Preparation of two glycoside hydrolases for use in micro-aqueous media

Article abstract of DOI:10.1016/j.molcatb.2014.06.009

Enzymatic synthesis of alkyl glycosides using glycoside hydrolases is well studied, but has yet to reach industrial scale, primarily due to limited yields. Reduced water content should increase yields by limiting the unwanted hydrolytic side reaction. However, previous studies have shown that a reduction in water content surprisingly favors hydrolysis over transglycosylation. In addition, glycoside hydrolases normally require a high degree of hydration to function efficiently. This study compares six enzyme preparation methods to improve resilience and activity of two glycoside hydrolases from Thermotoga neapolitana (TnBgl3B and TnBgl1A) in micro-aqueous hexanol. Indeed, when adsorbed onto Accurel MP-1000 both enzymes increasingly favored transglycosylation over hydrolysis at low hydration, in contrast to freeze-dried or untreated enzyme. Additionally, they displayed 17-70× higher reaction rates compared to freeze-dried enzyme at low water activity, while displaying comparable or lower activity for fully hydrated systems. These results provide valuable information for use of enzymes under micro-aqueous conditions and build toward utilizing the full synthetic potential of glycoside hydrolases.

Full text of DOI:10.1016/j.molcatb.2014.06.009

Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Preparation of two glycoside hydrolases for use in micro-aqueous
media
Pontus Lundemo∗, Eva Nordberg Karlsson, Patrick Adlercreutz
Lund University, Department of Chemistry, Biotechnology, P.O. Box 124, SE-221 00 Lund, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Enzymatic synthesis of alkyl glycosides using glycoside hydrolases is well studied, but has yet to reach
industrial scale, primarily due to limited yields. Reduced water content should increase yields by limiting
the unwanted hydrolytic side reaction. However, previous studies have shown that a reduction in
water content surprisingly favors hydrolysis over transglycosylation. In addition, glycoside hydrolases
normally require a high degree of hydration to function efficiently. This study compares six enzyme
preparation methods to improve resilience and activity of two glycoside hydrolases from Thermotoga
neapolitana (TnBgl3B and TnBgl1A) in micro-aqueous hexanol. Indeed, when adsorbed onto Accurel MP-
1000 both enzymes increasingly favored transglycosylation over hydrolysis at low hydration, in contrast
to freeze-dried or untreated enzyme. Additionally, they displayed 17–70× higher reaction rates com-
pared to freeze-dried enzyme at low water activity, while displaying comparable or lower activity for fully
hydrated systems. These results provide valuable information for use of enzymes under micro-aqueous
conditions and build toward utilizing the full synthetic potential of glycoside hydrolases.
© 2014 Elsevier B.V. All rights reserved.
Received 6 February 2014
Received in revised form 27 May 2014
Accepted 21 June 2014
Available online 30 June 2014
Keywords:
Transglycosylation
Hydrolase
Immobilization
Organic solvent
Alkyl glycoside
1. Introduction
hydrolysis the amount of water directly influences the equilibrium
yield, while enzyme properties are a significant factor for transg-
Alkyl glycosides are a group of attractive surfactants. They
exhibit antimicrobial activity, biodegradability and low toxicity [1],
and find their use in cosmetics, biochemistry and pharmaceutical
industry [2–4]. Currently, they are produced using conventional
chemistry, which lead to a mixture of anomers [4], and require
complicated separation techniques for purification. Enzymatic
synthesis using glycoside hydrolases (ˇ-glycosidases) is an attrac-
tive alternative, as it provides an anomerically pure product and
reduced waste, thereby constituting a more environmentally sus-
tainable option [5]. There are two possible enzymatic strategies: the
thermodynamically controlled reverse hydrolysis or the kinetically
controlled transglycosylation reaction [6].
lycosylation. The catalytic mechanism involves a glycosyl-enzyme
intermediate, which can be deglycosylated either by water or by
alcohol, yielding hydrolysis or alkyl glycoside respectively. There-
fore, the yield is determined by the acceptor specificity of the
enzyme, often quantified as the ratio of transferase over hydro-
lase activity (rs/rh). Several previous studies have been aimed at
increasing alkyl glycoside yield by improving rs/rh through protein
engineering [7–10].
