Production of galacto-oligosaccharides by the β-galactosidase from kluyveromyces lactis: Comparative analysis of permeabilized cells versus soluble enzyme

Article abstract of DOI:10.1021/jf2022012

The transgalactosylation activity of Kluyveromyces lactis cells was studied in detail. Cells were permeabilized with ethanol and further lyophilized to facilitate the transit of substrates and products. The resulting biocatalyst was assayed for the synthesis of galacto-oligosaccharides (GOS) and compared with two soluble β-galactosidases from K. lactis (Lactozym 3000 L HP G and Maxilact LGX 5000). Using 400 g/L lactose, the maximum GOS yield, measured by HPAEC-PAD analysis, was 177 g/L (44% w/w of total carbohydrates). The major products synthesized were the disaccharides 6-galactobiose [Gal-β(1?6)-Gal] and allolactose [Gal-β(1?6)-Glc], as well as the trisaccharide 6-galactosyl-lactose [Gal-β(1?6)-Gal-β(1?4)-Glc], which was characterized by MS and 2D NMR. Structural characterization of another synthesized disaccharide, Gal-β(1?3)-Glc, was carried out. GOS yield obtained with soluble β-galactosidases was slightly lower (160 g/L for Lactozym 3000 L HP G and 154 g/L for Maxilact LGX 5000); however, the typical profile ith a maximum GOS concentration followed by partial hydrolysis of the newly formed oligosaccharides was not observed with the soluble enzymes. Results were correlated with the higher stability of β-galactosidase when permeabilized whole cells were used.

Full text of DOI:10.1021/jf2022012

ARTICLE
pubs.acs.org/JAFC
Production of Galacto-oligosaccharides by the β-Galactosidase from
Kluyveromyces lactis: Comparative Analysis of Permeabilized Cells
versus Soluble Enzyme
Barbara Rodriguez-Colinas,† Miguel A. de Abreu,‡ Lucia Fernandez-Arrojo,† Roseri de Beer,†,4 Ana Poveda,#
Jesus Jimenez-Barbero,^ Dietmar Haltrich,X Antonio O. Ballesteros Olmo,† Maria Fernandez-Lobato,‡ and
Francisco J. Plou*,†
†Departamento de Biocatꢀalisis, Instituto de Catꢀalisis y Petroleoquímica, CSIC, 28049 Madrid, Spain
‡Centro de Biología Molecular Severo Ochoa, Departamento de Biología Molecular (CSIC-UAM), Universidad Autꢀonoma de Madrid,
28049 Madrid, Spain
#Servicio Interdepartamental de Investigaciꢀon, Universidad Autꢀonoma de Madrid, 28049 Madrid, Spain
^Centro de Investigaciones Biolꢀogicas, CSIC, 28040 Madrid, Spain
XFood Biotechnology Laboratory, BOKU University of Natural Resources and Applied Life Sciences, Vienna, Austria
ABSTRACT: The transgalactosylation activity of Kluyveromyces lactis cells was studied in detail. Cells were permeabilized with
ethanol and further lyophilized to facilitate the transit of substrates and products. The resulting biocatalyst was assayed for the
synthesis of galacto-oligosaccharides (GOS) and compared with two soluble β-galactosidases from K. lactis (Lactozym 3000 L HP G
andMaxilact LGX 5000). Using 400 g/L lactose, the maximum GOS yield, measured by HPAEC-PAD analysis, was 177 g/L (44% w/w
of total carbohydrates). The major products synthesized were the disaccharides 6-galactobiose [Gal-β(1f6)-Gal] and allolactose
[Gal-β(1f6)-Glc], as well as the trisaccharide 6-galactosyl-lactose [Gal-β(1f6)-Gal-β(1f4)-Glc], which was characterized by
MS and 2D NMR. Structural characterization of another synthesized disaccharide, Gal-β(1f3)-Glc, was carried out. GOS yield
obtained with soluble β-galactosidases was slightly lower (160 g/L for Lactozym 3000 L HP G and 154 g/L for Maxilact LGX 5000);
however, the typical profile with a maximum GOS concentration followed by partial hydrolysis of the newly formed oligosaccharides
was not observed with the soluble enzymes. Results were correlated with the higher stability of β-galactosidase when permeabilized
whole cells were used.
KEYWORDS: glycosidase, galacto-oligosaccharides, prebiotics, transglycosylation, β-galactosidase, oligosaccharides
’ INTRODUCTION
The hydrolytic activity of K. lactis β-galactosidase is very high.19
The production of GOS using batch and continuous bioreactors
of K. lactis β-galactosidase has been reported.15,16 Several studies
on the immobilization of this enzyme with different carriers have
been also described.17,20,21
β-Galactosidases (β-D-galactoside galactohydrolases, EC
3.2.1.23) catalyze the hydrolysis of the galactosyl moiety from
nonreducing termini of oligosaccharides. These enzymes have
attracted attention from researchers and dairy product manu-
facturers primarily due to their ability to remove lactose from
milk.1 In addition, β-galactosidases are employed for oligosac-
charide synthesis,2,3 because the normal hydrolytic reaction of
glycosidases can be reversed toward glycosidic bond synthesis
under appropriate conditions.4À6 Thus, β-galactosidase catalyzes
transgalactosylation reactions in which lactose (as well as the re-
leased glucose and galactose) serve as galactosyl acceptors, yielding
a series of di, tri-, and tetrasaccharides (and eventually of higher
polymerization degree) called galacto-oligosaccharides (GOS).7,8
Depending on the source of β-galactosidase, the yield and compo-
sition of GOS vary notably.9À11 GOS constitute the major part of
oligosaccharides in human milk.12 Among other health benefits,
GOS are noncariogenic, exhibit prebiotic properties, reduce the
level of cholesterol in serum, and prevent colon cancer.13,14
The major commercial source of β-galactosidase by far is the
mesophile yeast Kluyveromyces lactis.15À17 Due to its intracellular
nature, the production of cell-free K. lactis β-galactosidase is limi-
ted by the high cost associated with enzyme extraction and down-
stream processing as well as the low stability of the enzyme.8,18
Biochemical reactions using whole cells have advantages over
purified enzymes in many industrial bioconversion processes,
allowing the reuse of the biocatalyst and continuous processing.
