Tailoring the spacer arm for covalent immobilization of Candida antarctica lipase B - Thermal stabilization by bisepoxide-activated aminoalkyl resins in continuous-flow reactors

Article abstract of DOI:10.3390/molecules21060767

An efficient and easy-to-perform method was developed for immobilization of CaLB on mesoporous aminoalkyl polymer supports by bisepoxide activation. Polyacrylate resins (100-300 μm; ～50 nm pores) with different aminoalkyl functional groups (ethylamine: EA and hexylamine: HA) were modified with bisepoxides differing in the length, rigidity and hydrophobicity of the units linking the two epoxy functions. After immobilization, the different CaLB preparations were evaluated using the lipase-catalyzed kinetic resolution (KR) of racemic 1-phenylethanol (rac-1) in batch mode and in a continuous-flow reactor as well. Catalytic activity, enantiomer selectivity, recyclability, and the mechanical and long-term stability of CaLB immobilized on the various supports were tested. The most active CaLB preparation (on HA-resin activated with 1,6-hexanediol diglycidyl ether - HDGE) retained 90% of its initial activity after 13 consecutive reaction cycles or after 12 month of storage at 4°C. The specific rate (r_flow), enantiomer selectivity (E) and enantiomeric excess (ee) achievable with the best immobilized CaLB preparations were studied as a function of temperature in kinetic resolution of rac-1 performed in continuous-flow packed-bed bioreactors. The optimum temperature of the most active HA-HDGE CaLB in continuous-flow mode was 60°C. Although CaLB immobilized on the glycerol diglycidyl ether (GDGE)-activated EA-resin was less active and less selective, a much higher optimum temperature (80°C) was observed with this form in continuous-flow mode KR of rac-1.

Full text of DOI:10.3390/molecules21060767

molecules
Article
Tailoring the Spacer Arm for Covalent
Immobilization of Candida antarctica Lipase
B—Thermal Stabilization by Bisepoxide-Activated
Aminoalkyl Resins in Continuous-Flow Reactors
Emese Abaházi , Dávid Lestál , Zoltán Boros and László Poppe 1,2,*
1
1
2
1
Department of Organic Chemistry and Technology, Budapest University of Technology and Economics,
M u˝ egyetem rkp. 3, Budapest H-1111, Hungary; abahazi.emese@mail.bme.hu (E.A.);
ldavka@gmail.com (D.L.)
SynBiocat LLC; Löv o˝ ház u. 19/1, Budapest H-1043, Hungary; zoltan.boros@synbiocat.com
2
*
Correspondence: poppe@mail.bme.hu; Tel.: +36-1-463-3299
Academic Editor: Roberto Fernandez-Lafuente
Received: 10 May 2016; Accepted: 8 June 2016; Published: 13 June 2016
Abstract: An efficient and easy-to-perform method was developed for immobilization of CaLB on
mesoporous aminoalkyl polymer supports by bisepoxide activation. Polyacrylate resins (100–300 µm
;
~50 nm pores) with different aminoalkyl functional groups (ethylamine: EA and hexylamine:
HA) were modified with bisepoxides differing in the length, rigidity and hydrophobicity of the
units linking the two epoxy functions. After immobilization, the different CaLB preparations were
evaluated using the lipase-catalyzed kinetic resolution (KR) of racemic 1-phenylethanol (rac-1) in
batch mode and in a continuous-flow reactor as well. Catalytic activity, enantiomer selectivity,
recyclability, and the mechanical and long-term stability of CaLB immobilized on the various
supports were tested. The most active CaLB preparation (on HA-resin activated with 1,6-hexanediol
diglycidyl ether—HDGE) retained 90% of its initial activity after 13 consecutive reaction cycles or
˝
after 12 month of storage at 4 C. The specific rate (rflow), enantiomer selectivity (E) and enantiomeric
excess (ee) achievable with the best immobilized CaLB preparations were studied as a function of
temperature in kinetic resolution of rac-1 performed in continuous-flow packed-bed bioreactors.
˝
The optimum temperature of the most active HA-HDGE CaLB in continuous-flow mode was 60 C.
Although CaLB immobilized on the glycerol diglycidyl ether (GDGE)-activated EA-resin was less
˝
active and less selective, a much higher optimum temperature (80 C) was observed with this form in
continuous-flow mode KR of rac-1.
Keywords: lipase; immobilization; covalent attachment; kinetic resolution; thermal stabilization;
continuous-flow reactor
1
. Introduction
Lipases are ubiquitous enzymes which can be found in animals, plants, fungi and bacteria [
Naturally, lipases catalyze the hydrolysis of esters formed of glycerol and long-chain fatty acids [
However, under particular conditions, they can catalyze esterification and transesterification [
1
3
,
,
2
4
].
].
5
] with
high activity, specificity and selectivity. Therefore, lipases represent a significant class of industrially
relevant enzymes [6].
Lipases represent a peculiar mechanism of action, the so called interfacial activation [
a loop containing -helicoidal elements that covers the active site, a large proportion of lipases have
two different conformations (the closed and the open forms) [ ]. In an aqueous environment,
the equilibrium is shifted towards the water soluble closed form and the active site is secluded
7,8]. Due to
α
4,9
Molecules 2016, 21, 767; doi:10.3390/molecules21060767
www.mdpi.com/journal/molecules

Molecules 2016, 21, 767
2 of 18
from reaction medium [
4
,10
,11]. Catalysis takes place in the open-lid form which can be stabilized
by interfacial activation at the interface of the drops of the oily phase containing hydrophobic
substrates [12]. These structural changes related to the interfacial activation are crucial to provide the
substrates access to the active site. The performance of enzymes in organic solvents facilitated their
application as biocatalysts [13], even in supercritical medium [14].
The lipase B from Candida antarctica (CaLB) is one of the most frequently used lipase in
biotechnology due to its excellent biochemical properties such as resistance to organic solvents,
high stereo- and enantioselectivity [
resolved by Uppenberg [16 17]. CaLB—having a molecular weight of 33 kDa and isoelectric point of
.8—contains only a small lid covering the active site [18]. Therefore, it was thought that the typical
interfacial activation of lipases [ ] is not a characteristic feature of this enzyme [19]. However, recent
2,9,15]. The three-dimensional structure of CaLB has been
,
6
7,8
structural investigations revealed the open and closed states of CaLB and the mechanism of interfacial
activation [20].
Replacing soluble enzymes with immobilized preparations in synthetic processes can significantly
reduce the cost of enzyme production and complex downstream processing [21,22]. Immobilization of
enzymes is a powerful biochemical tool with a potential for improving enzyme activity, temperature
stability, and selectivity [23]. Retaining the active conformation during immobilization is very
important for enzyme activity [24]; therefore, finding an immobilization protocol that increases
enzyme activity and stability should be a target of real importance [25]. Several immobilization
techniques are available for protein immobilization: adsorption, covalent attachment onto solid
supports, cross-linking and entrapment of protein in matrices. Generally, a support for enzyme
immobilization must meet two requirements: it should contain an adequate amount of functional
groups on the surface and it should be mechanically stable for repeated use or in application
in continuous processes [21]. Adsorption of lipases on hydrophobic surfaces is easy to perform
because they resemble the surface of their natural substrates [
serious disadvantage [22 26]. After adsorption on the hydrophobic surface, post-immobilization with
glutaraldehyde (GA) is frequently used [ ]. GA modifies primary amino groups of proteins and it is
also an effective cross linker used to produce CLEAs (Cross-linked enzyme aggregate) [27]. GA can be
used for activating the primary amino groups of the support followed by covalent immobilization [ 25].
4], but enzyme leaching could be a
,
2
3,
Entrapment could be a proper immobilization method for enzymes which can be easily deactivated
as in this way the three dimensional structure of enzyme could be stabilized via rigidification by the
entrapping matrix [28].
In the past few decades, intensive studies were performed on development of supports for
lipase immobilization. The most straightforward strategy was to utilize hydrophobic supports
for lipase immobilization in their active conformation by simple adsorption advancing interfacial
activation. Hydrophobic organic materials such as octyl-sepharose or octadecyl-polyacrylate [4]
and polyethyleneimine-agarose [29] were successfully applied in adsorptive immobilization
where lipase stayed fully active even after a very long incubation in organic solvents. Polymer
supports with core-shell morphology could tune lipase properties during immobilization [12].
Similarly, surface-grafted mesoporous silica gel as inorganic material proved to beneficial for adsorptive
immobilization of lipases [2,30–32].
Although lipases after adsorptive immobilization can retain their activity in organic media but
under aqueous conditions they can release the enzyme causing decrease of activity and product
contamination. To circumvent the problem of enzyme leaching, covalent immobilization offered a
solution. For example, activated natural polymers (epichlorohydrin- or glutaraldehyde-activated
chitosan or agarose gels) proved to be suitable carriers for multipoint covalent immobilization of
proteins [
Epoxy or multifunctional epoxy-supports were suitable for covalent immobilization of
enzymes [32 35 36]. Epoxy supports—while being stable at neutral pH values for a long time in
wide temperature range—could react with different nucleophilic groups (such as amine, thiol or
3,33] or silica gel grafted with glyoxyl groups enabled covalent enzyme immobilization [34].
,
,

