to usage. The following solvents were
explored; methyl tert-butyl ether (MTBE), isooctane or toluene.
Samples were withdrawn at 24 and 48 h and analyzed on a Hewlett
Packard 5890 Series II gas chromatograph with a Chrompack CP-SIL
5CB column (25 m × 0.32 mm). The inlet and the detector were
set to 250 ◦C. The following method was used: Initial temperature
50 ◦C, 5 ◦C/min to 100 ◦C, 20 ◦C/min until 225 ◦C and hold 5 min.
2.10. Blue Native PAGE (BN-PAGE)
BN-PAGE was performed on NativePAGETM Novex® 4–16%
Bis–Tris Gels from Novex. Samples were prepared with a total vol-
ume of 10 ␮L, consisting of; 2.5 ␮L NativePAGETM sample buffer
(4×), 0.4 ␮L Coomassie G250, 2 ␮L enzyme (1.5 mg/mL) and 5.1 ␮L
HEPES buffer (50 mM, pH 8.2). For electrophoresis, dark blue cath-
ode buffer (50 mM Tricine, 15 mM Bis Tris 0.02% (w/v) of Coomassie
G250) and anode buffer (Bis–Tris, 50 mM, pH 7.0) were used.
Fig. 1. Melting points differences (ꢀTm) of Cv-ATA determined by comparing mea-
surements of enzyme in buffer with measurements when adding PLP, different
substrates, additives or co-solvents (glycerol, DMSO or methanol).
which may be a result from PMP (the reduced amine form of the
cofactor PLP) formation. PMP is not covalently bound to the enzyme
active site and may therefore promote enzyme destabilization or
denaturation.
3. Results
3.1. The effects of substrates, additives and co-solvents on
enzyme melting point
The effects of substrates, additives or co-solvents on the sta-
bility of Cv-ATA were explored by melting point measurements
(Fig. 1). The melting point differences (ꢀTm) compared to the mea-
surement without additives (buffer + enzyme), gives an indication
on the stabilization effects of individual additives or co-solvents
on the enzyme. The addition of glycerol and sucrose results in
increased melting point differences, while methanol shows a desta-
bilizing effect that heavily decreases the melting points. Further,
the addition of NaCl or DMSO did not influence the ꢀTm signifi-
cantly. It should however be noted that only concentrations ≤20%
of DMSO could be measured due to an impairment of the measure-
ment method at higher concentrations. The influence of common
ATA substrates (acetophenone, (S)-1-PEA, pyruvate and alanine)
was also explored. Interestingly, pyruvate yields a slight increase
of the ꢀTm while the addition of 5 mM acetophenone does not sig-
nificantly affect the melting point. However, the addition of amine
substrates (l-alanine and (S)-1-PEA) displayed destabilizing effects,
3.2. The effects of additives and co-solvents on enzyme storage
time
In order to pro-long the enzyme storage time, the effects of
adding additives or co-solvents to an enzyme preparation of Cv-
ATA were explored. A buffered enzyme stock was supplemented by
either of the following; co-solvent (DMSO 5–50%, glycerol 5–50%
or methanol 5–50%), an additive (sucrose 5–30%), cofactor (PLP),
a substrate ((S)-1-PEA) or surfactants before storage at room tem-
perature. Surfactants were included since they may have stabilizing
effects on protein structure during lyophilization processes [43,44].
The remaining activities were measured using an enzyme activ-
ity assay at room temperature. As the results show in Fig. 2, the
enzyme stored in buffer only, completely lost its activity after 5 days
of incubation. In contrast, the addition of co-solvents significantly
increases the enzyme storage stability. Higher glycerol concen-
trations (20–50%) clearly target the stability, while lower glycerol

S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
23
Fig. 2. The storage stability at room temperature of Cv-ATA in; (A) HEPES buffer (50 mM, pH 8.2), co-solvents such as (B) DMSO, (C) methanol or (D) glycerol, additive solution
(E) sucrose (5–30%) or (F) PLP (1 mM), (S)-1-PEA (5 mM) or surfactants (Brij® C10 and octyl ␤-d-glycopyranoside, 2.5 mg/mL of each).