Another way to impair the hydrolytic side reaction, and
increase the alkyl glycoside yield, is to reduce the water content
in the reaction media. However, previous reports of transgly-
cosylation, catalyzed by a wide range of ˇ-glycosidases has,
counter-intuitively, shown reduced selectivity (rs/rh) at low aw
[9,11,12]. In addition, most enzymes are not well suited for anhy-
drous conditions [13]. ˇ-glycosidases in particular have been
reported to require a water activity (aw) as high as 0.6 [11,14], in
contrast e.g. lipases such as CALB, which has been shown to retain
activity at aw as low as 0.02 [15].
Currently, the alkyl glycoside yields from enzymatic synthe-
sis are too low for an economically feasible industrial process.
This issue is intimately linked to the presence of water. In reverse
In this paper, we attempt to increase the synthetic useful-
ness of two ˇ-glycosidases by improving their selectivity and
activity in micro-aqueous media. Six enzyme preparation meth-
ods are compared for synthesis of hexyl-ˇ-d-glucoside (HG) from
p-nitrophenol-ˇ-d-glucopyranoside (pNPG) in hexanol. As model
enzyme the ˇ-glucosidase with the highest reported rs/rh, from
Abbreviations: aw, water activity; rs, transglycosylation rate; rh, hydrolysis
rate; HG, hexyl-ˇ-d-glucoside; pNPG, p-nitrophenol-ˇ-d-glucopyranoside; pNP, p-
nitrophenol; AOT, dioctyl sodium sulfosuccinate.
∗
Corresponding author. Tel.: +46 462220858.
E-mail addresses: pontus.lundemo@biotek.lu.se
(P. Lundemo), eva.nordberg karlsson@biotek.lu.se (E.N. Karlsson),
patrick.adlercreutz@biotek.lu.se (P. Adlercreutz).
http://dx.doi.org/10.1016/j.molcatb.2014.06.009
1381-1177/© 2014 Elsevier B.V. All rights reserved.

2
P. Lundemo et al. / Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
Thermotoga neapolitana (TnBgl3B), is used. It belongs to the glyco-
side hydrolase family 3 and has been reported to have very low
activity in micro-aqueous media, when no enzyme preparation
method was used [16]. In contrast, ˇ-glycosidase from P. furio-
sus, belonging to glycoside hydrolase family 1, has been shown
to be most active in absence of a separate aqueous phase [7]. To
avoid bias from this potential discrepancy between the two glyco-
side hydrolase families, Thermotoga neapolitana enzymes from both
families are studied in parallel. For lipases, up to 400-fold activation
has been demonstrated by selecting a proper enzyme preparation
method [17], but to the best of our knowledge, no such attempt has
previously been made for ˇ-glycosidases.
2.4. Lyophilization
The glycosidases were diluted up to 1 ml in 0.1 M citrate phos-
phate buffer, pH 5.6 to roughly 0.25–0.30 mg/ml and centrifuged
to remove insoluble residues. The supernatants were immediately
frozen at −80 ◦C and then freeze-dried for 18 h.
2.5. Surfactant modification
A reverse micellar system was created according to a previ-
ously described method [20]. 150 ␮l suspensions of 0.6–0.85 mg/ml
glycosidase in 0.1 M citrate phosphate buffer, pH 5.6 was added
to 5 ml 100 mM dioctyl sodium sulfosuccinate (AOT) in 2,2,4-
trimethylpentane and shaken vigorously. The trimethylpentane
was removed by rotary evaporation and the residue was further
dried in a vacuum desiccator.
2. Materials and methods
2.1. Material
2.6. Factorial immobilization test
Hexyl-ˇ-d-glucoside (HG), p-nitrophenol (pNP) and p-
The influence of buffer strength, pH and incubation time for
adsorption and covalent immobilization of TnBgl1A and TnBgl3B
on the supports listed in sections 2.7 and 2.8 was tested using a 23
factorial design. The software Minitab® (Release 14.1) was used to
evaluate the data. Three factors were studied (buffer strength 0.05,
0.15 and 0.25 mM; buffer pH 4, 5.5 and 7 and incubation time 1,
7 and 24 h). Two replicates and 4 central points was used giving a
total number of 20 runs per enzyme.
nitrophenol-ˇ-d-glucoside
(pNPG)
were
obtained
from
Sigma–Aldrich (St Louis, Missouri, USA) and all other chemicals
from VWR International (Stockholm, Sweden).
