Performance of Alcalase formulations in near dry organic media: Effect of enzyme hydration on dipeptide synthesis

Article abstract of DOI:10.1016/j.molcatb.2012.01.022

The use of different Alcalase formulations for protease-catalyzed dipeptide synthesis was investigated by studying the coupling of the carbamoylmethyl ester of N-protected phenylalanine with phenylalanine amide in tetrahydrofuran in the presence of molecular sieves (i.e. under near dry conditions). Hydration prior to drying (with anhydrous tert-butanol and anhydrous tetrahydrofuran) of the Alcalase formulations resulted in a significant increase in rate of the subsequent dipeptide synthesis. Repeated use, in the presence of molecular sieves, without intermediate rehydration led to inactivation of the enzyme. For three enzyme formulations this inactivation could be counteracted by intermediate rehydration. Inactivation of another enzyme formulation was only partially reversible by hydration. Alcalase immobilized onto dicalite with glutaraldehyde was found to be the most active in dipeptide synthesis, i.e. the formulation that initially produces the largest amount of product per gram of total formulation per unit of time. Due to its small particle size and its lack of operational stability, this formulation may nevertheless not be the best choice for the synthesis of dipeptides in neat organic media on a large scale. The most promising enzyme formulation for this is Alcalase covalently immobilized onto macroporous acrylic beads due to its reasonable activity, its seemingly good operational stability, and its size and uniform shape.

Full text of DOI:10.1016/j.molcatb.2012.01.022

Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Performance of Alcalase formulations in near dry organic media:
Effect of enzyme hydration on dipeptide synthesis
P. Vossenberga,∗, H.H. Beeftinka, T. Nuijensb, P.J.L.M. Quaedfliegb, M.A. Cohen Stuartc, J. Trampera
a Bioprocess Engineering, Wageningen University, P.O. Box 8129, 6700 EV Wageningen, The Netherlands
b DSM Innovative Synthesis B.V., P.O. Box 18, 6160 MD Geleen, The Netherlands
c Laboratory of Physical Chemistry and Colloid Science, Wageningen University, P.O. Box 8038, 6700 EK Wageningen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
The use of different Alcalase formulations for protease-catalyzed dipeptide synthesis was investigated
by studying the coupling of the carbamoylmethyl ester of N-protected phenylalanine with phenylalanine
amide in tetrahydrofuran in the presence of molecular sieves (i.e. under near dry conditions). Hydration
prior to drying (with anhydrous tert-butanol and anhydrous tetrahydrofuran) of the Alcalase formula-
tions resulted in a significant increase in rate of the subsequent dipeptide synthesis. Repeated use, in the
presence of molecular sieves, without intermediate rehydration led to inactivation of the enzyme. For
three enzyme formulations this inactivation could be counteracted by intermediate rehydration. Inacti-
vation of another enzyme formulation was only partially reversible by hydration. Alcalase immobilized
onto dicalite with glutaraldehyde was found to be the most active in dipeptide synthesis, i.e. the formu-
lation that initially produces the largest amount of product per gram of total formulation per unit of time.
Due to its small particle size and its lack of operational stability, this formulation may nevertheless not be
the best choice for the synthesis of dipeptides in neat organic media on a large scale. The most promising
enzyme formulation for this is Alcalase covalently immobilized onto macroporous acrylic beads due to
its reasonable activity, its seemingly good operational stability, and its size and uniform shape.
© 2012 Elsevier B.V. All rights reserved.
Received 9 November 2011
Received in revised form 20 January 2012
Accepted 24 January 2012
Available online 6 February 2012
Keywords:
Immobilized protease
Peptide synthesis
Molecular sieves
Near-anhydrous
1. Introduction
of its positive effect on the coupling rate [7]. The coupling was
executed with different commercially available Alcalase formu-
Natural and non-natural peptides play an important role in the
fields of health care, nutrition and cosmetics [1–5]. These pep-
tides may be synthesized chemo-enzymatically for instance by
protease-catalyzed coupling of an amino acid amide and a chem-
ically synthesized C-terminally activated N-protected amino acid,
e.g. an N-protected carbamoylmethyl (Cam) ester [6–11]. To pre-
vent hydrolytic side reactions (hydrolysis of the activated amino
acid, of the formed dipeptide, and of the protease itself), the
availability of water in the system should be minimized, e.g. by
executing the enzymatic peptide synthesis step in a neat organic
solvent [12,13]. The system (reaction mixture and enzyme) should,
however, not be completely dry as enzymes need some water to
maintain their catalytically active conformation [10,12–21].
The present study focuses on the coupling of the carbamoyl-
methyl ester of phenylalanine (of which the amino group was
benzyloxycarbonyl-protected, Z-Phe-OCam) and phenylalanine
amide (H-Phe-NH2) (Fig. 1). A carbamoylmethyl ester was used
instead of the more regularly used methyl or ethyl esters because
lations (Alcalase is also referred to as subtilisin A or subtilisin
Carlsberg [22–25]) in tetrahydrofuran (THF) in the presence of
molecular sieves (i.e. under near dry conditions at low water activ-
ity), which were assumed to be stable in the process. In previous
work we found that the rate of dipeptide synthesis could not
be increased by increasing water activity values without signifi-
cantly increasing the rate of hydrolysis, i.e. without significantly
decreasing the synthesis/hydrolysis (S/H) ratio [10]. An appropri-
ate amount of molecular sieves effectively prevents hydrolysis but
still allows enzymatic activity [10].