concentrations (5–10%) show less significant effect. Enzyme stored
for six months in 50% glycerol at −20 ◦C was still fully active (114%)
and the same solution stored at room temperature maintained 77%
activity. Methanol was the only solvent displaying a negative effect
at high concentrations (>40%), which completely denatures (pre-
cipitates) the enzyme in about a day. DMSO shows an increasing
capacity to help the enzyme to maintain its activity with increasing
proportion in stocks. Sucrose has a slight influence on enzyme stor-
age stability, while (S)-1-PEA decreased the enzyme stability. Both
PLP and surfactants show positive effects on maintaining enzyme
activity, but still less than when the enzyme is stored in co-solvents
(Fig. 2).
time. When the enzyme is stored in the fridge (at 4 ◦C) it readily
loses about 90% activity. Storage of Cv-ATA in different concentra-
tions of co-solvents was performed to see the possible influence of
co-solvents on enzyme stability. Methanol was selected because
it is easy to remove by evaporation, and glycerol was selected
because of its low toxicity. After three weeks storage in 4 ◦C,
enzyme stored in buffer solution maintains 7% activity, while an
enzyme preparation supplemented by glycerol (20–50%) main-
tains full activity. Although, the low temperature reduces the
enzyme activity, it beneficially prevents irreversible enzyme denat-
uration. This was shown by incubating an enzyme preparation
(stored at 4 ◦C in buffer solution) at 37 ◦C for one hour followed
by an activity measurement. Almost 100% enzyme activity could
be recovered. However, the volume of the enzyme stock affects the
time required to recover the enzyme activity by incubation at 37 ◦C.
Also, methanol shows a good capacity to maintain enzyme activ-
ity, but the denaturation of enzyme with 40% methanol limited its
possibility to maintain 100% enzyme activity (Fig. 3).
3.3. The effect of co-solvents on enzyme stored at 4 ◦C
The storage time of Cv-ATA after purification and before usage
is a limiting factor, since the enzyme normally loses activity with

24
S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
Fig. 3. Residual activity of Cv-ATA after storage in fridge (4 ◦C) for 21 days in buffer solution or co-solvent, (A) methanol or (B) glycerol measured after activity recovery at
37 ◦C for 4 h.
Fig. 4. Residual activity (%) of Cv-ATA preparations after storage at 65 ◦C in buffer solution, co-solvents (methanol, DMSO or glycerol) or sucrose.
Fig. 5. Stability of Cv-ATA preparations after storage in buffer solution or in co-solvents (DMSO, methanol or glycerol) followed by activity measurements in the corresponding
concentration of co-solvents at room temperature.
3.4. The effects of temperature on enzyme stored in co-solvents or
sucrose
The enzyme preparation stored in 50% glycerol maintained about
20% of its initial activity after 24 h of storage at 65 ◦C, while enzyme
stored in buffer solution completely lost its activity after four hours.
Methanol, DMSO and sucrose did not have positive effects on
A Cv-ATA preparation was stored in: HEPES buffer (50 mM, pH
8.2), methanol (30%), DMSO (20%), glycerol (50%) or sucrose (30%).

S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
25
enzyme thermal stability (Fig. 4). This result corresponds to the
melting point result, and indicates that glycerol has the capacity to
be a co-solvent for Cv-ATA.
3.5. Enzyme stability and remaining activity in co-solvents
In the previous experiments, enzyme stability in different co-
solvents was determined by measuring residual activity in an
enzyme assay without any co-solvent. Here, to simulate industrial
conditions, enzyme stability as well as activity was examined in
combination in the three previously tested co-solvents (glycerol
(10–50%), methanol (10–50%), DMSO (10–50%)). The enzyme was
stored in the co-solvents and residual activity was measured in the
same conditions. All three co-solvents showed a similar pattern of
deactivation (Fig. 5) with a slow decrease in activity over 27 days.
DMSO and methanol showed a decrease in initial activity with an
increase in co-solvent concentration. In contrast, lower concentra-
tions of glycerol (10–20%) resulted in a higher initial activity than
in the sample with only HEPES buffer. As a result, the higher con-
centrations of glycerol also resulted in a higher initial activity than
in the corresponding concentrations of DMSO and methanol.
Fig. 6. Residual activity of Cv-ATA in increasing concentrations (up to 50%) of co-
solvent (DMSO, methanol or glycerol) measured at room temperature.