Molecules 2016, 21, 767
3 of 18
carboxylate) on the protein surface under mild conditions by forming stable covalent bonds which
prevents enzyme leakage [21 37 38]. The usefulness of epoxy-functionalized carriers [33 39] led to
,
,
–
the idea of using bisepoxides for enzyme immobilization. Thus, poly(ethylene glycol) diglycidyl
ether was applied for binding various oxidases onto biosensor microelectrode [40]. Later, glycerol
diglycidyl ether was used as a cross-linking agent for immobilization of lipases and phenylalanine
ammonia-lyase (PAL) as cross-linked enzyme aggregates (CLEA) [41] or binding enzymes such as
PAL [42] or CaLB [43] onto carbon nanotubes.
Besides enzyme immobilization, reactor technology can also affect the efficiency of
biotransformations [44]. Nowadays, continuous-flow systems are of ever-increasing importance
since they offer facile automation, reproducibility and safety [45,46]. In addition to the ease of studies
on substrate concentration, temperature and flow rate and how these factors effect enzyme-catalyzed
reactions in such systems, the catalytic efficiency of the immobilized enzymes proved to be higher in
packed-bed reactors operated in continuous-flow mode than in batch mode [47–49]. Noteworthy, the
mechanical damages reducing the operational stability of the immobilized biocatalysts in stirred or
shaken batch reactors is much less pronounced in packed bed reactors. The influence of various
immobilization conditions [48,50] and of ionic liquids [51] on biocatalytic properties of CaLB was
already evaluated in continuous-flow processes.
The aim of this study was to investigate the activation of mesoporous aminoalkyl polymer resins
using low-cost bisepoxides in covalent immobilization of lipase B from Candida antarctica (CaLB)
offering a less-toxic, inexpensive and easy-to-handle alternative over glutaraldehyde-activation of
such carriers. A major concern was to fine-tune the length, flexibility and hydrophobicity of the spacer
arm by varying aminoalkyl functions and bisepoxides, thereby resulting in significant variation in
the properties of the CaLB biocatalysts by altering the microenvironment surrounding the enzyme
molecule during and after immobilization. Variation of the spacer arm offers tuning and optimization
possibilities in biocatalyst design and development and can lead to improved properties useful for
extended use in recycled batch reactions or in continuous-flow systems even at elevated temperatures.
2
. Results and Discussion
Selection of the supports was a crucial point for this study on surface fine-tuning exploring
variations created by different bisepoxide spacer arms. Besides the functions on the surface,
pore size and particle size were important parameters influencing the final properties of resins in
various applications.
2
.1. Selection and Bisepoxide Activation of Carriers Used for Covalent Immobilization of CaLB
Because a final goal was to apply the immobilized CaLB derivatives in continuous-flow reactors,
particle size influencing pressure drop and pore diameter affecting specific area for immobilization and
mass transfer properties were important parameters. Thus, two commercially available macroporous
polymethyl methacrylate enzyme carriers functionalized with ethylamine (EA) and hexylamine (HA)
at their surface were selected as starting supports for this study. The properties of the resins and
approximate density of ethylamine and hexylamine functional groups on resin surfaces are presented
in Table 1. Worth mentioning is that the different lengths of ethylamine and hexylamine functional
groups on the two starting carriers can contribute to the different properties of the final spacer arms
between the surface of carriers and the target protein.
2
.2. Covalent Immobilization of CaLB on Bisepoxide-Activated Supports and Its Properties in Kinetic
Resolution of 1-Phenylethanol rac-1
Due to their quite high content of water (~70%), drying of the resins was required to determine
their real mass before surface modification (see Section 3.3.). In the first set of experiments, six different
types of bisepoxides with different lengths and hydrophobicities were selected as activating agents
for the two aminoalkyl carriers (Figure 1). As reference, glutaraldehyde activation of the two carriers

Molecules 2016, 21, 767
4 of 18
was also performed. After pre-activation, immobilization of CaLB was carried out in phosphate buffer
(pH 7.5) at room temperature (24 h).
Table 1. Properties of polymer resins with aminoalkyl functions applied for CaLB immobilization.
Ion Exchange
Resin Name
Abbreviation)
Pore Size
(nm)
Particle
Size (µm)
Water
Retention (%)
Polymer Resin
Capacity, Wet
(
´1
(
µmol¨ g
)
ReliZyme
™
EA
4
4
0–60
0–60
100–300
100–300
500
60–70
60–70
4
03 (EA)
ReliZyme™
HA 403 (HA)
600
Figure 1. Bisepoxide activation of supports for immobilization of CaLB.
Because the epoxide functions may form covalent bonds not only via the surface exposed amine
functions of Lys residues but also via the S- and O-atoms of Cys, Tyr, Asp, Glu, covalent bond formation
probability may be higher with the bisepoxide-activated supports than with the GA-activated ones,
especially in the case of proteins having only a few surface exposed lysine residues. These differences
may explain the higher activity of the immobilized CaLB achieved with several bisepoxide-activated
carriers compared to the GA-activated ones. Moreover, the C X bonds (X = NH, S, O) forming by
´
immobilization with the bisepoxide-activated carriers, unlike the C=N bonds forming in GA-activated
ones, are not susceptible to hydrolysis.
Worth mentioning is that GA-activated resins are not suitable for long-term storage and they
should be used for immobilization shortly after activation. On the other hand, the bisepoxide-activated
supports—similarly to the usual epoxy carriers—can be stored for a long time enabling real separation
of the carrier activation and enzyme immobilization steps in time.
To characterize the catalytic performance of the biocatalysts, kinetic resolution of 1-phenylethanol
(rac-1) was selected as the test reaction (Figure 2). This test enabled gaining information on the activity
and enantiomer selectivity of the CaLB variants. To distinguish between the fraction of CaLB retained
by only physical adsorption from the one retained by stronger covalent bonds, the activity of the
preparations was tested before and after washing with Triton X-100 non-ionic detergent solution which
could remove enzyme molecules attached only by hydrophobic interactions onto the surface. Table 2
shows the wash resistance representing the covalent ratio of binding onto activated EA- and HA-resins
and the biocatalytic performance of the covalently bound CaLB preparations.

Molecules 2016, 21, 767
5 of 18
Figure 2. Kinetic resolution of racemic 1-phenylethanol (rac-1) catalyzed by covalently immobilized
CaLB preparations in batch mode and in continuous-flow reactor (for details of experiments see
Sections 3.7 and 3.12).
Table 2. Kinetic resolution of 1-phenylethanol (rac-1) catalyzed by CaLB on bisepoxide-activated
´
1
polymer resins (molar amount of bisepoxide 5 mmol
¨
g
carrier) compared to the catalysis by CaLB on a
commercial epoxy resin, on unmodified and on GA activated EA-resins. Results with the various forms
of CaLB after washing with Triton X-100 are shown. Experiments were performed with preparations at
˝
3
0 C for 2 h, as described in Section 3.7.
Binding Function
Wash Resistance a (%) U (µmol¨min´1¨g´1
)
c (%)
ee(R)-2 (%)
E b
b
EP c
d
39.6
28.9
0.3
7.4
17.4
6.2
20.2
28.9
1.6
8.8
0.2
8.7
17.4
17.9
32.4
34.8
4.6
99.1
>200
e
f
f
EA
7
0.4
10.2
23.7
8.5
27.5
39.6
2.2
EA-GA
EA-GDGE
EA-BDGE
EA-HDGE
EA-CHDGE
EA-PEDGE
EA-NDGE
39
53
22
58
81
7
98.9
99.4
98.2
99.4
>200
»200
>200
>200
99.5
»200
f
f
30
11
51
64
51
82
92
17
49
12.0
0.2
98.6
>200
g
f
f
HA
HA-GA
HA-GDGE
HA-BDGE
HA-HDGE
HA-CHDGE
HA-PEDGE
HA-NDGE
a
11.9
23.8
24.5
44.2
47.6
6.3
98.9
99.6
99.2
99.5
>200
»200
»200
»200
99.4
»200
f
f
20.2
14.7
99.1
>200
Wash resistance (%) = activity of samples washed with Triton X-100/activity of samples after immobilization
b
c
ˆ
100, in kinetic resolution of rac-
epoxy-functionalized acrylic beads;
1
;
Enantiomeric ratio [52];
EP CaLB: CaLB immobilized onto
d
Activity of EP CaLB washed with Triton X-100/activity of EP
e
Ethylamine resin without activation; f ee and E values are not given for reactions with
CaLB ˆ 100 = 125%
;
g
c < 5% due to their high uncertainty; Hexylamine resin without activation.
For comparison, a commercial preparation (EP CaLB) was selected as a standard in which CaLB
was covalently attached to macroporous acrylic beads of 150–300
functions (EP). In the case of EP CaLB, washing the preparation with Triton X-100 solution resulted in
25% activation possibly due to a slight bioimprinting effect [53]. In case of EA CaLB, the 7% wash
µm particle size by the aid of epoxy
1
resistance was consistent with the fact that the non-activated resin did not contain groups for covalent
attachment and only ionic interactions could retain the enzyme molecules causing residual activity.
In case of HA CaLB, the alkyl chain of hexylamine function rendered the surface more hydrophobic
compared to EA-resin, thus a combination of the increased hydrophobicity and ionic interactions could
rationalize the slightly higher wash resistance (11%).
When polyethylene diglycidyl ether (PEDGE) was used as activating bisepoxide, the wash
resistance remained quite similar as observed for the unmodified resins (7% for EA-PEDGE CaLB and

