β-galactosidase from Lactobacillus plantarum WCFS1: Biochemical characterization and formation of prebiotic galacto-oligosaccharides

Article abstract of DOI:10.1016/j.carres.2010.03.028

Recombinant β-galactosidase from Lactobacillus plantarum WCFS1, homologously over-expressed in L. plantarum, was purified to apparent homogeneity using p-aminobenzyl 1-thio-β-D-galactopyranoside affinity chromatography and subsequently characterized. The enzyme is a heterodimer of the LacLMfamily type, consisting of a small subunit of 35 kDa and a large subunit of 72 kDa. The optimum pH for hydrolysis of its preferred substrates o-nitrophenyl-β-D-galactopyranoside (oNPG) and lactose is 7.5 and 7.0, and optimum temperature for these reactions is 55 and 60 °C, respectively. The enzyme is most stable in the pH range of 6.5-8.0. The K_m, k _cat and k_cat/K_m values for oNPG and lactose are 0.9 mM, 92 s^-1, 130 mM^-1 s^-1 and 29 mM, 98 s^-1, 3.3 mM^-1 s^-1, respectively. The L. plantarum β-galactosidase possesses a high transgalactosylation activity and was used for the synthesis of prebiotic galacto-oligosaccharides (GOS). The resulting GOS mixture was analyzed in detail, and major components were identified by using high performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD) as well as capillary electrophoresis. The maximal GOS yield was 41% (w/w) of total sugars at 85% lactose conversion (600 mM initial lactose concentration). The enzyme showed a strong preference for the formation of β-(1→6) linkages in its transgalactosylation mode, while β-(1→3)-linked products were formed to a lesser extent, comprising ～80% and 9%, respectively, of the newly formed glycosidic linkages in the oligosaccharide mixture at maximum GOS formation. The main individual products formed were β-DGalp-( 1→6)-D-Lac, accounting for 34% of total GOS, and β-D-Galp-(1→6)-D-Glc, making up 29% of total GOS.

Full text of DOI:10.1016/j.carres.2010.03.028

Carbohydrate Research 345 (2010) 1408–1416
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
b-Galactosidase from Lactobacillus plantarum WCFS1: biochemical
characterization and formation of prebiotic galacto-oligosaccharides
Sanaullah Iqbal a,b, Thu-Ha Nguyen a,c, Tien Thanh Nguyen a,d, Thomas Maischberger a,c, Dietmar Haltrich a,
*
a Food Biotechnology Laboratory, Department of Food Sciences and Technology, BOKU University of Natural Resources and Applied Life Sciences, Muthgasse 18, A-1190 Vienna, Austria
b University of Agriculture, Faisalabad, Pakistan
c Research Centre Applied Biocatalysis, Graz, Austria
d Institute of Biological and Food Technology, Hanoi University of Technology, Hanoi, Viet Nam
a r t i c l e i n f o
a b s t r a c t
Article history:
Recombinant b-galactosidase from Lactobacillus plantarum WCFS1, homologously over-expressed in L.
plantarum, was purified to apparent homogeneity using p-aminobenzyl 1-thio-b-D-galactopyranoside
affinity chromatography and subsequently characterized. The enzyme is a heterodimer of the LacLM-
family type, consisting of a small subunit of 35 kDa and a large subunit of 72 kDa. The optimum pH
Received 17 January 2010
Received in revised form 14 March 2010
Accepted 19 March 2010
Available online 21 March 2010
for hydrolysis of its preferred substrates o-nitrophenyl-b-
and 7.0, and optimum temperature for these reactions is 55 and 60 °C, respectively. The enzyme is most
stable in the pH range of 6.5–8.0. The Km, kcat and kcat/Km values for oNPG and lactose are 0.9 mM, 92 sÀ1
D-galactopyranoside (oNPG) and lactose is 7.5
Keywords:
b-Galactosidase
Lactobacillus plantarum WCFS1
Transgalactosylation
Galacto-oligosaccharides
HPAEC
,
130 mMÀ1 sÀ1 and 29 mM, 98 sÀ1, 3.3 mMÀ1 sÀ1, respectively. The L. plantarum b-galactosidase possesses
a high transgalactosylation activity and was used for the synthesis of prebiotic galacto-oligosaccharides
(GOS). The resulting GOS mixture was analyzed in detail, and major components were identified by using
high performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD) as
well as capillary electrophoresis. The maximal GOS yield was 41% (w/w) of total sugars at 85% lactose
conversion (600 mM initial lactose concentration). The enzyme showed a strong preference for the for-
mation of b-(1?6) linkages in its transgalactosylation mode, while b-(1?3)-linked products were formed
to a lesser extent, comprising ꢀ80% and 9%, respectively, of the newly formed glycosidic linkages in the
oligosaccharide mixture at maximum GOS formation. The main individual products formed were b-D-
Galp-(1?6)-
D-Lac, accounting for 34% of total GOS, and b-
D
-Galp-(1?6)-
D-Glc, making up 29% of total
GOS.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
their generally recognized as safe (GRAS) status. Hence, enzymes de-
rived from these organisms can be used without extensive purifica-
tion in various food-related applications.4,5 Lactobacillus plantarum
is a versatile lactic acid bacterium, which is encountered in a range
of environmental niches including dairy, meat and especially vege-
table fermentations. Moreover, it is commonly found in the human
gastrointestinal tract.6 Recently, the complete genome sequence of
L. plantarum WCFS1, a single colony isolate of L. plantarum NCIMB
8826 from human saliva, was determined and annotated7 and this
achievement has opened new ways in the comprehensive study of
a variety of biological processes of this organism.