3.6. Enzyme activity in co-solvents
The initial activity of Cv-ATA was determined in different
amounts of co-solvent (DMSO, methanol and glycerol) to exam-
ine their applicability in enzyme-catalyzed reactions (Fig. 6). All
investigated co-solvents resulted in decreased activity. DMSO and
methanol followed the same pattern and activity was almost com-
pletely diminished at 50% co-solvent. Glycerol, however, showed a
slower rate of inactivation and 20% activity remained at 50% glyc-
erol.
Fig. 7. Surfactants applied in co-lyophilization with enzyme; (a) Brij® C10 and (b)
octyl ␤-d-glycopyranoside.
1:1) [47]. Cv-ATA-catalyzed reactions (Scheme 1) were performed
using lyophilized enzyme directly in dry organic solvents (MTBE,
isooctane or toluene). For the reactions (1 mL), 20 mg lyophilized
Cv-ATA with or without surfactants, (S)-1-PEA (50 mM), methoxy-
acetone (150 mM) and decane (20 mM) were mixed. Samples were
taken after 24 and 48 h for GC-analysis. The lyophilized enzyme
preparation with surfactants showed 5-fold improved conversion
compared to lyophilized enzyme without surfactant (Table 1). Over
80% conversion was achieved in isooctane after 48 h. The surfactant
3.7. The effects of surfactants on enzyme performance in organic
solvents
A purified Cv-ATA preparation was lyophilized with or with-
out surfactants (Brij® C10 and octyl ␤-d-glycopyranoside (Fig. 7),
Fig. 8. BN-PAGE of a Cv-ATA preparation stored in different co-solvents for 20 days at room temperature. A concentration step was performed before running the gel for
the lower protein storage concentration. Abbreviations: HEPES (HEPES buffer 50 mM, pH 8.2), HEPES + PLP (HEPES buffer 50 mM, pH 8.2 with PLP 1 mM), M (methanol), D
(DMSO), G (glycerol) and S (sucrose). After storage at room temperature for 20 days, enzyme preparation stored in HEPES buffer with or without 1 mM PLP, 5% methanol, 5%
DMSO, 5% glycerol or 5% sucrose dissociated to monomer, while Cv-ATA stored in 20% methanol, 50% DMSO, 50% glycerol or 30% methanol maintained dimeric structure.

26
S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
Table 1
The effect of surfactant-treatment on the performance of Cv-ATA was explored
by performing enzyme-catalyzed reactionsa using dry lyophilized preparations of
Cv-ATA in organic solvents (MTBE, isooctane or toulene) with (+) or without (−)
surfactantsb.
Solvent
Surfactants
Conversion (%)
24 h
48 h
MTBE
MTBE
−
+
5.9
32
9.6
45
Isooctane
Isooctane
Toluene
Toluene
−
13
65
9.1
36
17
84
12
54
+
−
+
a
Enzyme-catalyzed reactions (1 mL), 20 mg lyophilized Cv-ATA with or without
surfactants, (S)-1-PEA (15 mM), methoxyacetone (150 mM) and decane (20 mM)
were mixed in organic solvent pre-treated with molecular sieves. Reactions were
put on shaker at room temperature.
Fig. 9. Residual activity (%) of various enzyme preparations stored at different pro-
tein concentrations. A preparation of Cv-ATA was stored in HEPES buffer (50 mM,
pH 8.2) at room temperature in different protein concentrations: 0.1; 0.2; 0.5; 1 or
5 mg/mL and samples were regularly withdrawn for residual activity (%) measure-
ments for 7 days.
b
Brij® C10 and octyl ␤-d-glycopyranoside (1:1) [47].
treatment clearly affected the reaction conversions and the enzyme
performance.
opens the “phosphate group binding cup” to the solvents and may
lead to a release of PLP. The release of PLP from the monomer
structure may be a reason for enzyme inactivation.