Molecules 2016, 21, 767
6 of 18
1
7% for EA-PEDGE CaLB). This result indicating good adsorption ability but only occasional covalent
bond formation could be explained by assuming that during the activation both epoxy functions of
the bisepoxide with a long spacer arm of high flexibility could react with the surface amine functions.
Thus, only a few epoxy functions of PEDGE remained intact after the activation step explaining the
low frequency of covalent bond formation of the PEDGE-activated resin with the enzyme.
In contrast, high wash resistance was obtained with CaLB immobilized on 1,4-cyclohexaned
imethanol diglycidyl ether (CHDGE)-, 1,6-hexanediol diglycidyl ether (HDGE)- and glycerol diglycidyl
ether (GDGE)-activated resins (81%, 58% and 53%, respectively). This result can be attributed
to the hydrophobic character of CHDGE, since cyclohexyl rings increased the hydrophobicity
of support. Barbosa et al. proved that proteins become immobilized on epoxy supports via a
two-step mechanism: physical adsorption occurs in the first place then covalent linkages are formed
between the nucleophilic residues on the enzyme molecule and epoxy functions of the support [15].
Thus, enhanced hydrophobicity of the spacer arm with cyclohexyl ring could result in stronger
hydrophobic activation and consequently lead to CaLB forms of higher activity. A comparison
of the activation of the alkylamino resins with bisepoxides to the widely applied glutaraldehyde
activation indicated the value of the activation by a properly selected bisepoxide. The usefulness of a
bisepoxide-activation-based approach was indicated clearly by the approximately four-fold activity of
´
1
´1
g for EA-CHDGE CaLB and
the CaLB attached to CHDGE-activated supports (U = 39.6
µmol
¨min
¨
b
´
1
´1
U = 47.6 µmol¨min ¨g for HA-CHDGE CaLB) compared to the activity of the corresponding CaLB
b
´
1
´1
g
form immobilized by the aid of the widely used glutaraldehyde-activation (U = 10.2
µmol
¨min
¨
b
´
1
´1
g for HA-GA CaLB). The importance of the length of
for EA-GA CaLB and U = 11.9
µmol¨min
¨
b
spacer arms became apparent by analysis of the differences between bisepoxide-activation by various
bisepoxides. Within both series (EA-resin-based or HA-resin-based), the highest degree of covalent
immobilization (wash resistance) was achieved after CHDGE activation (92% with HA-resin and
81% with EA-resin). HDGE and GDGE activation resulted in a somewhat lower degree of covalent
binding (82% and 64% with HA-resin and 58% and 53% with EA-resin, respectively). However, only a
modest degree of covalent binding was observed after BDGE-activation (51% with HA-resin and 22%
with EA-resin).
Analysis of the activities of the various forms of CaLB covalently immobilized on pre-activated
resins indicated a role of the alkylamino function of the resins influencing the properties of immobilized
CaLB. Residual activity of CaLB attached to the modified HA-resins after activation with any bisepoxide
or glutaraldehyde was higher than that attached to the corresponding pre-activated EA-resin.
Because the trends within the two series were the same, optimization of the bisepoxide-activation
conditions for CaLB immobilization on either alkylamino resins was performed only with the three
best bisepoxides (CHDGE, HDGE and GDGE).
2
.3. Optimization of the Bisepoxide Activation for Covalent Immobilization of CaLB
The microenvironment of enzymes after immobilization can be tuned by proper optimization
´
1
of the process parameters. Our preliminary results applying bisepoxides in 5 mmol
¨g
of carrier
amounts showed (Table 2) that aminoalkyl (EA or HA)-resin-activation by three bisepoxides were
efficient for immobilization of CaLB. In this series of experiments, the amount of bisepoxides applied
for surface modification of the two aminoalkyl resins was varied in order to investigate the effect
of final epoxy group density on enzyme activity and immobilization yield (Figure 3). Reduction of
carrier caused dramatic decrease in specific activity of
EA-HDGE CaLB and EA-GDGE CaLB. Increase of the molar amount of bisepoxides from 5 mmol
´
1
the amount of bisepoxides to 2.5 mmol
¨g
´
g
1
¨
´
1
carrier to 10 mmol
A further increase of the amount of bisepoxides to 20 mmol
of activity of the immobilized CaLB compared to that of achievable by CaLB on the carriers treated by
¨g
carrier increased the specific activity of the immobilized CaLB preparations.
´
1
¨g
carrier resulted in negligible increase
´
1
1
0 mmol bisepoxide g carrier. Thus, for both aminoalkyl resins, the optimal amount of bisepoxide
´
1
for surface modification was 10 mmol¨g carrier.

Molecules 2016, 21, 767
7 of 18
(
a)
EA-GDGE
EA-HDGE
EA-CHDGE
(b)
HA-GDGE
50
HA-HDGE
HA-CHDGE
5
4
3
2
1
0
0
0
0
0
0
40
30
20
10
0
2
.5
5
10
20
2.5
5
10
20
-1
Molar amount of bisepoxide (mmol g )
-1
Molar amount of bisepoxide (mmol g )
Figure 3. Optimization of the bisepoxide activation of ethylamine (EA) and hexylamine (HA) resins for
´
1) of CHDGE,
CaLB immobilization. Effect of molar amount of bisepoxide to mass resin ratio (mmol
¨
g
HDGE and GDGE on CaLB activity. Experiments were performed as described in Section 3.7.
This result was confirmed for the other types of bisepoxides by protein concentration assays
as well. Immobilization yields for 12 variants of bisepoxide-activated supports were determined
by protein concentration measurements in the supernatant before and after immobilization of CaLB
´
g
1
according to Bradford’s method. The results for EA- and HA-resins modified with 2.5 and 10 mmol
bisepoxides (CHDGE, HDGE and GDGE) are shown in Figure 4.
¨
1
00
80
60
40
20
0
Figure 4. Effect of molar amount of bisepoxide to mass resin ratio on immobilization yields. Supports
modified with 2.5 mmol¨g bisepoxide (˝); supports modified with 10 mmol¨g´1 bisepoxide (ꢀ).
´1
Our results indicated that immobilization yields with the other two bisepoxides (CHDGE, HDGE)
were also significantly higher after bisepoxide-activation with the optimal amount of bisepoxides
´
1
´
1
(
10 mmol
¨g
carrier) as compared to the resins modified with only 2.5 mmol bisepoxide g carrier.
Supports modified with CHDGE showed highest immobilization yields (EA-CHDGE CaLB: 95% and
HA-CHDGE CaLB: 92%) which also confirmed the role of proper hydrophobicity of the linker which is
beneficial for the first lipase adsorption step during the covalent immobilization on epoxy-activated
supports [15].
2
.4. Operational Stability of the CaLB Preparations
One of the great advantages of immobilized enzyme preparations is the possibility of re-use in
repeated batch reactions, thus making the process more cost-effective [54]. Five CaLB preparations
EA-GDGE CaLB, EA-CHDGE CaLB, HA-GDGE CaLB, HA-CHDGE CaLB, HA-HDGE CaLB) were
selected for studying their stability during recycling. To test their operational stability, the CaLB
(
˝
biocatalysts were tested by repetitive cycles of KRs of rac-1 30 C for 1 h in a shaken test tube (1000 rpm,
in hexane/t-butyl methyl ether/vinyl acetate 6/3/1). Between each cycle, the CaLB biocatalysts were
washed with n-hexane. The relative residual activities of repeated batch experiments compared to the