However, the permeability barrier of the cell envelope for sub-
strates and products often causes very low reaction rates, especially
in yeasts. To increase the volumetric activity, permeabilization
of yeast cells is an economical, simple, and safe process that usually
facilitates substrate access to the intracellular enzymes;22À24
the enzyme yields trisaccharides as the main transglycosylation
products.16,25,26 The permeabilizing agent may decrease the phos-
pholipid content in the membrane, thus allowing the transit of
low molecular weight compounds in and out of the cells.27
In this context, ethanol-permeabilized K. lactis cells have been
evaluated for lactose hydrolysis22,23,28,29 and for the bioconversion
Received: June 2, 2011
Revised:
September 1, 2011
Accepted: September 1, 2011
Published: September 02, 2011
r
2011 American Chemical Society
10477
dx.doi.org/10.1021/jf2022012 J. Agric. Food Chem. 2011, 59, 10477–10484
|

Journal of Agricultural and Food Chemistry
ARTICLE
of lactose and fructose to the disaccharide lactulose.30 However,
to our knowledge, the use of K. lactis permeabilized cells for the
synthesis of GOS has not been investigated in detail. In this
context, permeabilized cells of Kluyveromyces marxianus were
recently employed for GOS production, although the structure of
the synthesized products was not reported.27
In this work, we have studied the transgalactosylation activity
of ethanol-permeabilized cells of a strain of K. lactis. The syn-
thetic activity of the crude preparation was compared with that of
two soluble β-galactosidases from K. lactis (Lactozym 3000 L HP
G and Maxilact LGX 5000). Structural characterization of the
synthesized galacto-oligosaccharides was performed in detail.
’ EXPERIMENTAL PROCEDURES
Materials. The commercial β-galactosidases from K. lactis Lactozym
3000 L HP G (batch DKN08677) and Maxilact LGX 5000 (batch 409-
240001) were kindly supplied by Novozymes A/S and DSM Food
Specialties, respectively. As both preparations contained 50% (v/v) glycerol,
they were dialyzed prior to use in 0.1 M potassium phosphate buffer (pH
6.8). Glucose, galactose, lactose monohydrate, and o-nitrophenyl-β-D-
galactopyranoside (ONPG) were from Sigma-Aldrich. 6-Galactobiose,
4-galactobiose, 3-galactobiose, 6-O-β-galactosyl-glucose (allolactose),
and 4-O-β-galactosyl-lactose were from Carbosynth (Berkshire, U.K.).
All other reagents and solvents were of the highest available purity and
used as purchased.
Culture Conditions and Preparation of Permeabilized
Cells. The K. lactis CECT1931 strain was obtained from the Spanish
CETC (Colecciꢀon Espa~nola de Cultivos Tipo). Yeasts were grown on
YEPL [2% (w/v) lactose, 2% (w/v) bactopeptone, 1% (w/v) yeast
extract] at 26 ꢀC. Growth was monitored spectrophotometrically at a
wavelength of 660 nm (A660). For enzyme preparations, cells from 200 mL
cultures growing logarithmically at an A660 of 2.0À2.5 were harvested
(3000g during 5 min) and washed with distilled water. The cells were
permeabilized following the protocol described by Fontes et al.20 with
slight modifications. The cells were resuspended in 50% (v/v) ethanol,
stirred for 15 min at 4 ꢀC, washed with distilled water, and lyophilized. A
total protein extract was also obtained from the cell mass using Yeast
Buster Protein Extraction Reagent (Novagen) following the manufac-
turer's instructions.
Activity Assay. The enzymatic activity toward ONPG was mea-
sured at 40 ꢀC following o-nitrophenol release at 405 nm. For the per-
meabilized cells, the reaction was started by adding 1.5 mg of cell mass to
2 mL of 15 mM ONPG in 0.1 M potassium phosphate buffer (pH 6.8).
The mixture was incubated for 20 min with orbital shaking (600 rpm),
and the reaction was stopped by immersion in a water bath for 5 min at
95 ꢀC. For the soluble enzymes (Lactozym 3000 L HP G or Maxilact
LGX 5000), the reaction was started by adding 10 μL of the dialyzed
enzyme (conveniently diluted) to 190 μL of 15 mM ONPG in 0.1 M
potassium phosphate buffer (pH 6.8). The increase of absorbance at
405 nm was followed in continuous mode at 40 ꢀC during 5 min. In both
cases the absorbance was measured using a microplate reader (Versamax,
Molecular Devices). The molar extinction coefficient of o-nitrophenol
at pH 6.8 was determined (1627 MÀ1 cmÀ1). One unit (U) of activity
was defined as that corresponding to the hydrolysis of 1 μmol of ONPG
per minute.
Figure 1. SEM micrographs of permeabilized K. lactis cells at 1000Â
(top) and 2000Â (bottom).
(110 mg of permeabilized cell mass, 46 μL of Lactozym 3000 L HP G, or
14 μL of Maxilact LGX 5000) to adjust the β-galactosidase activity in the
reaction mixture in the range of 1.2À1.5 U/mL. The mixture was
incubated at 40 ꢀC in an orbital shaker (Vortemp 1550) at 200 rpm. At
different times, 750 μL aliquots were harvested from the reaction
mixture and incubated for 5 min at 95 ꢀC and 350 rpm in a Thermomixer
(Eppendorf) to inactivate the enzyme. Samples were filtered using
0.45 μm PVDF filters coupled to Eppendorf tubes (National Scientific)
during 5 min at 6000 rpm. For each sample, two dilutions with water
(1:400 and 1:4000) were prepared for high-performance anion-exchange
chromatography with pulsed amperometric detection (HPAEC-PAD)
analysis.