Prebiotics are defined as a ‘selectively fermented ingredient that
allows specific changes, both in the composition and/or activity in
the gastrointestinal microbiota that confers benefits upon host
well-being and health’,8 and typically these prebiotic compounds
are oligosaccharides of various composition. Galacto-oligosaccha-
rides (GOS) together with fructo-oligosaccharides are amongst
the most important and best-studied groups of prebiotic oligosac-
charides. GOS have been classified as one of the few proven
prebiotics fulfilling the three criteria (i) resistance to gastric
The lactose-hydrolyzing enzyme b-galactosidase (b-D-galacto-
side galactohydrolase, EC 3.2.1.23) has long been used as an impor-
tant biocatalyst in the dairy industry as well as in synthetic
glycochemistry. b-Galactosidase can catalyze two basic reactions,
hydrolysis of the milk sugar lactose and structurally related galac-
tosides and transglycosylation reactions, resulting, for example, in
a mixture of galacto-oligosaccharides when lactose is the starting
material for the latter reaction. Possible sources of the enzyme
are plants, animal organs, bacteria, yeasts, fungi and moulds.1
Amongst them microbial sources are clearly preferable and are
mostly used for both reactions because of their ease of fermenta-
tive production, high activities and generally good stability.2,3
Lactic acid bacteria (LAB), which constitute a diverse group of lac-
tococci, streptococci and lactobacilli, have been studied intensively
with respect to their enzymes for various different reasons including
* Corresponding author. Tel.: +43 1 47654 6141; fax: +43 1 47654 6251.
E-mail address: dietmar.haltrich@boku.ac.at (D. Haltrich).
0008-6215/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2010.03.028

S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
1409
acidity, hydrolysis by mammalian enzymes and gastrointestinal
absorption; (ii) fermentation by intestinal microflora and (iii)
selective stimulation of the growth and/or activity of intestinal
bacteria associated with health and well-being.8 GOS are princi-
pally formed from lactose by biocatalytic reactions employing
the transgalactosylation activity of various retaining b-galactosi-
dases.1 In the reaction of lactose hydrolysis, a galactosyl–enzyme
intermediate is formed, and instead of transferring this galactose
unit to water, the enzyme transfers it to a suitable nucleophile
acceptor such as a sugar. In the synthesis of GOS, acceptor sugars
can be lactose, especially at the start of the reaction when its
concentration is high, but also the various reaction products of
the b-galactosidase-catalyzed reaction, either glucose or galactose
formed as primary hydrolysis products or the various oligosaccha-
rides formed through transgalactosylation (Scheme 1). Further-
more, these oligomeric products are only transiently formed as
kinetic intermediates, and are in turn hydrolyzed by b-galactosi-
dase. As a result, GOS are very complex mixtures consisting of
numerous different oligosaccharides varying in their DP and glyco-
sidic linkage. These mixtures depend strongly on the source of the
enzyme used for their production as well as on the progress of the
reaction. Relatively few studies have dealt with a detailed analysis
of the composition of these GOS mixtures.9,10 GOS typically have a
degree of polymerization between 2 and 10 with a terminal glu-
cose moiety. Very little is known about the structure–function
relationship of prebiotic oligosaccharides, although there has been
recent interest in this area.11
purification procedure based on p-aminobenzyl 1-thio-b-
D
-galacto-
pyranoside affinity chromatography. The specific activity of the
purified enzyme was 154 U mgÀ1 of protein, using standard assay
conditions and o-nitrophenyl b-D-galactopyranoside (oNPG) as sub-
strate, which is about threefold higher than that of the crude extract.
A total of 265 UoNPG b-galactosidase activity were obtained from the
initial activity (1040 U) in the crude cell extract, corresponding to a
purification recovery of ꢀ25%. The homogeneity of the purified b-
galactosidase was verified by SDS–PAGE and native PAGE (Fig. 1).
2.2. Determination of molecular mass
The L. plantarum b-galactosidase is a heterodimeric enzyme of a
molecular mass of ꢀ107 kDa, consisting of a larger subunit of
ꢀ72 kDa and a smaller subunit of ꢀ35 kDa as estimated under dena-
turing conditions of SDS–PAGE (Fig. 1A) and by native PAGE (Fig. 1B).
These results are in agreement with the theoretical molecular
masses of 35,223 Da and 72,180 Da for the smaller and larger sub-
units, respectively, as calculated (http://www.expasy.ch/tools/pi_
A
250 kDa
150 kDa
75 kDa
50 kDa
GOS are stable at high temperatures in acidic environments and
the caloric value of these oligosaccharides is only 1.7 kcal gÀ1
,
which makes them of particular interest for the food and drink
industry, for both their prebiotic properties and their use as sweet-
eners, especially in confectionery, acidic beverages and fermented
milk products.12 Since the GOS products currently available on the
market were not designed rationally, there is still need to find good
microbial sources of b-galactosidases showing high transgalactosy-
lation activities, and to characterize the individual components of
the GOS mixtures obtained with these biocatalysts.13
25 kDa
10 kDa
In the present study, b-galactosidase from L. plantarum WCFS1
homologously over-expressed in its original source by using a
Sakacin P-based expression system14 was purified to apparent
homogeneity, characterized and analyzed with respect to the
transgalactosylation of lactose. Major components of the GOS
formed by this enzyme were analyzed by high performance anion
exchange chromatography with pulsed amperometric detection
and capillary electrophoresis.
1
2
3
4
B
669 kDa
440 kDa
2. Results and discussion
232 kDa
140 kDa
2.1. Purification of b-galactosidase
The b-galactosidase from L. plantarum WCFS1 was homologously
over-expressed using its original source strain as expression host to-
gether with a lactobacillal Sakacin P-based expression system,14 and
subsequently purified to apparent homogeneity using a single-step
66 kDa
1
2
3
Figure 1. SDS–PAGE (A) and native PAGE (B) analysis of b-galactosidase from L.
plantarum WCFS1, applying approx. 1–2 g of total protein per lane. (A) Lanes 1–3
l
were Commassie blue stained, showing the Dual colour standard (Amersham) (Lane
1), purified b-galactosidase (Lane 2) and crude cell extract (Lane 3) while Lane 4
shows active staining with 4-methylumbelliferyl b-
D-galactoside (MUG) of purified
Scheme 1. Hydrolysis and galactosyl transfer reactions, both intra- and intermo-
lecular, during the b-galactosidase-catalyzed conversion of lactose. E, enzyme; Lac,
lactose; Gal, galactose; Glc, glucose; Nu, nucleophile.1
b-galactosidase. (B) Lane shows active staining with MUG of purified b-
galactosidase, Lanes 2 and 3 are Commassie blue stained showing purified enzyme
and high molecular weight marker (Amersham), respectively.
