Immobilization of thermostable β-glucosidase variants on acrylic supports for biocatalytic processes in hot water

Article abstract of DOI:10.1016/j.molcatb.2012.01.004

Two variants of the thermostable β-glucosidase TnBgl1A (wt and N221S/P342L) from Thermotoga neapolitana were immobilized on acrylic supports (Eupergit C, Eupergit C250L, and cryogel) and evaluated at conditions close to the boiling point of water. Thermo-

Full text of DOI:10.1016/j.molcatb.2012.01.004

Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Immobilization of thermostable ␤-glucosidase variants on acrylic supports for
biocatalytic processes in hot water
Samiullah Khana, Sofia Lindahlb, Charlotta Turnerb, Eva Nordberg Karlssona,∗
a Biotechnology, Department of Chemistry, Lund University, P.O. Box 124, SE-221 00 Lund, Sweden
b Centre for Analysis and Synthesis, Department of Chemistry, Lund University, 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:
Two variants of the thermostable ␤-glucosidase TnBgl1A (wt and N221S/P342L) from Thermotoga neapoli-
tana were immobilized on acrylic supports (Eupergit® C, Eupergit® C250L, and cryogel) and evaluated at
conditions close to the boiling point of water. Thermo-gravimetric analysis showed the supports to be
stable <250 ◦C. Both wt and N221S/P342L were covalently bound to oxirane-groups respectively via glu-
taraldehyde spacers, and for coupling reactions 26 Lys and 20 Ser/Thr were surface-located. Immobilized
enzymes were active on all supports in the temperature range 80–95 ◦C, but the observed specific activity
was low (≤19 U mg−1) using the cryogel. More than 91% of the initial activity was maintained after ten
times recycling, and the same was recovered after 3 months storage at 4 ◦C for Eupergit® supports by sim-
ply incubating the preparation with bovine serum albumin. No storage loss was detectable on cryogels.
The glutaraldehyde spacer improved activity on cryogels, but not on Eupergit® supports. Immobiliza-
tion on Eupergit® C250L yielded the highest observed specific activity (254 U mg−1 for N221S/P342L)
in a procedure including blocking of free oxirane-groups by BSA. This biocatalyst was used for on-line
hydrolysis of quercetin-glucosides in a yellow onion extract at 80 ◦C, proving the immobilized biocatalyst
to be promising in on-line systems for extraction and hydrolysis using hot pressurized water.
© 2012 Elsevier B.V. All rights reserved.
Received 19 August 2011
Received in revised form
31 December 2011
Accepted 6 January 2012
Available online 16 January 2012
Keywords:
Immobilization
Glycoside hydrolase family 1
Eupergit®
C
Eupergit® C250L
Cryogel
1. Introduction
In many processes, it can be an advantage to use enzymes
that are operationally stable at high temperatures as many pro-
Biocatalysts are attractive alternatives to chemical catalysts in
industry for many reasons, including high substrate specificity,
and production of stereo-specific products. Consequently, much
research has been undertaken to develop enzyme-mediated reac-
tions [1,2]. In addition, with the discovery and exploitation of an
increasing number of enzymes from extremophiles, the window
of operation for biocatalysis is expanded allowing use at temper-
atures up to at least the boiling point of water [3]. Enzymes have
the potential to make biomass processes environmentally friendly,
being renewable and biodegradable (compared to chemical cata-
lysts) and able to catalyze specific conversions. In order to make
it economically feasible to use enzymes on a large scale, a recy-
cling system for the catalyst can be developed which will keep
down costs as well as simplify separation of the biocatalyst from
the obtained products.
cesses require a rise in temperature for more efficient substrate
release/conversion [4]. When renewable resources such as agri-
cultural crops or wood are utilized (for extraction of high value
products, or intermediates for direct bioconversion into chemicals,
commodities and fuels) thermostable enzymes have an obvious
advantage as higher temperatures often promote better enzyme
penetration and cell-wall disorganization of the raw materials
[5–7].
Much research has been aimed at investigating the use of ther-
mostable enzymes in applications where a rise in temperature
could be beneficial. Previous work in our laboratory [4] have for
example shown that thermostable ␤-glucosidases from Thermotoga
neapolitana could be utilized directly after hot pressurized water
extraction (HPWE) of quercetin glucosides from yellow onion waste
(at 80–90 ◦C), hydrolysing the glucosylated species of quercetin to
a uniform aglycone. Quercetin is an antioxidizing compound with
electron scavenging properties that may slow down or prevent
the development of cancer [8]. Variants of the same enzyme have
also been developed to further increase conversion efficiency of
quercetin glucosides [9]. ␤-Glucosidases hydrolyze the glycosidic
bond between two glucose residues or between a glucose residue
and non-carbohydrate aglycone, and the latter reaction is utilized in
∗
Corresponding author at: Department of Biotechnology, Lund University, P.O.
Box 124, SE-221 00 Lund, Sweden. Tel.: +46 46 222 4626; fax: +46 46 222 4713.
E-mail addresses: sami.khan@biotek.lu.se (S. Khan),
sofia.lindahl@organic.lu.se (S. Lindahl), charlotta.turner@organic.lu.se (C. Turner),
eva.nordberg karlsson@biotek.lu.se (E.N. Karlsson).
1381-1177/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2012.01.004

S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
29
the quercetin conversions. The possibility to use the thermostable
enzyme in immobilized form in connection with the PHWE is not
yet investigated. The focus of this work is hence to immobilize
the T. neapolitana ␤-glucosidase and evaluate some properties of
the resulting biocatalyst, relevant for its application in a hot water
process.
Immobilization, or attachment of the enzymes to inert car-
rier supports, allows them to be held in place throughout the
reaction, facilitating enzyme recycling through multiple cycles of
batch wise hydrolysis and during use in flow systems. In addition,
enzyme immobilization frequently leads to improved thermosta-
bility, improved shelf-life and resistance to shear inactivation,
resulting in increased process robustness, and a longer duration
of the activity of the biocatalyst [10,11].