HPAEC-PAD. Product analysis was performed by HPAEC-PAD on
a ICS3000 Dionex system (Dionex Corp., Sunnyvale, CA) consisting of
an SP gradient pump, an AS-HV autosampler, and an electrochemical
detector with a gold working electrode and Ag/AgCl as reference
electrode. All eluents were degassed by flushing with helium. A pellicular
anion-exchange 4 Â 250 mm Carbo-Pack PA-1 column (Dionex) con-
nected to a CarboPac PA-1 guard column was used at 30 ꢀC. For eluent
preparation, Milli-Q water and 50% (w/v) NaOH (Sigma-Aldrich) were
used. The initial mobile phase was 15 mM NaOH at 1.0 mL/min for
12 min. A mobile phase gradient from 15 to 200 mM NaOH was per-
formed at 1.0 mL/min in 15 min, and the latter was maintained as mobile
phase for 25 min. The separation of lactose and allolactose was achieved
withan isocratic run (35 min) with15 mMNaOH at 1.0 mL/min, followed
by a 25 min elution with 200 mM NaOH to remove the rest of the car-
bohydrates. The peaks were analyzed using Chromeleon software. Identi-
fication of the different carbohydrates was done on the basis of glucose,
Thermostability of K. lactis β-Galactosidase. Soluble enzymes
or permeabilized cells were incubated at 40 ꢀC in 100 mM Tris-HCl
buffer, pH 7.0. Aliquots were harvested at different times, and the remain-
ing activity toward ONPG was determined. The same experiments were
carried out in the presence of 200 g/L glucose or galactose.
Production of Galacto-oligosaccharides. The reaction mix-
ture (20 mL) contained 400 g/L lactose (34.7% w/w) in 0.1 M
potassium phosphate buffer (pH 6.8). The biocatalyst was then added
10478
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
Figure 3. HPAEC-PAD analysis of the reaction of lactose with (A) per-
meabilized K. lactis cells and (B) Lactozym 3000 L HP G. Peaks:
(1) galactose; (2) glucose; (3, 4) disaccharides (unknown); (5) 6-galacto-
biose; (6) lactose and allolactose; (7) 3-galactobiose; (8) 6-galactosyl-
lactose; (10) 3-galactosyl-glucose; (14) 4-galactosyl-lactose; (9, 11À13, 15)
other GOS (unknown). (Inset) Separation of allolactose (6a) and
lactose (6b) under isocratic conditions. The chromatograms correspond
to reaction times of 25 and 7 h for K. lactis cells and Lactozym 3000 L HP
G, respectively.
helium, was used as mobile phase (flow rate = 8.1 mL/min) for 8 min.
Then, a gradient to acetonitrile/water 75:25 (v/v) was performed in 2 min,
and this eluent was maintained during 6 min. Finally, a gradient from this
composition to acetonitrile/water 70:30 (v/v) was performed in 5 min and
maintained for 20 min. Peaks were detected using an evaporative light-
scattering detector DDL-31 (Eurosep) equilibrated at 60 ꢀC. A three-way
flow splitter (model Acurate, Dionex) and a fraction collector II (Waters)
were employed. The fractions containing the main peaks were pooled, the
solvent (acetonitrile/water) was eliminated by rotary evaporation, and the
resulting sugars were dissolved in a small amount of methanol and a droplet
of water. Ice-cold acetone was added to initiate precipitation. The resulting
suspension was filtered, and the crystals were dried overnight in vacuo.
Mass Spectrometry. Samples were analyzed by MALDI-TOF
mass spectrometry (Bruker, model Ultraflex III TOFTOF) using 2,5-
dihydroxybenzoic acid doped with sodium iodide as matrix, in positive
reflectron mode.
Figure 2. Thermoinactivation of K. lactis β-galactosidase at 40 ꢀC in
100 mM Tris-HCl buffer, pH 7.0: (A) in the absence of sugars; (B) in the
presence of 200 g/L galactose; (C) in the presence of 200 g/L glucose.
Initial activity was 9.2 U/mL for permeabilized cells, 8.9 U/mL for
the total protein extract, and 1.3 U/mL for Lactozym 3000 L HP G.
Remaining activity was determined at the indicated times using the
ONPG assay.
galactose, lactose, 6-O-β-galactosyl-glucose (allolactose), 3-galactobiose,
4-galactobiose, 6-galactobiose, and 4-O-β-galactosyl-lactose standards.
Purification of GOS by Hydrophilic Interaction Chroma-
tography. For the isolation of unknown GOS in the mixture (not
coeluting with any of the standards), the reaction was scaled up to 5 mL.
When the GOS yield reached the maximum value, the biocatalyst was
inactivated by boiling the solution for 5 min. The reaction mixture was
filtered, and the mixture was purified by semipreparative HPLC. A quater-
nary pump (Delta 600, Waters) coupled to a Lichrosorb-NH2 column
(5 μm, 10 Â 250 mm, Merck) was used. The column temperature was
kept constant at 25 ꢀC. Acetonitrile/water 85:15 (v/v), conditioned with
Nuclear Magnetic Resonance (NMR). The structure of the
oligosaccharides was elucidated using a combination of 1H, 13C, and 2D
NMR (COSY, TOCSY, NOESY, HSQC) techniques. The spectra of
the samples (ca. 10 mM), dissolved in deuterated water, were recorded
on a Bruker AVANCE DRX500 spectrometer equipped with a tunable
broadband 1H/X probe with a gradient in the Z axis, at a temperature of
10479
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
Figure 4. 2D NMR heteronuclear single quantum coherence (HSQC) analysis of two galacto-oligosaccharides obtained in the reaction of lactose with
K. lactis cells: (A) trisaccharide 6-galactosyl-lactose; (B) disaccharide 3-galactosylglucose.
298 K. Chemical shifts were expressed in parts per million with respect to
the 0 ppm point of DSS, used as internal standard. COSY, TOCSY,
NOESY, and HSQC sequences were provided by Bruker. COSY, TOCSY
(80 ms mixing time), and NOESY (500 ms mixing time) experiments
were performed with 16, 8, and 48 scans, respectively, with 256 increments
in the indirect dimension and with 1024 points in the acquisition dimension.
The spectral widths were 9 ppm in both dimensions. The HSQC ex-
periment (16 scans) also used 256 increments in the indirect dimension and
1024 points in the acquisition dimension. The spectral width was 120 ppm
in the indirect dimension and 9 ppm in the acquisition one.
Scanning Electron Microscopy (SEM). The permeabilized cells
were mounted on aluminum SEM stubs. The mounted samples were
sputter-coated with a thin layer of gold at completed Torr vacuum and
examined by SEM using an XL3 microscope (Philips) at an acceleration
potential of 20 kV.
not subjected to ethanol treatment was performed, and similar
micrographs (not shown) were obtained.