3.8. The effects of additives and co-solvents on enzyme dimer
dissociation
3.9. The effects of enzyme storage at different protein
concentrations
Normally, as an enzyme loses catalytic activity, because of pro-
tein degradation, one could identify an enzyme precipitate in the
solution. When the enzyme preparation was stored in high con-
centrations (>40%) of methanol at room temperature, the enzyme
was denatured and a precipitate was created. However, this phe-
nomenon was not visible in the other inactive enzyme solutions,
such as after storage in HEPES buffer at room temperature. To
explore this phenomenon further, BN-PAGE analysis was per-
formed to visualize enzyme dimerization. As shown in Fig. 8, the
same enzyme stock stored for 20 days at room temperature in
HEPES buffer resulted in dissociation of the dimer to monomer.
However, as the concentrations of co-solvents were increased the
enzyme retained its dimeric structure. According to the gel (Fig. 8),
enzyme with maintained activity (Fig. 2) is dimeric, while the inac-
tive enzyme sample is monomeric. This indicates that irreversible
unfolding is not the primary reason for the reduced activity in most
of the samples, which is well acknowledged in the inactivation of
multimeric enzymes [48,49]. Since the enzyme is only active as
a dimer, a dissociation of subunits from dimeric to monomeric
structure will lead to an open binding pocket lacking important
hydrogen bonding amino acid residues building up the “phosphate
group binding cup” (from the second subunit) that coordinates the
PLP phosphate group [50]. The dissociation of dimer to monomer
Protein concentration is a known factor to influence multimeric
enzyme stability, since it affects the monomer-dimer equilibrium
[51]. The monomer-dimer equilibrium effect on enzyme storage
stability was explored by storing aliquots of an enzyme prepara-
tion at different protein concentrations at room temperature and
regularly follow the residual activity (Fig. 9) and monomer-dimer
equilibrium on BN-PAGE analysis (Fig. 10). After 3–4 days of incu-
bation, the enzyme activity started to decrease at the same time as
monomers started to show up in the BN-PAGE. After one week of
incubation, most of the enzyme was in the monomeric form and
had completely lost its activity. The highest residual activities and
fraction of active dimers was shown in the enzyme aliquot stored
at the highest protein concentration (5 mg/mL).
3.10. The effects of PLP on monomer-dimer equilibrium
Additional BN-PAGE analysis was performed to elucidate how
PLP was affected during the enzyme storage. The over-expression
of Cv-ATA in E. coli may lead to a lack of PLP. According to Fig. 11,
all samples stored at room temperature (Well 1–3) showed a vary-
ing degree of dissociation from dimer to monomer, while enzyme
Fig. 10. BN-PAGEs of Cv-ATA at varying storage concentrations over time. Cv-ATA stored in HEPES buffer (50 mM, pH 8.2) at room temperature in different protein
concentrations: 0.2; 0.5; 1 or 5 mg/mL for 7 days.

S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
27
glycerol displays a higher enzyme activity. When it comes to sol-
ubilizing hydrophobic substrates, DMSO and methanol are better
choices than glycerol. If solubility in DMSO and methanol is similar,
methanol is the better choice due to its volatility, making it easier
to evaporate in the purification process.
From different kinds of stability experiments, the investigated
co-solvents show a capacity to prevent Cv-ATA from both dimer
dissociation and denaturation. By addition of PLP, the fraction of
active dimer could be increased. The mechanism of co-solvent sta-
bilization of Cv-ATA is however still not clear to us.
Glycerol is well known and applied for the long-term storage
of bacterial and protein stocks. Glycerol stabilizes bacterial stocks
by preventing cellular membrane damage and therefore bacterial
glycerol stocks can be stored for many years at −80 ◦C. Also, glycerol
is commonly applied in protein refolding and protein crystalliza-
tion and has been shown to affect subunit association by effectively
associating the four subunits of aldehyde dehydrogenase [52,53].
From this, glycerol may increase the activity of Cv-ATA by prevent-
ing unfavorable protein unfolding and dimer dissociation. We see
a possible application of glycerol in enzyme-catalyzed reaction as
well as in storage stocks due to its stability effects.
Finally, we show the first example of the use of an ATA enzyme
co-lyophilized with surfactants. The co-lyophilization of Cv-ATA
with surfactants clearly improved the enzyme performance in
organic solvents. These results will contribute to improve the
application of ATAs in organic solvent. Surfactants can however
influence the downstream processing of an industrial process in a
negative fashion and that should be considered [43]. The effect of
the surfactant treatment will be further explored in our laboratory.
