Systematic assessment of the stability of benzaldehyde lyase in aqueous-organic biphasic systems and its stabilization by modification with methoxy-poly(ethylene) glycol

Article abstract of DOI:10.1016/j.molcatb.2010.08.007

Benzaldehyde lyase from Pseudomonas fluorescens Biovar I [BAL; E.C.4.1.2.38] catalyzes the stereoselective formation of C-C bonds coupling aldehydes to generate alpha-hydroxy ketones. A broad range of poorly water-soluble substrates are accepted in forward and reverse reactions. In this study, the stability of BAL in aqueous-organic biphasic systems as promising reaction media was systematically investigated using methyl-tert-butylether, 2-octanone, and toluene as the organic phase. Surprisingly, a strong individual molecular toxicity of these water-immiscible solvents was observed along with the interfacial toxicity exerted by the aqueous-organic interfaces. They could be considerably reduced by covalent attachment of methoxy-poly(ethylene) glycol (mPEG₇₅₀ and mPEG₂₀₀₀) to the enzyme surface increasing the half-life by a factor of up to 18. However, under reactive conditions solvent effects were strongly superimposed by an additional deactivating effect, possibly caused by the aldehyde substrate, and no differences between unmodified and modified BAL were detectable. For technical application of the enzyme in aqueous-organic biphasic media additional strategies for stabilization will therefore be desirable.

Full text of DOI:10.1016/j.molcatb.2010.08.007

Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Systematic assessment of the stability of benzaldehyde lyase in aqueous–organic
biphasic systems and its stabilization by modification with
methoxy-poly(ethylene) glycol
Anne van den Wittenboera, Bernd Niemeijera, Sanjib Kumar Karmeeb, Marion B. Ansorge-Schumachera,c,∗
a Institute of Biotechnology, RWTH Aachen University, Worringerweg 1, D-52074 Aachen, Germany
b Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, Worringerweg 1, D-52074 Aachen, Germany
c Institute of Chemistry, Department of Enzyme Technology, Technical University of Berlin, Str. des 17. Juni 124, D-10623 Berlin, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 20 March 2010
Received in revised form 7 August 2010
Accepted 12 August 2010
Available online 18 August 2010
Benzaldehyde lyase from Pseudomonas fluorescens Biovar I [BAL; E.C.4.1.2.38] catalyzes the stereoselec-
tive formation of C–C bonds coupling aldehydes to generate alpha-hydroxy ketones. A broad range of
poorly water-soluble substrates are accepted in forward and reverse reactions. In this study, the stability
of BAL in aqueous–organic biphasic systems as promising reaction media was systematically investi-
gated using methyl-tert-butylether, 2-octanone, and toluene as the organic phase. Surprisingly, a strong
individual molecular toxicity of these water-immiscible solvents was observed along with the interfa-
cial toxicity exerted by the aqueous–organic interfaces. They could be considerably reduced by covalent
attachment of methoxy-poly(ethylene) glycol (mPEG750 and mPEG2000) to the enzyme surface increas-
ing the half-life by a factor of up to 18. However, under reactive conditions solvent effects were strongly
superimposed by an additional deactivating effect, possibly caused by the aldehyde substrate, and no dif-
ferences between unmodified and modified BAL were detectable. For technical application of the enzyme
in aqueous–organic biphasic media additional strategies for stabilization will therefore be desirable.
© 2010 Elsevier B.V. All rights reserved.
Keywords:
Benzaldehyde lyase
Organic solvent stability
Process stability
mPEG modification
1. Introduction
water-miscible co-solvents to a homogeneous aqueous medium
[10,11] or by introduction of water-immiscible solvents to form a
Benzaldehyde lyase from Pseudomonas fluorescens Biovar I [BAL;
E.C.4.1.2.38] catalyzes the stereoselective cleavage and formation
of C–C bonds on a broad range of ␣-hydroxy ketones and alde-
hydes, respectively [1–4]. This is of utmost synthetic interest as
chiral ␣-hydroxy ketones are important building blocks for various
pharmacologically active compounds such as antiallergics, antide-
pressants, fungicides, or anti-inflammatory drugs [5,6], but access
is still limited due to low yields and enantiomeric excess of alter-
native chemical routes [7–9].
biphasic system [12] in which the enzymes are dissolved in the
aqueous phase, while substrates are provided and products are
extracted via the solvent phase [13]. The use of biphasic systems
provides a couple of advantages over the use of homogeneous
media such as the constant supply of substrates [14], the increase of
product yields [15–17] and the facilitation of downstream process-
ing [17]. It is even envisaged, that biphasic systems will in future
be the clue to the direct integration of enzyme catalyzed reactions
in existing chemical processes [18]. Nevertheless, the transfer of
biphasic media to enzyme catalyzed reactions is still limited [13]
because of their often detrimental effects on the catalytic activity
[19,20]. It is generally assumed that deactivation in these systems
mainly results from the interaction of the protein structure with
the interface (‘interfacial toxicity’) [14], whereas deactivation by
direct interactions between enzyme and solvent molecules in the
aqueous phase (‘molecular toxicity’) only plays a minor role [21,22].