2.2. Mutagenesis
The genes encoding TnBgl1A and TnBgl3B were previously
cloned into PET22b(+) (Novagen, Madison, WI, USA) [16,18].
Mutagenesis for construction of the N220F mutant was performed
in a previous study, using a QuikChange site-directed mutagenesis
kit (Stratagene, La Jolla, CA), with the sequence with GenBank
accession number AF039487 as the template and the primer
5ꢀ-GGAAAGATAGGGATTGTTTTCTTCAACGGATACTTCGAACCTGC-3ꢀ
[10]. The resulting plasmid was transformed into Escherichia coli
Nova Blue cells for storage and into E. coli BL21 (Novagen) for
expression. The complete gene was sequenced by GATC Biotech
AG (Konstanz, Germany) to confirm the mutations.
2.7. Adsorption
Both glycosidases were immobilized by adsorption to
a
hydrophobic support (Accurel MP-1000) and an anion-exchange
resin (IRA-400). 7 ml 0.09–0.11 mg/ml enzyme in 0.1 mM citrate
phosphate buffer, pH 5.6 was added to 400 mg Accurel MP-1000,
which was pre-wetted with 3 ml ethanol/g support, and to 400 mg
IRA-400, which was pre-washed with 0.1 mM citrate phosphate
buffer, pH 5.6. The enzyme and support was incubated on a nutat-
ing mixer overnight and thereafter filtered and washed with buffer.
Finally, the preparations were dried in a vacuum desiccator. For the
MP-1000 support, a milder drying technique previously described
by Moore et al. was also evaluated [21]. After removing the aque-
ous enzyme solution, the support was washed three times with
n-propanol, the same volume as the original aqueous solution, set
to the desired water activity. This was followed by two washes with
the same volume of hexanol, set to the desired water activity. The
hexanol was removed immediately prior to addition of substrate.
2.3. Expression and purification
The enzymes were synthesized in 0.5 L cultivations of E. coli BL21
(Novagen) in Erlenmeyer flasks at 37 ◦C, pH 7 in Luria-Bertania
(LB) media containing 100 ␮g/ml Ampicillin, inoculated with 1%
over night precultures. After reaching an OD620 of 0.6 TnBgl1A and
TnBgl3B gene expression was induced by addition of 0.5 ml 100 mM
isopropyl-ˇ-d-1-thiogalactopyranoside (IPTG) and production was
continued for 20 h. Cells were harvested by centrifugation for
10 min (4 ◦C, 5500 × g), re-suspended in binding buffer (20 mM
imidazole, 20 mM Tris–HCl, 0.75 M NaCl, pH 7.5) and lysed by son-
ication 6 × 3 min at 60% amplitude and a cycle of 0.5 using a 14 mm
titanium probe (UP400 S, Dr. Hielscher). Heat treatment (70 ◦C,
30 min) and centrifugation (30 min, 4 ◦C, 15,000 × g) was used to
remove most of the native E. coli proteins before purification by
immobilized metal affinity chromatography using an ÄKTA prime
system (Amersham Biosciences, Uppsala, Sweden). The protein
slurry was applied to a Histrap FF crude column (GE Healthcare)
pretreated with 0.1 M Copper (II) sulphate. Bound proteins were
eluted using elution buffer (250 mM imidazole, 20 mM Tris–HCl,
0.75 M NaCl, pH 7.5). Fractions containing protein were pooled
and dialyzed against 50 mM citrate phosphate buffer, pH 5.6, over
night using a 3500 Da molecular weight cut-off dialysis membrane
(Spectrum laboratories, Rancho Dominguez, CA, USA) and stored
at −20 ◦C until use. Purity of the expressed proteins was estimated
using SDS-PAGE according to Laemmli [19].
