Optimization of immobilization conditions of Thermomyces lanuginosus lipase on styrene-divinylbenzene copolymer using response surface methodology

Article abstract of DOI:10.1016/j.molcatb.2010.01.013

Microbial lipase from Thermomyces lanuginosus (formerly Humicola lanuginosa) was immobilized by covalent binding on a novel microporous styrene-divinylbenzene polyglutaraldehyde copolymer (STY-DVB-PGA). The response surface methodology (RSM) was used to o

Full text of DOI:10.1016/j.molcatb.2010.01.013

Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Optimization of immobilization conditions of Thermomyces lanuginosus lipase on
styrene–divinylbenzene copolymer using response surface methodology
Önder Aybastıer, Cevdet Demir∗
University of Uludag, Faculty of Science and Arts, Department of Chemistry, Gorukle, 16059 Bursa, Turkey
a r t i c l e i n f o
a b s t r a c t
Article history:
Microbial lipase from Thermomyces lanuginosus (formerly Humicola lanuginosa) was immobilized
by covalent binding on a novel microporous styrene–divinylbenzene polyglutaraldehyde copolymer
(STY–DVB–PGA). The response surface methodology (RSM) was used to optimize the conditions for the
maximum activity and to understand the significance and interaction of the factors affecting the specific
activity of immobilized lipase. The central composite design was employed to evaluate the effects of
enzyme concentration (4–16%, v/v), pH (6.0–8.0), buffer concentration (20–100 mM) and immobilization
time (8–40 h) on the specific activity. The results indicated that enzyme concentration, pH and buffer
concentration were the significant factors on the specific activity of immobilized lipase and quadratic
polynomial equation was obtained for specific activity. The predicted specific activity was 8.78 ␮mol p-
NP/mg enzyme min under the optimal conditions and the subsequent verification experiment with the
specific activity of 8.41 ␮mol p-NP/mg enzyme min confirmed the validity of the predicted model. The
lipase loading capacity was obtained as 5.71 mg/g support at the optimum conditions. Operational sta-
bility was determined with immobilized lipase and it indicated that a small enzyme deactivation (12%)
occurred after being used repeatedly for 10 consecutive batches with each of 24 h. The effect of methanol
and tert-butanol on the specific activity of immobilized lipase was investigated. The immobilized lipase
was almost stable in tert-butanol (92%) whereas it lost most of its activity in methanol (80%) after 15 min
incubation.
Received 7 September 2009
Received in revised form 12 January 2010
Accepted 12 January 2010
Available online 20 January 2010
Keywords:
Thermomyces lanuginosus
Immobilization
Enzyme activity
Styrene–divinylbenzene
Response surface methodology
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
tions by the improvement of their catalytic properties such as
activity and stability. Such an improvement can be carried out by
Lipases (triacylglycerol ester hydrolases, EC 3.1.1.3) catalyze the
hydrolysis of triglycerides to free fatty acids, diacylglycerols, mono-
glycerols and glycerol [1]. The use of lipases as industrial catalysts
is a very promising alternative to those used in conventional indus-
trial chemistry. They are also efficient in various reactions such
as ester synthesis, transesterification, interesterification, acidoly-
sis and aminolysis in organic solvents or on support materials [2,3].
Therefore, microbial lipases are currently receiving much attention
for their potential industrial applications in the detergent, food,
flavor industries, biocatalytic resolution of pharmaceuticals, esters
and amino acid derivatives, making of fine chemicals, agrochem-
icals, use as biosensor, bioremediation and cosmetics, perfumery
and biodiesel production [4–6]. However, the drawbacks of the
extensive use of lipases as biocatalyst compared to classical chemi-
cal catalysts can be found in the relatively low stability of enzyme in
their native state. Consequently, there is a great interest in methods
trying to develop competitive biocatalysts for industrial applica-
chemical or physical modifications of the native enzyme.
Immobilized enzymes are versatile catalysts in the laboratory
and on an industrial scale. Enzyme immobilization is the most
commonly used strategy to impart the desirable features of con-
ventional heterogeneous catalysts onto biological catalysts [7,8].
There are different methods for the immobilization of enzymes.
Immobilization facilitates the efficient recovery and reuse of costly
enzymes [9]. Besides enhanced stability, enzymes can acquire
additional advantageous properties via immobilization: (1) immo-
bilized enzymes can be used repeatedly or continuously in a variety
of reactors and (2) they can be easily separated from soluble reac-
tion products and unreacted substrate, thus simplifying work-up
and preventing protein contamination of the final product [10].
Lipases have been successfully immobilized on many different
types of supports, such as celite [11], Nylon-6 [12], chitosan, agarose
[13], hydrotalcite, zeolites [14], and cross-linked enzyme aggre-
gates (CLEAs) [15]. Styrene–divinylbenzene (STY–DVB) copolymer
has peculiar physicochemical and hydrophobic characteristics. It
has been shown as a good potential support material for lipase
immobilization. Lipase from Candida rugosa has been successfully
immobilized on the STY–DVB copolymer by adsorption and had
∗
Corresponding author. Tel.: +90 224 2941727; fax: +90 224 2941899.
E-mail address: cevdet@uludag.edu.tr (C. Demir).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.01.013

Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
171
high activity to perform hydrolysis and esterification reactions [16].
The C. rugosa has also been immobilized by activating the hydroxyl
groups of chitosan using carbodiimide coupling agent. The process
has the advantage of high ability to activate carboxyl groups and
low toxicity to enzymes [17]. Lipases have been immobilized on
different supports by covalent binding, adsorption, entrapment or
cross-linking [18–21]. The scope of enzyme stabilization depends
on the support used and selection of the immobilization process.