Protease-catalyzed peptide synthesis in monophasic organic
solvents has been studied previously, with different proteases:
chymotrypsin [14,26–33], subtilisin Carlsberg [7,28,34–36], subtil-
isin BPN’ [26], subtilisin 72 [37], thermolysin [31,38–40], Alcalase
[6,41–43], and papain [44]. Nevertheless, the present contribution
is to our knowledge the first to investigate the coupling of Z-Phe-
OCam and Phe-NH2 in THF with different Alcalase formulations
(‘hydrated’ and ‘dry’) in the presence of molecular sieves (i.e. under
near dry conditions at low water activity) and subsequently com-
pare the formulations with respect to their initial peptide synthesis
activity and their applicability on a large scale.
∗
Corresponding author. Tel.: +31 317 482954.
E-mail addresses: petra.vossenberg@wur.nl, petravossenberg@gmx.de
The aim of this study was twofold: (1) to investigate the effect of
initial hydration state of selected Alcalase formulations, achieved
(P. Vossenberg).
1381-1177/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2012.01.022

P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
25
Fig. 1. Coupling of an N-protected, C-terminally activated amino acid with an amino acid amide to give a dipeptide (1), and hydrolysis of an N-protected, C-terminally
activated amino acid (2). X1 and X2 can be any amino acid side chain, but are in this study both a benzyl side chain. R is an activating group. In this study R is carbamoylmethyl.
by a well-defined pretreatment protocol, on the subsequent
dipeptide synthesis rate, and (2) to compare these formulations
with respect to their applicability in dipeptide synthesis.
To achieve these aims, several commercially available Alcalase
formulations, both ‘hydrated’ and ‘dry,’ were applied in the cou-
pling of Z-Phe-OCam and Phe-NH2. In addition, the repeated use
(with and without intermediate rehydration) and thereby the oper-
ational stability of the four most active Alcalase formulations were
compared. For this, the conversion of Z-Phe-OCam to dipeptide
was monitored in time in two subsequent batch reactions. Fur-
thermore, Alcalase formulation properties (particle size and shape,
enzyme leaching from the support, and cost of immobilization) that
are valuable for the synthesis of dipeptides in organic media were
evaluated.
expected to hydrate the enzyme, and to remove the excess water.
A washing step involved adding washing liquid (Milli-Q, t-BuOH, or
THF) to the Alcalase formulation, shaking the sample for 10 s, cen-
trifuging the sample for 2 min at 10,000 rpm in order to facilitate the
separation of the washing liquid and the enzyme formulation, and
removing the washing liquid with a pipette. This method was anal-
ogous to the procedure used to produce propanol-rinsed enzyme
preparations [47,48].
Native Alcalase was not pretreated when applied ‘dry,’ assum-
ing the freeze-dried powder to be dry. When native Alcalase was
applied ‘hydrated’, it was hydrated by water vapor through the gas
phase: it was placed for 96 h at 4 ◦C in an exsiccator containing
Milli-Q.
2.4. Enzymatic peptide synthesis (1-timepoint measurement to
compare Alcalase formulations)
2. Materials and methods
2.1. Enzymes
For the enzymatic peptide synthesis (1-timepoint measure-
ment), 150 ␮l THF containing 5.9 ␮mol H-Phe-NH2 and 150 ␮l THF
containing 3.5 ␮mol Z-Phe-OCam was added to the enzyme for-
mulation. The amount of enzyme formulation was chosen to reach
approximately 50% conversion of the Cam-ester in 1 h, with a
‘hydrated’ enzyme formulation. The following amounts of enzyme
formulation were used: 5.5 mg of native Alcalase, 3 mg of Cov,
2.6 mg of CLEA-OM, 2.6 mg of CLEA-tBu, 8.5 mg of CLEA-ST, 1 mg
of Dic, 30 mg of S-epo, and 3 mg of S-imi. To obtain near dry condi-
The Alcalase formulations used in the present investigation and
a description of the formulations are listed in Table 1.
The formulations from CLEA Technologies [45], ChiralVision
[46], and DSM contain Alcalase from Novozyme Corp. The formula-
tions from SPRIN Technologies contain Subtilisin A from Bacillus
species. In the formulation from ChiralVision, Alcalase is cova-
lently immobilized onto Immobeads 150 (cross-linked copolymer
of methacrylate carrying oxirane groups) produced by ChiralVision
in cooperation with FermentaBiotech Ltd. (Maharashtra, India).
˚
tions during peptide synthesis, 75 mg of 3 A molecular sieves, 8–12
mesh beads, were added. Samples (50 ␮l) were diluted with 700 ␮l
dimethyl sulfoxide (DMSO) to stop the reaction before HPLC anal-
ysis. The peptide synthesis was repeated 3 times for each enzyme
formulation and carried out at 25 ◦C in 2 ml Eppendorf safe-lock
tubes placed in a blood rotator spinning at 30 rpm.
The reaction was stopped after 1 h of incubation except for the
conversions catalyzed by ‘dry’ Cov, ‘dry’ C-OM, and ‘dry’ S-epo, and
native Alcalase. After 1 h of incubation, these formulations only
achieved minimal conversion. For practical reasons (preventing the
use of large amounts of enzyme), these formulations were incu-
bated for 24 h in order to achieve a better measurable conversion.
2.2. Chemicals
tert-Butanol
(t-BuOH)
molecular
and
sieves,
THF
were
8–12
dried
mesh
over
beads
˚
activated
3 A
(KnNa12−n[(AlO2)12(SiO2)12]·xH2O, Sigma Aldrich, 208582),
prior to use. t-BuOH was melted by heating to 40 ◦C prior to use.