Many reports have already mentioned the response of BAL
Most substrates of BAL are characterized by a very low solubil-
ity in water, i.e. organic solvents have to be employed to achieve
relevant concentrations. This can either be done by addition of
Abbreviations: BAL, benzaldehyde lyase; DMBA, 3,5-dimethoxybenzaldehyde;
FA, 2-furaldehyde; FO, 2,2ꢀ-furoin; mPEG, methoxy-poly(ethylene) glycol;
MTBE, methyl-tert-butylether; ThDP, thiamin diphosphate; TMB, 3,3ꢀ,5,5ꢀ-
tetramethoxybenzoin; TNBS, trinitrobenzene sulfonate.
stability to aqueous–organic biphasic systems [11,12], but
a
∗
Corresponding author at: Technical University of Berlin, Institute of Chemistry,
systematic investigation has not been performed to date. Conse-
quently, a rational approach to the stabilization of this versatile
catalyst has not been possible yet, though it has been found in pre-
vious works that an improvement of stability can be achieved by
Department of Enzyme Technology (TC4), Str. des 17. Juni 124, D-10623 Berlin,
Germany. Tel.: +49 30 314 22127; fax: +49 30 314 22261.
E-mail addresses: m.ansorge@chem.tu-berlin.de, m.ansorge@biotec.rwth-
aachen.de (M.B. Ansorge-Schumacher).
1381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.08.007

A. van den Wittenboer et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
209
the entrapment of the enzyme in an aqueous gel matrix [23–25].
This indicates that a direct contact between the enzyme and the
interface must indeed have an impact on stability.
was performed using an analytical SFC (Jasco, USA) equipped with
a Chiralpak OD-H column (J.T. Baker, Deventer, Netherlands) with
supercritical CO2 (scCO2) as mobile phase. Details of analysis:
pressure 120 bar, eluent scCO2:isopropanol (5:1), flow 3 mL/min,
temperature 40 ◦C, UV-detector 220 nm, typical retention times:
(R)-TMB 4.4 min, (S)-TMB 6.0 min.
In this study the stability of BAL in aqueous–organic biphasic
reaction systems was for the first time systematically investigated
taking into account the lifetime in reactive and non-reactive bipha-
sic systems, and upon incubation in solvent-saturated aqueous
phases. For optimal reproducibility and comparability an exper-
imental set-up was developed that enabled the measuring of
all three systems with high efficiency and minimum variation.
Finally, a stabilization of the enzyme by the covalent attach-
ment of methoxy-poly(ethylene) glycol (mPEG) to accessible lysine
residues on the protein surface was performed and evaluated with
regard to enhanced stability in the different systems.
2.3. Determination of water solubility of solvents
Solvent concentrations within the aqueous phase were deter-
mined from solvent-saturated buffer samples via GC-analysis using
a HP5890 Series II (Agilent, Germany) equipped with a FS-FFAP-CB-
0.5 column (CS-Chromatography, Germany). 1-Butanol was used as
internal standard and was added after sampling. Details of analysis:
N
2 0.25 bar; injector/detector (FID) 200 ◦C; 45 ◦C for 3.5 min, 200 ◦C
2. Experimental
(20 K/min); typical retention times: methyl-tert-butylether (MTBE)
2.1 min, toluene 3.7 min, 1-butanol 6.6 min, 2-octanone 8.2 min.
2.1. Chemicals and enzyme
2.4. Stability measurement under non-reactive conditions
DMBA was purchased from Alfa Aesar (Karlsruhe, Germany), all
other reagents and solvents were from Sigma–Aldrich (Steinheim,
Germany), of highest available purity and used without further
purification. BAL was expressed in Escherichia coli SG13009 pRep4
(Pharmacia, Sweden) from expression vector pBAL-his6 [26]. Cell
disruption was done by sonication of a cell solution (4 mL lysis
buffer per g cell wet weight; 0.05 M potassium phosphate, pH
8, 40 mM imidazole, 300 mM NaCl). Enzyme was purified from
crude extract supernatant via immobilized metal-ion affinity chro-
matography on Ni-Sepharose (GE Healthcare, 25 mL matrix, flow:
1 mL/min). Elution of enzyme from affinity matrix was achieved by
increasing imidazole concentration from 40 to 250 mM (in 0.05 M
potassium phosphate, pH 8, 300 mM NaCl). Eluted protein was
desalted using a G25-desalting column (GE Healthcare; 0.05 M
potassium phosphate buffer, pH 8, 10 mM NaCl), lyophilized and
stored at −20 ◦C until use.