2.8. Covalent immobilization
400 mg epoxy-activated matrix, Eupergit® C250L, was washed
with 0.1 mM citrate phosphate buffer, pH 5.6. 7 ml 0.09–0.11 mg/ml
enzyme in 0.1 mM citrate phosphate buffer, pH 5.6 was added, incu-
bated on a nutating mixer overnight and thereafter filtered, washed
with buffer and dried in a vacuum desiccator.
2.9. Protein determination
Total protein concentration was estimated at 595 nm by the
Bradford method [22] using bovine serum albumin as standard.
2.10. Water activity
Substrate solutions (34 mM pNPG in hexanol) were incubated
over saturated salt solutions to defined water activities. The salts
used for equilibration were KCH3CO2 (aw = 0.23), MgCl2 (aw = 0.33),
Mg(NO3)2 (aw = 0.53), NaCl (aw = 0.75), KCl (aw = 0.84) and K2SO4

P. Lundemo et al. / Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
3
Table 1
Immobilization yield, enzyme load based on protein measurements in non-bound fraction as well as total activity relative to untreated enzyme for transglycosylation and
hydrolysis of pNPG in water saturated hexanol. Data is presented as average 1ꢀ from duplicate reactions.
Support
TnBgl1A
TnBgl3B
Protein yield (%)
Load (mg/g)
Rel. activity (%)
Protein yield (%)
Load (mg/g)
Rel. activity (%)
Eupergit C250La
Eupergit C250Lb
Accurel MP1000a
Accurel MP1000b
Amberlite IRA-400a
Amberlite IRA-400b
93
42
98
94
29
4
3
3
1
1
1
3
1.3 0.0
0.6 0.1
1.8 0.0
1.4 0.5
0.5 0.0
0.1 0.4
98
64 11
96
59
22
1
5
1.7 0.2
1.1 0.3
1.7 0.1
1.0 0.1
0.4 0.1
0.0 0.1
1.7
12.4
2.9
3.2
48.4
2.5
1
3
2
4
a
Best results obtained in any of the tested conditions.
Immobilized under conditions for use in organic media, incubated 24 h in 0.1 M citrate phosphate buffer pH 5.6.
b
(aw = 0.97). Triplicate samples from each equilibrated hexanol sam-
ple were injected on a 899 Karl Fischer coulometer (Metrohm,
Herisau, Switzerland). The obtained relation between water activ-
ity and water amount was used to estimate water activity in
the transferase reactions. The procedure was repeated using
n-propanol, to allow setting the water activity for drying the MP-
1000, as described above.
support material (Eupergit C250L, Amberlite IRA-400 and Accurel
MP1000), parameters that can affect the immobilization yield
(buffer strength, buffer pH and incubation time) was studied
using a 3 factor 2 level factorial design. The estimated effects and
coefficients for each parameter are presented in Supplementary
Table S1. Buffer pH had the highest influence on immobilization
yield, and a low pH (pH 4) was preferred for all preparations. The
effect was, unsurprisingly, especially prominent for immobilization
on the ion-exchange resin Amberlite IRA-400. However, for use in
organic solvents, it is advisable to generate the enzyme prepara-
tions under the pH optimum for the enzyme. This is because the
protonation state of the enzyme is conserved when transferred into
an organic medium [23]. For incubation time and buffer strength,
the lower values (1 h incubation or 0.05 mM buffer strength) were
detrimental to the immobilization, while less difference was seen
between the high values (15 h incubation and 0.25 mM buffer
strength) and the center point (8 h incubation and 0.15 mM buffer
strength).
2.11. Transferase reaction
Support corresponding to 21 ␮g of immobilized enzyme, based
on the Bradford assay, was mixed with 2 ml 34 mM pNPG set
to desired water activity, based on the Karl-Fisher calibration
curve, and incubated in a ThermoMixer (HLC Biotech, Bovenden,
Germany) set to 70 ◦C, 700 rpm. 150 ␮l samples were taken at 0, 2,
4, 8, 24, 48, 72 h and diluted with 150 ␮l 2 mM NaOH and 300 ␮l
methanol before HPLC analysis. 40 ␮l was withdrawn to check
actual water activity after 1 h reaction by Karl Fischer coulometry.