1

1410
S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
tool.html) from the amino acid sequences of the encoding lacL and
lacM genes (GenBank accession number AL935262). Very similar
molecular masses have been reported for the heterodimeric b-galac-
tosidases of the LacLM type from Lactobacillus sakei 23 K,15 Lactoba-
cillus acidophilus,16 Lactobacillus reuteri17 and Lactobacillus
helveticus.18 Active staining directly on the SDS–PAGE gel, after
pre-incubation of the enzyme with denaturing Laemmli buffer at
60 °C for 5 min, showed that only one band, corresponding to the lar-
the standard assay mixture, are shown in Table 2. The addition of
the monovalent cations K+ and Na+ showed positive effects on b-
galactosidase activity, when added in the selected concentration
range. The divalent ions Mg2+ and Mn2+ showed strong activation
of b-galactosidase activity for all concentrations considered. The
lower concentrations (1–10 mM) of urea, DTT and 2-mercapto-
ethanol showed a stimulating effect, while at a higher concentra-
tion (100 mM) these compounds inhibit the enzyme activity to a
varying extent. EDTA was found to strongly inactivate the enzyme
even at the lowest concentration that was tested (1 mM), indicating
the necessity of various metal ions for the activity of this enzyme.
ger subunit LacL, exhibited activity with 4-methylumbelliferyl-b-D-
galactoside (MUG) while the smaller subunit LacM did not show any
activity. Active staining on the native PAGE gel, using MUG as the
substrate, showed one single band with activity of approximately
115 kDa, correspondingtotheintactenzyme. Asecond, smallerband
of ꢀ85 kDa that is visible on the native gel after Commassie blue
staining is apparently a degradation product of the intact protein
or corresponds to the larger subunit LacL.17 It was furthermore ob-
served that heating the enzyme with Laemmli buffer at 99 °C for
5 min results in the complete denaturation of the heterodimeric en-
zyme and loss of activity.
The isoelectric point (pI) of b-galactosidase from L. plantarum
WCFS1 was determined by isoelectric focusing and a compact band
was observed in the range of 4.7–4.9 (not shown) which is in excel-
lent agreement with the theoretical calculated value for the pI of
4.74.
2.5. Effect of pH and temperature
The pH dependence of b-galactosidase activity from L. planta-
rum WCFS1 was determined using both lactose and oNPG as sub-
strates (Fig. 2A) and the optimum pH values were found to be
7.0 and 7.5, respectively. The activity with lactose shows a rather
narrow optimum (pH 6.5–7.5), while a broader optimum (pH
6.0–8.5) was observed when oNPG was the substrate. The enzyme
was found to be most stable in the neutral pH range of 6.5–8.0, and
retained more than 75% of its activity when stored at these pH val-
ues at 30 °C for 5 h.
Optimum temperatures for the hydrolysis of oNPG and lactose
were found to be 55 and 60 °C, respectively (Fig. 2B) under other-
wise standard assay conditions. The enzyme needs MgCl2 for its
stability, retaining 73% and 85% of its activity after an incubation
at 37 °C for 5 h in the presence of 1 and 10 mM MgCl2, respectively,
as compared to residual 45% activity without MgCl2. The Arrhenius
plot of the temperature dependence of b-galactosidase from L.
plantarum WCFS1 was found to be linear in the range of 30–
55 °C for oNPG and 20–55 °C for lactose. The energy of activation
(Ea) was calculated as 39.6 and 42.7 kJ molÀ1 for oNPG and lactose,
respectively.
2.3. Kinetic measurements
The kinetic behaviour of b-galactosidase from L. plantarum
WCFS1 with its natural substrate lactose and the artificial substrate
oNPG was studied, and results are summarized in Table 1. The Km
values calculated for oNPG (0.9 mM) and lactose (29 mM) are in
accordance with b-galactosidases from a number of different
sources, for which it was shown that Michaelis constants for the
artificial substrate oNPG are significantly lower than those for lac-
tose.19 The Km value of 29 mM determined for lactose is compara-
ble with the reported values for other b-galactosidases from
Lactobacillus spp., for example, 31 mM for L. reuteri L461 and
13 mM for L. reuteri L10317 and 14 mM for Lactobacillus crispatus.20
The Km value from L. plantarum b-galactosidase compares further-
more very favourably with the reported values for some commer-
cial b-galactosidases (36–180 mM for Aspergillus oryzae; 54–
100 mM for Aspergillus niger; 15–52 mM for Kluyveromyces fragilis
and 35 mM for Kluyveromyces lactis)21,22 which is important when
turning over lactose in lower concentrations. The kcat values were
calculated on the basis of theoretical vmax values obtained by
non-linear regression analysis using Sigma Plot (SPSS Inc.). The
specificity constants (kcat/Km) for two substrates again indicate that
the hydrolytic activity of b-galactosidase towards oNPG is better
than that towards lactose. This is mainly the result of a lower
Michaelis constant (Km) for oNPG (Table 1).
2.6. Synthesis of galacto-oligosaccharides
Several batch conversion experiments were carried out at 2 mL
scale and 37 °C using an initial lactose concentration of 600 mM pre-
pared in 50 mM sodium phosphate buffer (pH 6.5) containing 1 mM
MgCl2 and 1 ULac of b-galactosidase activity from L. plantarum
WCFS1 per mL reaction mixture to investigate the formation of ga-
lacto-oligosaccharides (GOS) and individual components of this
GOS mixture in detail. During lactose conversion, the relative
amount of GOS increased constantly up to ꢀ85% lactose conversion,
and reached a maximum of 41% of total sugars at this conversion
rate. Subsequently, hydrolysis prevailed over transgalactosylation,
Table 2
Effect of various chemicals and cations on the activity of purified b-galactosidase from
L. plantarum WCFS1a
2.4. Effects of various chemicals and cations on b-galactosidase
activity
Reagents
Relative activity (%)
10 mM
1 mM
100 mM
The effects of various chemicals and cations on the activity of b-
galactosidase, each added in final concentrations of 1–100 mM to
Blank (none)
100
115
294
158
262
17
100
326
341
165
414
22
100
197
nd
145
399
13
K+
Mn2+
Table 1
Na+
Kinetic parameters for b-galactosidase from L. plantarum WCFS1 for the hydrolysis of
o-nitrophenyl-b-D-galactopyranoside (oNPG) and lactose
Mg2+
EDTA
Urea
107
110
157
105
113
148
48
91
84
Kinetic parameters
Substrates
2-Mercaptoethanol
1,4-Dithiothreitol
Lactose
29
12.9 0.5
98
oNPG
Km (mM)
1
0.9 0.4
153
92
nd, not determined.