It is well-known that activity in the permeabilized cell mass is
low compared with the soluble enzyme.23 In this work, the
hydrolytic activity of the permeabilized cells, using ONPG as
substrate, was 210 U/g of cell mass. The thermal stability of
K. lactis β-galactosidase in permeabilized cells was studied at 40 ꢀC
in 100 mM Tris-HCl buffer (pH 7.0) (Figure 2). This neutral pH
has been reported to be optimal for β-galactosidase activity.19,35
The results were compared with those of the total protein extract
obtained by cell lysis and with a soluble β-galactosidase from
K. lactis (Lactozym 3000 L HP G) (Figure 2). As shown in Figure 2A,
the soluble enzyme inactivates very quickly under such condi-
tions (especially Lactozym 3000 L), and most of its activity was
lost in <60 min. The permeabilized cells exert a protecting effect
on the intracellular β-galactosidase, as the activity remaining after
120 min was approximately 40%.
’ RESULTS AND DISCUSSION
To assay the enzyme stability under conditions similar to those
employed in the biotransformations, studies were also performed
in the presence of 200 g/L glucose or galactose, which are the two
main monosaccharides present in the reaction mixture when
using lactose. It is well-known that sugars stabilize proteins;36
therefore, the operational conditions are usually favorable for the
stability of glycosidases.37 In the presence of 200 mg/mL galactose
(Figure 2B) or glucose (Figure 2C), the stabilizing effect of whole
cells was even more pronounced; the activity of the permeabilized
cells remained significantly stable during 120 min, whereas the
soluble β-galactosidases displayed a significant loss of activity in
the same time.
Production of GOS by K. lactis Cells. GOS are simultaneously
synthesized and degraded by β-galactosidases; as a consequence,
the GOS concentration and the composition of the sample change
dramatically with reaction time and can hardly be predicted.8,9 The
time at which reaction is harvested has a crucial effect on GOS
yield.14 The maximum GOS yield is basically determined by the
intrinsic enzyme properties (transgalactosylation to hydrolysis
ratio) as well as lactose concentration.38,39
Activity and Stability of K. lactis Cells. Different methods of
permeabilizing cells that do not significantly affect enzymatic
activity have been described, including concentration, drying,
treatment with solvents or surfactants, lyophilization, ultrasonic
treatment, mechanical disruption, etc.29,31 In this work, K. lactis
cells growing logarithmically on lactose as carbon source at 26 ꢀC
were exposed to 50% (v/v) ethanol for 15 min at 4 ꢀC, washed,
and lyophilized. Solvent treatment is an inexpensive method that
can be applied on a large scale.32 Depending on the solvent
nature and concentration, as well as the incubation time, the
permeability of membranes may be increased enough to lead to
disruption of cells, with the concomitant release of enzyme
molecules into the supernatant.18,33,34 In our study, the ethanol
treatment was not so intense as to produce cell disruption.
Although cells treated with short-chain alcohols become non-
viable in many cases, intracellular enzymes may resist such treat-
ment and, as a consequence, are not inactivated. In addition, the
permeability barriers of cell walls are lowered, so the cells them-
selves can be utilized as a biocatalyst repeatedly.
Yeast morphology was examined by SEM (Figure 1). The
typical oval shape of budding yeast cells was visualized at 1000Â
(upper picture) and 2000Â (bottom picture). A control of yeasts
We studied the behavior of K. lactis cells in GOS synthesis
compared with two soluble β-galactosidases (Lactozym 3000 L
10480
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
HP G and Maxilact LGX 5000). The experiments were carried
out at 1.2À1.5 U/mL, which is a lower enzyme concentration
than typically used in similar experiments with K. lactis
β-galactosidase (e.g., 3À9 U/mL in ref 15, 2.9À8.7 U/mL in
ref 16, and 12.5 U/mL in ref 30). At higher enzyme concentrations
reactions are so fast (<240 min) that it is difficult to observe any
effect of enzyme stability on reaction progress. It has been demon-
strated that although the enzyme concentration exerts a marked
influence on the reaction time at which the maximum oligosacchar-
ide concentration is achieved, it does not affect the absolute value of
GOS yield.16,40 The rest of the conditions used in our experiments
were similar to those previously described for soluble K. lactis
β-galactosidase (400 mg/mL lactose, pH 7.0, 40 ꢀC).
Figure 3A shows the HPAEC-PAD chromatogram of the
reaction mixture with permeabilized cells after 25 h. Peaks 1, 2,
and 6 were assigned to galactose, glucose, and lactose, respec-
tively. However, we observed that the shape of peak 6 varied
throughout the reaction; in fact, by changing the elution condi-
tions, we were able to split peak 6 into 6a [the lactose isomer
allolactose, Gal-β(1f6)-Glc] and 6b (lactose), as shown in the
insets of Figure 3. We found that the most abundant GOS
synthesized by the permeabilized cells corresponded to peak 5
(the disaccharide 6-galactobiose [Gal-β(1f6)-Gal]), peak 6a
(allolactose), and peak 8. Purification of peak 8 was performed by
semipreparative HILIC. Its mass spectrum showed that it was a
1
trisaccharide. The 1D and 2D H NMR spectra displayed four
anomeric signals, two of them corresponding to the two α,β
anomers of the reducing glucose moiety and two more corre-
sponding to β-linked galactose residues. From the combination
of the signals from COSY, TOCSY, NOESY, and HSQC it could
be deduced that the 1H and 13C resonance signals belong to the
different residues. The NOESY cross peaks from the two ga-
lactose anomeric signals also allowed the connectivities to their
contiguous moieties to be distinguished. In fact, one of them was
clearly connected with two H-6 signals, the 13C resonance of
which was shifted downfield (HSQC, Figure 4A), whereas the
other one was connected to the H-4 of the glucose residue, the
13C resonance of which was unambiguously identified (HSQC).
Thus, peak 8 showed a galactosyl moiety β-(1f6)-linked to the
galactose ring of lactose.
As a consequence, the three major products synthesized by
K. lactis cells contained a β-(1f6) bond between two galactoses
or between one galactose and one glucose. It is worth emphasiz-
ing that the chemical structure of the obtained oligosaccharides
(composition, number of moieties, and types of linkages between
them) may affect their fermentation pattern by probiotic bacteria
in the gut.15 In this context, it was reported that β(1f6) linkages
are cleaved very quickly by β-galactosidases from bifidobacteria,15
a key factor in the prebiotic properties.