Molecules 2016, 21, 767
8 of 18
first experiment as 100% are presented in Figure 5. Although every biocatalyst used in this study lost
its initial activity cycle-by-cycle, HA-HDGE CaLB retained 90% of its initial activity after 13 cycles.
This demonstrated that spacer arms with proper length and hydrophobicity between enzyme molecule
and support could contribute not only to the initial activity but also to the long-term stability of the
immobilized enzyme. The EA-GDGE CaLB in which the enzyme was attached by the shortest carbon
chain retained only 70% of its initial activity after 13 cycles. Furthermore, CaLB immobilized on EA- or
HA-resin activated with the cyclohexyl ring containing CHDGE preserved 82% of their initial activity
regardless of which resin was used.
1
00
8
6
4
2
0
0
0
0
0
EA-CHDGE
EA-GDGE
HA-CHDGE
HA-GDGE
HA-HDGE
1
3
5
7
9
11
13
Reaction cycle
Figure 5. Operational stability of CaLB immobilized on ethylamine (
activated by CHDGE ( ), GDGE ( ) or HDGE (
in first cycle 100 was determined in kinetic resolution of rac-1. Experiments were performed as
(
) or hexylamine (
) resin
N
ꢀ). Relative activity (%) = activity in given cycle/activity
ˆ
described in Section 3.9.
2
.5. Long-Term Storage Stability of the CaLB Preparations
The long-term storage stability of CaLB preparations were studied by storing the dry biocatalysts
˝
in screw capped vials at 4 C for 12 months. After 12 months, the residual activity of the stored CaLB
preparations was compared to the activity of the freshly prepared biocatalysts (results are shown in
Figure 6). After 12-month storage, HA-CHDGE CaLB and HA-HDGE CaLB preserved most of their
initial activity (95% and 90%, respectively). Presence of the cyclohexyl ring in the linker was beneficial
to preserve enzyme activity in case of CaLB attached to modified EA-resin (88% residual activity with
EA-CHDGE CaLB), while EA-GDGE CaLB preserved only 67% of its initial activity.
100
8
6
4
2
0
0
0
0
0
Figure 6. Long-term storage stability of the CaLB biocatalysts. Percentage of the residual activity of
the biocatalysts after 12-month storage were compared to the activity of freshly prepared biocatalysts.
Experiments were performed as described in Section 3.10.
2
.6. Continuous-Flow Kinetic Resolution of Racemic 1-Phenylethanol (rac-1) Catalyzed by CaLB Preparations
on Bisepoxide-Activated Resins—Substrate Concentration and Temperature Effects
Immobilization of enzymes and retention of biocatalysts played a major role in development
of continuous-flow processes [42,45–47,55,56]. It was previously shown that the mode of

Molecules 2016, 21, 767
9 of 18
immobilization [46] and operation temperature [45
,50] could significantly influence KR processes in
continuous-flow bioreactors. Therefore, six different CaLB preparations from this study were selected
to investigate the effect of substrate concentration and temperature on KR of rac- catalyzed by the
1
covalently immobilized enzyme. As a reference, this study was extended with the commercially
available EP CaLB which contained CaLB covalently immobilized onto epoxy-functionalized
acrylic resin.
During the substrate concentration dependency investigations, stainless-steel columns filled
˝
with the actual immobilized CaLB preparation were thermostated at 30 C and a solution containing
vinyl acetate and the substrate at various concentrations was pumped through the column at a
´ ´1
1
constant flow rate. The productivity of the various CaLB preparations (rflow
,
µmol¨g
¨
min ) was
investigated as a function of substrate concentration (Figure 7a). The 2.5-fold higher activity of the
most active HA-HDGE CaLB than that of the commercially available EP CaLB indicated the significance
of fine-tuning the bisepoxide activation for covalent immobilization. Similar to KRs of rac-
1 in batch,
CaLB on EA- and HA-resins activated with GDGE showed the lowest activity (EA-GDGE CaLB
´
1
´1
´1 ´1
min ) among tested derivatives. The behavior
2
9.2
of CaLB on resin-surfaces grafted with epoxy function via a linker containing cyclohexyl ring
EA-CHDGE CaLB and HA-CHDGE CaLB derivatives) demonstrated that this lipase interacting
µ
mol
¨
g
¨
min ; HA-GDGE CaLB 46.1
µ
mol
¨
g
¨
(
with hydrophobic groups became more active in flow systems as well. After an initial quasi-linear
range, the productivity–substrate concentration curve showed the typical saturation with all seven
CaLB preparations (Figure 7a). The rapidly increasing range of rflow–[(R)-
1
] curves ended at around
´
1
40 mg¨mL ; therefore, the temperature effect studies were performed at this substrate concentration.
1
1
20
00
1
1
1
80
60
40
(
a)
(b)
(d)
80
60
40
20
0
120
00
1
80
60
4
0
0
0
2
0
25
50
75
100
125
150
175
0
20
40
60
80
100
-
1
Concentration (mg mL )
Temperature(°C)
1
00.0
800
(
c)
700
600
500
9
9
9
9
9
9
9.5
9.0
8.5
8.0
7.5
7.0
400
300
200
100
0
0
20
40
60
80
100
0
20
40
60
80
100
Temperature(°C)
Temperature(°C)
Figure 7. Continuous-flow kinetic resolution of racemic 1-phenylethanol (rac-
preparations on bisepoxide-activated resins. Biocatalytic properties (rflow, ee, E) of immobilized CaLB
biocatalysts on ethylamine ( ) or hexylamine ( ) resin activated by CHDGE ( ), GDGE ( ) or
HDGE ( ). ( ) Effect of the substrate concentration on specific
), compared to EP CaLB ((  in
reaction rate (rflow); ( ) Effect of temperature on specific reaction rate (rflow); ( ) Effect of temperature
on enantiomeric excess (ee); (d) Effect of temperature on enantiomeric ratio (E).
1) catalyzed by CaLB
(
N
ꢀ
a
b
c

Molecules 2016, 21, 767
10 of 18
A major goal of this study was to gain information on the thermal stability of CaLB linked to
polymer resins by spacer arms with different length and hydrophobicity. Thus, KR of rac- was carried
out in bioreactors filled with the selected CaLB preparations similarly as in the substrate-concentration
1
˝
˝
tests but varying the temperature in the range of 0–110 C in 10 C steps (Figure 7b–d).
Productivity (rflow) of the six selected CaLB derivatives on bisepoxide-activated resins as a function
of temperature was compared to EP CaLB containing the lipase immobilized on epoxy-functionalized
acrylic resin (Figure 7b). The thermal behavior of CaLB attached to bisepoxide activated resins differed
significantly from EP CaLB (linking the enzyme to the resin by a quite short linker). The r
of
flow
˝
EP CaLB increased only up to 70 C, then a significant drop of enzyme activity could be observed.
In contrast, productivity of CaLB immobilized on EA-resin activated by CHDGE and GDGE bisepoxide
˝
increased almost linearly with increasing temperature up to 110 C indicating the importance of
properly selected linkers to achieve significant improvement of thermal stability. The productivity of
CaLB immobilized via longer and more flexible linkers (the HDGE-activated EA-resin and the CHDHE-
˝
or GDGE-activated HA-resins) increased also uninterruptedly up to 110 C, although with a slowdown
˝
in the increase starting at around 70 C. HA-HDGE CaLB involving the longest and most flexible linker
proved to be the least thermostable of all the six CaLB forms attached to bisepoxide-activated resins
˝
with a drop of enzyme activity at 100 C.
Besides the productivity of the immobilized enzyme, dependence of enantiomer selectivity of the
KR (R)-
characterized as a function of temperature by the enantiomeric excess of product (R)-
enantiomeric ratio (E) (Figure 7c,d, respectively). The ee–temperature (Figure 7 ) and E–temperature
Figure 7d) curves exhibited maxima at various temperatures between 50 and 80 C. This behavior
was in agreement with previously found maxima of E–temperature curves in kinetic resolutions of
secondary alcohols [45 46 50], although different resins revealed these maxima at different temperatures
Figure 7d). Characteristic differences of enantiomer selectivity optima were found between CaLB on
1
on temperature is also an important issue. Thus, enantiomer selectivity of the process was
1
(ee) and by the
c
˝
(
,
,
(
˝
˝
)
bisepoxide activated EA-resins (~80 C) compared to CaLB on bisepoxide activated HA-resins (~60 C
˝
indicating that increased flexibility of the linker allows a lower degree of selectivity. At 60 C, CaLB
immobilized on HA-resin activated by HDGE showed the highest enantiomer selectivity among all
tested preparations (E ~ 700 with ee(R)-2 = 99.3%), while the HA-GDGE CaLB with shorter and less
hydrophobic spacer arm showed a lower degree of enantiomer selectivity (E ~ 490 with ee(R)-2 = 99.2%).
On the other hand, CaLB immobilized on EA-resins activated by the three bisepoxides (HDGE, CHDGE
˝
and GDGE) displayed maxima in enantioselectivity at around 80 C (E ~ 550, 530 and 380, respectively).
This is a remarkable shift compared to the enantiomer selectivity maxima of CaLB immobilized on
˝
HA-resins at a 20 C lower temperature. Worth mentioning is that enantiomer selectivity maxima
˝
of both bisepoxide-activated series exceeded the 50 C optimum found for EP CaLB without a long
spacer between the resin surface and the enzyme.
2
.7. Characterization of Support Morphology after Recycling and Application in Continuous-Flow Reactors
In addition to their biocatalytic properties, mechanical stability of the immobilized biocatalysts
is also an important issue. Thus, morphology of two selected CaLB forms (on the GDGE-activated
EA-resin and on the HDGE-activated HA-resin) representing both aminoalkyl-resin based series was
investigated by scanning electron microscopy (SEM) before their use, after the recycling study and
after the temperature dependence study (Figure 8).
SEM investigations of EA-GDGE CaLB and HA-HDGE CaLB before their use in biotransformations
revealed intact morphology of spherical beads with diameters 150–300 µm (Figure 8a,d, respectively),
indicating no mechanical damage during bisepoxide activation and enzyme immobilization steps.
Remarkably, both CaLB preparations proved to be mechanically quite durable and remained intact
even after 13 cycles of KR experiments in batch mode (Figure 8b,e). However, the SEM pictures after
the temperature dependency study of KRs, performed in packed-bed reactors using continuous-flow
˝
mode in the 0–110 C temperature range, indicated limitations in the temperature stability of these