On the other hand, by the selection of suitable immobilization
method, support material and the operational cost of biocatalytic
reaction can be significantly reduced [22]. Glutaraldehyde activa-
tionof supportsisone ofthemost populartechniques toimmobilize
enzymes. The glutaraldehyde is used to activate supports by phys-
ical adsorption and covalent binding, immobilizing the enzyme
on a glutaraldehyde pre-activated support [23–25]. Multipoint
covalent attachment of enzymes on highly activated supports pro-
motes a rigidification of the enzyme structure of the immobilized
enzyme [26]. This rigidification reduces any conformational change
involved in enzyme inactivation and increases the enzyme stabil-
ity.
In the presence of any hydrophobic support, the lid moves
to permit the interaction between its hydrophobic face and the
hydrophobic residues that usually surround the lipase active cen-
ter with this hydrophobic surface. This way, the lipase becomes
adsorbed onto this hydrophobic surface, and the active center is
exposed to the reaction medium. This mechanism of action is
called interfacial activation of the lipases. The other strategy that
has been quite successful in the immobilization of Thermomyces
lanuginosus is the use of the peculiar catalytic mechanism of the
lipases to adsorb the enzyme on different hydrophobic supports
via interfacial activation. Thus the immobilized enzyme will sta-
bilize its open conformation [5]. For interfacially active lipases, it
is evident that the selectivity is often conformation-controlled due
to the fact that the lid opening is often a prerequisite for higher
activity and selectivity for the carrier-bound immobilized enzyme
[27].
most successful factorial design for the optimization of parameters
with a limited number of experiments and estimates the response
surface [35–37].
In the present study, immobilization was carried out by circu-
lating the enzyme solution by a peristaltic pump through novel
STY–DVB–PGA beads in a cylindrical column made of stainless
steel. The lipase enzyme is a single chain protein consisting of
269 amino acids. Its molecular weight is 31,700 g/mol and its iso-
electric point is 4.4. The lid is an alpha-helical mobile surface
loop consisting of amino acids 86–93 that covers the active site
[5]. Although the STY–DVB has been already shown as a good
potential matrix for lipase immobilization by physical adsorption,
covalent attachment of lipase on the copolymer and full character-
ization of support before and after immobilization have not been
investigated to our knowledge in literature. We provide data on
the properties of lipase immobilized on the STY–DVB by cova-
lent attachment using scanning electron microscopy (SEM) and
Fourier transform infrared (FTIR) spectroscopy. Immobilization of
lipase is affected by many factors such as enzyme concentra-
tion, pH, buffer concentration and immobilization time. Most of
the studies on immobilization changed one separate factor at a
time. However, immobilization system can be influenced by simul-
taneously changing more than one factor. The main objective
of this work is the development and evaluation of a statisti-
cal approach to better understand the relationship between the
parameters and specific activity of the lipase immobilized on a
novel STY–DVB–PGA support. Furthermore, lipase stability as a
function of incubation time in methanol and tert-butanol was
investigated to understand the effect of these solvents on lipase
activity.
2. Experimental
2.1. Materials
Lipase from T. lanuginosus (lipozyme TL 100 L) was a gift
of Novozymes Enzim Dıs¸ Ticaret Ltd. S¸ ti. (Istanbul, Turkey).
Glutaraldehyde (25% aqueous solution), divinylbenzene, potas-
sium dihydrogen phosphate, dipotassium hydrogen phosphate,
Coomassie brilliant blue G 250 were obtained from Merck. Span 80
and potassium peroxidosulfate were purchased from Fluka. Triton
X-100, gum arabic, p-nitrophenyl palmitate (p-NPP), p-nitrophenol
(p-NP), and bovine serum albumin (BSA) were purchased from
Sigma. All other chemicals were of analytical grade.
T. lanuginosus lipase has been immobilized on hydrophobic
supports to produce biodiesel with canola oil and methanol. The
immobilized lipase proved to be stable and lost little activity
when subjected to repeated uses [6,28,29]. It has been demon-
strated thatenzymesshow higher activityin relatively hydrophobic
organic solvents such as n-hexane and petroleum ether [30,31].
However, short chain alcohols especially methanol have poor sol-
ubility in these hydrophobic solvents, so the negative effects on
lipase activity caused by methanol cannot be eliminated and lipase
still exhibits poor stability in such reaction medium. The effect of
various alcohols (methanol, ethanol, iso-propanol, iso-butanol, ace-
tone, chloroform and acetonitrile) as well as alkanes (n-pentane,
n-hexane and n-heptane) has been studied on the activity of immo-
bilized lipase [12,32]. Among alcohols, an exposure to iso-propanol
increased the activity of bound enzyme while ethanol, methanol
and iso-butanol inhibited the activity. A moderate polar solvent,
tert-butanol was adopted as the organic solvent in which lipase
expressed quite high catalytic activity and operational stability. It
has been confirmed that tert-butanol is inert in the immobiliza-
tion system. The toxicity of methanol on lipase activity can also be
eliminated in the tert-butanol system.
2.2. Synthesis of polyglutaraldehyde
10 mL of glutaraldehyde (25% aqueous solution) was diluted to
50 mL with distilled water. pH value was adjusted to 10.5 with 1 M
NaOH solution. The solution was then mixed at room temperature
with a magnetic stirrer for 30 min [38].