vmax
kcat (sÀ1
cat/Km (mMÀ1 sÀ1
(
l
mol minÀ1 mgÀ1
)
)
2
a
Enzyme activity was determined under standard assay conditions using oNPG
as the substrate with the respective chemical/cation added to give the stated final
concentration.
k
)
3.3 0.5
130 50

S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
1411
terium bifidum NCIMB 4117124) and very favourably with yields gi-
ven for fungal b-galactosidases (e.g., b-galactosidase from A. oryzae
36%25 or 22%26). In a recent review27 it has been stated that typical
optimized yields for the production of GOS are in the range of 30–
40%. Hence, the b-galactosidase from L. plantarum shows a very
pronounced transgalactosylation potential, especially when consid-
ering that the initial lactose concentration of 205 g LÀ1 (correspond-
ing to the solubility of lactose at ambient temperature) used in our
experiments is relatively low compared to that in other studies,
and that the lactose concentration is considered as one of the main
factors affecting maximum GOS concentrations that can be
achieved.27
100
80
60
40
20
0
3
4
5
6
7
8
9
10
In a previous study,23 we produced GOS using b-galactosidase
from L. reuteri L103, and major components of this GOS mixture
were characterized. In order to analyze the GOS components ob-
tained by transgalactosylation using b-galactosidase from L. planta-
rum in more detail, these GOS were similarly characterized using
authentic standards and the standard addition technique and
compared with the GOS mixture produced by b-galactosidase from
L. reuteri using thin layer chromatography (TLC), high performance
anion exchange chromatography with pulsed amperometric
detection (HPAEC-PAD) and capillary electrophoresis. The thin
layer chromatogram (Fig. 4) as well as the chromatograms from
HPAEC-PAD (Fig. 5) and capillary electrophoresis (Fig. 6) indicates
that the major GOS components produced by L. reuteri L103 and L.
plantarum WCFS1 are quite similar. Furthermore, a comparison by
TLC showed that the GOS mixture obtained with the lactobacillal
enzymes is considerably different from a commercially available
GOS preparation, Elix’or (Friesland Foods Domo), which is com-
posed mainly of b-(1?4)-linked oligosaccharide products.9 The L.
plantarum b-galactosidase shows a strong preference for the for-
mation of b-(1?6) linkages in its transgalactosylation mode, while
b-(1?3)-containing compounds are formed to a lesser extent, with
approximately 80% and 9% of the newly formed transgalactosyla-
tion products containing a b-(1?6) or b-(1?3) linkage, respec-
tively. The main components of this GOS mixture were identified
pH
100
80
60
40
20
0
10
20
30
40
50
60
70
80
Temperature (°C)
Figure 2. pH (A) and temperature optimum (B) of the activity of b-galactosidase
from L. plantarum WCFS1. Symbols: (s), oNPG as the substrate; (d), lactose as the
substrate.
which finally resulted in the complete degradation of these newly
formed oligosaccharide products as well as lactose. After 8 h of reac-
tion (corresponding to 85% lactose conversion), GOS reached a max-
imum level of 84 g LÀ1 of total sugars, and this GOS mixture was
composed of 19% non-lactose disaccharides, 21% trisaccharides
and 1.3% tetrasaccharides, relative to the total sugars in the reaction
mixture (Fig. 3). This yield of GOS (41% w/w of total sugars) produced
by b-galactosidase from L. plantarum compares favourably with the
yields reported for other bacterial b-galactosidases (36% for b-galac-
tosidase from L. reuteri L103,23 38.5% for b-galactosidase from
L. acidophilus R2216 and 36–43% for b-galactosidase from Bifidobac-
as
(1?6)-
(1?3)-Lac.
D
-Galp-(1?6)-
D
-Glc (allolactose),
D
-Galp-(1?3)-
D
-Glc,
D
-Galp-
-Galp-
D
-Gal, -Galp-(1?3)-
D
D
-Gal, -Galp-(1?6)-Lac and
D
D
As mentioned above, the GOS mixtures produced by the two
lactobacillal enzymes are quite similar. Distinct differences are that
the L. plantarum b-galactosidase has a stronger preference for the
formation of the main trisaccharide compounds
D-Galp-(1?6)-
Lac and -Galp-(1?3)-Lac (Fig. 5B, peaks 7 and 9, respectively,
D
and Table 3) as compared to L. reuteri b-galactosidase, while on
the other hand the yet unidentified component with retention time
(RT) of ꢀ25 min and the individual non-lactose disaccharides are
formed in significantly lesser amounts (Table 3). A careful study
of the products formed by individual b-galactosidases thus gives
the opportunity for more selective production of specific structural
motifs in GOS products through choice of an individual biocatalyst
as well as reaction engineering.27
Figure 7 shows a detailed analysis of the individual main com-
ponents formed during the transgalactosylation reaction of b-
galactosidase from L. plantarum using lactose as substrate. After
only 10% lactose conversion the two main transgalactosylation
40
30
20
10
0
products were
ing this initial phase of the reaction the concentrations of these
two sugars -Gal and -Glc, which serve as galactosyl acceptors,
D-Galp-(1?6)-D-Gal and D-Galp-(1?6)-D-Glc. Dur-
D
D
are relatively low. This indicates that both monosaccharides are
excellent acceptor molecules for transgalactosylation, or that at
0
20
40
60
80
100
Lactose conversion (%)
least
D
-Galp-(1?6)-
D
-Glc is formed by intramolecular transgalac-
-Gal moiety is transferred onto -Glc before
Figure 3. Analysis of total sugars during lactose conversion catalyzed by b-
galactosidase from L. plantarum WCFS1. The reaction was performed at 37 °C with
205 g LÀ1 lactose as substrate in 50 mM sodium phosphate buffer (pH 6.5) and
1 mM MgCl2. Symbols: (d), glucose; (s), galactose; (j), disaccharides other than
tosylation, that is, the
it can leave the active site of b-galactosidase and another acceptor
molecule or water can enter (Scheme 1). Up to a lactose conversion
D
D
*
lactose; (N), trisaccharides; (h), tetrasaccharides; ( ), total galacto-oligosaccharides.