The N-protected Cam-ester of phenylalanine (Z-Phe-OCam) and
N-protected phenylalanyl-phenylalanine amide (Z-Phe-Phe-NH2)
were a kind gift from DSM (Geleen, The Netherlands), and used
as supplied. Phenylalanine amide (H-Phe-NH2) and N-protected
phenylalanine (Z-Phe-OH) were purchased from Bachem (Weil am
Rhein, Germany). Deionized water (Milli-Q) was used throughout.
2.5. Enzymatic peptide synthesis (repeated use)
To investigate the repeated use of the Alcalase formulations,
900 ␮l THF containing 29.4 ␮mol H-Phe-NH2 and 900 ␮l THF con-
taining 18.0 ␮mol Z-Phe-OCam were added to a ‘hydrated’ enzyme
formulation. The peptide synthesis was carried out at 25 ◦C in 2 ml
Eppendorf safe-lock tubes placed in a blood rotator spinning at
30 rpm. The enzyme concentrations were identical to the concen-
trations for the 1-timepoint measurements. To obtain near dry
2.3. Enzyme pretreatment
The Alcalase formulations were pretreated in two different ways
(the amounts of enzyme used are specified below, in Section 2.4).
‘Dry’ enzyme was obtained by washing with successively 1 ml of
anhydrous t-BuOH and 1 ml of anhydrous THF. ‘Hydrated’ enzyme
was obtained by washing with successively 1 ml of Milli-Q, 1 ml of
anhydrous t-BuOH, and 1 ml of anhydrous THF. This treatment is
˚
conditions, 10 mg of 3 A molecular sieves (8–12 mesh beads) per
mg of enzyme formulation were added. Samples (30 ␮l), that were

26
P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
Table 1
Alcalase formulations: abbreviations as used in the present investigation, full name/description, particle diameter, and supplier.
Abbreviation
Full name/description
Particle diameter
Supplier
Native
Cov
Lyophilized native protease from Bacillus licheniformis
Alcalase covalently immobilized onto macroporous acrylic
beads (Immobeads 150)
Fine powder
150–300 ␮m [62]
Sigma–Aldrich (Zwijndrecht, The Netherlands)
ChiralVision (Leiden, The Netherlands)
C-OM
Cross-linked enzyme aggregates (CLEA) containing
Alcalase, CLEA-OM, formulation optimized for use in
organic media
1 ␮m, but forms
clusters [63]
CLEA Technologies (Delft, The Netherlands)
C-tBu
C-ST
Dic
Alcalase CLEA-tBu, formulation precipitated and
cross-linked in tert-butanol
Alcalase CLEA-ST, standard CLEA formulation
1 ␮m, but forms
clusters [63]
1 ␮m, but forms
clusters [63]
8 ␮m [64]
CLEA Technologies (Delft, The Netherlands)
CLEA Technologies (Delft, The Netherlands)
Alcalase adsorbed onto dicalite 478 and cross-linked with
glutaraldehyde
A kind gift from DSM (Geleen, The
Netherlands)
S-epo
S-imi
Alcalase covalently immobilized onto epoxy acrylic resin
Alcalase covalently immobilized onto amino acrylic resin
100–300 ␮m [61]
100–300 ␮m [61]
SPRIN Technologies (Trieste, Italy)
SPRIN Technologies (Trieste, Italy)
taken in time, were diluted with 450 ␮l dimethyl sulfoxide (DMSO)
before HPLC analysis.
2.8. Scanning electron microscopy
After 24 h the Alcalase formulations were recycled. In the recy-
cling procedure, the Alcalase formulation was first recovered from
the reaction liquid. Then, fresh substrates were either added imme-
diately to the Alcalase formulation, or the Alcalase formulation,
from which the molecular sieve beads were separated one by one
using tweezers, was first rehydrated by washing with successively
1 ml of each Milli-Q, anhydrous t-BuOH, and anhydrous THF, before
adding fresh substrates and molecular sieve beads.
The Alcalase formulations were placed onto aluminum hold-
ers with double-sided sticky carbon tape (EMS, Washington, USA).
The samples were sputtered with platinum (JEOL, JFC 1200) and
subsequently analyzed with a high resolution scanning electron
microscope (FEI, Magellan 400) at room temperature at a working
distance of 4.1–10.8 mm, with SE detection at 3.5 and 5 kV. Images
were optimized by Photoshop CS5.
3. Results and discussion
3.1. Effect of the initial enzyme hydration state on dipeptide
synthesis
2.6. HPLC analysis
The amounts of dipeptide (Z-Phe-Phe-NH2), Cam-ester (Z-
Phe-OCam), and amino acid (Z-Phe-OH) were analyzed by HPLC
(Thermo Separation Products P4000 pump and AS3000 autosam-
pler) using a reversed-phase column (Inertsil ODS-3, C18, 5 ␮m,
150 mm × 4.6 mm) at 40 ◦C. UV detection was performed at 220 nm
using an Ultimate 3000 Diode Array Detector (Dionex). The gra-
dient program was: 0–25 min linear gradient ramp from 5% to
98% eluent B, 25–29 min linear gradient ramp back to 5% eluent
B, 29–40 min 5% eluent B (eluent A: 0.5 ml l−1 methane sulfonic
acid (MSA) in Milli-Q, eluent B: 0.5 ml l−1 MSA in acetonitrile). The
flow was 1 ml min−1. Injection volumes were 20 ␮l. Quantitative
analysis was carried out using calibration curves of Z-Phe-Phe-NH2,
Z-Phe-OCam, and Z-Phe-OH.
The coupling of Z-Phe-OCam and H-Phe-NH2 was performed
with eight different Alcalase formulations (Table 1). The Alcalase
formulations were either applied ‘dry’ (pre-treated with anhydrous
t-BuOH and THF), or ‘hydrated’ (pre-treated with Milli-Q, anhy-
drous t-BuOH, and THF).