2.3. Synthesis of styrene–divinylbenzene copolymer containing
polyglutaraldehyde
Response surface methodology (RSM) has been a widely prac-
ticed approach for the production and optimization of different
industrially important biotechnological and biochemical products
such as enzymes and chemicals [33,34]. Optimization with RSM
not only allows quick screening of wide experimental field, but
also shows the role of each of the components. The graphical
representation of RSM function is called response surface, which
is used to describe the individual and cumulative effects of the
parameters on the response. Central composite design has been the
Polymerized glutaraldehyde solution with potassium persulfate
(1.4%, w/v) as starter was added to the organic mixture containing
59% (v/v) of styrene, 26% (v/v) of divinylbenzene, 15% (v/v) of span
80. The ratios of the aqueous and organic phases were 90% and 10%,
respectively. Polymerization was conducted at 80 ◦C for 3 h and
waited at 60 ◦C until dry. Polymerization procedure resulted in an
activated polyglutaraldehyde styrene–divinylbenzene copolymer.
The porosity of the resulting polymer was 90% since the aqueous
phase was removed from the polymer [38].

172
Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
2.4. Lipase immobilization
2.9. Characterization of the support
Lipase from T. lanuginosus was immobilized on the
STY–DVB–PGA (0.6–1.4 mm particle size) by covalent linkage
method in a column with continuous circulation by a peristaltic
pump. Synthesis of STY–DVB–PGA and immobilization are illus-
trated in Fig. 1. Different concentrations of lipase solutions were
prepared using different concentrations and pHs of phosphate
buffer solutions. Before immobilization, polymeric beads were
washed with 200 mL of distilled water. The enzyme solution was
circulated at 25 ◦C throughout the reactor with 5 mL/min flow
rate for up to 40 h within the experimental range (Table 2). The
reactor has 2.8 cm outer diameter, 2.3 cm inner diameter and
10 cm length. The immobilized enzymes onto polymeric beads
were finally washed with 200 mL buffer solution which is the same
phosphate buffer used during immobilization.
FTIR spectra of the STY–DVB–PGA and lipase immobilized poly-
meric beads were obtained using FTIR spectrophotometer (Thermo
Nicolet 6700). The dry support was thoroughly mixed with KBr, and
pressed into a tablet form, and the spectrum was then recorded in
the 4000–400 cm1 range with 32 scans at a resolution of 4 cm−1
.
The STY–DVB–PGA was imaged in dry state by SEM (Carl Zeiss
EVO 40). For this purpose, the support was coated with 80% gold and
20% palladium on argon atmosphere for enhancing conductivity.
The coated support was then imaged.
2.10. Experimental design
Five-level-four-factor central composite design was employed
in this study, requiring 30 experiments for the optimization of
immobilization parameters. The parameters and their levels are
enzyme concentration (4–16%, v/v), pH (6.0–8.0), buffer con-
centration (20–100 mM) and immobilization time (8–40 h). The
experimental and predicted data in terms of specific activity are
shown in Table 1.
2.5. Protein assay
Protein concentration was determined using the Bradford assay
method [39]. The amount of immobilized protein on the support
material was determined by measuring the initial and final concen-
trations of protein within the enzyme solutions. Coomassie brilliant
blue solution was used as a dye reagent. Bovine serum albumin
(BSA) was used as a standard to construct the calibration curve.
Second-order polynomial equation (1) which includes all inter-
action terms was used to calculate the predicted response:
4
4
3
4
ꢀ
ꢀ
ꢀꢀ
y = b0
+
bixi +
biixi2
+
bijxij
(1)
i=1
i=1
i=j j=i+1
2.6. Measurement of lipase activity
where y is the specific activity, b0 is the offset term, bi is the linear
effect, bii is the squared effect, bij is the interaction effect and xi is
the ith independent variable.
Spectrophotometric measurement was based on the capability
of lipases to cleave p-nitrophenyl palmitate (p-NPP), thus releasing
an amount of p-nitrophenol (p-NP) which has a yellow color that
can be measured by absorbance at 410 nm. The enzymatic reac-
tion mixture contained 1.5 mL of 20 mM of p-nitrophenyl palmitate
solution in 2-propanol and 13.5 mL of 60 mM phosphate pH 7.0
containing 0.2% (w/v) gum arabic and 0.6% (v/v) Triton X-100 as
detergent. The reaction was initiated by the addition of 70 mg lipase
immobilized support. The mixture was stirred with a magnetic
stirrer for 5 min. The support was separated from the reaction mix-
ture by filtering with filter paper. 0.4 mL of filtrate was diluted to
12-folds with 10% 2-propanol and 90% phosphate buffer solution
(60 mM, pH 7.0). The absorbance of p-nitrophenol was measured
at 410 nm. One lipase unit was expressed as the release of 1 ␮mol
p-nitrophenol per minute under the assay conditions. Specific
activity was defined as the number of enzyme units per milligram
protein.
2.7. Stability of immobilized lipase in organic solvents
Immobilized lipase (150 mg) was incubated in 15 mL of
methanol and tert-butanol each at 25 ◦C for 60 min. The residual
activity was assayed for different times following the same proce-
dure as described above.
2.8. Operational stability of immobilized lipase
Immobilized lipase (3.30 g) was used for transesterification of
crude canola oil with methanol. 100 g of crude canola oil and three-
step addition of methanol, 27 mL with 9 mL in each step (a 1:4
molar ratio of oil/methanol), were circulated into the reactor by
a peristaltic pump at 5 mL/min flow rate for 3 h at each successive
step, then the reaction was continued at 40 ◦C for 24 h. The reac-
tions were conducted with 10 batches. One batch reaction time
was 24 h. The immobilized lipases were rinsed with tert-butanol
between each batch. The residual activity determined after 24 h
was expressed as relative activity.
Fig. 1. Scheme of lipase immobilization on the STY–DVB–PGA beads.

Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
173
Table 1
Central composite design of factors with coded values and specific activity.