Residual activities upon incubation in solvent-saturated
monophasic systems and respective biphasic systems were mea-
sured using the standard activity measurements as mentioned
above. BAL variants were incubated in the different mono- and
biphasic systems at 25 ◦C in potassium phosphate buffer (0.05 M;
pH 8; 0.25 mM ThDP; 2.5 mM MgSO4). To prevent evaporation
of solvents from the saturated monophasic systems, incubation
was done in sealed glass vessels. To avoid dilution of protein due
to solvent partition, buffer and solvents were saturated with the
respective other prior to use. Saturation was performed in a sep-
arating funnel, separated phases were stored in sealed, unshaken
vessels.
2.5. Stability measurements under reactive conditions
2.2. Determination of activity and stereoselectivity
Reactions were run continuously with respect to the organic
phase at 25 ◦C. Both phases had a volume of 5 mL and were sat-
urated with the respective other prior to use. Solvent-saturated
aqueous phase contained the enzyme and cofactors (0.05 M potas-
sium phosphate buffer; pH 8; 2.5 mM MgSO4; 0.25 mM ThDP).
Substrate (100 mM FA) was supplied via the continuously fed
buffer-saturated organic phase using a dosing pump (BF414S32,
Telab, Germany). Flow of the organic phase through the system
was accomplished by an overflow mechanism (pressurized with
solvent-saturated nitrogen). Further details on the reaction set-up
are given in [27]. Samples were taken from the organic phase out-
let and analyzed via GC using a HP5890 Series II (Agilent, Germany)
equipped with a FS-Supreme-5 column (CS-Chromatography, Ger-
many) and dodecane as internal standard. Details of analysis:
N2 0.35 bar; injector/detector (FID) 275/290 ◦C; 60 ◦C for 2.2 min,
280 ◦C (30 K/min); typical retention times: FA 2.7 min, FO 7.6 min,
dodecane 5.8 min.
All activity measurements were performed in potassium phos-
phate buffer (0.05 M; pH 8; 0.25 mM ThDP; 2.5 mM MgSO4;
25% (v/v) dimethylformamid (DMF) was added to reactions
with 3,5-dimethoxybenzaldehyde (DMBA)) using either UV–vis-
or fluorescence-spectrometry at 25 ◦C. Substrate concentrations
were 2 mM 2-furaldehyde (FA) and 2.5 mM DMBA, respectively.
UV–vis spectrometry was performed using a Varian Cary 50
(Varian, USA) with UV-Quartz cuvettes of 0.5 mm path length
for measurement of both model reactions: (1) 2-furaldehyde
to 2,2ꢀ-furoin (FO) (277 nm) and (2) 3,5-dimethoxybenzaldehyde
to 3,3ꢀ,5,5ꢀ-tetramethoxybenzoin (TMB) (325 nm). Concentra-
tions of all reactants were calculated from the extinction data
using Lambert–Beer’s law with extinction coefficients determined
beforehand (εFA = 14.883 mM−1 cm−1
; ;
εFO = 12.484 mM−1 cm−1
εDMBA = 2.434 mM−1 cm−1; εTMB = 2.26 mM−1 cm−1) and the mass
balance of the reaction. Reaction of DMBA to TMB was also
measured using fluorescence spectrometry (LS 55 Fluorescence
Spectrometer, Perkin Elmer, USA) with an excitation wavelength of
360 nm and an emission wavelength of 470 nm. Determination of
residual activities was performed as a relative measure compared
to starting values.
2.6. Determination of free amino groups (TNBS reaction)
Free amino groups were detected by reaction with 2,4,6-
trinitrobenzene (TNBS) sulfonate according to Snyder and Sobocin-
ski [28]. Samples were dissolved and diluted to give different
protein concentrations in 0.1 M sodium borate buffer pH 9.3. The
reaction was started by addition of 15 ␮L 0.03 M TNBS-solution
(0.1 M sodium borate buffer pH 9.3) to 600 ␮L of these protein
samples with and without addition of 10 mM 2-furaldehyde and
followed spectrophotometrically at 420 nm.
Enantioselectivity was determined in batch reactions (2 mL) at
25 ◦C in potassium phosphate buffer (0.05 M; pH 8; 0.25 mM ThDP;
2.5 mM MgSO4; 25% DMF) with DMBA (10 mM) as substrate. After
reaction samples were taken and extracted with an equal volume
of dichloromethane. The organic phase was dried with MgSO4 and
analyzed via supercritical fluid chromatography (SFC). SFC-analysis

210
A. van den Wittenboer et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
Table 1
Physicochemical properties of methyl-tert-butylether (MTBE), 2-octanone, and toluene.
Solvent
Structure
log P [–]
Interfacial tension (mN/m)
10.5c
Solubility in water (mol/L)a
0.668 0.153
MTBE
0.94b
2-Octanone
2.37b
2.73b
14.1c
36.1c
0.006 0.0007
0.003 0.0012
Toluene
a
Determined as described under Section 2.3.