Methods traditionally used for enzyme immobilization include:
adsorption, ionic binding, entrapment or crosslinking. Immobiliza-
tion via crosslinking, by covalent bonds, is usually the method
of choice, if harsh conditions, such as high temperature are used
in the following process. Covalent crosslinking leads to minimal
protein leaching from the supports, absence of a barrier between
the enzyme and the reaction medium and rational control of
the enzyme properties may be achieved [12]. Epoxy (oxirane)-
activated supports allow easy immobilization of enzymes under
mild conditions at both laboratory and industrial scale, and are
highly stable during storage, chemically and mechanically, over
a pH range from 2 to 12. They form very stable covalent multi-
point attachment with amino, hydroxyl, thiol and phenolic groups
of amino acid side chains on the enzyme surface [13] and capping
of the remaining epoxy groups can be made to prevent unde-
sirable protein–support interactions using a variety of reagents
(bovine serum albumin, mercaptoethanol, ethanolamine, glycine,
etc.) [12,14]. The epoxy group can be used directly for covalent
attachment. However, because this sometimes results in steric
hindrance of the enzyme, a connecting spacer can be added, for
example by modification of the support to an amine form by
reacting the epoxy group with reagents such as ethylenediamine
followed by reaction with a dialdehyde (e.g. glutaraldehyde) to
produce an aldehyde activated support [15].
under vacuum to remove dissolved oxygen. The samples were kept
in an ice bath, tetramethylethylenediamine (26 ␮L) was added and
the reaction was initiated by the addition of an aqueous solution
of ammonium persulphate (19.3 mg/mL). Portions of 0.5 mL of
the reaction mixture were poured into cold glass tubes (diameter
7 mm) and were frozen at −12 ◦C, for 16 h in ethanol using a
cryostat (Lauda RK20KP, Königshofen, Germany). The samples
were defrosted at room temperature, washed with deionized
water and stored at 4 ◦C for further use [16].
2.2.2. Glutaraldehyde extension onto epoxy activated supports
All three matrices were washed with de-ionized water and
then with 0.2 M Na2CO3 solution. Ethylenediamine (0.5 M in 0.2 M
Na2CO3) was applied to supports in a tube and mixed on a shak-
ing table for 4 h. After washing with water until the pH was
close to neutral the supports were washed with 0.1 M sodium-
phosphate buffer, pH 7.2. A solution of glutaraldehyde (5%, V/V) in
0.1 M sodium-phosphate buffer, pH 7.2 was applied to Eupergit®
C, Eupergit® C250L and cryogels in a tube and the mixture was
left under gentle shaking for 5 h at room temperature (Method 2,
step I and II, Fig. 1). The support materials were washed with 0.1 M
sodium phosphate buffer pH 7.2 [16].
2.3. Scanning electron microscopy and thermogravimetic
analysis of supports
Cryogel was sliced into thin slices using a scalpel, approx. 2 mm.
Two milliliters of citrate phosphate buffer pH 5.0 was added to one
slice of cryogel or 20 mg of Eupergit® C and heated at 95 ◦C for 1 h.
The buffer was removed and the cryogel and the Eupergit® C beads
were mounted on holders using double sided tape and stored in an
exicator overnight. The samples were sputtered with gold and SEM
images were collected using the scanning electron microscope LEO
1530 (Cambridge, UK) at 3 kV.
Approximately 1 mg of cryogel (chopped) or Eupergit® C was
weighed in an alumina pan and loaded into the graphite furnace
of a TGA Q500 instrument (TA Instruments, New Castle, USA). The
temperature program was set to 10 ◦C/min from room temperature
up to 500 ◦C and the measurement was done under air.
In this investigation the thermostable ␤-glucosidase and a
variant of the same enzyme from T. neapolitana were covalently
immobilized with, as well as without, a spacer arm, on three differ-
ent epoxy-activated polyacrylic matrices (Eupergit® C, Eupergit®
C250L and cryogel), and the properties of the immobilized deriva-
tives were studied. Thermostability of the support materials, reuse
of the immobilized enzymes, long term storage stability, and
post-immobilization stabilization were studied, and finally their
application in flavonoid hydrolysis was tried.
2.4. Expression and purification of the TnBgl1A variants
The genes encoding the mutant (N221S/P342L) and wild-type
(wt) glucosidases were cloned in plasmid pET-22b under control of
the T7/lac promoter, including a C-terminal hexa-histidine tag [3].
The glucosidases were produced intracellularly in 2.5 L Escherichia
coli batch cultivations at 37 ◦C, pH 7, using a defined medium [17]
with 100 ␮g/mL ampicillin and a dissolved oxygen tension (DOT)
above 40%. Expression was induced at OD620 nm = 3, by the addi-
tion of 1 mM isopropyl-beta-d-thiogalactopyranoside (IPTG), and
continued for 3 h.
2. Materials and methods
2.1. Chemicals
The E. coli cells were harvested by centrifugation (10,000 × g,
10 min, 4 ◦C), suspended in binding buffer (20 mM imidazole,
20 mM Tris–HCl, 0.75 M NaCl, pH 7.5) and the ice-chilled cell
suspension was lysed by sonication for 5 × 3 min using a 14 mm
titanium probe with a sound intensity of 60% and a cycle of
0.5 (UP400S, Dr. Hielscher), centrifuged (30 min, 39,000 × g, 4 ◦C).
Heat-labile E. coli proteins were removed by heat treatment (70 ◦C,
30 min) and centrifugation (30 min, 39,000 × g, 4 ◦C). The super-
natant was passed through a 0.45 ␮m Minisart high-flow filter
(Sartorius, Göttingen, Germany) and purified on an ÄKTA prime
system (Amersham Biosciences, Uppsala, Sweden) by immobilized
metal ion affinity chromatography (IMAC) using copper as a ligand
as described elsewhere [18]. The fractions containing the purified
protein were pooled and dialyzed against 0.1 M sodium phosphate
buffer, pH 7.2, overnight using a Spectra/Por dialysis membrane
All the chemicals were of pro-analysis grade from Merck Euro-
labs (Darmstadt, Germany) unless otherwise stated.
2.2. Preparation of support materials
2.2.1. Epoxy-activated supports
The two epoxy-group-containing matrices were Eupergit®
C
(purchased from Sigma (Sigma Aldrich)) and Eupergit® C250L
(kindly provided by Evonik Röhm GmbH).
Cryogels containing epoxy-groups were prepared from
monomers (1.316 g acrylamide, 0.354 g N,N methylene bisacry-
lamide and 0.28 mL of allyl glycidyl ether) dissolved in deionized
water (final concentration 10%) and the solutions were degassed

30
S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
Method 1
O
OH
Protein
NH2
CH2
CH2
HN
Protein
CH
CH
Matrix
Matrix
H2N
Method 2
O
OH
CH
NH2
CH2
I
CH
CH2
Matrix
HN
HN
Matrix
NH2
OH
CH
O
O
OH
CH
CH
O
CH2
Matrix
CH2
HN
CH
Matrix
N
NH2
CH
II
OH
CH
OH
O
CH2
HN
Matrix
NH2
Protein
N
Protein
N
CH
CH2
HN
III
CH
Matrix
N
OH
CH
OH
CH
Protein
NaBH4
N
CH2
HN
Matrix
CH2
HN
N
Protein
IV
Matrix
NH
NH
Fig. 1. Schematic overview of the two main methods used for the covalent coupling of TnBgl1A (wt and N221S/P342L) to the support. Method 1 represents coupling via
oxirane groups and Method 2 shows coupling onto the glutaraldehyde groups.