With authentic standards, it was also possible to identify peak 7
as the disaccharide 3-galactobiose [Gal-β(1f3)-Gal] and peak
14 as the trisaccharide 4-galactosyl-lactose [Gal-β(1f4)-Gal-
β(1f4)-Glc]. Peak 10 was also purified, and mass spectro-
metry analysis indicated it was a disaccharide. From the combi-
nation of the signals from COSY, TOCSY, NOESY, and HSQC
(Figure 4B) the structure of peak 10 was attributed to the
disaccharide Gal-β-(1f3)-Glc.
Figure 5. Galacto-oligosaccharides production from lactose by
(A) permeabilized K. lactis cells, (B) Lactozym 3000 L HP G, and
(C) Maxilact LGX 5000. Experimental conditions: 400 g/L lactose in
0.1 M sodium phosphate buffer (pH 6.8), 1.2À1.5 U/mL, 40 ꢀC.
Minor peaks 3, 4, 9, 11À13, and 15 could not be identified. As the
response factors of GOS isomers in HPAEC-PAD are closely
similar, the quantification of the unknown products was carried
out usinglactose asstandard for the disaccharides and 4-galactosyl-
lactose for the trisaccharides.
Chockchaisawasdee et al.16 were the first to perform a pre-
liminary structural analysis of the GOS formed by soluble K. lactis
β-galactosidase (Maxilact L2000), concluding that the major
bonds were β(1f6). Maugard et al.17 and Cheng et al.41
detected the formation of disaccharides by K. lactis, but no
structural identification was done. Martinez-Villaluenga et al.15
The complete identification of the GOS synthesized by a
particular enzyme is a difficult task. In fact, considering the
formation of Gal-β(1f2)-, β(1f3)-, β(1f4)-, and β(1f6)-
bonds, the theoretical number of synthesized GOS accounts for 7
disaccharides, 32 trisaccharides, 128 tetrasaccharides, and so on.
10481
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
Table 1. Carbohydrate Composition (Percent w/w) of the Reaction Mixture Using Permeabilized Cells from K. lactisa
reaction time (h)
galactose
glucose
lactose
6-galactobiose
allolactose
6-galactosyl-lactose
other GOS
total GOS
0
0.5
0.75
1
0.0
2.1
0.0
3.0
100.0
91.7
87.1
77.8
69.3
43.6
24.2
3.6
0.0
0.3
0.0
0.0
0.0
0.0
0.0
6.1
10.0
6.5
4.3
4.3
0.0
2.9
0.0
0.0
0.0
2.2
6.2
8.9
12.1
9.4
8.2
8.8
0.0
3.2
3.4
4.2
0.8
4.5
5.3
5.6
5.7
1.2
7.5
10.8
16.1
31.1
44.2
25.3
27.3
26.1
2
7.5
7.1
1.4
8.6
4
10.8
13.9
33.0
33.2
35.0
14.4
17.7
38.1
38.5
38.1
3.8
12.3
15.7
3.6
6
6.4
25
30
51.5
5.9
1.1
10.1
4.7
0.8
10.3
2.7
a Experimental conditions: 400 g/L lactose in 0.1 M sodium phosphate buffer (pH 6.8), 1.2 U/mL (5.5 mg cell mass/mL), 40 ꢀC.
Table 2. Carbohydrate Composition (Percent w/w) of the Reaction Mixture Using Lactozym 3000 L HP Gs
reaction time (h)
galactose
glucose
lactose
6-galactobiose
allolactose
6-galactosyl-lactose
other GOS
total GOS
0
0.5
1
0.0
3.4
0.0
5.6
100.0
84.4
75.5
62.9
36.2
26.6
19.5
6.7
0.0
0.6
2.0
1.2
3.3
4.1
4.3
6.2
5.9
5.9
0.0
0.0
0.0
5.4
0.0
0.7
2.3
4.2
6.2
7.3
9.3
11.5
9.2
9.5
0.0
6.7
6.0
6.5
0.0
7.7
12.0
16.5
30.2
34.6
37.5
42.6
39.7
38.9
1.5
3.5
5
6.9
13.6
22.0
25.5
29.8
34.4
36.7
37.3
0.0
11.2
15.1
15.9
16.5
14.3
14.3
13.4
11.5
13.3
13.2
16.4
18.3
18.5
5.6
7.2
7
7.5
22
49.5
70
10.6
10.2
10.1
5.2
5.3
s Experimental conditions: 400 g/L lactose in 0.1 M sodium phosphate buffer (pH 6.8), 1.5 U/mL, 40 ꢀC.
Table 3. Carbohydrate Composition (Percent w/w) of the Reaction Mixture Using Maxilact LGX 5000a
reaction time (h)
galactose
glucose
lactose
6-galactobiose
allolactose
6-galactosyl-lactose
other GOS
total GOS
0
0.5
1
0.0
2.8
0.0
4.8
100.0
88.2
78.3
72.0
51.4
44.1
36.8
19.0
18.1
15.8
17.3
0.0
0.3
1.1
0.7
2.5
3.8
4.4
4.4
4.8
5.1
4.9
0.0
0.0
0.0
0.0
3.2
3.4
4.2
7.0
8.2
8.3
8.1
0.0
3.6
0.0
0.4
0.9
1.6
3.9
4.6
6.1
14.0
9.4
8.7
9.2
0.0
4.2
4.5
8.2
6.9
8.9
1.5
3.5
5
5.8
10.9
17.5
20.1
22.6
26.5
28.6
30.9
28.9
8.9
11.2
22.3
25.7
29.4
40.7
38.5
37.7
38.5
8.9
12.7
13.8
14.7
15.4
16.1
15.6
16.3
10.1
11.2
13.7
14.8
15.6
15.3
7
22
28
49.5
70
a Experimental conditions: 400 g/L lactose in 0.1 M sodium phosphate buffer (pH 6.8), 1.5 U/mL, 40 ꢀC.
reported the formation of 6-galactobiose, allolactose, and 6-
galactosyl-lactose with K. lactis β-galactosidase, in accordance
with our findings with permeabilized cells, but the formation of other
GOS was not mentioned. To our knowledge, the present study is the
most detailed of those published with K. lactis β-galactosidase in
terms of the chemical structure of the synthesized products.