of approximately 50%, the formation of D-Galp-(1?6)-Lac proceeds

1412
S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
Glc
Gal
Lac
GOS
A
Glc
Gal
Lac
GOS
B
E
LGG
1
1.5
8
E
0
0.5
2
3
5
10
24 LGG
Figure 4. Hydrolysis of lactose catalyzed by b-galactosidase from L. plantarum WCFS1 (A) and L. reuteri L103 (B) as analyzed by thin layer chromatography (TLC) on pre-
activated silica plates (eluent: n-butanol–n-propanol–ethanol–water = 2:3:3:2). The reaction was carried out at 37 °C with an initial lactose concentration of 205 g LÀ1 in
50 mM sodium phosphate buffer (pH 6.5) containing 1 mM MgCl2 and 1.0 ULac mLÀ1 of recombinant b-galactosidase. The samples were taken at regular time intervals during
the reaction. A commercially available GOS preparation, Elix’or (Friesland Foods Domo), was added for comparison (indicated by E). LGG indicates a standard mixture of
lactose, glucose and galactose.
almost linearly, while the formation of
pronounced and the potential galactosylation product
D
-Galp-(1?3)-Lac is far less
-Galp-
3.2. Bacterial strains and culture conditions
D
(1?4)-Lac was not identified at all. This again confirms the strong
preference of b-galactosidase from L. plantarum to form (1?6)
linkages in its transgalactosylation reaction. At maximal GOS pro-
L. plantarum WCFS1 was originally from the National Collection
of Industrial and Marine Bacteria (Aberdeen, UK), and the construc-
tion of recombinant L. plantarum WCFS1 carrying the expression
plasmid pEH3P for the homologous overexpression of the genes
lacLM encoding b-galactosidase from L. plantarum WCFS1 was re-
cently described.14 The strain was grown anaerobically overnight
at 37 °C in MRS broth containing 10 g LÀ1 peptone, 10 g LÀ1 meat
extract, 5 gÀL1À1 yeast extract, 2 g LÀ1 di-potassium hydrogen phos-
phate, 2 g L di-ammonium hydrogen citrate, 5 g LÀ1 sodium ace-
tate, 0.2 g LÀ1 magnesium sulfate, 1 g LÀ1 Tween 80 and 0.025 g LÀ1
manganese sulfate. Glucose 2% (w/v) served as the C-source (Lacto-
bacillus broth29). Whenever necessary, the MRS medium contained
duction, b-
ride formed comprising 34% of total GOS, followed by b-
(1?6)- -Glc as second most important GOS component formed
D
-Galp-(1?6)-
D
-Lac is the most abundant oligosaccha-
D-Galp-
D
(29% of total GOS). While current commercial GOS products con-
tain structures with predominant b-(1?4) linkages,9,10 the lacto-
bacilli enzymes form mainly b-(1?6) linkages in their
transgalactosylation products. It has been suggested that the for-
mation of a (1?6) bond is favoured since the C6–OH group is
the most sterically accessible alcohol group of the galactosyl accep-
tor, and that several other b-galactosidases also preferentially pro-
duce b-(1?6) linkages.27 In a study on the structure–function
relationship of various disaccharides with respect to their prebiotic
effect it was shown that amongst a group of galactose-containing
disaccharides, those containing a (1?6) linkage supported growth
of bifidobacteria best in mixed culture populations. Because of this
strong bifidogenic effect of these compounds, a GOS mixture pro-
duced with b-galactosidase from L. plantarum that is rich in these
b-(1?6)-linked oligosaccharides should therefore be of consider-
able interest as a prebiotic mixture.28
5 l
g mLÀ1 erythromycin to stably maintain the plasmid in recom-
binant L. plantarum. For the homologous over-expression of the b-
galactosidase genes lacLM from L. plantarum WCFS1, an overnight
culture of L. plantarum WCFS1 harbouring plasmid pEH3P was di-
luted in 50 mL of fresh MRS medium containing erythromycin to
an OD600 of <0.1 and incubated at 37 °C to an OD600 of 0.3. The cul-
tures were then induced with 25 ng mLÀ1 of the peptide phero-
mone IP-673 (supplied by the Molecular Biology Unit, University
of Newcastle-upon-Tyne, UK).
3.3. Purification of b-galactosidase from Lactobacillus
plantarum WCFS1
3. Experimental
Cells were harvested by centrifugation (8000g, 20 min at 4 °C)
when the stationary phase of microbial growth was reached (OD
at 600 nm of approx. 1.8). The cells were washed twice with
50 mM sodium phosphate buffer (pH 6.5) and resuspended in the
same buffer, containing 1 mM of both DTT and PMSF. Cell disrup-
tion was carried out with a French Press (Aminco, Maryland) and
cell debris was removed by centrifugation (25,000g, 25 min, 4 °C)
to obtain the crude cell extract. A single-step purification of
b-galactosidase was performed using affinity chromatography on
3.1. Chemicals and enzymes
All chemicals used in this study were purchased from Sigma
(St. Louis, MO, USA) unless otherwise stated, and were of the
highest quality available. Phenylmethanesulfonyl fluoride (PMSF)
99% and glucose oxidase (GOD) from A. niger (lyophilized,
205 U mgÀ1 enzyme preparation) were from Fluka (Buchs, Swit-
zerland), while 1,4-dithiothreitol (DTT) was from Roth (Karlsruhe,
Germany). The test kit for the determination of
D-glucose and
a 10-mL column of p-aminobenzyl 1-thio-b-
immobilized on cross-linked 4% beaded agarose (Sigma). The
column was pre-equilibrated with buffer (50 mM sodium
D-galactopyranoside
D-galactose was purchased from Megazyme (Bray, Ireland). All
chromatographic materials were from Amersham Biosciences
(Uppsala, Sweden).