Molecules 2016, 21, 767
11 of 18
CaLB forms on polymer beads (Figure 8c,f). Although data sheets of the basic EA- or HA-resins
˝
declared them to be stable up to 60 C, we tested CaLB on their bisepoxide-activated derivatives up to
1
˝
10 C in continuous-flow reactors. Thus, the result indicating fracture of EA-GDGE polymer beads at
this elevated temperature was not surprising (Figure 8c). Remarkably, HA-HDGE beads remained
˝
apparently unaltered even after the temperature-dependency study ended at 110 C.
Figure 8. Scanning electron microscopy (SEM) analysis of EA-GDGE CaLB: (
resolution of rac- ; ( ) after 13 cycles of KRs in batch; ( ) after the temperature-dependence
study in continuous-flow reactor up to 110 C. Scanning electron microscopy (SEM) analysis of
HA-HDGE CaLB: ( ) before kinetic resolution of rac- ; ( ) after 13 cycles of KRs in batch; ( ) after the
a) before kinetic
1
b
c
˝
d
1
e
f
˝
temperature-dependence study in continuous-flow reactor up to 110 C.
3
. Materials and Methods
3
.1. Materials
Bisepoxides [1,6-hexanediol diglycidyl ether (HDGE), neopentylglycol diglycidyl ether
(
(
(
NDGE), 1,4-cyclohexanedimethanol diglycidyl ether (CHDGE), 1,4-butanediol diglycidyl ether
BDGE), poly(ethylene glycol) diglycidyl ether (PEDGE)] were products of Ipox Chemicals Ltd.
Budapest, Hungary). Racemic 1-phenylethanol (rac-1), vinyl acetate, glutaraldehyde (GA)
solution (25% w/v in H O); hydrochloric acid, glycerol diglycidyl ether (GDGE) and Triton
™
2
X-100 [4-(1,1,3,3-tetramethylbutyl)phenyl-polyethylene glycol] were purchased from Sigma-Aldrich
Milwaukee, WI, USA). NaH PO , Na HPO and sodium hydroxide were purchased from Merck
(
2
4
2
4
KGaA (Darmstadt, Germany).
®
Solvents (n-hexane, t-butyl methyl ether, ethanol, Patosolv (mixture of ethanol/isopropanol,
t-butanol), propan-1-ol and isopropanol from Molar Chemicals Kft. (Budapest, Hungary) were dried
and/or freshly distilled prior to use.
Ethylamine (EA)- and hexylamine (HA)-functionalized methacrylic polymer resins (ReliZyme
EA 403 and ReliZyme HA 403; polymethyl methacrylate supports, particle size 150–300 m, pore
size 400–600 Å) were purchased from Resindion S.r.l. (Rome, Italy).
™
™
µ

Molecules 2016, 21, 767
12 of 18
Lipase B from Candida antarctica (CaLB, lyophilized) was purchased from c-LEcta GmbH
Leipzig, Germany). EP CaLB (Candida antarctica lipase B, T2-150, covalently attached to dry acrylic
(
beads with a 150–300 µm particle size) was the product of ChiralVision BV (Leiden, The Netherlands).
3
.2. Analytical Methods
GC analyses were carried out on Agilent 4890 instrument equipped with FID detector using
˝
˝
H carrier gas (injector: 250 C, FID: 250 C, head pressure: 12 psi, split ratio: 50:1) on a
2
Hydrodex
t-butyldimethylsilyl)-β-cyclodextrine; Macherey & Nagel GmbH). tr (min): for 1 and 2 (oven: 120 C,
min), 3.62 [(S)- ], 3.97 [(R)- ], 5.09 [(R)- ], 5.35 [(S)- ]. Conversion (c), enantiomeric excess (ee)
β-6TBDM column (25 m
ˆ
0.25 mm
ˆ
0.25 µm film with heptakis-(2,3-di-O-methyl-6-O-
˝
8
2
2
1
1
and enantiomeric ratio (E) were determined by GC measurements with base-line separations of the
enantiomers of 1 and 2 using precise integration methods.
Enantiomeric ratio (E) was calculated from the conversion (c) and enantiomeric excess of the
product (ee(R)-2) using the equation E = ln[1
to experimental errors, E values calculated in the 100–200 range are given as >100, in the 200–500 range
as >200 and above 500 as >>200.
´
c(1 + ee(R)-2)]/ln[1
´
c(1
´
ee(R)-2)] [52]. Due to sensitivity
To characterize the productivity of the biocatalysts, specific reaction rate (or specific biocatalyst
activity) in batch reactions before (U ) and after (U ) washing with Triton X-100 solution (for details of
b0
b
washing see Section 3.6) were calculated using the equation U = n(R)-2/(t
ˆ
m ) (where n
[µmol] is
b
B
(R)-2
the amount of the product, t [min] is the reaction time and m [g] is the mass of the applied biocatalyst).
B
Specific reaction rates in continuous-flow systems (rflow) were calculated using the equation
´
1
rflow = [(R)-2] ˆ v/m (where [(R)-
2] [µmol mL ] is the molar concentration of the product (R)-2, v
B
´
1
[
mL¨min ] is the flow rate and m [g] is the mass of the applied biocatalyst) [45].
B
The surface morphology of the samples was investigated with a JEOL JSM-5500LV scanning
electron microscope (SEM). The samples of immobilized CaLB were coated with gold prior to analysis.
Electron beam energy of 25 kV was used for the investigations.
3
.3. Drying of the Aminoalkyl Polymer Resins
The original water content (~60%–70%) of the ethylamine- (EA) and hexylamine-functionalized
(
HA) beads was removed prior to further use according to the following procedure: In a proper-sized
flask, 50.0 g of resins were swirled with ethanol (2 100 mL; 2 5 min) and hexane (1 100 mL;
ˆ
ˆ
ˆ
2
ˆ 5 min), including filtration between the washing steps. After the final filtration, resins were dried
at room temperature for 2 h followed by further drying in a VDL 23 vacuum drying chamber, Binder
GmbH (Tuttlingen, Germany) at room temperature until the vacuum level dropped below 2 mbar.
3
.4. Surface Activation of Aminoalkyl Polymer Resins with Bisepoxides or Glutaraldehyde
Dry polymer support (EA- or HA-resin, 500.0 mg) was added to a solution of bisepoxide (2.5 mmol,
GDGE, HDGE, NDGE, CHDGE, BDGE or PEDGE) or glutaraldehyde (2.5 mmol) in propan-1-ol (15 mL).
The suspension of polymer support in activating agent solution was shaken at 400 rpm for 24 h at
˝
®
2
5 C. The activated support was filtered off on a glass filter (G3), washed with Patosolv (3
ˆ
10 mL),
˝
dried at room temperature (4 h) and stored at 4 C (in case of the GA-activated support, under argon).
3
.5. Immobilization of CaLB on Bisepoxide-Activated Polymer Resins
CaLB (40.0 mg, lyophilized powder) was dissolved in phosphate buffer (10.0 mL, 100 mM, pH 7.5),
and then the support (200.0 mg) was added to the solution. The enzyme-support suspension was
˝
shaken at 400 rpm for 24 h at 25 C. The immobilized CaLB preparation was filtered off on a glass filter
G3), washed with isopropanol (2 15 mL) and hexane (15 mL), dried for 4 h at room temperature and
(
ˆ
˝
stored at 4 C. Protein concentration of the CaLB solution before immobilization and in supernatant
after immobilization were determined according to Bradford’s method [57]. As a standard, BSA
was used.