Figure 5A shows the reaction progress with permeabilized
cells. The maximum GOS concentration [177 g/L, which repre-
sents 44% (w/w) of the total carbohydrates in the reaction
mixture] was observed at 6 h. After that, the amount of GOS
diminished by the hydrolytic action of β-galactosidase until
reaching an equilibrium GOS concentration of 105 g/L. Inter-
estingly, lactose disappeared almost completely at the end of the
process. Table 1 summarizes the carbohydrate composition (in
weight) at different reaction times. At the point of maximal GOS
concentration (6 h), the mixture was formed by monosaccharides
(32%), lactose (24%), 6-galactobiose (6%), allolactose (10%),
6-galactosyl-lactose (16%), and other GOS (12%). Martinez-
Villaluenga et al.,15 using a soluble K. lactis β-galactosidase,
reported a mixture formed by monosaccharides (49%), residual
lactose (21%), the disaccharides 6-galactobiose and allolactose
(13%), and 6-galactosyl-lactose (17%).
10482
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
The upper range of GOS yield is close to 40À45%,9,42,43 which
is lower than that reported for other prebiotics such as fructoo-
ligosaccharides (approximately 65%).44,45 In this work, GOS
yield was slightly higher than that reported in previous studies
with K. lactis β-galactosidase.15À17 The lower yield reported in
such studies could be related with the overestimation of lactose
due to its tendency to coelute with allolactose or by not consider-
ing the contribution of other minor GOS to the total yield.
Production of GOS by Soluble K. lactis β-Galactosidase.
The activities of Lactozym 3000 L HP and Maxilact LGX 5000
toward ONPG were 645 and 2145 U/mL, respectively. The
reaction profile of both enzymes in the GOS synthesis is shown
in Figure 5, panels B and C, respectively. Several differences were
found in the behavior of soluble β-galactosidases compared with
permeabilized cells. First, panels B and C of Figure 5 do not show
the typical pattern in which a maximum GOS yield is followed by
the hydrolysis of a part of the synthesized products. Second,
lactose is not completely consumed at the equilibrium, with
remaining concentrations of 21 and 65 g/L for Lactozym and
Maxilact LGX 5000, respectively, whereas only 3 g/L was
determined for permeabilized cells. Maximum GOS yield was
slightly lower for soluble β-galactosidases (160 g/L for Lactozym
and 154 g/L for Maxilact LGX 5000, which represent 42 and 41%
w/w of the total carbohydrates present in the mixture) compared
with permeabilized cells (177 g/L, 44% w/w).
2, 28049 Madrid, Spain. Fax: +34-91-5854760. E-mail: fplou@
icp.csic.es.
Present Addresses
4Institute for Molecules and Materials, Radboud University
Nijmegen, Heyendaalseweg 135, 6525 AJ Nijmegen, The
Netherlands.
Funding Sources
Projects BIO2007-67708-C04-01, BIO2007-67708-C04-03,
BIO2010-20508-C04-01, and BIO2010-20508-C04-04 from
the Spanish Ministry of Science and Innovation supported this
research. B.R.-C. and M.A.d.A. weresupportedbyfellowships from
the Spanish Ministries of Science and Innovation (FPI program)
and Education and Culture (FPU program), respectively.
’ ACKNOWLEDGMENT
We thank Ramiro Martínez (Novozymes A/S, Madrid, Spain)
for supplying Lactozym and for useful suggestions. We also thank
DSM Food Specialties for supplying Maxilact LGX 5000.
’ REFERENCES
(1) Adam, A. C.; Rubio-Texeira, M.; Polaina, J. Lactose: the milk
sugar from a biotechnological perspective. Crit. Rev. Food Sci. 2004, 44,
553–557.
Figure 3B shows the HPAEC-PAD chromatogram obtained
with Lactozym 3000 L HP close to the point of maximum GOS
concentration. As illustrated, the profile was similar to that
obtained with permeabilized cells. The main transglycosylation
products were again 6-galactobiose (peak 5), allolactose (peak
6a), and 6-galactosyl-lactose (peak 8). The main difference of the
soluble enzyme compared with the permeabilized cells was the
absence of several minor peaks (peaks 3, 4, 13, and 14 in
Figure 3A); however, some of them appeared briefly at different
reaction times, but they quickly disappeared. A very similar
chromatogram was obtained with Maxilact LGX 5000 (data
not shown).
Tables 2 and 3 summarize the progress of carbohydrate
distribution for both enzymes. The above results seem to indicate
that the reaction is stopped before reaching the final equilibrium
due to the low stability of soluble β-galactosidases. In contrast,
the higher stability of permeabilized cells is manifested by the
characteristic reaction profile, with a maximum, when a competi-
tion exists between hydrolysis and transglycosylation (kinetic
control).46
The maximum GOS concentration with permeabilized cells
was obtained when lactose conversion was 76%. Lactozym 3000
L HP and Maxilact LGX 5000 reached 95 and 83% of lactose
conversion, without displaying a maximum in GOS concentra-
tion. This could indicate that the microenvironment of the
permeabilized cells also exerts an influence on the transglycosy-
lation to hydrolysis ratio. Thus, in the initial stages of the reac-
tion, permeabilized cells show a higher tendency to form GOS
than the soluble enzymes (Figure 5); however, when the GOS
accumulate in the system, the whole cells also have a higher
tendency to hydrolyze the GOS formed.
(2) Hsu, C. A.; Lee, S. L.; Chou, C. C. Enzymatic production of
galactooligosaccharides by β-galactosidase from Bifidobacterium longum
BCRC 15708. J. Agric. Food Chem. 2007, 55, 2225–2230.
(3) Matella, N. J.; Dolan, K. D.; Lee, Y. S. Comparison of galactoo-
ligosaccharide production in free-enzyme ultrafiltration and in immobi-
lized-enzyme systems. J. Food Sci. 2006, 71, C363–C368.
(4) Seibel, J.; Jordening, H. J.; Buchholz, K. Glycosylation with
activated sugars using glycosyltransferases and transglycosidases. Bioca-
tal. Biotransform. 2006, 24, 311–342.