(Eupergit® C or Eupergit® C250L) or a piece of cryogel (42.2 mg) was
pre-incubated at the specific reaction temperature for 10 min and
in parallel the pNPGlc solution (1.46 mM) was pre-heated at the
same conditions. Pre-heated substrate solution was added to the
pre-incubated immobilized enzyme beads (1 mL) or cryogel piece
(10 mL). The sample was incubated at the selected temperature
for 1 min and put on ice for 5 min. After 5 min, the samples were
removed from the ice and left at room temperature for another
5 min. The absorbance was measured spectrophotometrically at
405 nm (Perkin Elmer). pNPGlc solution incubated at the same con-
ditions was used as blank.
with a 3500 Da molecular weight cut-off (Spectrum Laboratories,
Rancho Dominguez, CA, USA). The dialysed protein fractions were
stored at 4 ◦C until use.
2.5. Protein and enzyme analyses
2.5.1. Surface exposed amino acids for immobilization
A three-dimensional structure model of TnBgl1A [19] was used
to display surface exposed amino acids. Figures were prepared
using PyMol (Delano Scientific LLC).
All measurements were performed in triplicate, and specific
activity (U/mg) was determined using the extinction coefficients
2.5.2. Total protein estimation
The concentration of protein was estimated by the bicinchoninic
acid (BCA) method (Sigma, Steinheim, Germany) according to
the manufacturer’s instruction, using bovine serum albumin
(0.2–1 mg/mL, Sigma–Aldrich) as standard.
The amount of protein bound to the support was calculated as:
(concentration in solution before immobilization) − (concentration
in the filtrate after immobilization).
of pNP (para-nitrophenol) ε80 C, 405 nm = 2.4639 × 103 mL/mM cm,
◦
ε85 C, 405 nm = 2.5588 × 103 mL/mM cm, ε90 C, 405 nm = 2.9527 × 103
◦
◦
mL/mM cm, ε95 C, 405 nm = 3.0965 × 103 mL/mM cm. 1 U was 1 ␮mol
◦
of p-nitrophenol formed per minute at the respective reaction tem-
perature.
2.6. Immobilization of the TnBgl1A variants
2.5.3. Estimation of ˇ-glucosidase activity
Enzyme activity was estimated every 5 ◦C in the interval
80–95 ◦C by determining release of p-nitrophenol using pNPGlc
(para-nitrophenyl-␤-d-glucopyranoside) as substrate in 20 mM
citrate phosphate buffer pH 5.6. A certain quantity of immobilized
enzyme beads equivalent to 2.4 ␮g of wt and 1.2 ␮g of N221S/P342L
In order to immobilize ␤-glucosidase variants on Eupergit®
C, Eupergit® C250 L and the cryogel, two immobilization meth-
ods were used. In the first method enzymes were immobilized
directly via oxirane groups and in the second method a deriva-
tized matrix with functional aldehyde groups was used for

S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
31
Table 1
Specific activity of immobilized TnBgl1A variants at different temperatures using different supports. The preparation with the highest specific activity (for the respective
variant) is shown in bold. Vmax for the free enzyme at 80 ◦C was 540 U/mg (wt), and 1890 U/mg (N221S/P342L).
TnBgl1A variants
Support
Additive
Reactive group
Specific activity
Specific activity
Specific activity
Specific activity
Immob. efficiency
80 ◦C (%)
80 ◦C (U mg−1
)
85 ◦C (U mg−1
)
90 ◦C (U mg−1
)
95 ◦C (U mg−1
)
N221S/
P342L
EupC
EupC
EupC
EupC250L
Cryogel
Cryogel
Cryogel
–
Epoxy
Epoxy
57.8 5.5
77.0 14.7
41.0 2.1
243.7
58.4 2.6
80.3 3.0
47.9 6.4
254.0 12.0
1.3
49.5 2.8
65.8 6.0
36.1 0.8
242.1 24.6
1.1
48.2 2.8
67.3 4.9
35.5 1.9
238.9 3.7
0.6
3.1
4.1
2.2
1% glu
–
1% glu, 1% BSA
–
Aldehyde
Aldehyde
Epoxy
Aldehyde
Aldehyde
12.9
1.5
0.1
0.7
0.5
–
12.8 0.6
10.2 0.004
16.9 1.8
10.9 0.001
16.6 2.1
11.7 0.004
19.0 1.8
10.0 0.009
1% glu, 1% BSA
Wild type
EupC
EupC250L
Cryogel
–
Epoxy
Aldehyde
Aldehyde
30.7 4.9
110.4 14.1
2.7 0.4
33.2 2.3
131.3 26.6
4.0 0.4
32.5 5.6
93.8 10.6
3.9 0.0
25.7 6.7
93.0 2.5
3.8 0.6
5.7
20.4
0.5
1% glu, 1% BSA
–
immobilization (Fig. 1). Immobilization efficiency (Table 1) was
calculated based on specific activity and is per milligram protein:
(specific activity of immobilized enzyme on any support/enzyme
activity added using 1 mg enzyme) × 100.
80–95 ◦C in the presence of pNPGlc followed by spectrophotometric
measurement of released pNP (during 1 min) under the conditions
given above (see Section 2.5.3). After each activity measurement,
the bound enzyme was washed three times with 20 mM citrate
phosphate buffer, pH 5.6. The Eupergit® C and Eupergit® C250L
beads were then centrifuged at 13,000 rpm for 5 min, the super-
natant was decanted and the recycled supports subjected to the
activity assay in a second cycle and so on. The washing step and
activity measurement were the same as above for the cryogel
pieces except that the centrifugation step was replaced by man-
ually pressing the gel against a tissue paper followed by washing
with buffer. Finally the absorbance of wash fractions from both
Eupergit® and cryogel were measured spectrophotometrically at
405 nm.