Treatment
Factors
Specific activity (␮mol
p-NP/mg enzyme min)
Enzyme concentration (%, v/v), x1
pH, x2
Buffer concentration (mM), x3
Immobilization time (h), x4
Experimental
Predicted
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
−1
1
−1
−1
1
−1
−1
−1
−1
1
1
1
1
−1
−1
−1
−1
1
1
1
1
0
−1
−1
−1
−1
−1
−1
−1
−1
1
1
1
1
1
1
1
1
0
0
0
0
0
3.45
2.59
4.58
4.34
5.00
4.55
5.23
5.54
3.30
3.39
4.62
4.32
4.40
3.82
6.14
4.66
7.93
5.88
3.12
5.20
2.33
3.83
3.29
3.41
4.59
4.21
4.66
4.64
4.18
4.07
3.73
3.10
4.82
4.19
4.71
4.08
5.80
5.17
3.70
3.06
4.79
4.15
4.68
4.04
5.77
5.13
7.53
6.27
3.30
5.48
2.10
4.06
3.38
3.32
4.39
4.39
4.39
4.39
4.39
4.39
−1
1
1
−1
1
−1
−1
1
−1
1
1
−1
1
−1
−1
1
−1
1
1
−1
1
−1
−1
1
1
0
−1
1
−2
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
−2
2
0
0
0
0
0
0
0
0
−2
2
0
0
0
0
0
0
0
0
0
−2
2
0
0
0
0
0
0
0
0
0
2.11. Data analysis
The data were analyzed using Design Expert program (version
7.1.4) and the coefficients were interpreted using F-test. Three main
analytical steps: analysis of variance (ANOVA), regression analy-
sis and plotting of contour plot were performed to establish the
optimum conditions for specific activity.
3. Results and discussion
3.1. Characterization of STY–DVB–PGA
The STY–DVB–PGA beads were characterized before and after
immobilization using lipase from T. lanuginosus by FTIR spec-
troscopy and SEM. FTIR spectra of the polymeric support before
and after immobilization of lipase are shown in Fig. 2. The
STY–DVB–PGA beads show OH absorption band of water at
3436 cm−1 formed during polymerization step. The spectrum
shows aromatic C–H stretching at 3059 and 3025 cm−1, aliphatic
C–H stretching at 2924 and 2853 cm−1, aromatic C C stretching
at 1603, 1492 and 1451 cm−1. Aromatic C–H in-plane bending at
1064, 1029 and 902 cm−1, aromatic out-plane bending at 829, 758
and 699 cm−1 for p-substitute benzene. The peaks at 1740, 1725
and 1707 cm−1 show C O stretching corresponding to aldehyde
group that indicates the presence of polyglutaraldehyde in the
copolymer (Fig. 2a). The lipase immobilized on the STY–DVB–PGA
beads shows an absorption peak at 1678 cm−1 due to the for-
mation of imine bond (–C N) during covalent immobilization
(Fig. 2b). The peaks at 1640 and 1630 cm−1 are for amide I and
amide II, respectively, indicating the presence of enzyme in the
polymeric beads [40]. The support was washed after immobiliza-
tion process with buffer solution to remove unbound enzyme.
Hence, amide I and amide II peaks originate covalent bound
enzyme.
Fig. 2. FTIR spectra of STY–DVB–PGA beads (a) before and (b) after immobilization
of lipase enzyme.

174
Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
Fig. 3. SEM picture of STY–DVB–PGA beads (a) before and (b) after immobilization of lipase enzyme.
Scanning electron microscopy (SEM) allowed the verification
cant terms was given below:
of morphological differences of STY–DVB–PGA beads in the non-
immobilized and immobilized states of the enzyme. A typical
polyHIPE (high internal phase emulsion) structure, which is the
open cellular architecture consisting of voids and connecting pores
through the walls of the polymerized continuous phase of the emul-
sion, can be seen on the SEM pictures, which was used to examine
the surface morphology of STY–DVB–PGA beads before and after
immobilization of lipase from T. lanuginosus (Fig. 3). The enzyme
immobilization by means of polyglutaraldehyde leads to notable
surface morphology changes as illustrated in Fig. 3b. The changes
of the surface morphology of copolymer indicate the presence of
the enzyme attached on the STY–DVB–PGA beads.
y = 4.39 − 0.32x1 + 0.55x2 + 0.49x3 − 0.02x4 + 0.63x12
− 0.33x32 − 0.26x42
(2)
where x1 is the enzyme concentration, x2 is the pH, x3 is the buffer
concentration, and x4 is the immobilization time.
The quadratic polynomial model was highly significant and suf-
ficient to represent the actual relationship between the response
and significant parameters with very low p-value (0.0001) from the
ANOVA (Table 3). The computed model F-value of 16.52 was higher
than the tabular value of F0.05(14,15) = 2.42, implying the model is sig-
nificant at 95% confidence level. The model also showed statistically
insignificant lack of fit, as is evident from the computed F-value
of 2.67 which is lower than the tabular value of F0.05(10,5) = 4.74 at
95% confidence level. Furthermore, the value of pure error (0.35)
is low which indicates good reproducibility of the data obtained
with a small p-value (0.0001) from the ANOVA and a satisfac-
tory coefficient of determination (R2 = 0.9391). The coefficient of
determination also revealed that there are excellent correlations
between the independent variables.