(5) Bojarova, P.; Kren, V. Glycosidases: a key to tailored carbohy-
drates. Trends Biotechnol. 2009, 27, 199–209.
(6) Desmet, T.; Soetaert, W. Enzymatic glycosyl transfer: mechan-
isms and applications. Biocatal. Biotransform. 2011, 29, 1–18.
(7) Torres, D. P.; Goncalves, M.; Teixeira, J. A.; Rodrigues, L. R.
Galacto-oligosaccharides: production, properties, applications, and sig-
nificance as prebiotics. Compr. Rev. Food Sci. F 2010, 9, 438–454.
(8) Park, A. R.; Oh, D. K. Galacto-oligosaccharide production using
microbial β-galactosidase: current state and perspectives. Appl. Micro-
biol. Biotechnol. 2010, 85, 1279–1286.
(9) Splechtna, B.; Nguyen, T. H.; Steinbock, M.; Kulbe, K. D.;
Lorenz, W.; Haltrich, D. Production of prebiotic galacto-oligosacchar-
ides from lactose using β-galactosidases from Lactobacillus reuteri. J.
Agric. Food Chem. 2006, 54, 4999–5006.
(10) Maischberger, T.; Leitner, E.; Nitisinprasert, S.; Juajun, O.;
Yamabhai, M.; Nguyen, T. H.; Haltrich, D. β-Galactosidase from Lacto-
bacillus pentosus: purification, characterization and formation of galacto-
oligosaccharides. Biotechnol. J. 2010, 5, 838–847.
(11) Iqbal, S.; Nguyen, T. H.; Nguyen, T. T.; Maischberger, T.;
Haltrich, D. β-Galactosidase from Lactobacillus plantarum WCFS1:
biochemical characterization and formation of prebiotic galacto-oligo-
saccharides. Carbohydr. Res. 2010, 345, 1408–1416.
(12) Shadid, R.; Haarman, M.; Knol, J.; Theis, W.; Beermann, C.;
Rjosk-Dendorfer, D.; Schendel, D. J.; Koletzko, B. V.; Krauss-Etschmann,
S. Effects of galactooligosaccharide and long-chain fructooligosaccharide
supplementation during pregnancy on maternal and neonatal microbiota
and immunity À a randomized, double-blind, placebo-controlled study.
Am. J. Clin. Nutr. 2007, 86, 1426–1437.
(13) Rastall, R. A.;Gibson, G. R.;Gill, H. S.; Guarner, F.; Klaenhammer,
T. R.; Pot, B.; Reid, G.; Rowland, I. R.; Sanders, M. E. Modulation of the
microbial ecology of the human colon by probiotics, prebiotics and
’ AUTHOR INFORMATION
Corresponding Author
*Postal address: Departamento de Biocatꢀalisis, Instituto de
Catꢀalisis y Petroleoquímica, CSIC, Cantoblanco, Marie Curie
10483
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484

Journal of Agricultural and Food Chemistry
ARTICLE
synbiotics to enhance human health: an overview of enabling science and
potential applications. FEMS Microbiol. Ecol. 2005, 52, 145–152.
(14) Gosling, A.; Stevens, G. W.; Barber, A. R.; Kentish, S. E.; Gras,
S. L. Recent advances refining galactooligosaccharide production from
lactose. Food Chem. 2010, 121, 307–318.
(34) Schwab, C.; Lee, V.; Sorensen, K. I.; Ganzle, M. G. Production
of galactooligosaccharides and heterooligosaccharides with disrupted
cell extracts and whole cells of lactic acid bacteria and bifidobacteria. Int.
Dairy J. 2011, 21, 748–754.
(35) Matioli, G.; De Moraes, F. F.; Zanin, G. M. Operational stability
and kinetics of lactose hydrolysis by β-galactosidase from Kluyveromyces
fragilis. Acta Sci. Health Sci. 2003, 25, 7–12.
(36) Prasad, S.; Roy, I. Effect of disaccharides on the stabilization of
bovine trypsin against detergent and autolysis. Biotechnol. Prog. 2010,
26, 627–635.
(37) Ghazi, I.; Fernandez-Arrojo, L.; Garcia-Arellano, H.; Ferrer, M.;
Ballesteros, A.; Plou, F. J. Purification and kinetic characterization of a
fructosyltransferase from Aspergillus aculeatus. J. Biotechnol. 2007, 128,
204–211.
(38) Plou, F. J.; Martin, M. T.; Gomez de Segura, A.; Alcalde, M.;
Ballesteros, A. Glucosyltransferases acting on starch or sucrose for the
synthesis of oligosaccharides. Can. J. Chem. 2002, 80, 743–752.
(39) Plou, F. J.; Gomez de Segura, A.; Ballesteros, A. Application of
glycosidases and transglycosidases for the synthesis of oligosaccharides.
In Industrial Enzymes: Structure, Function and Application; Polaina, J.,
MacCabe, A. P., Eds.; Springer: New York, 2007; pp 141À157.
(40) Buchholz, K.; Kasche, V.; Bornscheuer, U. T. Biocatalysts and
Enzyme Technology; Wiley-VCH Verlag: Weinheim, Germany, 2005.
(41) Cheng, C. C.; Yu, M. C.; Cheng, T. C.; Sheu, D. C.; Duan, K. J.;
Tai, W. L. Production of high-content galacto-oligosaccharide by
enzyme catalysis and fermentation with Kluyveromyces marxianus.
Biotechnol. Lett. 2006, 28, 793–797.
(42) Rabiu, B. A.; Jay, A. J.; Gibson, G. R.; Rastall, R. A. Synthesis and
fermentation properties of novel galacto-oligosaccharides by β-galacto-
sidases from Bifidobacterium species. Appl. Environ. Microbiol. 2001,
67, 2526–2530.
(43) Hansson, T.; Adlercreutz, P. Optimization of galactooligosac-
charide production from lactose using β-glycosidases from hyperther-
mophiles. Food Biotechnol. 2001, 15, 79–97.
(15) Martinez-Villaluenga, C.; Cardelle-Cobas, A.; Corzo, N.; Olano,
A.; Villamiel, M. Optimization of conditions for galactooligosaccharide
synthesis during lactose hydrolysis by β-galactosidase from Kluyveromyces
lactis (Lactozym 3000 L HP G). Food Chem. 2008, 107, 258–264.