Fig. 11. BN-PAGE of a preparation of Cv-ATA 5 days after purification. Well 1: Addi-
tion of PLP (3 mM) and stored at room temperature for 5 days. Well 2: Addition of
PLP (3 mM), incubation overnight followed by desalting and stored at room temper-
ature for 5 days. Well 3: No additional PLP added and stored at room temperature
for 5 days. Well 4: Addition of PLP (3 mM) and stored at 4 ◦C for 5 days. Well 5:
Addition of PLP (3 mM), incubation overnight followed by desalting and stored at
4
◦C for 5 days. Well 6: No additional PLP added and stored at 4 ◦C for 5 days. Well
5. Conclusions
7: Supernatant before purification stored at 4 ◦C for 5 days.
The operational stability of Cv-ATA has been improved by the
addition of co-solvents, additives and surfactants. E.g., the addi-
tion of 50% DMSO, 20% methanol or 50% glycerol to an enzyme
solution stored in room temperature kept the enzyme 100% active
(or higher) during the investigated 25 days. Also, 50% glycerol was
investigated for storage over 6 months and enzyme stored at −20 ◦C
maintained full activity whereas enzyme stored at room tempera-
ture maintained 77% activity. In comparison, the enzyme solution
stored without co-solvent lost all activity after 5 days of storage.
Furthermore, the addition of glycerol also resulted in an increased
thermostability. We have also shown that storing Cv-ATA in higher
enzyme concentrations make it more stable, suggesting the optimal
storage conditions for Cv-ATA. The reason for activity loss over time
was shown to be the dissociation of the enzyme dimer complex.
Lyophilized Cv-ATA was successfully applied in dry organic sol-
vents and its performance was further increased by the addition of
surfactants prior to lyophilization. A 5-fold improvement of con-
version was observed over a 48 h long reaction.
samples stored at 4 ◦C (Well 4–7) maintained dimeric structure.
For samples stored in room temperature, the enzyme sample with-
out additional PLP (Well 3) displays more monomer over dimer,
compared to the enzyme samples incubated with additional PLP
(Well 1 and 2). As the enzyme sample stored at room tempera-
ture is incubated with additional PLP followed by desalting and a
second incubation for 5 days, the fraction of active dimer could be
increased. The BN-PAGE shows that PLP has a crucial role in dimer
stabilization.
4. Discussion
The stability of Cv-ATA in aqueous solution has been greatly
improved by the addition of co-solvents such as DMSO, methanol
and glycerol. These three co-solvents show the capacity to enhance
enzyme stability when added to stock solutions of Cv-ATA. Addi-
tions of 10–50% DMSO, 10–20% methanol or 20–50% glycerol
maintained the enzyme’s maximum residual activity, respectively,
after 3 weeks. Both melting point measurements and residual activ-
ity of enzyme stored at 65 ◦C shows that glycerol have stabilizing
effects on the enzyme at higher temperatures. The possibility to
utilize co-solvents in reactions with the aim to improve enzyme
stability during long-term catalysis processes was explored by a
continuous activity assay. The residual activities show that Cv-
ATA maintains 60–80% activity if 15% DMSO, 15% methanol or
30% glycerol is applied as reaction co-solvent. This means that
adding co-solvents could improve the long-term catalysis pro-
cesses in buffer solution in large scales. Activity would have to be
sacrificed for stability, but the increase in stability is larger than
the decrease in activity. The choice of co-solvent for an industrial
process depends on if stability, activity, solubility or a simple purifi-
cation is wanted. Stability is similar in the three co-solvents while
These findings will all help in the future application of Cv-ATA
in industrial processes.
Acknowledgements
This work was funded by the China Scholarship Council (CSC)
and KTH Royal Institute of Technology.
The authors would like to express their gratefulness to Dr.
Rosaria Gandini for experimental guidance regarding the use of
BN-PAGE.
References
[1] K.M. Polizzi, A.S. Bommarius, J.M. Broering, J.F. Chaparro-Riggers, Curr. Opin.
Chem. Biol. 11 (2007) 220–225.
[2] P.V. Iyer, L. Ananthanarayan, Process Biochem. 43 (2008) 1019–1032.