2.6.1. Immobilization via oxirane groups (Method 1, Fig. 1)
Purified ␤-glucosidase (0.350 mg/mL) in 0.1 M sodium phos-
phate buffer pH 7.2 (30 mL) was added to 2.0 g Eupergit® C or
15 mL of the above enzyme solution was added to 8 pieces of cryo-
gel where one piece of cryogel approximately weighs 42.2 mg. The
support (Eupergit® C or cryogel) and enzyme solution were gently
incubated at room temperature on a shaking table for 24 h. After
incubation, the Eupergit® C beads were collected by vacuum filtra-
tion using a porous glass filter and rinsed thoroughly on the same
filter with plenty of MilliQ water and then with 20 mM citrate phos-
phate buffer, pH 5.6. In a similar fashion, the cryogel pieces were
collected, washed thoroughly with plenty of MilliQ water and then
with 20 mM citrate phosphate buffer, pH 5.6 [20].
2.9. Pressurized hot water extraction
Yellow onion was bought from a local supermarket in Sweden
and the onion was chopped, both skin and bulb, and kept in
a freezer (−20 ◦C). Approximately 15 g of chopped onion was
weighed into 33 mL stainless steel extraction vessels, with a glass
fiber filter in the bottom, and loaded into a Dionex ASE®-200
instrument (Dionex, Sunnyvale, CA, USA). Water was used as
extraction solvent as described by Turner et al. [4]. The extrac-
tion was done at a pressure of 50 bar, at a temperature of 120 ◦C,
with an initial heating of 6 min and the extraction time was
15 min (5 min × 3 cycles). After each extraction cycle the cell was
purged with N2 (g) and the extract was collected in 100 mL clear
glass vials. The extraction was repeated six times and the col-
lected extracts were pooled and diluted to 500 mL with MilliQ
water
2.6.2. Immobilization onto the glutaraldehyde linker (Method 2,
Fig. 1)
The derivatized matrix with functional aldehyde groups was
used for immobilization of a TnBgl1A variant. A 65 mL solution of
␤-glucosidase (0.225 mg/mL) in 0.1 M sodium phosphate buffer, pH
7.2, was mixed with 20 g of either Eupergit® C or Eupergit® C250L,
or with 72 pieces of cryogel and incubated at room temperature
for 24 h. Finally, 70 mL of freshly prepared NaBH4 solution, 0.1 M in
0.2 M sodium carbonate buffer pH 9.2, was applied to the cryogels
for 3 h to reduce the Schiff-base formed between the enzyme and
the aldehyde containing matrix [16].
2.7. Addition of glucose and bovine serum albumin in the
immobilization procedure
2.10. On-line hydrolysis of yellow onion extract and
quantification of quercetin
As glucose is the hydrolysis product in ␤-glucosidase catalyzed
reaction and can be bound to the active site, the effect of its presence
during the immobilization procedure was evaluated by the addition
of 1% glucose (w/v) to the enzyme solution prior to the addition of
support (Eupergit® C, Eupergit® C250L or cryogel).
The importance of blocking unreacted epoxy-groups was eval-
uated by incubating freshly washed Eupergit® C, Eupergit® C250L
and cryogels after the enzyme binding step with 1% (w/v) bovine
serum albumin (BSA) in 0.1 M phosphate buffer pH 7.2 at room
temperature for 24 h.
A home built continuous flow system was used for the on-
line hydrolysis of the yellow onion extract (Fig. 2). A 0.2 g portion
(with added glucose during the immobilization procedure and with
BSA blocking) of Eupergit® C250L with immobilized N221S/P342L,
(according to Method 2, Fig. 1) was loaded into a home built
peek holder and the peek holder was loaded into a stainless steel
guard column from Agilent Technologies and placed in the GC-
oven. Twenty milliliters of yellow onion extract was mixed with
20 mL of 40 mM citrate-phosphate buffer, pH 5.5 (Solvent A). The
temperature of the GC-oven was set to 80 ◦C and the flow of the
pump was 1 mL/min. A gradient between 20 mM citrate-phosphate
buffer, pH 5.5, (Solvent B) and the onion-buffer solution (Solvent
A) was used; 0–5 min 100% B, 5–5.5 min 100% A, 5.5–15.5 min 100%
A, 15.5–16 min 100% B and 16–21 min 100% B. Between 0–5 min
2.8. Recycling of the immobilized enzymes
The immobilized preparations were repeatedly incubated
(10 min residence time) at a given temperature in the interval

32
S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
Fig. 2. Instrument set-up of the continuous flow system used for the on-line hydrolysis of the yellow onion extract. The system was composed of one HPLC-pump (HP1050,
Agilent Technologies, Walbronn, Germany), a GC-oven (HP 5890, Agilent Technologies, Walbronn, Germany), a fraction collector (Fraction Collector III, Waters, Milford, US)
and 1/16 in. stainless steel tubings were used. Upstream the cell with enzyme, a 2 m coil of tubing was used, to ensure enough time to heat up the solvent before reaching
the enzymes.
and 16–21 min the fraction collection time was 2.5 min/sample
and between 5 and 16 min the fraction collection time was 1 min
per sample. To the collected fractions, morin (internal standard),
methanol and formic acid were added, to a final concentration
of 0.03 g/L of morin, 40% of methanol and 0.5% of formic acid.
Methanol and formic acid were added to avoid precipitation of
quercetin (Q) and to lower the pH to ensure that hydroxyl groups
of Q are protonated. 500 ␮L of the collected fractions were diluted
with 1000 ␮L of mobile phase (see below) and analyzed with HPLC-
UV.
The injection volume was 10 ␮L and detection was accom-
plished at 350 nm. Quantification of Q, quercetin-3-glucoside (Q-3),
quercetin-4ꢀ-glucoside (Q-4ꢀ), quercetin-3,4ꢀ-diglucoside (Q3,4ꢀ)
and morin (internal standard) was done using a five-point calibra-
tion curve (standards of the respective compound at concentrations
between 0.5 and 25 mg/mL). Each vial taken to analysis had a total
volume of 1500 ␮L.
3. Results
The chromatographic system used was an UltiMate 3000 from
Dionex (Germering, Germany) and the column used was an Agilent
Zorbax SB-C18 column (100 mm × 2.1 mm, 3.5 ␮m), with an Agi-
lent Zorbax SB-C8 pre-column (2.1 mm × 12.5 mm, 5 ␮m). Isocratic
separation was used, with a methanol–water (40:60) and formic
acid (0.13 M, 0.5 vol%) mobile phase at a flow rate of 0.15 mL/min.