The effects of the initial enzyme hydration state on Z-Phe-OCam
conversion (Fig. 2) were studied. Comparing Cam-ester conversion
with and without initial enzyme hydration, it can be concluded
that initial enzyme hydration causes a significant increase (i.e. no
overlap in error bars) in the rates of Z-Phe-OCam conversion in
case of Cov, C-OM, C-ST, Dic, S-epo, and S-imi (Fig. 2). Without ini-
tial hydration by Milli-Q washing, these enzymes seem to lack the
minimal amount of water needed to maintain their catalytically
active conformation [10,12–21]. This lack of water and therefore
lack of enzyme activity is especially significant for Cov and C-OM.
Initial hydration of C-tBu does not cause an increase in the rate
of Z-Phe-OCam conversion. Apparently, the crude untreated C-tBu
formulation was already sufficiently hydrated.
Only with ‘hydrated’ C-ST and S-epo, Z-Phe-OCam hydrolysis
was observed (in the present system only the hydrolysis of sub-
strate, and not of the product, occurs [10]), albeit minimal. Due to
the low activity of C-ST and S-epo, a large amount of these enzyme
formulations had to be added in order to achieve around 50% con-
version of the Cam-ester in 1 h. The ratio of molecular sieves to
enzyme formulation was therefore the lowest for C-ST and S-epo
(respectively 9 and 2.5 mg molecular sieve beads per mg of enzyme
formulation). Apparently, the amount of molecular sieves present
was not sufficient to adsorb all the water left after washing the
enzyme formulations with anhydrous t-BuOH and THF, and thus
to prevent Cam-ester hydrolysis. For practical reasons (preventing
the use of large amounts of molecular sieves) the amount of molec-
ular sieves was not adjusted. With the other enzyme formulations,
the Cam-ester was exclusively converted to dipeptide without
2.7. Aqueous Alcalase activity
The aqueous activity of the Alcalase formulations was deter-
mined by monitoring the hydrolysis of 25% (v/v) ethyl lactate at
40 ◦C and pH 6.8 (10 ml of 100 mM sodium phosphate buffer pH
6.8, 20 ml of Milli-Q, and 10 ml of ethyl lactate). The formed lactic
acid was titrated with 0.1 mol l−1 sodium hydroxide using pH-stat
equipment (719 Stat Titrino Metrohm; Herisau, Switzerland). The
pH-stat equipment was connected to a computer that logged the
consumption of sodium hydroxide every 2 s. The method was based
on a protocol obtained from ChiralVision [49]. The blanc consump-
tion of sodium hydroxide was monitored for 10 min. Subsequently
the Alcalase formulation was added and the consumption of sodium
hydroxide was monitored for another 30 min. The Alcalase activity
is defined by the rate of sodium hydroxide consumption (corrected
for the blanc consumption of sodium hydroxide). The rate was
determined after equilibration and was based on 100 data points
in an interval of 200 s in total.

P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
27
Fig. 2. Effect of initial hydration state of different Alcalase formulations on the conversion rate of Z-Phe-OCam (left y-axis) and the aqueous activity of the Alcalase formulations
(right y-axis). The amount of converted Cam-ester was measured after 1 h of incubation at 25 ◦C, except for ‘dry’ Cov, ‘dry’ C-OM, ‘dry’ S-epo, and native Alcalase. These
formulations were incubated for 24 h. The error bars are equal to the standard deviation of 3 independent measurements.
hydrolysis. In these cases, the amount of molecular sieve beads
(>14 mg per mg of enzyme formulation) was apparently high
enough to prevent Z-Phe-OCam hydrolysis.
either be reversible or irreversible. A possible cause for reversible
inactivation is dehydration [50]: enzymes are known to require
some essential water to maintain their catalytically active confor-
mation [10,12–21]. Because molecular sieves remove water very
efficiently, the enzyme formulation may slowly lose essential water
and thus activity, as water is transferred from the enzyme to the
molecular sieves. Indeed, in previous work, C-OM that had been
pre-incubated with molecular sieve powder for 24 h before addi-
tion of substrates, showed minimal activity [10]. If such inactivation
by dehydration is reversible, one should be able to restore activity
by rehydrating the enzyme formulations before proceeding with
the next batch.
Indeed, a nearly full reversal of inactivation was observed for
Cov (Fig. 3A), C-OM (Fig. 3B), and C-tBu (Fig. 3C). With intermediate
rehydration, comparable activities were achieved for Cov, C-OM,
and C-tBu, in the first and second batch. Apparently, the activity
loss observed without intermediate rehydration for Cov, C-OM, and
C-tBu, is caused by dehydration.
3.2. Comparison of Alcalase formulations
In terms of Cam-ester conversion after 1 h of incubation, Dic,
both in ‘dry’ and ‘hydrated’ form, is the most active enzyme for-
mulation (Fig. 2), followed by C-tBu (independent of the hydration
state), ‘hydrated’ Cov, ‘hydrated’ C-OM, and ‘hydrated’ S-imi.
The least active enzyme formulations for the synthesis of Z-
Phe-Phe-NH2 are native Alcalase, C-ST, and S-epo (Fig. 2). Although
hydration of native Alcalase, by water vapor through the gas phase,
increased its synthetic activity, it was probably still too dry for sig-
nificant enzymatic peptide synthesis in THF. The reason for the low
activity of C-ST and S-epo is not clear, especially because the aque-
ous activity (i.e. the hydrolysis of ethyl lactate) of C-ST is 1.7 times
higher than the activity of C-OM; the aqueous activities of S-epo
and S-imi are similar (Fig. 2). One might think that the time for the
hydration step (10 s) may not have been sufficient to hydrate C-ST
and S-epo prior to use in organic solvent. We therefore washed C-
ST en S-epo for 30 min with Milli-Q. The longer hydration step did,
however, not remedy the low synthetic activity of C-ST and S-epo
in organic solvent at all.