A

S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
1413
Gradient 1
75.0
nC
A
60.0
50.0
40.0
30.0
20.0
10.0
0.0
2
1
5
4
3
1
2
1
2
min
-10.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
Gradient 2
250
200
150
100
50
nC
B
4+5
7
6
3
1+2
9
8
21
21
min
-20
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
50.0
55.0
60.0
65.0
0.0
Figure 5. Separation and quantification by HPAEC-PAD of individual galacto-oligosaccharides (GOS) produced by transgalactosylation with lactose as substrate catalyzed by
b-galactosidase from L. plantarum (blue/black) and compared with well-characterized GOS10 produced by L. reuteri L103 b-galactosidase (red/grey). Two different gradients
were used in HPAEC-PAD as give in Section 3. Initial lactose concentration was 205 g LÀ1 in 50 mM sodium phosphate buffer (pH 6.5). The extent of substrate conversion was
approximately 85%. The identified compounds are indicated: (1), galactose; (2), glucose; (3), D-Galp-(1?6)-D-Gal; (4), -Galp-(1?6)-D-Glc (allolactose); (5), lactose; (6), D-
Galp-(1?3)-
D
-Gal; (7),
D
-Galp-(1?6)-Lac; (8),
D
-Galp-(1?3)-
D
-Glc; and (9),
D
-Galp-(1?3)-Lac. Peaks marked with an  Dare not-yet-identified GOS. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)
phosphate buffer, pH 6.5) and then the enzyme was eluted with
buffer B (50 mM sodium phosphate and 2 M NaCl, pH 6.5) using
a linear gradient from 0% to 100% buffer B in 15 column volumes.
The flow rate was adjusted to 0.5 mL minÀ1. The purified enzyme
was desalted, concentrated (Amicon Ultra Centrifugal filter tubes;
10 kDa cut-off; Millipore, Billeria, MA, USA) and stored at 4 °C in
50 mM sodium phosphate buffer, pH 6.5.
solution was added to 480 lL of 22 mM oNPG prepared in
50 mM sodium phosphate buffer (pH 6.5) to start the reaction,
which was stopped after 10 min of incubation at 30 °C and shaking
at 600 rpm by adding 750 lL of 0.4 M Na2CO3. The release of o-
nitrophenol (oNP) was measured by determining the absorbance
at 420 nm. One unit of oNPG activity (UoNPG) was defined as the
amount of enzyme liberating 1 lmol o-nitrophenol (oNP) per min-
ute under the described conditions.
3.4. Enzyme assays
When lactose was used as the substrate, 20 lL of enzyme solu-
tion was added to 480 lL of 600 mM lactose solution in 50 mM so-
b-Galactosidase activity was determined using both o-nitro-
dium phosphate buffer, pH 6.5. After 10 min of incubation at 30 °C
and shaking at 600 rpm, the reaction was stopped by heating the
reaction mixture at 99 °C for 5 min. The reaction mixture was cooled
phenyl-b-
mogenic oNPG was used as the substrate, 20
D
-galactopyranoside (oNPG) and lactose.17 When chro-
lL of the enzyme

1414
S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
Monosaccharides
1
mAU
70
A
60
50
40
30
20
10
2
Disaccharides
3
Trisaccharides
7
5
6
Tetrasaccharides
4
8
min
14
16
18
20
22
24
26
28
mAU
70
Monosaccharides
1
B
60
50
40
30
20
10
2
Disaccharides
3
5
6
Trisaccharides
7
8
Tetrasaccharides
4
min
14
16
18
20
22
24
26
28
Figure 6. Separation and quantification by capillary electrophoresis of individual galacto-oligosaccharides produced during the lactose conversion catalyzed by b-
galactosidase from L. plantarum WCFS1 (A) and compared with GOS produced by the enzyme from L. reuteri L103 (B). The samples are mixtures of sugars obtained after the
reaction of b-galactosidase with 205 g LÀ1 initial concentration of lactose. The extent of substrate conversion was approximately 85%. The identified compounds are indicated
as follows: (1), glucose; (2), galactose; (3), lactose; (4),
D-Galp-(1?3)-D-Glc; (5), D-Galp-(1?6)-D-Glc (allolactose) with D-Galp-(1?3)-D-Gal; (6), D-Galp-(1?6)-D-Gal; (7), D-
Galp-(1?6)-Lac; and (8),
tetrasaccharides.
D-Galp-(1?3)-Lac. Products marked with an  are minor components and were not identified. Peaks appearing at a retention time ꢀ27 min are
Table 3
Concentrations of individual GOS components produced by the transgalactosylation reaction with lactose as substrate and catalyzed by b-galactosidases from L. plantarum WCFS1
(A) and L. reuteri L103 (B)
GOS components
Concentration of individual GOS components (g LÀ1) at different lactose conversion levels
15% 35% 50% 66% 85%
8%
95%
A
B
A
B
A
B
A
B
A
B
A
B
A
B
6.10
2.90
nd
0.81
nd
1.17
nd
2.90
2.10
5.80
1.60
0.35
6.54
1.90
4.62
0.36
4.52
3.78
0.33
8.18
1.78
1.54
0.76
5.76
5.32
8.04
0.69
10.6
11.32
0.83
29.3
1.88
11.2
1.52
11.4
14.5
8.82
14.3
1.37
29.7
0.54
12.9
22.1
13.5
24.1
18.8
27.4
4.24
nd
1.78
3.37
3.94
1.29
5.02
0.60
0.08
0.38
D
D
D
D
D
D
-Galp-(1?6)-
-Galp-(1?6)-
-Galp-(1?3)-
D
D
D
-Gal
5.26
0.50
-Glc (allolactose)
-Gal
nd
3.97
1.10
nd
0.55
1.04
19.6
4.70
1.59
24.9
1.42
1.98
28.3
1.02
2.18
22.3
3.02
21.5
17.5
2.42
5.26
0.88
0.49
0.24
-Galp-(1?6)-Lac
-Galp-(1?3)- -Glc
-Galp-(1?3)-Lac
2.12
9.79
0.95
D
2.97
1.93
7.33
3.00
9.18
2.57
4.85
4.41
6.15
3.06
4.59
4.51
Unidentified compounda
nd, not determined due to low concentration.
a
Unidentified compound, corresponding to the peak with retention time ꢀ25 min in Figure 5.

S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
1415
3.6. Steady-state kinetic analysis
75
60
45
30
15
0
All steady-state kinetic measurements were performed at 30 °C
using oNPG and lactose as the substrates in 50 mM sodium phos-
phate buffer, pH 6.5, with concentrations ranging from 0.1 to
22 mM for oNPG and from 0.1 to 600 mM for lactose. The kinetic
parameters were calculated by non-linear regression and the ob-
served data were fit to the Henri–Michaelis–Menten equation
using SigmaPlot (SPSS Inc., Chicago, IL, USA).