Table 1 shows the immobilization yield of TnBgl1A and TnBgl3B
onto Eupergit C250L, Amberlite IRA-400 and Accurel MP1000,
based on protein measurements of the initial enzyme slurry and
the filtrate after preparation. Both data at the best immobilization
conditions from the factorial design and at the conditions selected
for use in the transferase reaction are presented. It is clear from
the table, that Amberlite IRA-400 is not as well suited for immo-
bilization of ˇ-glycosidases as the other two. For TnBgl1A, Accurel
MP1000 maintain high immobilization yield at pH optimum for the
enzyme (pH 5.6). For the remaining two methods, freeze-drying
and AOT modification, 100% protein yields are assumed. Table 1
also shows the relative specific total activity (hydrolysis and trans-
glycosylation) of pNPG for each enzyme preparation, compared
to untreated enzyme in water-saturated hexanol. Amongst the
three preparations, both enzymes retain the most activity in water-
saturated hexanol when adsorbed onto Accurel MP1000. However,
at high hydration, untreated enzyme still has higher specific activ-
ity, which is not the case at lower hydration.
2.12. HPLC analysis
Transferase reactions were monitored using RP-HPLC
(LaChrom; pump L-7100, interface L-7000, autosampler L-
7250 with a 20 ml injection loop, UV-detector L7400, Hitachi
Ltd. Tokyo, Japan) equipped with an evaporative light scattering
detector (Alltech 500 ELSD, Alltech Associates Inc., Deer-field,
USA) with evaporator temperature 94 ◦C, a nebulizer gas flow of
2.5 standard liters per minute and a Kromasil 100 5C18 column
(4.6 ␮m × 250 mm, Kromasil, EkaChemicals AB, Separation Prod-
ucts, Bohus, Sweden). A gradient was applied from 50% to 70%
methanol in 0.1% acetic acid in MQ H2O over 5 min and kept at 70%
for 1 min before returning to initial conditions for re-equilibration.
A constant flow rate of 1.0 ml/min was used. pNPG elutes after
3.5 min and is followed at 405 nm as well as with ELSD. HG and
pNP both have a retention time of 7.5 min, but HG does not absorb
at 405 nm and pNP is too volatile to be detected by ELSD. Con-
centrations were determined by use of 8 point external standard
curves.
3.2. Improved catalytic activity at low hydration
3. Results and discussion
To determine which enzyme preparation method is most suited
for low water activities, transglycosylation reactions were followed
at four different hydration states for each enzyme preparation. In
the transglycosylation reaction monitored, pNPG is converted into
pNP and either glucose or HG, visualized in Fig. 1. Although all of
these components were detected using HPLC, formation of HG was
used for comparing the different preparation methods. This is pri-
marily since pNP binds to Amberlite IRA-400 (data not shown), and
can therefore not be used to follow all reactions. Table 2 shows
the initial formation rates of HG at the various aw levels based
on the amount of water added to the substrate mixture before
adding the enzyme preparations. The enzyme preparations influ-
enced the water activity to different extent, and in Figs. 2 and 3 the
3.1. Enzyme preparation
Two ˇ-glycosidases from Thermotoga neapolitana (TnBgl1A and
TnBgl3B) were modified by six different enzyme preparation meth-
ods; adsorption onto Accurel MP1000 (porous polypropylene)
or Amberlite IRA-400 (ion-exchange resin), covalent linking to
Eupergit C250L, freeze-drying, surfactant encapsulation using AOT
as well as application of a gentle drying method on adsorp-
tion onto MP1000. Before formulating the enzyme preparations,
protein purity was established to above 95% using SDS-PAGE (Fig-
ure S1). For the methods involving adsorption or binding to a

4
P. Lundemo et al. / Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
reaction rates are plotted versus the experimentally determined
water activities. The first figure, plotted on a logarithmic scale,
shows that both TnBgl1A and TnBgl3B display exponential increase
of reaction rates with increasing aw. This correlates well with pre-
vious characterizations of ˇ-glycosidases in micro-aqueous media
[7,14]. Moreover, Table 2 demonstrates that enzyme adsorbed to
Accurel MP1000 and dried using 1-propanol, retains the most
activity at low aw (0.83 nmol min−1 g−1 and 1.65 nmol min−1 g−1
for TnBgl1A and TnBgl3B respectively at aw ≈ 0.7), while freeze-
dried enzymes are amongst the least suited for low hydration
(0.049 nmol min−1 g−1 and 0.023 nmol min−1 g−1 for TnBgl1A and
TnBgl3B respectively at aw ≈ 0.7). However, the opposite relation
applies at high aw for TnBgl1A, where the freeze-dried preparation
outcompetes the MP1000 adsorbed preparation, as illustrated in
Fig. 3. The figure shows the specific initial formation rates, normal-
ized against the activity of untreated enzyme for each approximate
aw, and plotted against the experimentally determined water activ-
ities for the untreated enzyme used for normalization. As can
be seen in the figure, the trend is the same for TnBgl3B as for
Fig. 1. Schematic representation of the enzymatic conversion of p-nitrophenyl-ˇ-
d-glucoside to hexyl-ˇ-glucoside at the rate rs and to glucose at the rate rh catalyzed
by a retaining ˇ-glucosidase.