Molecules 2016, 21, 767
13 of 18
Immobilization yield (IY) was calculated according to equation IY(%) = P/P
ˆ
100 (where P0
0
´
1
´1
[
mg
¨
mL ] is the initial protein concentration before immobilization, and P [mg
¨
mL ] is the protein
concentration in supernatant after immobilization).
3
.6. Lipase Desorption Tests
Samples of the immobilized CaLB preparations (80.0 mg of each) were washed at room
temperature for 1.5 h with a 1% (v/v) solution of Triton X-100 in phosphate buffer (7.5 mL; 100 mM;
pH 7.5) in a shaker at 400 rpm. The samples were then filtered off on a glass filter (G3) and resuspended
in distilled water (15 mL) and incubated for 30 minutes at 400 rpm, filtered off and washed several
times with distilled water, isopropanol (2 ˆ 10 mL) and hexane (5 mL).
The “Wash resistance” was calculated using the equation Wash resistance (%) = U /U
ˆ
100
b
b0
´
1
´1
] is the specific reaction rate achieved with the immobilized CaLB right
(
where Ub0
after immobilization and Ub [
X-100 solution. For a definition of U , see Section 3.2).
[µmol min
g
´
1
´1
g ] is the specific reaction rate after washing with Triton
µmol min
b
3
.7. Kinetic Resolution of 1-Phenylethanol (rac-1) Catalyzed by the CaLB Preparations in Batch Mode
In a screw capped glass vial (4 mL), the immobilized CaLB biocatalyst (25.0 mg) was added
to a solution of 1-phenylethanol (rac-
methyl ether/vinyl acetate 6/3/1 (2.0 mL) and the mixture was shaken (1000 rpm) at 30 C for
h. The reactions were analyzed by GC after 1, 2, and 4 h as described in Section 2.2. All test reactions
1: 50.0 mg, 0.409 mmol) in a mixture of hexane/t-butyl
˝
4
were performed in triplicates. Standard deviations of conversion were below 5%, standard deviations
of enantiomeric excess were below 0.4%.
3
.8. Optimization of the Bisepoxide Activation of Alkylamino Polymer Resins
Bisepoxide (GDGE, HDGE or CHDGE) in various molar amounts (1.25 mmol, 2.5 mmol,
mmol or 10 mmol) was dissolved in propan-1-ol (15 mL) and then polymer support (EA or
HA, 500.0 mg) was added to the solution. The mixture was shaken (400 rpm) at 25 C for 24 h.
5
˝
®
The bisepoxide-activated support was filtered off on a glass filter (G3), washed with Patosolv
˝
(
3 ˆ 10 mL), dried at room temperature (4 h) and stored at 4 C. Catalytic properties of the CaLB
on the bisepoxide-activated support was tested (for details of immobilization, washing and testing see
Sections 3.5–3.7, respectively).
3
.9. Recycling of the Immobilized CaLB Biocatalysts
Immobilized CaLB biocatalyst (25.0 mg) was added to a solution of 1-phenylethanol (rac-1, 50.0 mg,
0
.409 mmol) in hexane/t-butyl methyl ether/vinyl acetate 6/3/1 (1.0 mL) in an Eppendorf tube and
˝
the mixture was shaken (1000 rpm) for 1 h at 30 C. After 1 h, the reaction mixture was centrifuged,
the immobilized CaLB biocatalyst was washed with hexane (2 1.0 mL), then fresh solution of rac-
50.0 mg, 0.409 mmol) in hexane/t-butyl methyl ether/vinyl acetate 6/3/1 (1.0 mL) was added to
ˆ
1
(
˝
biocatalysts and samples were shaken again (1000 rpm) at 30 C for 1 h. In this way, immobilized CaLB
biocatalysts were tested in 13 consecutive cycles.
˝
3
.10. Long-Term Stability of CaLB Derivatives (Storage at 4 C)
Residual activities of enzyme preparations (EA-GDGE CaLB, EA-HDGE CaLB, EA-CHDGE CaLB,
˝
HA-GDGE CaLB, HA-HDGE CaLB, HA-CHDGE CaLB) stored at 4 C in glass vials closed with
screw caps under argon for 12 months in kinetic resolution of rac-1 were determined as described in
Section 3.7.

Molecules 2016, 21, 767
14 of 18
3
.11. Packed-Bed Columns Filled with Immobilized CaLB
Immobilized CaLB preparations [EP CaLB, EA-GDGE CaLB, EA-HDGE CaLB, EA-CHDGE CaLB,
HA-GDGE CaLB, HA-HDGE CaLB, HA-CHDGE CaLB) were packed into stainless steel CatCart
™
columns (stainless steel, inner diameter: 4 mm; total length: 70 mm; packed length: 65 mm;
inner volume: 0.816 mL) according to the filling process of ThalesNano Inc (Budapest, Hungary).
The columns were settled by silver metal [Sterlitech Silver Membrane from Sigma-Aldrich (Milwaukee,
WI, USA), Z623237, pore size 0.45
Milwaukee, WI, USA, WHA10411311, pore size 0.45
per enzyme preparation (separate columns for temperature and substrate concentration test) were
µ
m; purity: 99.97%] and PTFE [Whatman® Sigma-Aldrich,
m] filter membranes. Two CatCart columns
µ
™
packed (filling weights: EP CaLB, 250
˘
10 mg; HA-GDGE CaLB, 230
15 mg; EA-GDGE CaLB, 229
˘
13 mg; HA-HDGE CaLB,
5 mg; EA-HDGE CaLB,
2
2
20 ˘ 2 mg; HA-CHDGE CaLB, 230 ˘
˘
12 ˘ 9 mg; EA-CHDGE CaLB, 214 ˘ 14 mg).
3
.12. Kinetic Resolution of 1-Phenylethanol (rac-1) Catalyzed by the CaLB Preparations in
Continuous-Flow Mode
Continuous-flow kinetic resolutions were performed in a laboratory flow reactor built from a
Knauer K-120 isocratic HPLC pump attached to CatCart™columns filled with the immobilized CaLB
biocatalysts in an in–house made aluminum metal block column holder with precise temperature
control. The CaLB-filled columns were washed with a dry hexane/t-butyl methyl ether 2/1 mixture
´
1
(
1 mL¨min , 20 min) before reactions with rac-1.
To study the effect of substrate concentration on the biocatalytic properties of immobilized CaLB
biocatalysts, the solution of 1-phenylethanol (rac-
1
) at different concentrations (1, 2.5, 5, 9.5, 22, 40,
´
1
6
8, 103, 148, 174 mg
¨
mL ) and vinyl acetate (5 equiv.) in dry hexane/t-butyl methyl ether 2/1 was
˝
´1
min . Samples were
pumped through the enzyme-filled columns set to 30 C at a flow rate of 0.2 mL
¨
analyzed by GC every 10 min up to 40 min to allow the establishment of a stationary state. Samples
were collected during stationary operation (40 min after changing the parameters) and analyzed as
described in Section 3.2.
3
.13. Effect of Temperature on Kinetic Resolution of 1-Phenylethanol (rac-1) Catalyzed by the CaLB
Preparations in Continuous-Flow Mode
The solution of 1-phenylethanol rac-
1
¨
mL 1) and vinyl acetate (5 equiv.) in
´
(40.0 mg
hexane/t-butyl methyl ether 2/1 was pumped through the biocatalyst-filled columns thermostated to
˝
´1
min . Samples were collected during
various temperatures (0–110 C) at a flow rate of 0.2 mL
¨
stationary operation (40 min after changing the parameters) and analyzed as described above.
˝
˝
The experiments were performed with 10 C steps in a temperature range between 0 and 110 C.
After completing experiments at different substrate concentrations and temperatures, the columns
´
1
were routinely washed with hexane/t-butyl methyl ether 2/1 mixture (0.5 mL
then analyzed by scanning electron microscope (for details on SEM, see Section 3.2).
¨
min , 20 min) and
4
. Conclusions
Covalent immobilization of Candida antarctica lipase B (CaLB) was successfully carried out on
bisepoxide-activated aminoalkyl polymer resins. Our study with six bisepoxides of different chemical
features as activating agents for macroporous acrylate resins with ethylamine and hexylamine functions
on their surface revealed that, by selecting the proper bisepoxide activation, tuning of the biocatalytic
properties of the immobilized CaLB was feasible. Re-use and long-term stability studies indicated that
bisepoxide-activated polymer supports provided a suitable micro-environment for CaLB molecule
during catalysis and storage. Bisepoxide 1,4-cyclohexanedimethanol diglycidyl ether—CHDGE proved
to be especially suitable in activating alkylamino resins for covalent immobilization of CaLB by
providing a hydrophobic surface and epoxy function suitable for covalent immobilization in a single
˝
step. The thermal stability of covalently immobilized biocatalysts was studied between 0 and 110 C in