28
S. Chen et al. / Journal of Molecular Catalysis B: Enzymatic 124 (2016) 20–28
[3] V. Stepankova, S. Bidmanova, T. Koudelakova, Z. Prokop, R. Chaloupkova, J.
Damborsky, ACS Catal. 3 (2013) 2823–2836.
[28] D. Pressnitz, C.S. Fuchs, J.H. Sattler, T. Knaus, P. Macheroux, F.G. Mutti, W.
Kroutil, ACS Catal. 3 (2013) 555–559.
[4] A.M. Klibanov, Nature 409 (2001) 241–246.
[5] M.C. Manning, D.K. Chou, B.M. Murphy, R.W. Payne, D.S. Katayama, Pharm.
Res. 27 (2010) 544–575.
[6] F. Steffen-Munsberg, C. Vickers, H. Kohls, H. Land, H. Mallin, A. Nobili, L.
Skalden, T. van den Bergh, H.J. Joosten, P. Berglund, M. Höhne, U.T.
Bornscheuer, Biotechnol. Adv. 33 (2015) 566–604.
[29] N. Schmidt, R.C. Simon, W. Kroutil, Adv. Synth. Catal. 357 (2015)
1815–1821.
[30] N. Richter, R.C. Simon, H. Lechner, W. Kroutil, J.M. Ward, H.C. Hailes, Org.
Biomol. Chem. 13 (2015) 8843–8851.
[31] D. Deszcz, P. Affaticati, N. Ladkau, A. Gegel, J.M. Ward, H.C. Hailes, P.A. Dalby,
FEBS J. 282 (2015) 2512–2526.
[7] D. Schiroli, A. Peracchi, Biochim. Biophys. Acta 1854 (2015) 1200–1211.
[8] S. Mathew, H. Yun, ACS Catal. 2 (2012) 993–1001.
[9] M. Fuchs, J.E. Farnberger, W. Kroutil, Eur. J. Org. Chem. (2015) 6965–6982.
[10] J.-K. Shin, B.-G. Kim, D.-H. Shin, Enzyme Microb. Technol. 29 (2001) 232–239.
[11] S.-S. Yi, C.-W. Lee, J. Kim, D. Kyung, B.-G. Kim, Y.-S. Lee, Process Biochem. 42
(2007) 895–898.
[32] M.S. Humble, K.E. Cassimjee, M. Håkansson, Y.R. Kimbung, B. Walse, V. Abedi,
H.-J. Federsel, P. Berglund, D.T. Logan, FEBS J. 279 (2012) 779–792.
[33] M.S. Humble, K.E. Cassimjee, V. Abedi, H.-J. Federsel, P. Berglund,
ChemCatChem 4 (2012) 1167–1172.
[34] K.E. Cassimjee, M.S. Humble, V. Miceli, C.G. Colomina, P. Berglund, ACS Catal. 1
(2011) 1051–1055.
[12] D. Koszelewski, N. Müller, J.H. Schrittweiser, K. Faber, W. Kroutil, J. Mol. Catal.
B Enzym. 63 (2010) 39–44.
[13] M. Cárdenas-Fernández, W. Neto, C. López, G. Álvaro, P. Tufvesson, J.M.
Woodley, Biotechnol. Prog. 28 (2012) 693–698.
[14] K. Ni, X. Zhou, L. Zhao, H. Wang, Y. Ren, D. Wei, PLoS One 7 (2012) 1–8.
[15] G. Rehn, C. Grey, C. Branneby, L. Lindberg, P. Adlercreutz, Process Biochem. 47
(2012) 1129–1134.
[35] J.-K. Shin, B.-G. Kim, D.-H. Shin, Enzyme Microb. Technol. 29 (2001) 232–239.
[36] M.D. Truppo, H. Strotman, G. Hughes, ChemCatChem 4 (2012) 1071–1074.
[37] F.G. Mutti, W. Kroutil, Adv. Synth. Catal. 354 (2012) 3409–3413.
[38] E. Siirola, F.G. Mutti, B. Grischek, S.F. Hoefler, M.F.W. Fabian, G. Grogan, W.
Kroutil, Adv. Synth. Catal. 355 (2013) 1703–1708.
[39] C.S. Fuchs, R.C. Simon, W. Riethorst, F. Zepeck, W. Kroutil, Bioorg. Med. Chem.