Lide [31].
Freitas et al. [32] (in pure water systems).
b
c
2.7. Chemical modification of BAL with methoxy-poly(ethylene)
glycol (mPEG)
area in this reactor was kept constant and independent stirring of
both phases ensured optimum mixing and mass transfer. The evap-
oration of volatile organic solvents was prevented by use of airtight
constructions; operation was equally possible in continuous and
batch mode.
Mixing, reaction temperature, pH and buffer type and strength
were identical in all experiments. Methyl-tert-butylether (MTBE),
2-octanone, and toluene, respectively, were selected to form the
organic phase because of their suitable solvation capacity for sub-
strates and products and their different polarity, hydrophobicity
and functionality. The relevant physicochemical properties of these
solvents are summarized in Table 1.
Activation of mPEG derivatives was performed following the
procedure of Hernáiz et al. [29] using p-nitrophenylchloroformate
as activating agent. The final precipitate was dried under vacuum
and stored at −20 ◦C. The yield of precipitated mPEG in respect
to the initially applied mPEG amount was >95%. The activation
degree of the purified mPEG derivatives (>99%) was determined
spectrophotometrically by measuring p-nitrophenol (ꢀ = 400 nm)
released upon alkaline hydrolysis. p-nitrophenyl-mPEG2000
-
carbonate 1H NMR (CDCl3, 300 MHz): ı (ppm) = 8.32–8.24 (m, 2H,
Ar-H), 7.43–7.35 (m, 2H, Ar-H), 3.9–3.4 (m, all other H), 3.36 (s,
3H, OMe); p-nitrophenyl-mPEG750-carbonate 1H NMR (CDCl3,
300 MHz): ı (ppm) = 8.32–8.24 (m, 2H, Ar-H), 7.43–7.35 (m, 2H,
Ar-H), 3.9–3.4 (m, all other H), 3.36 (s, 3H, OMe).
Enzyme activity and stereoselectivity were measured with
regard to the carboligation of either 2-furaldehyde (Fig. 1A) or 3,5-
dimethoxybenzaldehyde (Fig. 1B).
Modification of BAL with mPEG was performed for up to 24 h at
25 ◦C in potassium phosphate buffer (0.05 M; pH 9; 0.25 mM ThDP;
2.5 mM MgSO4) with a 500-fold molar excess of activated mPEG750
or mPEG2000, respectively. Modified protein was purified from the
reaction mixture with desalting columns (PD-10, GE Healthcare)
allowing simultaneous buffer exchange (0.05 M; pH 8; 0.25 mM
ThDP; 2.5 mM MgSO4). Samples of purified mPEG-variants were
applied to activity measurements, SDS-PAGE and size exclusion
chromatography for characterization.
3.2. Stability of unmodified BAL under non-reactive conditions
The half-life of BAL in non-reactive systems was determined by
measuring the residual carboligation activity after different times
of incubation in the respective mono- and biphasic systems in the
absence of substrate. The corresponding deactivation rates and
half-lives were estimated assuming a time-dependent first order
rate law.
The half-life of BAL in the presence of the investigated sol-
vents, but without formation of distinct solvent phases was found
to be 17.3 h ( 6.3), 5.6 h ( 2.5) and 3.2 h ( 0.8) for saturation
with MTBE, 2-octanone, and toluene, respectively (Table 2). Control
experiments in pure aqueous phase without solvent confirmed that
deactivation in such systems did not result from stirring and/or con-
tact with the aqueous–gaseous interface (half-life 82.1 6.8). In the
respective non-reactive biphasic systems half-lives of 14.9 h ( 2.9),
0.6 h ( 0.5) and 0.2 h ( 0.1) were calculated for MTBE, 2-octanone
and toluene as organic phase (Table 2).
2.8. SDS-PAGE and size exclusion chromatography of mPEG-BAL
Denaturating polyacrylamide gel electrophoresis was per-
formed according to Lämmli [30]. Separating gels had
a
concentration of 10% (w/v) polyacrylamide, stacking gels a concen-
tration of 5% (w/v). Size exclusion chromatography was performed
on a fast protein liquid chromatography system equipped with
an UV-detector (Pharmacia, Sweden) using 240 mL Sephacryl HR-
300 matrix in a 102 cm × 0.8 cm (height/width) column at a flow
rate of 0.5 mL/min (0.1 M TEA/HCl pH 8; 0.15 M sodium chloride).
Exclusion volume was determined using Blue Dextran (2000 kDa)
and calibration was done with carboanhydrase from erythrocytes
(29 kDa), bovine serum albumin (66 kDa), alcohol dehydrogenase
from Saccharomyces cerevisiae (150 kDa) and apoferritin from horse
liver (443 kDa).