3.7. Temperature and pH profile of b-galactosidase activity
The temperature optimum of the hydrolytic activity of b-galac-
tosidase from L. plantarum was determined by assaying the en-
zyme samples over the range of 20–70 °C for 10 min using oNPG
and lactose as substrates. Thermal stability was determined by
incubating aliquots of the enzyme in 50 mM sodium phosphate
buffer (pH 6.5) at the desired temperatures. Aliquots were re-
moved at various times and assayed under standard oNPG assay
conditions to determine the residual activity. The effect on stability
of different concentrations of MgCl2 was also studied at different
temperatures. Arrhenius plots were prepared for the determina-
tion of energy of activation (Ea) for both substrates.
For the determination of the optimum pH of the enzyme activ-
ity, oNPG and lactose assays were performed for a pH range of 4–9
using Britton–Robinson buffer (containing 20 mM each of phos-
phoric, acetic and boric acids adjusted to the required pH with
NaOH). To determine the pH stability of L. plantarum b-galactosi-
dase, enzyme samples were incubated at pH values ranging from
4.0 to 9.0 at 30 °C for up to 24 h, and the remaining enzyme activity
was measured at different time intervals using oNPG standard as-
say conditions.
0
20
40
60
80
100
Lactose conversion (%)
30
25
20
15
10
5
0
0
20
40
60
80
100
Lactose conversion (%)
Figure 7. Formation of individual galacto-oligosaccharides during lactose conver-
sion by L. plantarum WCFS1 b-galactosidase. The reaction was performed at 37 °C at
an initial concentration of 205 g LÀ1 lactose in 50 mM sodium phosphate buffer (pH
6.5) + 1 mM MgCl2. Symbols: (j),
D
-Galp-(1?3)-
D
-Glc; (N),
D-Galp-(1?6)-D-Glc;
-Galp-(1?3)-Lac.
3.8. Effect of various cations and reagents
(d), -Galp-(1?6)- -Gal; (h), -Galp-(1?6)-Lac; (s), D
D
D
D
To study the effect of various cations and reagents on b-galacto-
sidase activity, the enzyme samples were assayed at 30 °C for
10 min with 22 mM oNPG (in 50 mM sodium phosphate buffer,
pH 6.5) as the substrate in the presence of various cations and re-
agents added in final concentrations of 1, 10 and 100 mM. The
measured activities were compared with the activity blank of the
enzyme solution determined under identical conditions but with-
out added cations/reagents.
to room temperature, and the release of D-glucose was determined
colourimetrically using the glucose oxidase–peroxidase (GOD–
POD) assay.30 One unit of lactase activity (ULac) was defined as the
amount of enzyme releasing 1 lmol of D-glucose per minute under
the given conditions. Protein concentrations were determined by
the dye-binding method according to Bradford using the Biorad re-
agent and bovine serum albumin (BSA) as a standard.
3.5. Gel electrophoresis analysis and active staining
3.9. Enzymatic synthesis of galacto-oligosaccharides
Native polyacrylamide gel electrophoresis (PAGE) and denatur-
ing sodium dodecyl sulfate polyacrylamide gel electrophoresis
(SDS–PAGE) were performed on a PhastSystem unit (Amersham,
Uppsala, Sweden) using precast polyacrylamide gels (PhastGel 8–
25, Amersham). For SDS–PAGE, the enzyme was pre-incubated
with Laemmli buffer for 5 min at 90 °C for Commassie blue staining
and 60 °C for active staining. Commassie brilliant blue R-250 was
used for the visualization of protein bands, and active staining
A 600-mM lactose solution was prepared in 50 mM sodium
phosphate buffer, pH 6.5, containing 1 mM MgCl2. Transgalactosy-
lation reactions were performed on a 2-mL scale at 37 °C using
300 rpm agitation and 1 ULac mLÀ1 final concentration of a homog-
enous preparation of b-galactosidase from L. plantarum. Samples
were withdrawn at specific time intervals and immediately trans-
ferred to 99 °C for 5 min to inactivate the enzyme. The samples
were stored at À18 °C for subsequent analysis. High performance
anion exchange chromatography with pulsed amperometric detec-
tion (HPAEC-PAD), capillary electrophoresis (CE) and thin layer
chromatography (TLC) were used to determine the composition
of GOS mixtures, following previously described methods.10 Two
different combinations of four eluents were used in HPAEC–PAD
for effective separation of GOS. Eluent A (100 mM NaOH), eluent
B (100 mM NaOH and 1 M NaOAc), eluent C (water) and eluent D
(100 mM NaOH and 50 mM NaOAc) were mixed to form two differ-
ent gradients. Gradient 1, 15% A and 85% C from 0 to 50 min and
gradient 2, 100% A from 0 to 20 min and from 0 to 100% D from
20 to 70 min. For eluent preparation, MilliQ water, 50% (w/v) NaOH
and NaOAc (BDH, Prolabo, Belgium) were used. After each run, the
was performed by using 4-methylumbelliferyl-b-D-galactoside
(MUG)17 as substrate, which was added directly onto the gel. Sam-
ples that were applied to the individual lanes contained approxi-
mately 1–2 lg of the total protein.
Isoelectric focusing was performed in the range of pH 3–10
using the Multiphor System (Amersham) and precast, dry poly-
acrylamide gels (CleanGel IEF, Amersham) rehydrated with carrier
ampholytes as recommended by the supplier with some modifica-
tions. The broad pI marker protein kit, pH 3–10, (Amersham) was
used to compare and determine the pI value. Protein bands were
visualized with Coomassie blue staining following the instructions
of the manufacturer.

1416
S. Iqbal et al. / Carbohydrate Research 345 (2010) 1408–1416
10. Splechtna, B.; Nguyen, T. H.; Steinböck, M.; Kulbe, K. D.; Lorenz, W.; Haltrich, D.
J. Agric. Food Chem. 2006, 54, 4999–5006.
11. Barboza, M.; Sela, D. A.; Pirim, C.; Locascio, R. G.; Freeman, S. L.; German, J. B.;
Mills, D. A.; Lebrilla, C. B. Appl. Environ. Microbiol. 2009, 75, 7319–7325.