The effects of immobilization parameters such as enzyme con-
centration, pH, buffer concentration, and immobilization time were
investigated on the specific activity. The p-values mark the sig-
nificance of coefficients and are also important for understanding
the pattern of the mutual interactions between the parameters. A
value of Prob > F less than 0.05 indicates that the model terms are
significant. x1 (enzyme concentration), x2 (pH), x3 (buffer concen-
tration), x12, x32, and x42 are the most significant parameters (Prob > F
less than 0.05). However, x4 (immobilization time), x1x2, x1x3, x1x4,
x2x3, x2x4, x3x4 and x22 have less effect (Prob > F more than 0.05)
on the specific activity of the immobilized lipase (Table 3). The
relationship between predicted and experimental specific activities
is shown in Fig. 4. It can be seen that there is a high correla-
tion (R2 = 0.9226) between the predicted and experimental specific
activities.
3.2. Optimization of immobilization parameters
The selection of suitable immobilization conditions is critical
to maximize the multipoint covalent attachment. Immobilization
conditions should favor the enzyme–support reaction [26]. The
main objective of this work is the development and evaluation
of a statistical approach to better understand the relationship
between the parameters of the lipase immobilization. Optimiza-
tion of immobilization parameters was performed using central
composite design procedure. The coded values, experimental spe-
cific activity and predicted specific activity are given in Table 1
and the experimental domain is shown in Table 2. Among the
30 experiments including 5 replicates, experiment 17 (enzyme
concentration 4%, pH 7, buffer concentration 60 mM and immo-
bilization time 24 h) had the greatest specific activity (7.93 ␮mol
p-NP/mg enzyme min) and experiment 21 (enzyme concentra-
tion 10%, pH 7, buffer concentration 20 mM and immobilization
time 24 h) had the smallest specific activity (2.33 ␮mol p-NP/mg
enzyme min). The effects of each factor and its interactions were
calculated using a Design Expert program (version 7.1.4). Fitting
of the data with various models and the subsequent analysis of
variance (ANOVA) showed that immobilization of lipase was most
suitably described with quadratic polynomial model. From the
Design Expert, the quadratic polynomial equation (2) with signifi-
The relationship between immobilization parameters and spe-
cific activity was investigated by contour plots. Fig. 5 shows the
Table 2
Range of coded and actual values for central composite design.
Coded value
Enzyme concentration (%, v/v)
pH
Buffer concentration (mM)
Immobilization time (h)
−2
4
7
10
13
16
6.0
6.5
7.0
7.5
8.0
20
40
60
80
100
8
16
24
32
40
−1
0
1
2

Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
175
Table 3
Analysis of variance for the fitted quadratic polynomial model for optimization of immobilization parameters.
Source
Sum of squares
Degree of freedom
Mean square
F-value
p-value (Prob > F)
Model
Enzyme (x1)
pH (x2)
Buffer concentration (x3)
Immobilization time (x4)
x1x2
x1x3
x1x4
34.14
2.40
7.14
5.77
0.01
0.00
0.05
0.07
0.11
0.03
0.24
10.79
0.10
2.96
1.87
2.21
1.86
0.35
36.35
14
1
1
1
1
1
1
1
1
1
1
1
1
1
1
15
10
5
2.44
2.40
7.14
5.77
0.01
0.00
0.05
0.07
0.11
0.03
0.24
10.79
0.10
2.96
1.87
0.15
0.19
0.07
16.52
16.25
48.35
39.06
0.04
0.00
0.33
0.46
0.74
0.23
1.62
73.09
0.65
20.03
12.68
<0.0001a
0.0011a
<0.0001a
<0.0001a
0.8354b
0.9490b
0.5745b
0.5097b
0.4039b
0.6389b
0.2220b
<0.0001a
0.4324b
0.0004a
0.0028a
x2x3
x2x4
x3x4
x12
x22
x32
x42
Residual
Lack of fit
Pure error
Cor total
2.67
0.1453b
29
R
2 = 0.9391
a
Significant at “Prob > F” less than 0.05.
Insignificant at “Prob > F” more than 0.05.
b
yields high specific activity (6.58 ␮mol p-NP/mg enzyme min) at
pH 7.5. The effect of buffer concentration, enzyme concentration
and their mutual interaction on the specific activity of immobilized
lipase is illustrated in Fig. 6. Apparently, low enzyme concentra-
tion (5%, v/v) increased the specific activity (6.85 ␮mol p-NP/mg
enzyme min) at a higher buffer concentration (75 mM). Fig. 7 rep-
resents the effects of varying buffer concentrations and pHs on
specific activity. An increase in specific activity was observed
with the increasing of buffer concentration at first, then the trend
was stable when the buffer concentration reached to 75 mM at
pH 7.0. There is no significant interaction between the immo-
bilization parameters. However, higher interaction was observed
between enzyme concentration and pH than between other fac-
tors. Although immobilization may be performed at neutral pH in
many cases, incubation at alkaline pH values, where the reactivity
of the nucleophiles of the protein may be improved, is convenient
to reach a high enzyme–support reaction.
Fig. 4. Predicted specific activity versus experimental specific activity.
The optimum specific activity of lipase enzyme from the model
was determined as 4% (v/v) for enzyme concentration, pH 8.0
for buffer, 75 mM for buffer concentration and 24 h for immobi-
lization time within the experimental region. The pH and buffer
effect of enzyme concentration, pH and their mutual interaction on
specific activity of immobilized lipase. Enzyme concentration of 4%
(v/v) and pH of 7.5 led to the maximum specific activity (7.47 ␮mol
p-NP/mg enzyme min). An enzyme concentration of 16% (v/v) also
Fig. 5. Contour plot of the combined effects of enzyme concentration and pH on the
Fig. 6. Contour plot of the combined effects of enzyme concentration and buffer
specific activity.
concentration on the specific activity.

176
Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
Fig. 8. Stability of immobilized Thermomyces lanuginosus lipase as a function of
incubation time in methanol and tert-butanol at 25 ◦C.
Fig. 7. Contour plot of the combined effects of pH and buffer concentration on the
specific activity.