3.1. Thermostability of support materials
Two different experiments were set up to confirm the heat
stability of the cryogel and Eupergit® C support materials at
temperatures relevant for the process conditions. In the first
Fig. 3. Scanning electron microscope images of (a) non-heat treated cryogel, (b) heat treated cryogel (1 h at 95 ◦C), (c) non-heat treated Eupergit® C and (d) heat treated
Eupergit® C (1 h at 95 ◦C).

S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
33
3.2. Surface-exposed amino acid side-chains on TnBgl1A variants
100
90
80
70
60
50
40
30
20
10
0
a
b
Immobilization via epoxy- and aldehyde groups allows direct
chemical coupling to amino-groups ( NH2), and via the epoxy
groups also hydroxyl-groups ( OH) on the amino acid side-chains
[13]. Lys-residues are the most commonly utilized anchoring points
as they are typically present on the exterior of proteins [13,22].
Inspection of the 3D-structure model of TnBgl1A showed that 26
of the total 27 Lys residues in the sequence were located on the
surface (Fig. 5a–c), evenly spread, and available for covalent inter-
actions with aldehyde- and epoxy groups on the support materials.
Five of the Lys-residues surrounded the active site (Fig. 5a), which
can lead to covalent bonds that obstruct activity of the immobilized
enzyme. In addition, serine (Ser) and threonine (Thr) residues were
analyzed. Ten of the 18 Ser and 10 of totally 15 Thr were surface
located, of which 3 Ser-residues flank the active site (Fig. 5a).
Cryogel
Eupergit® C
3.3. Immobilization of TnBgl1A wt and N221S/P342L
Three epoxy-activated acrylic polymer matrices with differ-
ent texture properties were used and included particulate beads
of Eupergit® C and Eupergit® C250L (differing in pore size and
oxirane number) and a monolithic cryogel. Firstly, the enzyme vari-
ant N221S/P342L was used, and selected immobilization conditions
repeated for the wt enzyme. The N221S/P342L was immobilized
both directly onto the epoxy group (Method 1, Fig. 1) and via
an aldehyde spacer arm (Method 2, Fig. 1). It was found that the
0
100
200
300
400
500
Temperature, °C
Fig. 4. Thermogravimetric analysis of cryogel and Eupergit® C. The change of weight
of the sample with temperature is given. The profile shows: (a) weight loss due to
loss of water and (b) weight loss due to decomposing of the polymer.
aldehyde had little effect on specific activity using Eupergit®
C
(1.3× specific activity decrease was observed using the spacer
arm, Table 1). However the spacer arm had a drastic effect on
the cryogel support, where the specific activity was approximately
50 times lower compared to Eupergit® C when the enzyme was
directly linked to the epoxy group, but increased one order of
magnitude upon introduction of the spacer arm (Table 1). In this
set of experiments, no effort was made to protect the active site
or to block oxirane groups remaining after incubation with the
enzyme.
experiment, the materials were heated in buffer at 95 ◦C for 1 h and
SEM images were collected of non heat-treated material and heat-
treated material. Fig. 3 shows that there is no obvious conformation
change of the support materials after heat treatment.
In the second experiment the temperature range was extended,
to study at which temperature the materials start to decompose,
using thermogravimetric analysis (TGA). The result of the TGA
shows that Eupergit® C and cryogel starts to decompose around
250 ◦C (Fig. 4). These results confirm that the support materials are
stable for use in systems operating up to the current upper limit for
protein stability (150 ◦C) [21].
3.3.1. Effect of additives on ˇ-glucosidase immobilization
According to Tu et al. [23], application of 1% glucose and 1% BSA
in the immobilization procedure of a ␤-glucosidase improves the
immobilization efficiency. To investigate this for the N221S/P342L
Fig. 5. Surface view of the 3D structure model of TnBgl1A, constructed using ␤-glucosidase BglA from Thermotoga maritima (TmGH1 (PDB code 2WC4)) as template. From
left to right: (a) front view (with the active site pocket), (b) side view, and (c) view of the side opposing the active site of TnBgl1A. Lysine residues are labelled blue, and serine
or threonine residues are shown in orange.

34
S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
Fig. 6. Reusability of immobilized ␤-glucosidase variants (immobilized directly to the epoxy group (a-b) or with spacer arm (c)) on Eupergit® C and Eupergit® C250L was
monitored in the temperature range 80–95 ◦C without a noticeable loss of specific activity (U/mg). (a) Immobilized TnBgl1A (wt) on Eupergit® C recycled three times (b)
immobilized (N221S/P342L) on Eupergit® C recycled three times (c) immobilized (N221S/P342L) on Eupergit® C250L recycled ten times. The numerical numbers on the
x-axis shows the cycle number.
leads to an approximately 5 times decrease using Eupergit®
C250L, while on the cryogel the specific activity decrease is
almost two orders of magnitude, using the current evaluation
methodology.
variant, glucose (1%, w/v) was added to the enzyme solution prior to
immobilization by direct linkage to the epoxy groups of Eupergit®
C (Method 1, Fig. 1). Glucose additions led to a slight improve-
ment in specific activity (see Table 1) while addition of bovine
serum albumin (BSA) after immobilization of the enzyme to block
unreacted oxirane or aldehyde groups showed stronger effect on
the Eupergit® support materials. In these experiments Eupergit®
C
3.4. Batch recycling and storage of immobilized TnBgl1A variants
250L was used instead of Eupergit® C, as the latter was no longer
available at the supplier. Moreover, the glutaraldehyde spacer arm
was used, as complementary experiments had shown this to be
better for long term use (see below). In accordance with the results
by Tu et al. [23], blocking by BSA resulted in a 5 times increase of
the specific activity of the thermostable TnBgl1A variants (Table 1).
A contribution to this effect may however also be the larger pore
size of Eupergit® C250L which previously has been suggested to
improve enzyme diffusion into the macroporous matrix [24]. How-
ever, when BSA was added to a stored preparation immobilized of
Eupergit® C with the lower pore size (see below), also in this case
specific activity was increased, indicating that BSA is indeed needed
to reduce substrate/support interactions.
The possibility to use the enzyme in repeated cycles (repeated
batches) to obtain more product is a most attractive characteristic
of immobilized enzyme applications. The ␤-glucosidase variants
immobilized on Eupergit® C without spacer arm (Method 1, Fig. 1)
was recycled 3 times and the activity monitored after each cycle
(Fig. 6a and b). N221S/P342L immobilized on Eupergit® C250L with
glutaraldehyde spacer arm (Method 2, Fig. 1) was recycled 10 times
at the temperatures 80, 85, 90 and 95 ◦C (Fig. 6c). The residence time
was 10 min in each cycle followed by a cooling period of the same
length of time. The preparations were washed and centrifuged
between the trials and no activity could be detected in the wash
fractions (data not shown). In these experiments, no significant loss
of activity was observed.