3.3. Stability of unmodified BAL under process conditions
The half-life of BAL under process conditions in biphasic media,
i.e. during reaction was determined under steady-state conditions
in continuous operation mode. Substrate and product were contin-
uously supplied and extracted, respectively, via the organic phase
while keeping the phase volume and the substrate concentration
constant; the enzyme was kept in the discontinuous aqueous phase.
The activity decrease of BAL in the presence of MTBE, 2-octanone
and toluene as organic phase is illustrated in Fig. 2. The correspond-
ing deactivation rates were estimated via non-linear regression
assuming a time-dependent first order rate law. With all three
solvents, MTBE, 2-octanone, and toluene, very similar half-lives of
1.5 h ( 0.1), 1.4 h ( 0.1) and 2.0 h ( 0.1), respectively, were found.
3. Results and discussion
3.1. Investigation system
As a general set-up for measuring stability a purpose-built
biphasic mini-reactor allowing the consumption of minimum
amounts of solvent and catalyst was chosen [27]. The interfacial

A. van den Wittenboer et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
211
Fig. 1. BAL-catalyzed carboligation of (A) 2-furaldehyde to (R)-2,2ꢀ-furoin and (B) 3,5-dimethoxybenzaldehyde to (R)-3,3ꢀ,5,5ꢀ-tetramethoxybenzoin.
Table 2
Stability of native BAL, mPEG750-BAL and mPEG2000-BAL in non-reactive mono- and biphasic as well as reactive biphasic systems with MTBE, 2-octanone and toluene (protein
concentration 100 ␮g/mL). The large deviations of the measurements can be explained by the inhomogeneous modification extent observed for the mPEG-BAL preparations
[42].
Solvent
System
Half-life (h)
Native BAL
82.1 6.8
mPEG750-BAL
–
mPEG2000-BAL
–
Without
MTBE
Non-reactive, monophasic
Non-reactive, monophasic
Non-reactive, biphasic
Reactive, biphasic
17.3 6.3
14.9 2.9
1.5 0.1
44.0 16.8
36.3 11.4
2.3 0.2
69.7 1.3
6.4 1.4
1.9 0.2
2-Octanone
Toluene
Non-reactive, monophasic
Non-reactive, biphasic
Reactive, biphasic
5.6 2.5
0.6 0.5
1.4 0.1
4.6 0.4
2.0 0.2
1.3 0.1
10.5 3.8
0.7 0.1
1.3 0.1
Non-reactive, monophasic
Non-reactive, biphasic
Reactive, biphasic
3.2 0.8
0.2 0.1
2.0 0.1
17.3 2.6
0.5 0.1
2.1 0.2
57.8 7.4
3.4 0.3
1.8 0.2
This was a decrease of at least 80 h compared to the half-life of the
catalyst in non-reactive, purely aqueous medium (see Table 2).
ing effect on BAL, which demonstrates that the molecular toxicity of
hydrophobic solvents cannot be neglected, as is generally assumed
[21,22]. Among the investigated solvents the lowest impact on
stability was not even correlated with the lowest solvent concen-
tration in water, as the best stability was found for MTBE, which
has a considerably higher solubility in water than 2-octanone and
toluene.
In the respective non-reactive biphasic systems, which com-
bine the molecular effects of solvent molecules within the aqueous
phase and interactions at the interface between aqueous and
organic phase, only minor additional impact was found for the
MTBE-interface, whereas the half-life of BAL was again strongly
affected by the interfaces towards 2-octanone and toluene. This is
partly in accordance with the findings of Ghatorae et al. [19], who
related the degree of interfacial deactivation in biphasic systems
to the interfacial tension. In the investigated systems, the interface
3.4. Conclusions on the stability of unmodified BAL
An overview of the results of all stability measurements, includ-
ing the half-life of BAL in stirred aqueous phases without addition
of any solvent or substrate, is given in Table 2. They clearly demon-
strate two things: 1. Any of the investigated solvents, whether
dissolved or forming a second phase, has a strong negative effect
on the enzyme stability; half-life decreases by a minimum of 65 h
(∼79%). 2. The enzyme stability is not only, not even mainly,
affected by the formation of interfaces, but reacts on several inter-
related influences.
Even very small concentrations of dissolved solvent molecules
in the monophasic systems (see Table 1) exert a strong deactivat-
Fig. 2. Formation of (R)-2,2ꢀ-furoin in continuous reactions using BAL in biphasic systems with MTBE, 2-octanone and toluene as respective organic phases.