12. Ouwehand, A. C.; Derrien, M.; de Vos, W.; Tiihonen, K.; Rautonen, N. Curr. Opin.
Biotechnol. 2005, 16, 212–217.
column was washed with 100% B for 10 min and re-equilibrated for
15 min with the starting conditions of the employed gradient. Indi-
vidual GOS compounds were identified and quantified by using
authentic standards and the standard addition technique.10
13. Rastall, R. A.; Maitin, V. Curr. Opin. Biotechnol. 2002, 13, 490–496.
14. Halbmayr, E.; Mathiesen, G.; Nguyen, T.-H.; Maischberger, T.; Peterbauer, C. K.;
Eijsink, V. G. H.; Haltrich, D. J. Agric. Food Chem. 2008, 56, 4710–4719.
15. Chaillou, S.; Champomier-Verges, M. C.; Cornet, M.; Crutz-Le Coq, A. M.; Dudez,
A. M.; Martin, V.; Beaufils, S.; Darbon-Rongere, E.; Bossy, R.; Loux, V.; Zagorec,
M. Nat. Biotechnol. 2005, 23, 1527–1533.
16. Nguyen, T. H.; Splechtna, B.; Krasteva, S.; Kneifel, W.; Kulbe, K. D.; Divne, C.;
Haltrich, D. FEMS Microbiol. Lett. 2007, 269, 136–144.
17. Nguyen, T. H.; Splechtna, B.; Steinböck, M.; Kneifel, W.; Lettner, H. P.; Kulbe, K.
D.; Haltrich, D. J. Agric. Food Chem. 2006, 54, 4989–4998.
Acknowledgements
S.I. thanks the Government of Pakistan (Higher Education Com-
mission—HEC) and T.T.N. thanks the Government of Austria (Tech-
nology Grant Southeast Asia) for a scholarship enabling the studies
in Austria.
References
18. Fortina, M. G.; Ricci, G.; Mora, D.; Guglielmetti, S.; Manachini, P. L. Appl.
Environ. Microb. 2003, 69, 3238–3243.
19. Samoshina, N. M.; Samoshin, V. V. Enzyme Microb. Tech. 2005, 36, 239–251.
20. Kim, J. W.; Rajagopal, S. N. Folia Microbiol. 2000, 45, 29–34.
21. de Roos, A. Industrial Enzymes: Enzymes in Dairy Applications. In Enzymes in
Industry; Aehle, W., Ed., 2nd ed.; Wiley-VCH: Weinheim, Germany, 2004; p
144.
22. Jurado, E.; Camacho, F.; Luzon, G.; Vicaria, J. M. Enzyme Microb. Tech. 2002, 31,
300–309.
23. Splechtna, B.; Nguyen, T. H.; Haltrich, D. J. Agric. Food Chem. 2007, 55, 6772–
6777.
1. Nakayama, T.; Amachi, T. b-Galactosidase, Enzymology. In Encyclopedia of
Bioprocess Technology: Fermentation, Biocatalysis, and Bioseparation; Flickinger,
M. C., Drew, S. W., Eds.; John Wiley: New York, 1999; Vol. 3, pp 1291–1305.
2. Sani, R. K.; Chakraborti, S.; Sobti, R. C.; Patnaik, P. R.; Banerjee, U. C. Folia
Microbiol. 1999, 44, 367–371.
3. Park, A. R.; Oh, D. K. Appl. Microbiol. Biotechnol. 2010, 85, 1279–1286.
4. Vasiljevic, T.; Jelen, P. Innov. Food Sci. Emerg. Technol. 2002, 3, 175–184.
5. Somkuti, G. A.; Dominiecki, M. E.; Steinberg, D. H. Curr. Microbiol. 1998, 36,
202–206.
24. Goulas, A.; Tzortzis, G.; Gibson, G. R. Int. Dairy J. 2007, 17, 648–656.
25. Reuter, S.; Rusborg, N. A.; Zimmermann, W. Enzyme Microb. Tech. 1999, 25,
509–516.
26. Matella, N. J.; Dolan, K. D.; Lee, Y. S. J. Food Sci. 2006, 71, C363–C368.
27. Gosling, A.; Stevens, G. W.; Barber, A. R.; Kentish, S. E.; Gras, S. L. Food Chem.
2010, 121, 307–318.
28. Sanz, M. L.; Gibson, G. R.; Rastall, R. A. J. Agric. Food Chem. 2005, 53, 5192–5199.
29. De Man, J. C.; Rogosa, M.; Sharpe, M. E. J. Appl. Bacteriol. 1960, 23, 130–135.
30. Kunst, A.; Draeger, B.; Ziegenhorn, J. Colorimetric Methods with Glucose
Oxidase and Peroxidase. In Methods of Enzymatic Analysis; Bergmeyer, H. U.,
Bergmeyer, J., Graßl, M., Eds.; VCH: Weinheim, 1988; pp 178–185.
6. De Vries, M. C.; Vaughan, E. E.; Kleerebezem, M.; de Vos, W. M. Int. Dairy J. 2006,
16, 1018–1028.
7. Kleerebezem, M.; Boekhorst, J.; van Kranenburg, R.; Molenaar, D.; Kuipers, O.
P.; Leer, R.; Tarchini, R.; Peters, S. A.; Sandbrink, H. M.; Fiers, M. W. E. J.;
Stiekema, W.; Lankhorst, R. M. K.; Bron, P. A.; Hoffer, S. M.; Groot, M. N. N.;
Kerkhoven, R.; de Vries, M.; Ursing, B.; de Vos, W. M.; Siezen, R. J. Proc. Natl.
Acad Sci. U.S.A. 2003, 100, 1990–1995.
8. Gibson, G. R.; Probert, H. M.; Van Loo, J.; Rastall, R. A.; Roberfroid, M. B. Nutr.
Res. Rev. 2004, 17, 259–275.
9. Coulier, L.; Timmermans, J.; Bas, R.; van den Dool, R.; Haaksman, I.; Klarenbeek,
B.; Slaghek, T.; van Dongen, W. J. Agric. Food Chem. 2009, 57, 8488–8495.