Neither glucose nor BSA had any significant effect on the spe-
cific activity using the cryogel support (Table 1). In all cases,
the specific activity of the immobilized enzyme is significantly
lower than that of free enzyme in solution. Immobilization
In order to investigate the industrial practicability of an immo-
bilized enzyme, the loss of enzyme activity, known as storage
stability, is an important parameter to be taken into account. Hence

S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
35
Fig. 7. Reusability and storage stability of the ␤-glucosidase variants on Eupergit® C250L, immobilized via the glutaraldehyde spacer arm, after storage at 4 ◦C for six months,
followed by recycling ten times at 80–95 ◦C. (a) specific activity (U/mg) of TnBgl1A (wt) and (b) specific activity (U/mg) of (N221S/P342L).
Fig. 8. Effect of the addition of 1% bovine serum albumin (BSA) on the specific activity (U/mg) of ␤-glucosidase variants immobilized on Eupergit® C via the epoxy group. (a)
Specific activity (U/mg) of TnBgl1A (wt) at 80–95 ◦C after three months storage (4 ◦C) in the absence of BSA and with 1% BSA. (b) Specific activity (U/mg) of (N221S/P342L) at
80–95 ◦C after three months storage (4 ◦C) in the absence of BSA and with 1% BSA.
the stability of the immobilized enzymes on cryogels, Eupergit®
C and Eupergit® C250L was studied after storage at 4 ◦C. In a
combined experiment on storage and reusability, the enzyme
immobilized on Eupergit® C250L (Method 2, Fig. 1) was stored at
4 ◦C for 6 months and a 10–30% decrease in specific activity was
observed both in wt and N221S/P342L enzyme (Fig. 7). The decrease
in specific activity was further measured by recycling the enzymes
for 10 times at the selected temperatures in the 80–95 ◦C interval.
As shown in Fig. 7, the immobilized enzyme retained more than 75%
(wt) and 50% (N221S/P342L) of its original activity after 10 reuses,
respectively. The wash fractions were also analyzed, but again no
activity loss to the supernatant could be observed, using the pNPGlc
substrate.
After 3 months there was no loss in activity of enzymes immo-
bilized by Method 2 (see Fig. 1) on cryogels but on the other hand
the Eupergit® C preparation immobilized by Method 1 had lost
approximately 50% of its initial activity (Fig. 8). The specific activ-
ity of TnBgl1A immobilized on Eupergit® C250L (immobilized by
Method 2, Fig. 1) was measured after 6 months storage, and the
enzyme still retained more than 75% of its original activity (Fig. 7).
This clearly indicated that the enzyme preparations are very stable
when immobilized by the glutaraldehyde method.
Fig. 9. Comparison of storage stability of ␤-glucosidase variants immobilized directly to the epoxy groups of Eupergit® C and cryogel. (a) Specific activity (U/mg) of freshly
immobilized TnBgl1A (wt) and after three months storage at 4 ◦C. (b) Specific activity (U/mg) of (N221S/P342L) before three months and after three months.

36
S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
4. Discussion
Previous results in our laboratory have shown that the ther-
mostable ␤-glucosidase A from T. neapolitana, TnBgl1A, as well as
a mutated variant thereof, can be used as a catalyst for the con-
version of quercetin glucosides to quercetin and glucose in yellow
onion extracts obtained from pressurized hot water extractions
[4,9]. Use of the free thermostable enzyme in solution connected
to the high temperature extraction methodology allowed catalysis
without extensive cooling, but use of free enzyme can be expen-
sive as it necessitates batch wise additions. Immobilization was
thus an interesting alternative allowing recovery and reuse of the
biocatalyst. Due to the high temperature conditions in the extrac-
tion process, covalent immobilization methods were chosen, and
two different approaches explored: (a) direct enzyme binding to
the matrices via the oxirane groups, and (b) enzyme binding to the
support via an ethylene diamine/glutaraldehyde spacer.
Three epoxy-activated acrylic polymer matrices with different
texture properties were used in the study and included particulate
beads of Eupergit® C and Eupergit® C250L (differing in pore size
and oxirane number) and a monolithic cryogel. The suitability of the
support material for application in the high temperature method
was confirmed, as the support materials were stable and can be
used in future on-line flow system where temperatures up to 150 ◦C
(the current upper limit for protein stability [21]) are applicable.
Coupling to epoxy- and glutaraldehyde groups on the supports for
immobilization to amino and hydroxyl groups on the enzyme sur-
face, exposed to the aqueous medium [13] was also successful, and
judged to be due to the Lys, Thr and Ser residues evenly distributed
on the enzyme surface as shown by inspection of the 3D-structure
model of TnBgl1A [19]. In principle, a number of different groups
e.g. amino, hydroxyl, thiol and phenolic can react with epoxy- and
aldehyde groups [25] but the amino-group of Lys and hydroxyl of
Thr/Ser are most favored [13] for immobilization.
Fig. 10. Concentration of quercetin-3-glucoside (Q-3) and quercetin in fractions
(were Q-3 and quercetin was detected) collected from on-line hydrolysis of onion
extract using the immobilized enzyme variant N221S/P342L. The fractions were
numbered from the first fraction where quercetin was detected.
In another experiment the activity of an immobilized prepara-
tion of ␤-glucosidase variants on Eupergit® C (Method 1, Fig. 1)
which had lost approximately 50% activity after 3 months storage
at 4 ◦C (Fig. 9) was incubated with 1% BSA for 24 h after which
activity again was measured. By this BSA treatment, the prepa-
ration regained approximately the original activity of the freshly
immobilized enzymes (Fig. 8).
All three types of support materials (Eupergit® C; Eupergit®
C250L and cryogels) used in this study are examples of epoxy acti-
vated acrylic supports, but have some differences. Eupergit® C and
Eupergit® C250L are micro-porous, epoxy-activated acrylic beads
commercially available with a diameter of 100–250 ␮m and a pore
size of 10–100 nm [11], while the cryogels are produced through
gelation at subzero temperature with properties such as elastic and
sponge-like morphology and with a pore size of 1–100 ␮m, giving
a macroporous gel reported to be highly flow permeable [26,27].