For Dic (Fig. 3D), the inactivation could not be completely
reversed by intermediate rehydration. With intermediate rehydra-
tion, the activity was higher than without, but still significantly
lower than the initial activity in the first batch. Apparently, Dic is
partially irreversibly inactivated in THF in the presence of molec-
ular sieves. Also, some Dic may have been lost during the washing
steps.
It should be realized that, initially, the ‘hydrated’ enzyme for-
mulations are clearly not at equilibrium with the dry solvent
containing molecular sieves. During the incubation with the sub-
strates and molecular sieves, the system will therefore change
and move towards equilibrium, which involves dehydration of the
enzyme formulation and thereby loss of enzymatic activity.
The kinetics of de- and re-hydration of the Alcalase formulations
as well as the kinetics of the dipeptide synthesis will be studied in
more detail in future work.
3.3. Repeated use of Alcalase formulations
For an Alcalase formulation to be used for dipeptide synthesis
in organic media for longer periods of time, the operational stabil-
ity of the enzyme formulation is important. We therefore studied
the repeated use, with and without intermediate rehydration, of
the four most active Alcalase formulations, i.e. ‘hydrated’ Cov, C-
OM, C-tBu, and Dic (Fig. 3A–D). Z-Phe-OCam was converted only
to Z-Phe-Phe-NH2 (thus was not hydrolyzed) for all 4 enzyme for-
mulations. In the current system, data points requiring extended
incubation and handling may have been affected somewhat by a
certain amount of solvent evaporation (e.g. the concentration of Z-
Phe-Phe-NH2 after about 24 h of incubation is seen to be higher
than the initial concentration of Z-Phe-OCam; Fig. 3A–D).
3.4. Evaluation of Alcalase formulations
From the above results it is concluded that initial hydration of
Cov, C-OM, C-ST, Dic, S-epo, and S-imi, significantly increases the
subsequent rate of Z-Phe-OCam conversion to Z-Phe-Phe-NH2. In
addition, Alcalase immobilized onto dicalite with glutaraldehyde
(Dic) is the most active enzyme formulation for dipeptide synthesis
Without intermediate rehydration, the rate of Z-Phe-OCam con-
version in the second batch was significantly lower than in the first
one, for all four Alcalase formulations. This activity loss in time may

28
P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
Fig. 3. Effect of repeated use, with and without intermediate rehydration, of Cov (A), C-OM (B), C-tBu (C), and Dic (D), on the conversion of Z-Phe-OCam. Conversion of
Z-Phe-OCam (᭹, ꢀ) and, for illustration, formation of Z-Phe-Phe-NH2 (ꢁ) with an initially ‘hydrated’ enzyme formulation; subsequent conversion of Z-Phe-OCam, with (ꢀ)
and without (ꢁ) intermediate rehydration.
(i.e. the formulation that initially produces the largest amount of
product per gram of total formulation per unit of time).
Although Dic is most active, there are other properties to be con-
sidered when selecting an appropriate Alcalase formulation for the
synthesis of dipeptides in neat organic media, such as particle size
and shape, leaching of enzyme from the support, cost of immobi-
lization method, enzyme loading, and operational stability [51,52].
We discuss each of these properties.
channeling, large pressure drops, and clogging [57–59]. Cov, S-epo,
and S-imi, are the most promising formulations to be used in a
packed-bed bioreactor due to their uniformity in shape and size
(Fig. 4).
Enzyme leaching from a support may occur when an enzyme
is physically adsorbed onto a carrier [53,60]. Leaching is a minor
problem in organic solvents in which the enzymes are insoluble
[52,53]. Hydration of the enzyme, however, may aggravate leach-
ing. Dic is prepared by physical adsorbtion of Alcalase onto a carrier
and then cross-linking it with glutaraldehyde. Cross-linked enzyme
molecules may be trapped inside the pores of the dicalite, and
leaching may be modest but not entirely suppressed. Leaching
should not occur with Cov, the CLEAs, S-epo, and S-imi, as these
enzyme formulations are covalently bound.
The costs of immobilization increase when a carrier is involved
[55]. Therefore, the CLEAs are expected to have lower production
costs on industrial scale than Cov, Dic, S-epo, and S-imi. On a small
scale, the selling price of the CLEAs, (D 225 for 5 g; information
obtained from supplier) is, however, the highest compared to Cov
(D 99 for 5 g [46]), S-epo and S-imi (D 25 for 5 g [61]).
An enzyme formulation with
a particle size exceeding
100–300 ␮m [52–54] can be easily separated from the reaction
mixture, which is important when recycling and reusing a costly
enzyme formulation and when enzyme contamination of the prod-
uct should be minimized [53–58]. Apart from native Alcalase, Dic is
the finest enzyme formulation (Table 1, Fig. 4). Its small size ham-
pers separation. Separation of Dic from the reaction mixture will
therefore be costly (e.g. centrifugation). The economic advantage
of the good activity of Dic should thus be weighed against its poor
separability. Filtration with common sieve plates will work for Cov,
S-epo, and S-imi (Table 1, Fig. 4). The economic disadvantage of
a lower activity of Cov and S-imi, compared to Dic may thus be
compensated by their easier separation.
The particle shape of the enzyme formulation is important when
the use of a packed-bed bioreactor is desired. The best candidates
for the formation of a bed with good and stable flow proper-
ties are rigid and uniform particles that exceed 100–200 ␮m [58].