Molecules 2016, 21, 767
15 of 18
continuous-flow packed-bed reactors and was compared to commercially available CaLB immobilized
on macroporous acrylic epoxy resin. Spacer arms of various lengths, hydrophobicity and rigidity
markedly altered the biocatalytic properties of CaLB attached to polymer resins. It seemed that
hydrophobic stabilization of the immobilized enzyme by a relatively rigid and hydrophobic spacer
arm contributed significantly to preservation of enzyme activity and selectivity at higher temperatures.
Acknowledgments: This work was supported by the New Széchényi Development Plan (“Development
of quality oriented and harmonized R+D+I strategy and functional model at BME” project; TÁMOP-4.2.1/
B-09/1/KMR-2010-0002) and by Gedeon Richter pharmaceutical company. Thanks are also due to ThalesNano
Nanotechnology Inc. for the fillable CatCart
valuable comments.
™ columns and to Prof. Mihály Nógrádi (Budapest, Hungary) for his
Author Contributions: Emese Abaházi, Zoltán Boros and László Poppe conceived and designed the experiments;
Dávid Lestál performed the experiments; Emese Abaházi, Dávid Lestál and László Poppe analyzed the data;
Emese Abaházi and László Poppe wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations
The following abbreviations are used in this manuscript:
CaLB
EA
Candida antarctica lipase B
ethylamine functionalized macroporous acrylate resin
hexamethylamine functionalized macroporous acrylate resin
glycerol diglycidyl ether
HA
GDGE
HDGE
CHDGE
BDGE
NDGE
PEDGE
GA
1,6-hexanediol diglycidyl ether
cyclohexanedimethanol diglycidyl ether
1,4-butanediol diglycidyl ether
neopentylglycol diglycidyl ether
polyethyleneglycol diglycidyl ether
glutaraldehyde
References
1
2
.
.
Sandoval, G. Lipases and Phospholipases, Methods and Protocols; Humana Press: New York, NY, USA, 2012.
Forde, J.; Vakurov, A.; Gibson, T.D.; Millner, P.; Whelehan, M.; Marison, I.W.; Ó’Fágáin, C. Chemical
modification and immobilisation of lipase B from Candida antarctica onto mesoporous silicates. J. Mol. Catal.
B Enzym. 2010, 66, 203–209. [CrossRef]
3.
4.
5.
6.
Rodrigues, D.S.; Mendes, A.A.; Adriano, W.S.; Goncalves, L.R.B.; Giordano, R.L.C. Multipoint covalent
immobilization of microbial lipase on chitosan and agarose activated by different methods. J. Mol. Catal.
B Enzym. 2008, 51, 100–109. [CrossRef]
Palomo, J.M.; Muñoz, G.; Fernández-Lorente, G.; Mateo, C.; Fernández-Lafuente, R.; Guisán, J.M. Interfacial
adsorption of lipases on very hydrophobic support (octadecyl-Sepabeads): Immobilization, hyperactivation
and stabilization of the open form of lipases. J. Mol. Catal. B Enzym. 2002, 19–20, 279–286. [CrossRef]
Poppe, J.K.; Costa, A.P.O.; Brasil, M.C.; Rodrigues, R.C.; Ayub, M.A.Z. Multipoint covalent immobilization
of lipases on aldehyde-activated support: Characterization and application in transesterification reaction.
J. Mol. Catal. B Enzym. 2013, 94, 57–62. [CrossRef]
Hasan, F.; Shah, A.A.; Hameed, A. Industrial applications of microbial lipases. Enz. Microb. Technol. 2006, 39,
2
35–251. [CrossRef]
7
8
.
.
Verger, R. Interfacial activation of lipases: Facts and artifacts. Trends Biotechnol. 1997, 15, 32–38. [CrossRef]
Reis, P.; Holmberg, K.; Watzke, H.; Leser, M.E.; Miller, R. Lipases at interfaces: A review. Adv. Colloid
Interface Sci. 2009, 147–148, 237–250. [CrossRef] [PubMed]
9
.
Bornscheuer, U.T.; Kazlauskas, R.J. Hydrolases in Organic Synthesis. In Regio- and Stereoselective
Biotransformations, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2006; Chapter 5; pp. 61–183.
1
0. Palomo, J.M.; Fernández-Lorente, G.; Mateo, C.; Fuentes, M.; Fernández-Lafuente, R.; Guisan, J.M.
Modulation of the enantioselectivity of Candida antarctica B lipase via conformational engineering: Kinetic
resolution of (˘)-α-hydroxy-phenylacetic acid derivatives. Tetrahedron 2002, 13, 1337–1345. [CrossRef]

Molecules 2016, 21, 767
16 of 18
1
1
1. Brady, L.; Brzozowski, A.M.; Derewenda, Z.S.; Dodson, E.; Dodson, G.; Tolley, S.; Turkenburg, J.P.;
Christiansen, L.; Huge-Jensen, B.; Norskov, L.; et al. A serine protease triad forms the catalytic center
of a triacylglycerol lipase. Nature 1990, 343, 767–770. [CrossRef] [PubMed]
2. Cunha, A.G.; Besteti, M.D.; Manoel, E.A.; da Saliva, A.A.T.; Almeida, R.V.; Simas, A.B.C.;
Fernandez-Lafuente, R.; Pinto, J.C.; Freire, D.M.G. Preparation of core-shell polymer supports to immobilize
lipase B from Candida antarctica. Effect of the support nature on catalytic properties. J. Mol. Catal. B Enzym.
2
014, 100, 59–67. [CrossRef]
1
1
3. Kramer, M.; Cruz, J.C.; Pfromm, P.H.; Rezac, M.E.; Czermak, P. Enantioselective transesterification by Candida
antarctica Lipase B immobilized on fumed silica. J. Biotechnol. 2010, 150, 80–86. [CrossRef] [PubMed]
4. Lozano, P.; Víllora, G.; Gómez, D.; Gayo, A.B.; Sánchez-Conesa, J.A.; Rubio, M.; Iborra, J.L. Membrane
reactor with immobilized Candida antarctica lipase B for ester synthesis in supercritical carbon dioxide.
J. Supercrit. Fluids 2004, 29, 121–128. [CrossRef]
1
1
1
1
5. Barbosa, O.; Ortiz, C.; Torres, R.; Fernandez-Lafuente, R. Effect of immobilization protocol on the properties
of lipase B from Candida antarctica in organic media: Enantiospecific production of atenolol acetate. J. Mol.
Catal. B Enzym. 2011, 71, 124–132. [CrossRef]
6. Uppenberg, J.; Hansen, M.T.; Patkar, S.; Jones, T.A. The sequence, crystal structure determination and
refinement of two crystal forms of lipase B from Candida antarctica. Structure (Oxford, UK) 1994, 2, 293–308.
[CrossRef]
7. Uppenberg, J.; Ohrner, N.; Norin, M.; Hult, K.; Kleywegt, G.J.; Patkar, S.; Waagen, V.; Anthonsen, T.;
Jones, T.A. Crystallographic and molecular-modeling studies of lipase B from Candida antarctica reveal a
stereospecificity pocket for secondary alcohols. Biochemistry 1995, 34, 16838–16851. [CrossRef] [PubMed]
8. Salis, A.; Svensson, I.; Monduzzi, M.; Solinas, V.; Adlercreutz, P. The atypical lipase B from Candida antarctica
is better adapted for organic media than the typical lipase from Thermomyces lanuginosa. Biochim. Biophys.
Acta 2003, 1646, 145–151. [CrossRef]
19. Martinelle, M.; Holmquist, M.; Hult, K. On the interfacial activation of Candida antarctica lipase A and B as
compared with Humicola lanuginosa lipase. Biochim. Biophys. Acta 1995, 1258, 272–276. [CrossRef]
20. Stauch, B.; Fisher, S.J.; Cianci, M. Open and closed states of Candida antarctica lipase B: Protonation and the
mechanism of interfacial activation. J. Lipid Res. 2015, 56, 2348–2358. [CrossRef] [PubMed]
21. Mihailovic, M.; Stojanovic, M.; Banjanac, K.; Carevic, M.; Prlainovic, N.; Milosavic, N.; Bezbradica, D.
Immobilization of lipase on epoxy-activated Purolite A109 and its post-immobilization stabilization.
Process Biochem. 2014, 49, 637–646. [CrossRef]
2
2
2
2
2. Kumar, D.; Nagar, S.; Bhushan, I.; Kumar, L.; Parshad, R.; Gupta, V.K. Covalent immobilization of organic
solvent tolerant lipase on aluminium oxide pellets and its potential application in esterification reaction.
J. Mol. Catal. B Enzym. 2013, 87, 51–61. [CrossRef]
3. Barbosa, O.; Ariza, C.; Ortiz, C.; Torres, R. Kinetic resolution of (R/S)-propranolol (1-isopropylamino-3-
(1-naphtoxy)-2-propranol) catalyzed by immobilized preparations of Candida antarctica lipase B (CAL-B).
New Biotechnol. 2010, 27, 844–850. [CrossRef] [PubMed]
4. Bukhari, A.; Idris, A.; Atta, M.; Loong, T.C. Covalent immobilization of Candida antarctica lipase B on
nanopolystyrene and its application to microwave-assisted esterification. Chin. J. Catal. 2014, 35, 1555–1564.
[CrossRef]
5. Barbosa, O.; Torres, R.; Ortiz, C.; Fernandez-Lafuente, R. Versatility of glutaraldehyde to immobilize lipases:
Effect of the immobilization protocol on the properties of lipase B from Candida antarctica. Process Biochem.
2
012, 47, 1220–1227. [CrossRef]
2
2
2
2
6. Poojari, Y.; Clarson, S.J. Thermal stability of Candida antarctica lipase B immobilized on macroporous acrylic
resin particles in organic media. Biocatal. Agric. Biotechnol. 2013, 2, 7–11. [CrossRef]
7. Sheldon, R.A. Characteristic features and biotechnological applications of cross-linked enzyme aggregates
(CLEAs). Appl. Microbiol. Biotechnol. 2011, 92, 467–477. [CrossRef] [PubMed]
8. Avnir, D.; Braun, S.; Lev, O.; Ottolenghi, M. Enzymes and Other Proteins Entrapped in Sol-Gel Materials.
Chem. Mater. 1994, 6, 1605–1614. [CrossRef]
9. Torres, R.; Ortiz, C.; Pessela, B.C.C.; Palomo, J.M.; Mateo, C.; Guisán, J.M.; Fernández-Lafuente, R.
Improvement of the enantioselectivity of lipase (fraction B) from Candida antarctica via adsorption on
polyethyleneimine-agarose under different experimental conditions. Enzyme. Microb. Technol. 2006, 39,
1
67–171. [CrossRef]