(45 ␮mol p-NP/g support min) that is immobilized on porous sil-
ica granulates. Different immobilization preparations of the same
lipase, when acting at different experiment conditions, may exhibit
a very different activity. T. lanuginosus lipase was immobilized on
different hydrophobic supports and their properties were com-
pared with those found in literature [32,42]. The results suggested
that selection of different supports yielded very different results
in terms of activity and stability. On the other hand, the type of
carbonyl group together with the immobilization method has an
important effect on the catalytic properties of T. lanuginosus lipase.
Microbial lipases have been found to exhibit a tendency to
form bimolecular aggregates in the solution at low concentra-
tions [43,44]. Detergents have been suggested to be able to shift
the open/close equilibrium of lipase toward the open conforma-
tion, by coating the hydrophobic areas of the lipase that surround
the active center of the enzyme. In this paper, we focused on the
evaluation of a statistical approach to better understand the rela-
tionship between the immobilization parameters in order to obtain
the highest lipase activity. The mechanism of aggregation of lipase,
stabilization of the open form of lipase by covalent linking with
STY–DVB–PGA and the effect of detergent need to be studied in
detail with this particular support and lipase.
concentration have an effect on the ionic state of the lipase
molecules. At pH 8.0 and high buffer concentration, the enzymatic
polarity might be weakened, which could enhance the lipase bind-
ing onto the hydrophobic surface. A similar study was shown by
Huang and Cheng [41] who reported the effect of pH on the lipase
from Penicillium expansum immobilization process. The activity of
immobilized enzyme on a bimodal ceramic foam was changed with
the increase of the immobilization pH value (6–10), and lipase
immobilized at pH 8.0 showed maximum activity. Lower enzymatic
activity was obtained at both lower and higher pH values. Similar
optimum pH was obtained with our study although different types
of lipase enzyme and support were used.
The accuracy of the model was validated under the optimal
conditions obtained from central composite design. Theoretical
specific activity was calculated as 8.78 ␮mol p-NP/mg enzyme min
from the model according to the limit criterion of specific activity
maximization. The immobilization was repeated at the opti-
mum conditions and experimental specific activity was found
as 8.41 ␮mol p-NP/mg enzyme min. The specific activity of the
free lipase enzyme was determined as 10.15 ␮mol p-NP/mg
enzyme min. Verification experiments confirmed the validity of the
predicted model. As a result, the model from central composite
design was considered to be accurate and reliable for predicting the
specific activity for immobilization of T. lanuginosus lipase by cova-
lent binding on the STY–DVB–PGA. The lipase loading capacity and
immobilization yield were determined as 5.71 mg/g support and
59%, respectively. The recovered activity was obtained as 83% after
immobilization on the polymeric beads at optimum conditions.
Similar results were obtained with literature [38]. The authors
compared the loading capacity, immobilization yield and recov-
ered activity of the immobilized lipase on the STY–DVB–PGA and
STY–DVB. In addition, under different experimental conditions,
the amount of protein immobilized onto STY–DVB–PGA beads
(11.81 mg/g support) was higher than that of protein immobi-
lized onto STY–DVB (10.79 mg/g support). The results indicate that
STY–DVB–PGA beads are more efficient than STY–DVB in terms of
these parameters and covalent attachment has some real advan-
tages on lipase immobilization.
3.3. Enzyme stability in organic solvents
The objective of studying the effect of organic solvents on
the specific activity of lipase immobilized on the STY–DVB–PGA
was its potential application in biodiesel production. Studies have
shown that the stability of commercial immobilized lipase from
T. lanuginosus (lipozyme TL IM) in biodiesel production could be
significantly enhanced in the tert-butanol system [30,31]. In the
tert-butanol system, both methanol and by-product glycerol are
soluble, so the negative effect caused by methanol and glycerol
could be eliminated totally. The effect of methanol and tert-butanol
on the specific activity of lipase immobilized on the STY–DVB–PGA
is shown in Fig. 8. The immobilized enzyme lost 80% of its activity
in 15 min in the presence of methanol. The reason is that the sup-
ports might trap and prevent the disruption of the enzyme-bound
water essential to maintain the three-dimensional structure of the
enzyme for catalysis as the polar solvents tend to strip water from
the enzyme molecule. Therefore, the enzyme may not be stable
in the presence of excessive methanol in the enzymatic reaction
medium. On the other hand, the specific activity of lipase was
almost stable in tert-butanol (Fig. 8). It can be seen that the immo-
bilized enzyme was stable with 90% of its relative specific activity
after 15 min incubation in the presence of tert-butanol. Therefore,
the incubation of the immobilized lipase in tert-butanol helped
The activities of both the immobilized lipase on the
STY–DVB–PGA and the commercial immobilized lipase (lipozyme
TL IM) were measured with the same procedure as described
above. The lipase slightly increased its activity (48 ␮mol p-NP/g
support min) when it was immobilized on the STY–DVB–PGA while
the commercial immobilized lipase displayed slightly less activity

Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
177
binding. RSM proved to be a powerful tool for the optimization of
immobilization parameters. A second-order model was obtained
to describe the relationship between the specific activity and the
parameters of enzyme concentration, pH, buffer concentration and
immobilization time. The results indicated that enzyme concen-
tration, pH and buffer concentration were the significant factors on
the specific activity of immobilized lipase. The optimum conditions
of immobilized lipase were enzyme concentration 4% (v/v), pH 8.0,
buffer concentration 75 mM and immobilization time 24 h. The pre-
dicted specific activity was 8.78 ␮mol p-NP/mg enzyme min under
the optimal conditions and the subsequent verification experiment
with the specific activity of 8.41 ␮mol p-NP/mg enzyme min con-
firmed the validity of the predicted model. Operational stability was
determined with immobilized lipase and it indicated that a small
enzyme deactivation occurred after being used repeatedly for 10
consecutive batches with each of 24 h. Lipase stability as a function
of incubation time in organic solvent indicated that the enzyme lost
most of its activity in a very short time in methanol, whereas lipase
could maintain relatively high activity in tert-butanol and there
was no obvious loss in lipase activity even after being incubated
for 2 h.