Small, deformable, and irregularly shaped particles can result in
The costs of an immobilized enzyme formulation increases if
a large amount of native enzyme is needed to achieve a certain
specific activity of the formulation. The exact enzyme loading can
only be quantified during the immobilization procedure, and is
unknown for the immobilized Alcalase formulations used in this
investigation.

P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
29
Fig. 4. Scanning electron microscope images of a Cov bead, a CLEA, Dic particles, and an S-imi bead (an S-epo bead looks identical to an S-imi-bead).
Operational stability of the enzyme formulation in organic sol-
vents is essential when using an enzyme formulation for longer
periods of time. Preferably, the activity of the enzyme formula-
tion should not change significantly. If the activity does change,
the enzyme formulation needs to be continuously replenished. In
this work, the issue of operational stability was investigated for
Cov, C-OM, C-tBu, and Dic. When reusing the enzyme formulations,
Cov, C-OM, and C-tBu, could reach similar activities in two sub-
sequent batch reactions if subjected to intermediate rehydration.
In contrast, Dic is either partially irreversibly inactivated in THF
in the presence of molecular sieves or is lost during the washing
steps.
To conclude, although Dic is the most active enzyme formula-
tion, it may not be the best choice for dipeptide synthesis in neat
organic media on a large scale, due to its small size, cost of immobi-
lization, and lack of operational stability. Cov, C-OM, and C-tBu, may
be better choices due to their good operational stability and cova-
lent binding. For application in a packed-bed bioreactor, Cov may
be the best choice due to the reasonable size and uniform shape.
Therefore, Cov will be the focus of our future work on the kinet-
ics of de- and re-hydration as well as the kinetics of the dipeptide
synthesis.
Z-Phe-Phe-NH2 synthesis under dry conditions. Without such ini-
tial hydration, the enzymes seem to lack the water needed to
maintain their catalytically active conformation.
When reusing Cov, CLEA-OM, CLEA-tBu, and Dic, without inter-
mediate rehydration, a significant decrease in the rate of dipeptide
synthesis is observed. The activity loss observed without inter-
mediate rehydration for Cov, C-OM, and C-tBu, was caused by
dehydration. Dic is either partially irreversibly inactivated in THF
in the presence of molecular sieves or is lost during the washing
steps.
Dic is the most active enzyme formulation (the formulation that
initially produces the largest amount of product per gram of total
formulation per unit of time) for the synthesis of Z-Phe-Phe-NH2.
Nevertheless, it may not be the best choice to synthesize dipeptides
in neat organic media on a large scale mainly due to its small size
and operational instability.
Acknowledgements
This work is part of the IBOS-2 research project “chemo-
enzymatic peptide synthesis,” which is financially supported by
the NWO-ACTS (The Netherlands). The authors thank T. Franssen-
Verheijen of the Wageningen Electron Microscopy Centre for her
help with the scanning electron microscope analyses.
4. Conclusions
References
Initial hydration of Alcalase covalently immobilized onto macro-
porous acrylic beads (Cov), Alcalase CLEA-OM (C-OM), Alcalase
CLEA-ST (C-ST), Alcalase immobilized onto dicalite using glu-
taraldehyde (Dic), and Alcalase covalently immobilized onto acrylic
beads (S-imi and S-epo) causes a significant increase in the rate of
[1] F. Guzmán, S. Barberis, A. Illanes, Electronic Journal of Biotechnology 10 (2007)
279–314.
[2] K. Lintner, O. Peschard, International Journal of Cosmetic Science 22 (2000)
207–218.

30
P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
[3] Vivó-Sesé, M.D. Pla, Bioactive ingredients in cosmetics, in: A. Salvador, A.
Chisvert (Eds.), Analysis of Cosmetic Products, Elsevier B.V, 2007, p. 380.
Nonaqueous Solvents: Methods and Protocols, Humana Press Inc., Totowa, NJ,
2001, pp. 97–104.
[4] E.N. Vulfson, Trends in Food Science & Technology 4 (1993) 209–215.
[5] C.L. Stevenson, Current Pharmaceutical Biotechnology 10 (2009) 122–137.
[6] T. Nuijens, C. Cusan, A.C.H.M. Schepers, J.A.W. Kruijtzer, D.T.S. Rijkers, R.M.J.
Liskamp, P.J.L.M. Quaedflieg, Journal of Molecular Catalysis B: Enzymatic 71
(2010) 79–84.
[49] ChiralVision, 2009, http://www.chiralvision.com/pdf/protease ELU activity assay.pdf.
[50] J.F.A. Fernandes, P.J. Halling, Biotechnology Progress 18 (2002) 1455–1457.
[51] J.M. Guisan, Immobilization of enzymes as the 21st century begins, in: J.M.
Guisan (Ed.), Immobilization of Enzymes and Cells, Humana Press, 2006, pp.
1–13.
[7] T. Miyazawa, M. Hiramatsu, T. Murashima, T. Yamada, Letters in Peptide Science
9 (2002) 173–177.
[8] T. Nuijens, C. Cusan, A.C.H.M. Schepers, P.J.L.M. Quaedflieg, Journal of Peptide
Science 16 (2010) 44–47.
[9] P.J.L.M. Quaedflieg, T. Nuijens, C. Cusan, A.C.H.M. Schepers, World Pat. Appl.
2010/057961 (A1), (2010).
[10] P. Vossenberg, H.H. Beeftink, T. Nuijens, M.A. Cohen Stuart, J. Tramper, Journal
of Molecular Catalysis B: Enzymatic 75 (2012) 43–49.