(16) Chockchaisawasdee, S.; Athanasopoulos, V. I.; Niranjan, K.;
Rastall, R. A. Synthesis of galacto-oligosaccharide from lactose using
β-galactosidase from Kluyveromyces lactis: studies on batch and continuous
UF membrane-fitted bioreactors. Biotechnol. Bioeng. 2005, 89, 434–443.
(17) Maugard, T.; Gaunt, D.; Legoy, M. D.; Besson, T. Microwave-
assisted synthesis of galacto-oligosaccharides from lactose with immobilized
β-galactosidase from Kluyveromyces lactis. Biotechnol. Lett. 2003, 25, 623–629.
(18) Pinho, J. M. R.; Passos, F. M. L. Solvent extraction of
β-galactosidase from Kluyveromyces lactis yields a stable and highly
active enzyme preparation. J. Food Biochem. 2011, 35, 323–336.
(19) Pal, A.; Pal, V.; Ramana, K. V.; Bawa, A. S. Biochemical studies
of β-galactosidase from Kluyveromyces lactis. J. Food Sci. Technol. 2009,
46, 217–220.
(20) Zhou, Q. Z. K.; Chen, X. D. Effects of temperature and pH on
the catalytic activity of the immobilized β-galactosidase from Kluyver-
omyces lactis. Biochem. Eng. J. 2001, 9, 33–40.
(21) Zhou, Q. Z.; Chen, X. D.; Li, X. Kinetics of lactose hydrolysis by
β-galactosidase of Kluyveromyces lactis immobilized on cotton fabric.
Biotechnol. Bioeng. 2003, 81, 127–133.
(22) Siso, M. I. G.; Cerdan, E.; Picos, M. A. F.; Ramil, E.; Belmonte,
E. R.; Torres, A. R. Permeabilization of Kluyveromyces lactis cells for milk
whey saccharification: a comparison of different treatments. Biotechnol.
Tech. 1992, 6, 289–292.
(23) Fontes, E. A. F.; Passos, F. M. L.; Passos, F. J. V. A mechanistical
mathematical model to predict lactose hydrolysis by β-galactosidase in a
permeabilized cell mass of Kluyveromyces lactis: validity and sensitivity
analysis. Process Biochem. 2001, 37, 267–274.
(44) Ghazi, I.; Fernandez-Arrojo, L.; Gomez de Segura, A.; Alcalde,
M.; Plou, F. J.; Ballesteros, A. Beet sugar syrup and molasses as low-cost
feedstock for the enzymatic production of fructo-oligosaccharides.
J. Agric. Food Chem. 2006, 54, 2964–2968.
(45) Alvaro-Benito, M.; De Abreu, M.; Fernandez-Arrojo, L.; Plou,
F. J.; Jimenez-Barbero, J.; Ballesteros, A.; Polaina, J.; Fernandez-Lobato,
M. Characterization of a β-fructofuranosidase from Schwanniomyces
occidentalis with transfructosylating activity yielding the prebiotic 6-kestose.
J. Biotechnol. 2007, 132, 75–81.
(24) Kondo, A.; Liu, Y.; Furuta, M.; Fujita, Y.; Matsumoto, T.;
Fukuda, H. Preparation of high activity whole cell biocatalyst by per-
meabilization of recombinant flocculent yeast with alcohol. Enzyme
Microbial Technol. 2000, 27, 806–811.
(25) Boon, M. A.; Janssen, A. E. M.; Van’t Riet, K. Effect of tem-
perature and enzyme origin on the enzymatic synthesis of oligosacchar-
ides. Enzyme Microbial Technol. 2000, 26, 271–281.
(26) Prenosil, J. E.; Stuker, E.; Bourne, J. R. Formation of oligosac-
charides during enzymatic lactose hydrolysis and their importance in a
whey hydrolysis process: part II. Experimental. Biotechnol. Bioeng. 1987,
30, 1026–1031.
(46) Rodriguez-Alegria, M. E.; Enciso-Rodriguez, A.; Ortiz-Soto,
M. E.; Cassani, J.; Olvera, C.; Lopez-Munguia, A. Fructooligosaccharide
production by a truncated Leuconostoc citreum inulosucrase mutant.
Biocatal. Biotransform. 2010, 28, 51–59.
(27) Manera, A. P.; De Almeida, F. A.; Rodrigues, M. I.; Kalil, S. J.;
Filho, F. M. Galacto-oligosaccharides production using permeabilized
cells of Kluyveromyces marxianus. Int. J. Food Eng. 2010, 6, 1–13.
(28) Genari, A. N.; Passos, F. V.; Passos, F. M. L. Configuration of a
bioreactor for milk lactose hydrolysis. J. Dairy Sci. 2003, 86, 2783–2789.
(29) Panesar, R.; Panesar, P. S.; Singh, R. S.; Kennedy, J. F.; Bera,
M. B. Production of lactose-hydrolyzed milk using ethanol permeabi-
lized yeast cells. Food Chem. 2007, 101, 786–790.
(30) Lee, Y. J.; Kim, C. S.; Oh, D. K. Lactulose production by
β-galactosidase in permeabilized cells of Kluyveromyces lactis. Appl.
Microbiol. Biotechnol. 2004, 64, 787–793.
(31) Panesar, R.; Panesar, P. S.; Singh, R. S.; Kennedy, J. F. Hydro-
lysis of milk lactose in a packed bed reactor system using immobilized
yeast cells. J. Chem. Technol. Biotechnol. 2011, 86, 42–46.
(32) Stred’ansky, M.; Tomaska, M.; Sturdik, E.; Kremnicky, L.
Optimization of β-galactosidase extraction from Kluyveromyces marx-
ianus. Enzyme Microbial Technol. 1993, 15, 1063–1065.
(33) Puri, M.; Gupta, S.; Pahuja, P.; Kaur, A.; Kanwar, J. R.; Kennedy,
J. F. Cell disruption optimization and covalent immobilization of β-D-
galactosidase from Kluyveromyces marxianus YW-1 for lactose hydrolysis
in milk. Appl. Biochem. Biotechnol. 2010, 160, 98–108.
10484
dx.doi.org/10.1021/jf2022012 |J. Agric. Food Chem. 2011, 59, 10477–10484