212
A. van den Wittenboer et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
The chemical modification of proteins by the attachment of
poly(ethylene) glycol (PEG) and derivatives such as methoxy-
poly(ethylene) glycol (mPEG) to free amino functions is a well
established method, primarily for the increase of solubility in var-
ious fluids [38–41]. Applied to BAL, the initial activity decreased
to approximately 60%, when mPEG750 was employed, and to
70%, when a longer-chain mPEG2000 was used; enantioselectivity
was unaffected (e.e.P > 99%). SDS-PAGE analysis revealed that all
enzyme molecules were successfully modified. The modification
extent varied between 2 and 4 mPEG molecules per BAL subunit,
i.e. 8–16 mPEG molecules were attached to each BAL-tetramer. The
tetrameric structure was not affected as verified by size exclusion
chromatography.
Interestingly, the half-life of modified BAL in the investigated
systems changed rather incoherently; in parts a considerable
increase was observed (Table 2).
Stability against dissolved solvent molecules in the aqueous
phase, especially against MTBE and toluene, was clearly enhanced;
Fig. 3. Course of TNBS reactions with BAL at two different protein concentrations
in the absence and presence of 2-furaldehyde (FA): (—) 0.8 mg/mL BAL without
the stabilization was higher with mPEG2000 compared to mPEG750
.
It has been described that mPEG forms folded domain struc-
tures when interacting with hydrophobic surface regions [43,33].
Such structures could prevent interactions between the solvent
molecules and hydrophobic surface regions and thus account for
stabilization. They also increase the rigidity of the protein molecule
thereby hindering the unfolding events underlying deactivation
[42,44–47]. The different effects of the two mPEG species at the
same modification extent could then be explained by the larger
surface shielding and higher rigidity implied by the longer-chain
mPEG.
In the corresponding biphasic systems stability of mPEG-BAL
variants varied with mPEG species as well as solvents. This could
be explained by two competing effects resulting from mPEG mod-
ification: enhanced adsorption onto the interface on the one hand
and reduced structural rearrangements on the other hand. Both
would be stronger with the longer-chain mPEG2000 as explained in
the preceding paragraph. In the biphasic system with toluene the
increase in adsorption might not be very pronounced since it can be
assumed that the unmodified protein is already adsorbed to a great
extent. Consequently, the major effect of the modification would be
the shielding against structural rearrangements, where the modi-
fication with the larger mPEG2000 has the more beneficial effect.
In contrast, increased adsorption to the interface probably plays
a bigger role in the biphasic systems with MTBE and 2-octanone,
FA; (– –) 0.8 mg/mL BAL with 10 mM FA; (
FA; ( ) 0.3 mg/mL BAL with 10 mM FA. Reactions were followed at a
wavelength of 420 nm.
) 0.3 mg/mL BAL without
towards toluene exhibits the highest interfacial tension and also
results in the highest deactivation of BAL, while the least interfa-
cial effect was measured for MTBE, which also displays the lowest
interfacial tension. The results for 2-octanone, however, are not
consistent with this theory, which indicates that there must be
more influencing factors than the interfacial tension and it also
demonstrates that the evaluation of solvent effects is not easily
accessed.
Finally, the findings for BAL stability in reactive biphasic sys-
tems suggest that this enzyme is not only affected by the solvent
system, but also by its substrates. This influence interferes with
the observed solvent effects and can even superimpose. It has been
described in literature, that aldehydes can form Schiff bases with
amino functions of proteins thereby leading to structural rear-
rangements [34–36]. These Schiff bases are usually not stable, but
might induce irreversible unfolding of the protein and thus lead
to sustainable deactivation. An interaction of 2-furaldehyde with
free amino functions of BAL was indeed observed, i.e. when amino
groups of BAL were reacted with trinitrobenzene sulfonate (TNBS)
[28] in the absence and presence of 2-furaldehyde, the presence of
2-furaldehyde clearly decelerated the reaction with TNBS (Fig. 3).
Formation of Schiff bases with a subsequent denaturation of BAL
might therefore account for the observed deactivation of BAL in
reactive systems.
which leads to an unchanged or decreased stability of mPEG2000
-
BAL variants in these media. Overall, the modification with mPEG
led to good stabilization of BAL under non-reactive conditions,
particularly against deactivation by dissolved solvent molecules.
Nevertheless, the additional deactivation by the interface in the
respective biphasic systems was still high. The strongest effect was
achieved against toluene, where the stability of mPEG2000-BAL sur-
passed the stability of unmodified BAL by factor 18. In contrast,
the modification with mPEG750 and mPEG2000 had no pronounced
effect on BAL stability in reactive biphasic systems, which indicates
that amino functions susceptible for or affected by the formation
of Schiff bases were not modified by reaction with mPEG.
3.5. Stabilization of BAL by mPEG modification
In the respective literature it is generally assumed that enzyme
deactivation by dissolved solvent molecules results from the dehy-
dration of the protein surface and the concomitant break-down of
non-covalent interactions that are essential for maintenance of the
correct three-dimensional structure [37], while aqueous–organic
interfaces promote the adsorption of protein with concomi-
tant structural rearrangements [19,20]. Consequently, stabilization
against both effects should be possible by reducing the surface area
susceptible for the interaction with dissolved solvent molecules
and/or upon adsorption at the interface. On the other hand, block-
ing of accessible lysine residues on the surface of the protein should
reduce the formation of Schiff bases with aldehydes, i.e. stabi-
lize against the supposed detrimental effects of such reactants.