[11] P. Vossenberg, H.H. Beeftink, M.A. Cohen Stuart, J. Tramper, Biocatalysis and
Biotransformation, 29, 2011, 288–298.
[52] S. Pedersen, M.W. Christensen, Immobilized biocatalysts, in: A.J.J. Straathof, P.
Adlercreutz (Eds.), Applied Biocatalysis, Harwood Academic Publishers, 2000,
pp. 213–226.
[53] U. Hanefeld, L. Gardossi, E. Magner, Chemical Society Reviews 38 (2009)
453–468.
[54] W. Tischer, V. Kasche, Trends in Biotechnology 17 (1999) 326–335.
[55] R.A. Sheldon, Advanced Synthesis & Catalysis 349 (2007) 1289–1307.
[56] U. Kragl, L. Greiner, C. Wandrey, in: M.C. Flickinger, S.W. Drew (Eds.), Encyclo-
pedia of Bioprocess Technology, John Wiley & Sons, 1999, pp. 1064–1074.
[57] K.v.t. Riet, J. Tramper, Basic Bioreactor Design, Marcel Dekker, Inc, New York,
1991.
[12] A.M. Klibanov, Nature 409 (2001) 241–246.
[13] G. Ljunger, P. Adlercreutz, B. Mattiasson, Enzyme and Microbial Technology 16
(1994) 751–755.
[58] K. Buchholz, V. Kasche, U.T. Bornscheuer, Biocatalysts and Enzyme Technology,
Wiley-VCH Verlag GmbH & Co, Weinheim, 2005.
[14] H. Gaertner, A. Puigserver, European Journal of Biochemistry 181 (1989)
207–213.
[15] F. Bordusa, Chemical Reviews 102 (2002) 4817–4867.
[16] P. Adlercreutz, European Journal of Biochemistry 199 (1991) 609–614.
[17] A. Zaks, A.M. Klibanov, The Journal of Biological Chemistry 263 (1988)
8017–8021.
[59] X. Xu, European Journal of Lipid Science and Technology 105 (2003) 289–304.
[60] G.F. Bickerstaff, Immobilization of enzymes and cells: some practical consider-
ations, in: G.F. Bickerstaff (Ed.), Immobilization of Enzymes and Cells, Humana
Press Inc., Totowa, 1997, pp. 1–11.
[61] SPRINTechnologies, 2011, http://www.sprintechnologies.com/sprin/index.
jsp?idPagina=11.
[18] A. Zaks, A.M. Klibanov, The Journal of Biological Chemistry 263 (1988)
3194–3201.
[19] D.B. Volkin, A. Staubli, R. Langer, A.M. Klibanov, Biotechnology and Bioengi-
neering 37 (1991) 843–853.
[20] P.J. Halling, Philosophical Transactions of the Royal Society B 359 (2004)
1287–1297.
[62] ChiralVision, 2011, http://www.chiralvision.com/immobead.htm.
[63] R.A. Sheldon, R. Schoevaart, L.M.v. Langen, Biocatalysis and Biotransformation
23 (2005) 141–147.
[64] O.S. Lin’kova, S.A. Zhukovskaya, N.S. Davidova, Pharmaceutical Chemistry Jour-
nal 10 (1976) 696–700.
[21] M. Paravidino, M.J. Sorgedrager, R.V.A. Orru, U. Hanefeld, Chemistry - A Euro-
pean Journal 16 (2010) 7596–7604.
[22] M. Ottesen, I. Svendsen, The subtilisins, in: L.L. Gertrude, E. Perlmann (Eds.),
Methods in Enzymology, Academic Press, 1970, pp. 199–215.
[23] R.J. DeLange, E.L. Smith, The Journal of Biological Chemistry 243 (1968)
2134–2142.
Petra Vossenberg graduated “Cum Laude” in Biotech-
nology with the specialization process technology at
Wageningen University. After obtaining her M.Sc., she
started her Ph.D. at the department of Bioprocess
Engineering at Wageningen University, investigating
chemo-enzymatic peptide synthesis.
[24] Sigma-Aldrich,
2012,
http://www.sigmaaldrich.com/life-science/
metabolomics/enzyme-explorer/analytical-enzymes/subtilisin.html.
[25] R. Gupta, Q.K. Beg, P. Lorenz, Applied Microbiology and Biotechnology 59 (2002)
15–32.
[26] H. Kise, A. Hayakawa, Enzyme and Microbial Technology 13 (1991) 584–588.
[27] Å. Jönsson, P. Adlercreutz, B. Mattiasson, Biotechnology and Bioengineering 46
(1995) 429–436.
[28] M.V. Sergeeva, V.M. Paradkar, J.S. Dordick, Enzyme and Microbial Technology
20 (1997) 623–628.
[29] P. Clapés, P. Adlercreutz, B. Mattiasson, Journal of Biotechnology 15 (1990)
323–338.
[30] P. Clapés, G. Valencia, P. Adlercreutz, Enzyme and Microbial Technology 14
(1992) 575–580.
[31] Y. Kimura, T. Yoshida, K. Muraya, K. Nakanishi, R. Matsuno, Agricultural and
Biological Chemistry 54 (1990) 1433–1440.
[32] M. Capellas, M.D. Benaiges, G. Caminal, G. González, J. López-Santín, P. Clapés,
Biocatalysis and Biotransformation 13 (1996) 165–178.
[33] A.O. Richards, I.S. Gill, E.N. Vulfson, Enzyme and Microbial Technology 15 (1993)
928–935.
[34] J.U. Klein, A. Prykhodzka, V. Cerovsky, Journal of Peptide Science 6 (2000)
541–549.