Fig. 9. Stability of Thermomyces lanuginosus lipase immobilized by covalent attach-
ment on the STY–DVB–PGA and physical adsorption on the STY–DVB beads with
repeated use for transesterification of canola oil. The reaction conditions: 20%
enzyme based on oil weight (100 g); oil/alcohol molar ratio 1:4; reaction tempera-
ture 40 ◦C and reaction time 24 h.
improve the enzyme activity and stability. Immersion of lipases
in tert-butanol was claimed as a pretreatment method to increase
lipase activity in the synthesis of methyl ester [31]. The activity
of the lipase immersed in the tert-butanol was found to be higher
than that of the enzyme immersed in methyl ester [34]. When the
enzyme was immersed in tert-butanol, the yield of methyl ester
was about 7–10 times higher than that of the immobilized enzyme
without pretreatment. The reason is that the methanol adsorbed on
the immobilized enzyme can be dissolved in tert-butanol. Hence,
it is possible to propose that washing with tert-butanol is an effi-
cient way to regenerate the deactivated immobilized lipase and
remove the contaminant of glycerol in lipase-catalyzed biodiesel
production.
Acknowledgement
The authors thank Uludag University Research Foundation for
providing financial support for this project (Project No. 2004/43).
References
[1] P. Villeneuve, N. Barouh, B. Barea, G. Piombo, M.C. Figueroa-Espinoza, F. Turon,
M. Pina, R. Lago, Food Chem. 100 (2007) 1443–1452.
[2] M.H. Sörensen, J.B.S. Ng, L. Bergström, P.C.A. Alberius, J. Colloid Interface Sci.
343 (2010) 359–365.
[3] F. Hasan, A.A. Shah, A. Hameed, Enzyme Microb. Technol. 39 (2006)
235–251.
[4] L. Fjerbaek, K.V. Christensen, B. Norddahl, Biotechnol. Bioeng. 102 (2009)
1298–1315.
[5] R.F. Lafuente, J. Mol. Catal. B: Enzym. 62 (2010) 197–212.
[6] M.S. Antczak, A. Kubiak, T. Antczak, S. Bielecki, Renew. Energy 34 (2009)
1185–1194.
3.4. Stability of immobilized lipase on reuse
[7] J.M. Palomo, Curr. Org. Synth. 6 (2009) 1–14.
The stability of the immobilized enzyme is of great impor-
tance for industrial applications [15,26,45]. The T. lanuginosus lipase
immobilized on the STY–DVB–PGA was stable for 10 cycles of
canola oil transesterification retaining more than 88% residual
activity (Fig. 9), thereby indicating its robust biocatalytic potential.
The enzymatic activity of lipase immobilized onto STY–DVB–PGA
was compared with celite 545, silica gel, and STY–DVB supports.
The immobilized lipase lost its whole activity after five repeated
reuses with these supports. Decrease in the enzyme activity of the
lipase immobilized on the STY–DVB could be due to leaching of
enzyme by washing the solvent with tert-butanol and transesteri-
fication reaction mixture from support pores or due to inactivation
of enzyme. The high reusability of immobilized lipase has been
reported for cross-linked enzyme aggregates (CLEAs) [15] and
cross-linked lipase on silica gel [46]. On the other hand, commercial
lipase (lipozyme TL IM) displayed 86.2% activity after being used for
10 batches when waste oil was used for transesterification reaction
[37]. The lipase immobilized on the electrospun fibrous membranes
retained 30% residual activity after reuses [18]. Our results suggest
that the polyglutaraldehyde enzyme–support cross-link is able to
produce an intense enzyme–support reaction that leads to a great
increase in the enzyme stability.
[8] U. Hanefeld, L. Gardossi, E. Magner, Chem. Soc. Rev. 38 (2009) 453–468.
[9] R.A. Sheldon, Adv. Synth. Catal. 349 (2007) 1289–1307.
[10] K. Won, S. Kim, K.J. Kim, H.W. Park, S.J. Moon, Process Biochem. 40 (2005)
2149–2154.
[11] S.F. Chang, S.W. Chang, Y.H. Yen, C.J. Shieh, Appl. Clay Sci. 37 (2007)
67–73.
[12] S. Pahujani, S.S. Kanwar, G. Chauhan, R. Gupta, Bioresour. Technol. 99 (2008)
2566–2570.
[13] D.S. Rodrigues, A.A. Mendes, W.S. Adriano, L.R.B. Goncalves, R.L.C. Giordano, J.
Mol. Catal. B: Enzym. 51 (2008) 100–109.
[14] F. Yagiz, D. Kazan, A.N. Akin, Chem. Eng. J. 134 (2007) 262–267.
[15] P. Gupta, K. Dutt, S. Misra, S. Raghuwanshi, R.K. Saxena, Bioresour. Technol. 100
(2009) 4074–4076.
[16] P.C. Oliveira, G.M. Alves, H.F. Castro, Biochem. Eng. J. 5 (2000) 63–71.
[17] S.H. Chiou, W.T. Wu, Biomaterials 25 (2007) 197–204.
[18] X.J. Huang, A.G. Yu, Z.K. Xu, Bioresour. Technol. 99 (2008) 5459–5465.