In order to reduce both susceptible surface area and formation of
Schiff bases, methoxy-poly(ethylene) glycol (mPEG) was covalently
bound to BAL.
4. Conclusion
In aqueous–organic biphasic systems the stability of benzalde-
hyde lyase is affected by solvent molecules present in the aqueous
phase (“molecular toxicity”), aqueous–organic interfaces (“inter-
facial toxicity”), and substrate molecules. Their effects strongly
overlap. Molecular toxicity and interfacial toxicity mainly depend
on the solvent system and can be reduced by covalent binding of
mPEG to the protein surface. However, under reactive conditions

A. van den Wittenboer et al. / Journal of Molecular Catalysis B: Enzymatic 67 (2010) 208–213
213
both effects are superimposed by another deactivating effect, most
likely by the aldehyde substrate. Stabilization of BAL for technical
application will therefore require the development of additional
strategies to prevent this deactivation.
[21] V.V. Mozhaev, Y.L. Khmelnitsky, M.V. Sergeeva, A.B. Belova, N.L. Klyachko, A.V.
Levashov, K. Martinek, Eur. J. Biochem. 184 (1989) 597–602.
[22] Y.L. Khmelnitsky, A.B. Belova, A.V. Levashov, V.V. Mozhaev, FEBS Lett. 284
(1991) 267–269.
[23] T. Hischer, S. Steinsiek, M.B. Ansorge-Schumacher, Biocatal. Biotransform. 24
(2006) 437–442.
[24] M.B. Ansorge-Schumacher, L. Greiner, F. Schroeper, S. Mirtschin, T. Hischer,
Biotechnol. J. 1 (2006) 564–568.
Acknowledgements
[25] M.B. Ansorge-Schumacher, Mini Rev. Org. Chem. 4 (2007) 243–245.
[26] H. Iding, T. Dünnwald, L. Greiner, A. Liese, M. Müller, P. Siegert, P. Grötzinger,
A.S. Demir, M. Pohl, Chem. Eur. J. 6 (2000) 1483–1495.
[27] A. van den Wittenboer, T. Schmidt, P. Müller, M.B. Ansorge-Schumacher, L.
Greiner, Biotechnol. J. 4 (2009) 44–50.
We gratefully acknowledge the financial support by the German
Research Foundation DFG within the DFG research training group
GRK 1166 (Biocatalysis in Non-Conventional Media–BioNoCo).
[28] S.L. Snyder, P.Z. Sobocinski, Anal. Biochem. 64 (1975) 284–288.
[29] M.J. Hernáiz, J.M. Sánchez-Montero, J.V. Sinisterra, Biotechnol. Bioeng. 55
(1997) 252–260.
References
[30] U.K. Lämmli, Nature 227 (1970) 680–685.
[31] D.R. Lide (Ed.), CRC Handbook of Chemistry and Physics, 89th ed., CRC Press,
Taylor & Francis Group, Boca Raton, 2008.
[32] A.A. Freitas, F.H. Quina, F.A. Carroll, J. Phys. Chem. B 101 (1997) 7488–7493.
[33] M.V. Filho, T. Stillger, M. Müller, A. Liese, C. Wandrey, Angew. Chem. Int. Ed.
Engl. 42 (2003) 2993–2996.
[34] M. Biselli, U. Kragl, C. Wandrey, in: K. Drauz, H. Waldmann (Eds.), Enzyme
Catalysis in Organic Synthesis, VCH-Wiley, Weinheim, 2002.
[35] M. Di Lorenzo, A. Hidalgo, R. Molina, J.A. Hermoso, D. Pirozzi, U.T. Bornscheuer,
Appl. Environ. Microbiol. 73 (2007) 7291–7299.
[36] H.K. Weber, J. Zuegg, K. Faber, J. Pleiss, J. Mol. Catal. B: Enzym. 3 (1997)
131–138.
[37] Y.L. Khmelnitsky, V.V. Mozhaev, A.B. Belova, M.V. Sergeeva, K. Martinek, Eur. J.
Biochem. 198 (1991) 31–41.
[38] H. Ogino, H. Ishikawa, J. Biosci. Bioeng. 91 (2001) 109–116.
[39] G. DeSantis, J.B. Jones, Curr. Opin. Biotechnol. 10 (1999) 324–330.
[40] Q. Jene, J.C. Pearson, C.R. Lowe, Enzyme Microb. Technol. 20 (1997) 69–74.
[41] Y. Inada, K. Takahashi, T. Yoshimoto, A. Ajima, A. Matsushima, Y. Saito, Trends
Biotechnol. 4 (1986) 190–194.