22 (2014) 5558–5562.
[16] M.D. Truppo, H. Strotman, G. Hughes, ChemCatChem 4 (2012) 1071–1074.
[17] H. Mallin, U. Menyes, T. Vorhaben, M. Höhne, U.T. Bornscheuer,
ChemCatChem 5 (2012) 588–593.
[40] A.M. Klibanov, Trends Biotechnol. 15 (1997) 97–101.
[41] J.F. Carpenter, S.J. Pretrelski, T. Arakawa, Arch. Biochem. Biophys. 303 (1993)
456–464.
[18] A.A. Halim, N. Szita, F. Baganz, J. Biotechnol. 168 (2013) 567–575.
[19] H. Mallin, J. Muschiol, E. Byström, U.T. Bornscheuer, ChemCatChem 5 (2013)
3529–3532.
[42] J.H. Crowe, J.F. Carpenter, L.M. Crowe, T.J. Anchordoguy, Cryobiology 27
(1990) 219–231.
[43] F. Bordusa, Chem. Rev. 102 (2002) 4817–4867.
[20] L.H. Andrade, W. Kroutil, T.F. Jamison, Org. Lett. 16 (2014) 6092–6095.
[21] M. Gustavsson, M.N. Muraleedharan, G. Larsson, Appl. Environ. Microbiol. 80
(2014) 2293–2298.
[44] K. Imamura, K. Murai, T. Korehisa, N. Shimizu, R. Yamahira, T. Matsuura, H.
Tada, H. Imanaka, N. Ishida, K. Nakanishi, J. Pharm. Sci. 103 (2014)
1628–1637.
[22] K.E. Cassimjee, M. Kadow, Y. Wikmark, M.S. Humble, M.L. Rothstein, D.M.
Rothstein, J.-E. Bäckvall, Chem. Commun. 50 (2014) 9134–9137.
[23] W. Neto, M. Schürmann, L. Panella, A. Vogel, J.M. Woodley, J. Mol. Catal. B
Enzym. 117 (2015) 54–61.
[24] A.R. Martin, R. DiSanto, I. Plotnikov, S. Kamat, D. Shonnard, S. Pannuri,
Biochem. Eng. J. 37 (2007) 246–255.
[25] C.K. Savile, J.M. Janey, E.C. Mundorff, J.C. Moore, S. Tam, W.R. Jarvis, J.C.
Colbeck, A. Krebber, F.J. Fleitz, J. Brands, P.N. Devine, G.W. Huisman, G.J.
Hughes, Science 329 (2010) 305–309.
[45] F.H. Niesen, H. Berglund, M. Vedadi, Nat. Protoc. 2 (2007) 2212–2221.
[46] S. Schätzle, M. Höhne, E. Redestad, K. Robins, U.T. Bornscheuer, Anal. Chem. 81
(2009) 8244–8248.
[47] L. Borén, B. Martín-Matute, Y. Xu, A. Córdova, J.-E. Bäckvall, Chem. Eur. J. 12
(2006) 225–232.
[48] K.E. Neet, D.E. Timm, Protein Sci. 3 (1994) 2167–2174.
[49] K. Sakurai, M. Oobatake, Y. Goto, Protein Sci. 10 (2001) 2325–2335.
[50] A.I. Denesyuk, K.A. Denessiouk, T. Korpela, M.S. Johnson, J. Mol. Biol. 316
(2002) 155–172.
[26] K. Deepankumar, M. Shon, S.P. Nadarajan, G. Shin, S. Mathew, N. Ayyadurai,
B.-G. Kim, S.-H. Choi, S.-H. Lee, H. Yun, Adv. Synth. Catal. 356 (2014) 993–998.
[27] U. Kaulmann, K. Smithies, M.E.B. Smith, H.C. Hailes, J.M. Ward, Enzyme
Microb. Technol. 41 (2007) 628–637.
[51] R. Fernandez-Lafuente, Enzyme Microb. Technol. 45 (2009) 405–418.
[52] J.F. Clark, W.B. Jakoby, J. Biol. Chem. 245 (1970) 6072–6077.
[53] S.L. Bradbury, W.B. Jakoby, Proc. Natl. Acad. Sci. U. S. A. 69 (1972) 2373–2376.