Table 2
Specific initial formation of hexyl-ˇ-d-glucopyranoside (␮mol min−1 mg−1) at each approximated water activity (aw). Data is presented as average 1ꢀ from duplicate
reactions.
aw
TnBgl1A
TnBgl3B
0.7
0.8
0.9
1.0
0.7
0.8
0.9
1.0
Untreated
Freeze-dried
AOT
Eupergit C250L
Accurel MP1000
Vacuum dried
Propanol dried
Amberlite IRA-400
0.24 0.05
0.05 0.05
0.10 0.06
0.00 0.01
0.79 0.04
0.33 0.19
0.40 0.00
0.02 0.01
4.75 0.90
4.53 1.35
0.42 0.05
0.11 0.03
14.73 0.21
6.78 2.35
1.84 0.21
0.25 0.08
0.49 0.05
0.02 0.03
0.15 0.08
0.01 0.01
0.90 0.17
0.04 0.02
0.38 0.00
0.03 0.03
8.49 0.87
0.91 0.64
1.14 1.18
0.11 0.20
25.45 1.76
9.65 0.72
6.15 0.00
0.61 0.23
0.16 0.02
0.83 0.15
0.00 0.00
0.38 0.06
1.03 0.15
0.21 0.19
0.84 0.09
2.03 0.16
0.45 0.23
1.34 0.22
4.52 0.30
1.08 0.90
0.41 0.03
1.65 0.37
0.03 0.11
0.92 0.14
1.93 0.13
0.05 0.05
2.60 0.46
3.69 1.09
0.16 0.00
13.95 2.60
9.38 0.13
0.98 0.44
100
10
100
10
1
1
0.1
0.01
0.1
0.01
0.001
0.001
0.6
0.8
1.0
1.2
0.6
0.8
1.0
1.2
aw
aw
Fig. 2. Initial transglycosylation activity of TnBgl1A (left) and TnBgl3B (right) plotted versus experimentally determined water activities (aw). The enzyme was freeze-dried
(♦), deposited on Accurel MP1000; vacuum dried (ꢀ) or propanol dried (ꢁ), covalently linked to Eupergit C250L (ꢁ) or added as aqueous solution (ꢂ). Error bars represent
1ꢀ, based on triplicate measurements.
4
3.5
3
2.5
2
4
3.5
3
2.5
2
1.5
1
0.5
0
1.5
1
0.5
0
0.6
0.8
1.0
0.6
0.8
1.0
aw
aw
Fig. 3. Initial transglycosylation activity of TnBgl1A (left) and TnBgl3B (right) normalized against untreated enzyme at each approximate water activity (aw). The compared
preparation methods are freeze-drying (♦), deposition on Accurel MP1000; vacuum dried (ꢀ) or propanol dried (ꢁ), and covalent linking to Eupergit C250L (ꢁ).

P. Lundemo et al. / Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
5
1.4
1.2
1
3.5
3
2.5
2
1.5
1
0.5
0
0.8
0.6
0.4
0.2
0
-0.5
0.6
0.8
1.0
1.2
0.6
0.8
1
1.2
aw
aw
Fig. 4. Selectivity for transglycosylation (rs/rh) of TnBgl1A (left) and TnBgl3B (right) as a function of water activity (aw). The enzyme was freeze-dried (♦), deposited on Accurel
MP1000; vacuum dried (ꢀ) or propanol dried (ꢁ), and covalently linked to Eupergit C250L (ꢁ). Error bars represent 1ꢀ, based on triplicate measurements.