H.H. (Rik) Beeftink has an M.Sc. in Microbiology from
the University of Groningen and obtained his Ph.D. in
Mathematics and Natural Sciences at the department
of Chemical Engineering at the University of Amster-
dam. Before joining the Bioprocess Engineering group at
Wageningen University in 1989, he worked at the Uni-
versity of Utrecht (Department of Molecular Cell Biology)
and the Technical University of Delft (Department of Bio-
process Engineering). Main interests are the design of
bacterium- or enzyme-catalyzed processes, multi-phase
systems and engineering education.
[35] S.-y. Okazaki, M. Goto, S. Furusaki, Enzyme and Microbial Technology 26 (2000)
159–164.
[36] A. Ferjancic, A. Puigserver, H. Gaertner, Applied Microbiology and Biotechnol-
ogy 32 (1990) 651–657.
[37] A.V. Belyaeva, A.V. Bacheva, E.S. Oksenoit, E.N. Lysogorskaya, V.I. Lozinskii, I.Y.
Filippova, Russian Journal of Bioorganic Chemistry 31 (2005) 529–534.
[38] I.Y. Filippova, E.N. Lysogorskaya, Russian Journal of Bioorganic Chemistry 29
(2003) 496–501.
[39] K. Oyama, S. Nishimura, Y. Nonaka, K.-i. Kihara, T. Hashimoto, Journal of Organic
Chemistry 46 (1981) 5242–5244.
Timo Nuijens was born in 1981 in Utrecht, The
Netherlands. In 2007 he graduated in Drug Innovation at
the University of Utrecht. Currently he is working at DSM
in the Biocatalysis R&D Department within DSM Phar-
maceutical Products innovative synthesis B.V. as a Ph.D.
student (Geleen, The Netherlands). His current research is
focussed on the enzymatic synthesis and modification of
amino acids and peptides.
[40] K. Nakanishi, A. Takeuchi, R. Matsuno, Applied Microbiology and Biotechnology
32 (1990) 633–636.
[41] S.-T. Chen, S.-Y. Chen, K.-T. Wang, The Journal of Organic Chemistry 57 (1992)
6960–6965.
[42] S.-J. Li, J.-A. Wang, L. Xu, X.-Z. Zhang, J. Li, D. Suo, Preparative Biochemistry and
Biotechnology 38 (2008) 334–347.
[43] R.-Z. Hou, Y. Yang, G. Li, Y.-B. Huang, H. Wang, Y.-J. Liu, L. Xu, X.-Z. Zhang,
Biotechnology and Applied Biochemistry 44 (2006) 73–80.
[44] X.-z. Zhang, X. Wang, S. Chen, X. Fu, X. Wu, C. Li, Enzyme and Microbial Tech-
nology 19 (1996) 538–544.
[45] CLEATechnologies, 2010, http://www.cleatechnologies.com/PDF/AlcalaseCLEA.
pdf.
[46] ChiralVision, 2010, http://www.chiralvision.com/immozymes.htm#Proteases.
Peter Quaedflieg graduated from Eindhoven University
(The Netherlands) in Chemical Engineering in 1988 and
obtained a Ph.D. degree in Organic Chemistry with Prof.
Jacques van Boom at Leiden University (The Netherlands)
in 1993. In that year he joint DSM where he held vari-
ous positions in R&D and process development. Currently
he is Principal Scientist Route Scouting and Science Team
Leader Organic Chemistry at DSM Innovative Synthesis
in Geleen (The Netherlands). His research interests cover
the synthesis of chiral pharmaceutical intermediates and
peptides and the use of enzymes in organic and polymer
synthesis.
[47] J. Partridge, P.J. Halling, B.D. Moore, Chemical Communications
841–842.
7 (1998)
[48] B.D. Moore, J. Partridge, P.J. Halling, Very high activity biocatalysts for low-
water systems, in: E.N. Vulfson, P.J. Halling, H.L. Holland (Eds.), Enzymes in

P. Vossenberg et al. / Journal of Molecular Catalysis B: Enzymatic 78 (2012) 24–31
31
Martien Cohen Stuart graduated from Groningen State
University with a degree in polymer chemistry and a
minor in physical chemistry. In 1975, he joined Wagenin-
gen University, doing a Ph.D. in colloid science under
supervision of Gerard Fleer and Bert Bijsterbosch, inves-
tigating adsorption of flexible polymers at solid/liquid
interfaces. He obtained his degree in 1980 and then was
appointed assistant professor in the physical chemistry
Johannes (Hans) Tramper has academic degrees from the
Technical University Delft, NL (Ir. 1973), from Purdue Uni-
versity, USA (M.Sc. 1974), from Wageningen University,
NL (Ph.D. 1979) and an honorary degree from the Tech-
nical University of Lodz, PL (Doctor Honoris Causa, 2000).
He is one of the founders of the study Biotechnology at
Wageningen University, which started in 1990. In March
2007 he resigned from being head of the Bioprocess Engi-
neering chair, but remains in this chair group as professor
with special attention for educational renewal. Presently
he is also coordinator of the educational program within
the national BE-Basic research program.
&
colloid science group. Following a sabbatical/post-
doctoral year in the group of Pierre Gilles de Gennes in
Paris, where he worked on wetting and spreading experi-
ments, he became associate professor in Wageningen and
in 1996 he succeeded Hans Lyklema in the Chair of Phys-
ical Chemistry and Colloid Science at Wageningen University. His interests cover a
wide spectrum, from statics and dynamics of macromolecules, interfaces, wetting,
phase behaviour, self-assembly, networks and gels, emulsions and coalescence, and
the physics of macromolecules in living cells. He is a member of the Royal Dutch
Academy of Sciences (KNAW).