[42] J.A. Rodriguez-Martinez, R.J. Sola, B. Castillo, H.R. Cintron-Colon, I.
Rivera-Rivera, G. Barletta, K. Griebenow, Biotechnol. Bioeng. 101 (2008)
1142–1149.
[43] B.N. Manjula, A. Tsai, R. Upadhya, K. Perumalsamy, P.K. Smith, A. Malavalli, K.
Vandegriff, R.M. Winslow, M. Intaglietta, M. Prabhakaran, J.M. Friedman, A.S.
Acharya, Bioconjug. Chem. 14 (2003) 464–472.
[44] D.I. Svergun, F. Ekstrom, K.D. Vandegriff, A. Malavalli, D.A. Baker, C. Nilsson,
R.M. Winslow, Biophys. J. 94 (2008) 173–181.
[45] H. Garcia-Arellano, B. Valderrama, G. Saab-Rincon, R. Vazquez-Duhalt, Biocon-
jug. Chem. 13 (2002) 1336–1344.
[46] A.L. Soares, G.M. Guirmaraes, B. Polakiewicz, R.N.D. Pitombo, J. Abrahao-Neto,
Int. J. Pharm. 237 (2002) 163–170.
[1] A.S. Demir, Ö. Sesenoglu, E. Eren, B. Hosrik, M. Pohl, E. Janzen, D. Kolter, R.
Feldmann, P. Dünkelmann, M. Müller, Adv. Synth. Catal. 344 (2002) 96–103.
[2] A.S. Demir, Ö. Sesenoglu, P. Dünkelmann, M. Müller, Org. Lett.
2047–2050.
5 (2003)
[3] T. Hischer, D. Gocke, M. Fernández, P. Hoyos, A.R. Alcántara, J.V. Sinisterra, W.
Hartmeier, M.B. Ansorge-Schumacher, Tetrahedron 61 (2005) 7378–7383.
[4] P. Domínguez de María, M. Pohl, D. Gocke, H. Gröger, H. Trauthwein, T. Stillger,
L. Walter, M. Müller, Eur. J. Org. Chem. (2007) 2940–2944.
[5] A.S. Demir, H. Hamamci, O. Sesenoglu, F. Aydogan, D. Capanoglu, R. Nesli-
hanoglu, Tetrahedron Asymmetry 12 (2001) 1953–1956.
[6] U. Schörken, G.A. Sprenger, Biochim. Biophys. Acta 1385 (1998) 229–243.
[7] D. Enders, K. Breuer, J.H. Teles, Helv. Chim. Acta 79 (1996) 1217–1221.
[8] C.A. Dvorak, V.H. Rawal, Tetrahedron Lett. 39 (1998) 2925–2928.
[9] J. Pesch, K. Harms, T. Bach, Eur. J. Org. Chem. (2004) 2025–2035.
[10] N. Nemeria, L. Korotchkina, M.J. McLeish, G.L. Kenyon, M.S. Patel, F. Jordan,
Biochemistry 46 (2007) 10739–10744.
[11] P. Domínguez de María, T. Stillger, M. Pohl, S. Wallert, K. Drauz, H. Gröger, H.
Trauthwein, A. Liese, J. Mol. Catal. B: Enzym. 38 (2006) 43–47.
[12] T. Stillger, Doctoral thesis, University of Bonn, 2004.
[13] P. Fernandes, J.M. Cabral, in: G. Carrea, S. Riva (Eds.), Organic Synthesis with
Enzymes in Non-Aqueous Media, VCH-Wiley, Weinheim, 2008.
[14] M.H. Vermue, J. Tramper, Pure Appl. Chem. 67 (1995) 345–373.
[15] K. Martinek, A.N. Semenov, I.V. Berezin, Biochim. Biophys. Acta 658 (1981)
76–89.
[16] D.K. Eggers, H.W. Blanch, J.M. Prausnitz, Enzyme Microb. Technol. 11 (1989)
84–89.
[17] M.F. Eckstein, M. Peters, J. Lembrecht, A.C. Spiess, L. Greiner, Adv. Synth. Catal.
348 (2006) 1591–1596.
[18] D. Pollard, B. Kosjek, in: G. Carrea, S. Riva (Eds.), Organic Synthesis with Enzymes
in Non-Aqueous Media, VCH-Wiley, Weinheim, 2008.
[19] A.S. Ghatorae, M.J. Guerra, G. Bell, P.J. Halling, Biotechnol. Bioeng. 44 (1994)
1355–1361.
[47] J.I. Lopez-Cruz, G. Viniegra-Gonzalez, A. Hernandez-Arana, Bioconjug. Chem.
17 (2006) 1093–1098.
[20] A.C. Ross, G. Bell, P.J. Halling, J. Mol. Catal. B: Enzym. 8 (2000) 183–192.