10
1
4
3.5
3
2.5
2
1.5
1
0.5
0
0.1
0.6
0.8
1.0
0.6
0.8
1.0
aw
aw
Fig. 5. Initial transglycosylation activity (left) and selectivity for transglycosylation (rs/rh) for mutant N220F of TnBgl1A as a function of water activity (aw). The enzyme was
deposited on Accurel MP1000 and dried by propanol wash. Error bars represent 1ꢀ, based on triplicate measurements.
TnBgl1A, although the freeze-dried preparation only reached an
activity equal to the MP1000 adsorbed TnBgl3B at water satura-
tion. The results emphasize the importance of choosing enzyme
treatment based on intended hydration condition. For example,
TnBgl1A adsorbed to Accurel MP1000 had 17× higher transglyco-
sylation activity than freeze-dried enzyme at the lowest studied
aw, but only 0.45× the rate at aw = 0.9. The reason for the higher
enzyme activity when adsorbed on Accurel MP1000 is likely due
to an improved dispersion of the catalyst and thereby an increased
surface accessibility [24]. The additional increase in activity when
using the propanol drying procedure could be from avoiding the
removal of essential water molecules from the protein [21].
alterations to the enzyme. Why the effects of immobilization are
less prominent for TnBgl3B is, however, not clear.
3.4. Proper enzyme preparation in combination with TnBgl1A
mutation N220F
In a previous study, we found that the single mutation N220F
of TnBgl1A increased rs/rh 7-fold (up to 1.38 0.20) for HG
synthesis in a biphasic water:hexanol system (15:85) without
enzyme preparation [10]. We can now show that immobilizing
onto Accurel MP1000 enables further increase in the rs/rh up to
3.16 0.06 at aw = 0.85, as demonstrated in Fig. 5. However, this
increase in specificity comes at the cost of a severely reduced
reaction rate (2.4 0.3 ␮mol min−1 mg−1 at aw = 0.85 compared to
87.8 3.2 ␮mol min−1 mg−1 with 15% water).
3.3. Increased ratio of transglycosylation/hydrolysis at low
hydration
Previous reports of transglycosylation catalyzed by a wide range
of ˇ-glycosidases have shown reduced selectivity (rs/rh) at low aw
[9,11,12]. As seen in Fig. 4, this unwanted and counter-intuitive
trend was observed for untreated, freeze-dried and covalently
bound enzyme preparations of both TnBgl1A and TnBgl3B. Never-
theless, for TnBgl3B adsorbed onto Accurel MP1000 the rs/rh ratio
was maintained at low hydration. Furthermore, TnBgl1A on Accurel
MP1000 even displayed increased selectivity at low water activity.
Consequently, we have successfully showed that reduced hydration
can be used as means to improve alkyl glycoside yields from trans-
glycosylation catalyzed by TnBgl1A, when immobilized on Accurel
MP1000. We believe the, previously observed, reduced selectivity
at low hydration for ˇ-glycosidases has been due to detrimental
protein-protein interactions when the enzymes are transferred to
an organic solvent. Gentle immobilization onto a suitable support
can reduce these interactions without introducing new harmful
4. Conclusions
Previous studies on ˇ-glycosidase catalyzed transglycosyla-
tion in micro-aqueous media have, counter-intuitively, shown
increased ratio of undesired hydrolysis at low water activity. This
paper demonstrates, for the first time, that reduced water content
can be a viable method of reducing the hydrolytic side reaction, pro-
vided the enzyme is properly prepared for use in micro-aqueous
media. We show that deposition onto Accurel MP-1000 is a suit-
able preparation method, especially when dried using the propanol
washing method described by Moore et al. [21]. However, the reac-
tion rates at low hydration still require significant enhancement
to reach an enzymatic method competitive to the classical chem-
ical routes used for synthesis of alkyl glycosides today. Feasible
routes to get there include screening for ˇ-glycosidases with desir-
able properties in micro-aqueous media or alternatively protein

6
P. Lundemo et al. / Journal of Molecular Catalysis B: Enzymatic 108 (2014) 1–6
engineering of promising candidates such as TnBgl3B. Neverthe-
less, deposition on MP1000 provides one step toward unlocking
the full synthetic potential of ˇ-glycosidases.