Molecules 2016, 21, 767
17 of 18
3
3
0. Lee, C.H.; Lin, T.S.; Mou, C.Y. Mesoporous materials for encapsulating enzymes. Nano Today 2009, 4, 165–179.
CrossRef]
1. Boros, Z.; Abaháziová, E.; Oláh, M.; Sátorhelyi, P.; Erdélyi, B.; Poppe, L. Novel hydrophobic silica gels as
[
carriers for lipases—Separation of lipase A and lipase B from Candida antarctica. Chim. Oggi/Chem. Today
2
012, 30, 28–31.
3
3
2. Blanco, R.M.; Terreros, P.; Munoz, N.; Serra, E. Ethanol improves lipase immobilization on a hydrophobic
support. J. Mol. Catal. B Enzym. 2007, 47, 13–20. [CrossRef]
3. Mendes, A.A.; de Castro, H.F.; Andrade, G.S.S.; Tardioli, P.W.; Giordano, R.L.C. Preparation and application
of epoxy-chitosan/alginate support in the immobilization of microbial lipases by covalent attachment.
React. Funct. Polym. 2013, 73, 160–167. [CrossRef]
3
3
4. Bernal, C.; Urrutia, P.; Illanes, A.; Wilson, L. Hierarchical meso-macroporous silica grafted with glyoxyl
groups: Opportunities for covalent immobilization of enzymes. New Biotechnol. 2013, 30, 500–506. [CrossRef]
[PubMed]
5. Grazú, V.; Abian, O.; Mateo, C.; Batista-Viera, F.; Fernández-Lafuente, R.; Guisán, J.M. Stabilization
of Enzymes by Multipoint Immobilization of Thiolated Proteins on New Epoxy-Thiol Supports.
Biotechnol. Bioeng. 2005, 90, 597–605. [CrossRef] [PubMed]
3
6. Mateo, C.; Grazu, V.; Palomo, J.M.; López-Gallego, F.; Fernández-Lafuente, R.; Guisán, J.M. Immobilization
of enzymes on heterofunctional epoxy supports. Nat. Protoc. 2007, 2, 1022–1033. [CrossRef] [PubMed]
7. Barbosa, O.; Torres, R.; Ortiz, C.; Berenguer-Murcia, Á.; Rodrigues, R.C.; Fernandez-Lafuente, R.
Heterofunctional Supports in Enzyme Immobilization: From Traditional Immobilization Protocols to
Opportunities in Tuning Enzyme Properties. Biomacromolecules 2013, 14, 2433–2462.
3
3
3
4
4
4
8. Garcia-Galan, C.; Berenguer-Murcia, Á.; Fernandez-Lafuente, R.; Rodrigues, R.C. Potential of Different
Enzyme Immobilization Strategies to Improve Enzyme Performance. Adv. Synt. Catal. 2011, 353, 2885–2904.
[CrossRef]
9. Mateo, C.; Torres, R.; Fernández-Lorente, G.; Ortiz, C.; Fuentes, M.; Hidalgo, A.; López-Gallego, F.; Abian, O.;
Palomo, J.M.; Betancor, L.; et al. Epoxy-amino groups: A new tool for improved immobilization of proteins
by the epoxy method. Biomacromolecules 2003, 4, 772–777. [CrossRef] [PubMed]
0. Vasylieva, N.; Barnych, B.; Meiller, A.; Maucler, C.; Pollegioni, L.; Lin, J.S.; Barbier, D.; Marinesco, S. Covalent
enzyme immobilization by poly(ethylene glycol) diglycidyl ether (PEGDE) for microelectrode biosensor
preparation. Biosens. Bioelectron. 2011, 26, 3993–4000. [CrossRef] [PubMed]
1. Weiser, D.; Varga, A.; Kovács, K.; Nagy, F.; Szilágyi, A.; Vértessy, B.G.; Paizs, C.; Poppe, L. Bisepoxide
Cross-Linked Enzyme Aggregates—Novel Immobilized Biocatalysts for Selective Biotransformations.
ChemCatChem 2014, 6, 1463–1469.
2. Bartha-Vári, J.H.; To s¸ a, M.I.; Irimie, F.D.; Weiser, D.; Boros, Z.; Vértessy, B.G.; Paizs, C.; Poppe, L.
Immobilization of phenylalanine ammonia-lyase on single-walled carbon nanotubes for stereoselective
biotransformations in batch and in continuous-flow modes. ChemCatChem 2015, 7, 1122–1128. [CrossRef]
[PubMed]
4
4
3. Bencze, L.C.; Bartha-Vári, J.H.; Katona, G.; To s¸ a, M.I.; Paizs, C.; Irimie, F.D. Nanobioconjugates of Candida
antarctica lipase B and single-walled carbon nanotubes in biodiesel production. Bioresour. Technol. 2016, 200,
8
53–860. [CrossRef] [PubMed]
4. Boros, Z.; Hornyánszky, G.; Nagy, J.; Poppe, L. Stereoselective Hydrolase-Catalyzed Processes in
Continuous-Flow Mode. In Cascade Biocatalysis: Integrating Stereoselective and Environmentally Friendly
Reactions; Wiley-VCH: Weinheim, Germany, 2014; pp. 199–230.
4
4
4
4
5. Mak, X.Y.; Laurino, P.; Seeberger, P.H. Asymmetric reactions in continuous flow. Beilstein. J. Org. Chem.
2
009, 5. [CrossRef] [PubMed]
6. Yuryev, R.; Strompen, S.; Liese, A. Coupled chemo(enzymatic) reactions in continuous flow. Beilstein. J.
Org. Chem. 2011, 7, 1149–1467. [CrossRef] [PubMed]
7. Csajági, C.; Szatzker, G.; T o˝ ke, E.R.; Ürge, L.; Darvas, F.; Poppe, L. Enantiomer selective acylation of racemic
alcohols by lipases in continuous-flow bioreactors. Tetrahedron Asymmetry 2008, 19, 237–246. [CrossRef]
8. Boros, Z.; Falus, P.; Márkus, M.; Weiser, D.; Oláh, M.; Hornyánszky, G.; Nagy, J.; Poppe, L. How the mode of
Candida antarctica lipase B immobilization affects the continuous-flow kinetic resolution of racemic amines at
various temperatures. J. Mol. Catal. B Enzym. 2013, 85–86, 119–125.

Molecules 2016, 21, 767
18 of 18
4
5
9. Falus, P.; Boros, Z.; Kovács, P.; Poppe, L.; Nagy, J. Lipase-Catalyzed Kinetic Resolution of
-(2-Hydroxycyclohexyl)Indoles in Batch and Continuous-Flow Systems. J. Flow Chem. 2014, 4, 125–134.
CrossRef]
0. Silva, M.V.M.; Bassut, J.F.; Junior, I.I.; de Souza, S.P.; Estrada, M.L.G.; Miranda, L.S.M.; de Souza, R.O.M.A.
Lipase immobilization towards improved productivity on kinetic resolutions by a continuous-flow process.
RSC Adv. 2015, 5, 102409–102415. [CrossRef]
1
[
5
5
5
5
1. García-Verdugo, E.; Altava, B.; Burguete, M.I.; Lozano, P.; Luis, S.V. Ionic liquids and continuous flow
processes: A good marriage to design sustainable processes. Green Chem. 2015, 17, 2693–2713. [CrossRef]
2. Chen, C.S.; Fujimoto, Y.; Girdaukas, G.; Sih, C.J. Quantitative analyses of biochemical kinetic resolutions of
enantiomers. J. Am. Chem. Soc. 1982, 104, 7294–7299. [CrossRef]
3. Hellner, G.; Boros, Z.; Tomin, A.; Poppe, L. Novel sol-gel lipases by designed bioimprinting for
continuous-flow kinetic resolutions. Adv. Synth. Catal. 2011, 353, 2481–2491. [CrossRef]
4. Poppe, J.K.; Garcia-Galan, C.; Matte, C.R.; Fernandez-Lafuente, R.; Rodrigues, R.C.; Ayub, M.A.Z.
Optimization of synthesis of fatty acid methyl esters catalyzed by lipase B from Candida antarctica
immobilized on hydrophobic supports. J. Mol. Catal. B Enzym. 2013, 94, 51–56. [CrossRef]
5. Abaháziová, E.; Boros, Z.; Poppe, L. Additives enhancing the catalytic properties of lipase from Burkholderia
cepacia immobilized on mixed-function-grafted mesoporous silica gel. Molecules 2014, 19, 9818–9837.
5
[CrossRef] [PubMed]
5
6. Boros, Z.; Weiser, D.; Márkus, M.; Abaháziová, E.; Magyar, Á.; Tomin, A.; Koczka, B.; Kovács, P.;
Poppe, L. Hydrophobic adsorption and covalent immobilization of Candida antarctica lipase B on
mixed-function-grafted silica gel supports for continuous-flow biotransformations. Process Biochem. 2013, 48,
1
039–1047. [CrossRef]
5
7. Bradford, M.M. A rapid and sensitive method for the quantitation of microgram quantities of protein
utilizing the principle of protein dye binding. Anal. Biochem. 1976, 72, 248–254. [CrossRef]
Sample Availability: Samples of CaLB on bisepoxide-activated EA- or HA-resins are available from the authors.
©
2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
(