[19] N. Öztürk, S. Akgöl, M. Arısoy, A. Denizli, Sep. Purif. Technol. 58 (2007)
83–90.
[20] H. Noureddini, X. Gao, R.S. Philkana, Bioresour. Technol. 96 (2005) 769–777.
[21] N. Hilal, R. Nigmatullin, A. Alpatova, J. Membr. Sci. 238 (2004) 131–141.
[22] B.K. Vaidya, G.C. Ingavle, S. Ponrathnam, B.D. Kulkarni, S.N. Nene, Bioresour.
Technol. 99 (2008) 3623–3629.
[23] L. Betancor, F.L. Gallego, A. Hidalgo, N.A. Morales, G.D. Ortiz, J.M. Guisan, R.F.
Lafuentene, J. Biotechnol. 121 (2006) 284–289.
[24] F.L. Gallego, L. Betancor, A. Hidalgo, N. Alonso, G.F. Lorente, J.M. Guisan, R.F.
Lafuentene, Enzyme Microb. Technol. 37 (2005) 750–756.
[25] L. Betancor, F.L. Gallego, A. Hidalgo, N.A. Morales, G.D.O.C. Mateo, R.F. Lafuente,
J.M. Guisan, Enzyme Microb. Technol. 39 (2006) 877–882.
[26] C. Mateo, J.M. Palomo, G.F. Lorente, J.M. Guisan, R.F. Lafuente, Enzyme Microb.
Technol. 40 (2007) 1451–1463.
[27] L. Cao, Carrier-Bound Immobilized Enzyme, Wiley-VCH, Weinheim, 2005.
[28] L. Wang, W. Du, D. Liu, L. Li, N. Dai, J. Mol. Catal. B: Enzym. 43 (2006)
29–32.
[29] Y. Shimada, H. Watanabe, A. Sugihara, Y. Tominaga, J. Mol. Catal. B: Enzym. 17
(2002) 133–142.
4. Conclusion
Lipase from T. lanuginosus was successfully immobilized on
polyglutaraldehyde activated styrene–divinylbenzene by covalent

178
Ö. Aybastıer, C. Demir / Journal of Molecular Catalysis B: Enzymatic 63 (2010) 170–178
[30] L. Li, W. Du, D. Liu, L. Wang, Z. Li, J. Mol. Catal. B: Enzym. 43 (2006) 58–62.
[31] D. Royon, M. Daz, G. Ellenrieder, S. Locatelli, Bioresour. Technol. 98 (2007)
648–653.
[32] Z. Cabrera, J.M. Palomo, G.F. Lorente, R.F. Lafuente, J.M. Guisan, Enzyme Microb.
Technol. 40 (2007) 1280–1285.
[33] O. Prakash, M. Talat, S.H. Hasan, R.K. Pandey, Bioresour. Technol. 99 (2008)
7565–7572.
[34] X. Yuan, J. Liu, G. Zeng, J. Shi, J. Tong, G. Huang, Renew. Energy 33 (2008)
1678–1684.
Önder Aybastıer is a research assistant at the Faculty
of Science and Arts, Uludag University, Bursa, Turkey.
He received his BS in Chemistry from Uludag University
(Bursa). Since 2006, he has been doing his MSc at the
Chemistry Department of Uludag University where he has
been focusing on the immobilization of lipase enzymes on
various supports and production of biodiesel by enzymatic
transesterification using vegetable cooking oil.
[35] G. Singh, N. Ahuja, M. Batish, N. Capalash, P. Sharma, Bioresour. Technol. 99
(2008) 7472–7479.
[36] W. Huang, Z. Li, H. Niu, D. Li, J. Zhang, J. Food Eng. 89 (2008) 298–302.
[37] Y. Wang, H. Wu, M.H. Zong, Bioresour. Technol. 99 (2008) 7232–7237.
[38] N. Dizge, B. Keskinler, A. Tanriseven, Biochem. Eng. J. 44 (2009) 220–225.
[39] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254.
[40] P.S. Kalsi, Spectroscopy of Organic Compounds, 2nd ed., New Age International
Limited, India, 1995.
[41] L. Huang, Z.M. Cheng, Chem. Eng. J. 144 (2008) 103–109.
[42] G.F. Lorente, Z. Cabrera, C. Godoy, R.F. Lafuente, J.M. Palomo, J.M. Guisan, Process
Biochem. 43 (2008) 1061–1067.
Cevdet Demir is a Professor of Analytical Chemistry at
the Faculty of Science and Arts, Uludag University, Bursa,
Turkey. He received his BS. in Chemistry from Hacettepe
University (Ankara). In 1997, he received his PhD degree
in Chemistry from Bristol University (UK) in the field
of chromatography and chemometrics. His recent scien-
tific interest is in chemometric method development for
enzyme immobilization on various supports and synthe-
sis of biodiesel from vegetable waste oil. He has also been
focusing on the determination of phenolic compounds in
foods and plants, and development of multivariate cal-
ibration methods for the determination of antioxidant
activity of phenolic compounds in plant.
[43] J.M. Palomo, M. Fuentes, G.F. Lorente, C. Mateo, J.M. Guisan, R.F. Lafuente,
Biomacromolecules 4 (2003) 1–6.
[44] G.F. Lorente, J.M. Palomo, C. Mateo, R. Munilla, C. Ortiz, Z. Cabrera, J.M. Guisan,
R.F. Lafuente, Biomacromolecules 7 (2006) 2610–2615.
[45] P.V. Iyer, L. Ananthanarayan, Process Biochem. 43 (2008) 1019–1032.
[46] D.H. Lee, C.H. Park, J.M. Yeo, S.W. Kim, J. Ind. Eng. Chem. 12 (2006)
777–782.