Ligands can be coupled directly or via spacer arms to a reactive
group, e.g. epoxy-groups, to both Eupergit® C and cryogels [15]. In
all cases, the specific activity (U/mg) of the immobilized TnBgl1A-
variants did not decrease in the temperature interval 80–95 ◦C,
which is an improvement as the soluble counterpart was only stable
up to 90 ◦C [9]. Some differences in enzyme activity were however
observed between the supports (even though all had the same reac-
tive groups) which are likely due to their different nature [28] and
will be discussed further.
3.5. On-line hydrolysis of quercetin glucosides, using TnBgl1A
(N221/P342L) immobilized on Eupergit® C 250L
With knowledge in hand that the immobilized enzyme was
functional in batch trials using the model substrate pNPGlc, the
experiments were expanded to include a trial using a flow system,
to investigate possibilities to deglycosylate quercetin-glucosides
(desirable to extract and deglycosylates with the combined PHWE
and biocatalysis methodology [4,9]). An enzyme preparation of
N221/P342L, immobilized on Eupergit® C250L (with glutaralde-
hyde linker, immobilized in presence of glucose and pretreated
with BSA), was used. Yellow onion extract mixed with citrate-
phosphate buffer was pumped through a cell with the enzyme at
80 ◦C and fractions were collected at different time intervals. In
Fig. 10 the concentration of Q-3 and quercetin in collected fractions
(where Q-3 and quercetin were detected) of the on-line hydroly-
sis can be seen. The quercetin derivatives identified in the yellow
onion extract-buffer solution (Solvent A) before hydrolysis were
Q-4ꢀ, Q-3,4ꢀ and quercetin (data not shown). The results of the on-
line hydrolysis of the onion extract is that all Q-4ꢀ is hydrolyzed
to quercetin and that Q-3,4ꢀ is firstly hydrolyzed to Q-3 and sec-
ondly to quercetin. Hydrolysis of Q-3 is incomplete. In summary,
the immobilized enzyme appears successful for the deglycosylation
step of quercetin glucosides in hot yellow onion water extract, but
the method have to be further optimized if the goal is to achieve
complete hydrolysis of the quercetin glucosides.
Higher specific activity was observed on Eupergit® C250L than
on Eupergit® C, despite having the same composition. The reasons
for the lower specific activity of Eupergit® C could be, either its
smaller pore diameter which may cause diffusion limitation [14],
or the higher content of the oxirane groups of Eupergit® C that
could result in the formation of multiple covalent bonds between
the support and enzyme, which might disturb the enzyme structure
and cause some loss in activity [24].
The low activity on the cryogel compared to the Eupergit® sup-
ports could be explained if it is assumed that a substantially large
portion of enzymes immobilized on the cryogel were entrapped in
the intrinsic section of the carrier, limiting access of the substrate
to the active site due to diffusion limitation. This explanation is
supported by the fact that the specific activity was higher when

S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
37
binding to Eupergit® C, Eupergit® C250L and cryogel. This is the
first report of covalent immobilization of ␤-glucosidase variants
from T. neapolitana on acrylic supports, and shows that immobiliza-
tion of the enzymes allow recycling, improved the thermostability
and resulted in a preparation with good operational stability
allowing storage (up to 6 months thus far tested). This makes
the enzyme a potential candidate for high temperature process-
ing, successfully used for the hydrolysis of quercetin-glucosides
(with potential for other by-product modifications) in hot water
extractions, and can be a great candidate for various industrial
applications.
the reactive groups in the support were secluded from its surface
through short spacer arms. Thus, the introduction of spacer arms
helps in avoiding steric hindrance effects as it guarantees an addi-
tional distance between the support and the enzyme and places the
enzyme in a better position for catalytic action after immobilization
[29,30]. The ideal situation for enzymes immobilized on cryogel is
hence to pass the substrate through the monolith (e.g. combined
with an on-line flow system), and this way the substrate could
access to the active site of the enzyme entrapped in the intrinsic
section of the carrier and offer a higher activity [27].
One problem with enzymes directly immobilized onto the epoxy
groups of both Eupergit® C and Eupergit® C250L (Method 1, Fig. 1)
was a more rapid loss of enzymatic activity, as previously described
by [30], so the technique that led to higher storage stability was con-
sidered superior and further experiments were carried out using
enzyme immobilized via the glutaraldehyde linker (Method 2,
Fig. 1). The enzyme immobilized by this method retained more than
75% specific activity even after six months (on Eupergit® C250L)
while no loss in activity was observed on cryogels.
To maintain high and stable activity of immobilized enzymes
(which are most important factors to be considered) two meth-
ods, using additives, have been used. In the first method glucose
was added to the enzyme solution before immobilization, which
according to Tu et al. [23] changes the microenvironment and pro-
tects the active site of the enzyme during immobilization, leading
to increased specific activity. Glucose has however previously been
shown to activate TnBgl1A [19,31], and the slight improvement
seen in specific activity (1.3× higher) in our experiments, is in the
same range as the previously reported activating effect (1.1× higher
specific activity). Hence, even though specific activity increased
we cannot conclude that glucose has any significant role in pro-
tecting the active site during immobilization. In another method,
the microenvironment of the enzyme molecules has been changed
by the addition of BSA, which can easily react with active func-
tionalities, such as epoxy rings or aldehyde groups of the support
without the use of additional activating chemistry [32]. Addition of
BSA can significantly increase the specific activity and stability, and
this effect was observed when stored preparations of immobilized
enzyme on Eupergit® C were incubated with BSA. This has been
explained as a hydrophilic environment created into the proxim-
ity of the enzyme enhancing the hydrolytic activity and stability,
and has previously been observed in consecutive modifications
of ␤-glucosidase [23] and penicillin V acylase [33] immobilized
on Eupergit® C. It should however be noted that with our ther-
mostable enzyme, the most promising activities were found when
immobilized enzyme was stored long term in the absence of BSA,
and conditioned with the BSA immediately prior to use. In con-
trast to BSA, blocking of residual groups with 2-mercaptoethanol
had no effect on enzyme activity, strongly suggesting that it is the
introduction of the hydrophilic environment that is favorable for
enzyme stabilization [33]. Upon immobilization, the ␤-glucosidase
could be used for many cycles of hydrolysis without any significant
loss in activity. This makes the immobilized preparation feasible
for use in a flow system, where long stability is desired as the
substrate then can be pumped in and converted in a continuous
way. The high stability (up to 95 ◦C) and good reusability also indi-
cate that longer residence times (applicable in a flow system) could
be used. The possibility to use this type of methodology was also
briefly shown in this work, where quercetin-glucosides from yel-
low onion extracts were converted to their aglycone forms using
the immobilized enzyme preparation.