[6] F. van Rantwijk, M. Woudenberg-van Oosterom, R.A. Sheldon, J. Mol. Catal. B:
Enzym. 6 (1999) 511–532.
[7] T. Hansson, P. Adlercreutz, Biotechnol. Bioeng. 75 (2001) 656–665.
[8] F.M. Kone, M. Le Bechec, J.P. Sine, M. Dion, C. Tellier, Prot. Eng. Des. Sel. 22
(2009) 37–44.
[9] I. Mladenoska, C.E. Grey, E. Winkelhausen, S. Kuzmanova, P. Adlercreutz, Bio-
catal. Biotrans. 25 (2007) 382–385.
Acknowledgements
[10] P. Lundemo, P. Adlercreutz, E.N. Karlsson, Appl. Environ. Microbiol. 79 (2013)
3400–3405.
[11] T. Hansson, M. Andersson, E. Wehtje, P. Adlercreutz, Enzyme Microb. Technol.
29 (2001) 527–534.
This work was supported by the Swedish Research Council (VR)
and the EU FP7 program AMYLOMICS.
[12] I. Mladenoska, E. Winkelhausen, S. Kuzmanova, Food Technol. Biotechnol. 46
(2008) 311–316.
[13] A.M. Klibanov, Trends Biotechnol. 15 (1997) 97–101.
[14] G. Ljunger, P. Adlercreutz, B. Mattiasson, Enzyme Microb. Technol. 16 (1994)
751–755.
[15] A.E. Petersson, P. Adlercreutz, B. Mattiasson, Biotechnol. Bioeng. 97 (2007)
235–241.
[16] P. Turner, D. Svensson, P. Adlercreutz, E.N. Karlsson, J. Biotechnol. 130 (2007)
67–74.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.
2014.06.009.
References
[17] M. Persson, I. Mladenoska, E. Wehtje, P. Adlercreutz, Enzyme Microb. Technol.
31 (2002) 833–841.
[18] C. Turner, P. Turner, G. Jacobson, K. Almgren, M. Waldeback, P. Sjoberg, E.N.
Karlsson, K.E. Markides, Green Chem. 8 (2006) 949–959.
[19] U.K. Laemmli, Nature 227 (1970) 680.
[20] A. Blinkovsky, in, Novo Nordisk Biotech, Inc., USA. 1998, 5 pp., Cont.-in-part of
U.S. 5,693,516.
[1] S. Matsumura, K. Imai, S. Yoshikawa, K. Kawada, T. Uchibor, J. Am. Oil Chem.
Soc. 67 (1990) 996–1001.
[2] B.A.L. Elias Meezan, B.A.L. Dennis J. Pillion, in, US 4868289 (Sep, 1989) Mag-
nusson et al., 36/4.1; US 5182258 (Jan, 1993) Chiou 514/3; US 5369095 (Nov,
1994) Kee et al., 514/24, US, 1997.
[21] B. Moore, J. Partridge, P. Halling, in: E. Vulfson, P. Halling, H. Holland (Eds.),
Enzymes in Nonaqueous Solvents, Humana Press, 2001, pp. 97–104.
[22] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[23] A. Zaks, A.M. Klibanov, Proc. Natl. Acad. Sci. U.S.A. 82 (1985) 3192–3196.
[24] D.G. Rees, P.J. Halling, Enzyme Microb. Technol. 28 (2001) 282–292.
[3] M. le Maire, P. Champeil, J.V. Moller, Biochim. Biophys. Acta 1508 (2000)
86–111.
[4] W. von Rybinski, K. Hill, Angew. Chem. Int. Ed. 37 (1998) 1328–1345.
[5] B.M. de Roode, M.C. Franssen, A. van der Padt, R.M. Boom, Biotechnol. Prog. 19
(2003) 1391–1402.