Acknowledgments
Oskar Werner is thanked for help with SEM analysis and the
Department of Medical Cell Biology, Uppsala University for use of
SEM instrument, Prof. Patric Jannasch and Carlos Rodriguez Arza
(Centre for Analysis and Synthesis, Dept Chemistry, Lund Univer-
sity) are thanked for assistance on TGA analysis and Dr. Fatima
Plieva (Biotechnology, Department of Chemistry, Lund University)
for help with the cryogel work.
The authors wish to thank Formas (2006-1346 (Sustain-X-
Enz), and 2009-1527 (SuReTech)), STINT (YR2009-7015), Oscar &
Lili Lamm’s Foundation and “Kungliga Fysiografiska Sällskapet i
Lund” for funding, and the Higher Education Commission (HEC) of
Pakistan for support to Samiullah Khan.
References
[1] S. Panke, M. Held, M. Wubbolts, Curr. Opin. Biotechnol. 15 (2004)
272–279.
[2] S.B. Rubin-Pitel, H.M. Zhao, Comb. Chem. High Throughput Screen. 9 (2006)
247–257.
[3] P. Turner, G. Mamo, E.N. Karlsson, Microb. Cell Fact. 6 (2007), Art. No. 9.
[4] C. Turner, P. Turner, G. Jacobson, K. Almgren, M. Waldeback, P. Sjoberg, E.N.
Karlsson, K.E. Markides, Green Chem. 8 (2006) 949–959.
[5] S. Fernando, S. Adhikari, C. Chandrapal, N. Murali, Energy Fuels 20 (2006)
1727–1737.
[6] G. Paes, M.J. O’Donohue, J. Biotechnol. 125 (2006) 338–350.
[7] B. Kamm, M. Kamm, Appl. Microbiol. Biotechnol. 64 (2004) 137–145.
[8] S. De, R.N. Chakraborty, S. Ghosh, A. Sengupta, S. Das, J. Exp. Clin. Cancer Res.
23 (2004) 251–258.
[9] S. Lindahl, A. Ekman, S. Khan, C. Wennerberg, P. Börjesson, P.J.R. Sjöberg, E.N.
Karlsson, C. Turner, Green Chem. 12 (2010) 159–168.
[10] J. Torres-Bacete, M. Arroyo, R. Torres-Guzmán, I. de la Mata, M.P. Castillón, C.
Acebal, J. Chem. Technol. Biotechnol. 76 (2001) 525–528.
[11] T. Boller, C. Meier, S. Menzler, Org. Process Res. Dev. 6 (2002) 509–519.
[12] C. Mateo, J.M. Palomo, G. Fernandez-Lorente, J.M. Guisan, R. Fernandez-
Lafuente, Enzyme Microb. Technol. 40 (2007) 1451–1463.
[13] F. Rusmini, Z.Y. Zhong, J. Feijen, Biomacromolecules 8 (2007) 1775–1789.
[14] P.-Y. Wang, S.-W. Tsai, T.-L. Chen, J. Chem. Technol. Biotechnol. 83 (2008)
1518–1525.
[15] M. Petro, F. Svec, J.M.J. Frechet, Biotechnol. Bioeng. 49 (1996) 355–363.
[16] A. Kumar, F.M. Plieva, I.Y. Galaev, B. Mattiasson, J. Immunol. Methods 283 (2003)
185–194.
[17] S.O. Ramchuran, O. Holst, E.N. Karlsson, J. Biosci. Bioeng. 99 (2005)
477–484.
[18] P. Turner, O. Holst, E.N. Karlsson, Protein Expr. Purif. 39 (2005) 54–60.
[19] S. Khan, T. Pozzo, M. Megyeri, S. Lindahl, A. Sundin, C. Turner, E.N. Karlsson,
BMC Biochem. 12 (2011) 11.
[20] M.T. Martin, F.J. Plou, M. Alcalde, A. Ballesteros, J. Mol. Catal. B: Enzym. 21 (2003)
299–308.
[21] T. Tanaka, M. Sawano, K. Ogasahara, Y. Sakaguchi, B. Bagautdinov, E. Katoh,
C. Kuroishi, A. Shinkai, S. Yokoyama, K. Yutani, FEBS Lett. 580 (2006)
4224–4230.
[22] S.V. Rao, K.W. Anderson, L.G. Bachas, Biotechnol. Bioeng. 65 (1999)
389–396.
[23] M. Tu, X. Zhang, A. Kurabi, N. Gilkes, W. Mabee, J. Saddler, Biotechnol. Lett. 28
(2006) 151–156.
[24] Y.J. Cho, O.J. Park, H.J. Shin, Enzyme Microb. Technol. 39 (2006) 108–113.
[25] I. Migneault, C. Dartiguenave, M.J. Bertrand, K.C. Waldron, Biotechniques 37
(2004) 790–802.
5. Conclusion
[26] G. Massolini, E. Calleri, J. Sep. Sci. 28 (2005) 7–21.
[27] D. Josic, A. Buchacher, A. Jungbauer, J. Chromatogr. B: Analyt. Technol. Biomed.
Life Sci. 752 (2001) 191–205.
In conclusion the present work demonstrates immobiliza-
tion of active thermostable ␤-glucosidase variants by covalent

38
S. Khan et al. / Journal of Molecular Catalysis B: Enzymatic 80 (2012) 28–38
[28] C. Mateo, R. Torres, G. Fernandez-Lorente, C. Ortiz, M. Fuentes, A. Hidalgo, F.
Lopez-Gallego, O. Abian, J.M. Palomo, L. Betancor, B.C.C. Pessela, J.M. Guisan, R.
Fernandez-Lafuente, Biomacromolecules 4 (2003) 772–777.
[29] Q. Luo, H. Zou, Q. Zhang, X. Xiao, J. Ni, Biotechnol. Bioeng. 80 (2002) 481–489.
[30] Z. Knezevic, N. Milosavic, D. Bezbradica, Z. Jakovljevic, R. Prodanovic, Biochem.
Eng. J. 30 (2006) 269–278.
[31] D.A. Yernool, J.K. McCarthy, D.E. Eveleigh, J.D. Bok, J. Bacteriol. 182 (2000)
5172–5179.
[32] L. Cao, Curr. Opin. Biotechnol. 9 (2005) 217–226.
[33] J. Torres-Bacete, M. Arroyo, R. Torres-Guzman, I. de la Mata, M.P. Castillon, C.
Acebal, J. Chem. Technol. Biotechnol. 76 (2001) 525–528.