Catalytic properties of Talaromyces thermophilus α-l- arabinofuranosidase and its synergistic action with immobilized endo-β-1,4-xylanase

Article abstract of DOI:10.1016/j.molcatb.2010.11.003

When grown on wheat bran, Talaromyces thermophilus produces a wide spectrum of hemicellulases, mainly endo-β-1,4-xylanase, α-l- arabinofuranosidase and β-xylosidase. The extracellular α-l-arabinofuranosidase was purified to homogeneity by sequential opera

Full text of DOI:10.1016/j.molcatb.2010.11.003

Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
Contents lists available at ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Catalytic properties of Talaromyces thermophilus ␣-l-arabinofuranosidase and its
synergistic action with immobilized endo-␤-1,4-xylanase
∗
Mohamed Guerfali, Ali Gargouri, Hafedh Belghith
Laboratoire de Génétique Moléculaire des Eucaryotes, Centre de Biotechnologie de Sfax, PB “1177” 3038 Sfax, Tunisia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 July 2010
Received in revised form 22 October 2010
Accepted 2 November 2010
Available online 10 November 2010
When grown on wheat bran, Talaromyces thermophilus produces a wide spectrum of hemicel-
lulases, mainly endo-␤-1,4-xylanase, ␣-l-arabinofuranosidase and ␤-xylosidase. The extracellular
␣
-l-arabinofuranosidase was purified to homogeneity by sequential operation of ammonium sulfate
precipitation, Q-sepharose column chromatography, gel filtration on Sephacryl S-200 and MonoQ col-
umn. The pure ␣-l-arabinofuranosidase had a specific activity of 49 U/mg of protein and was purified
2
6.7-fold. The molecular mass of the enzyme was estimated to be 35 kDa, determined by SDS-PAGE and
Keywords:
by gel filtration. The ␣-l-arabinofuranosidase exhibited maximal activity at pH 6.0–7.0 and an optimal
temperature at 55 C. The half-life of the ␣-l-arabinofuranosidase at 60 C was approximately 2 h and it
was very stable over a wide pH range for 24 h at 4 C. The apparent Michaelis constant Km value of the
␣-l-arabinofuranosidase was 0.77 mM for p-nitropenyl-␣-l-arabinofuranoside. The turnover number
(Kcat) and catalytic efficiency (Kcat/Km) were found to be 14.3 s and 1.8 104 M
ions such as Hg and Cu inhibited enzyme activity, whereas it was strongly activated by Mn . The
␣
-l-Arabinofuranosidase
◦
◦
Wheat arabinoxylan
Talaromyces thermophilus
Hemicellulases
◦
−1
−1 −1
Bioreactor
s
, respectively. Metal
2
+
2+
2+
␣
-l-arabinofuranosidase was specific for the ␣-linked arabinoside in the furanoside configuration and
can also retain 52% of its activity in the presence of p-nitrophenyl-␤-d-xylopyranoside as substrate. ␣-l-
arabinofuranosidase acted synergistically with the immobilized endo-␤-1,4-xylanase for the breakdown
of alkali-extracted arabinoxylan and in the improvement of xylobiose and monosaccharide production.
©
2010 Elsevier B.V. All rights reserved.
1
. Introduction
depolymerization relies on endo-1,4-␤-xylanase (EC 3.2.1.8) and
␤-xylosidase (EC 3.2.1.37) activities [5,6].
Plant cell wall arabinoxylans have complex structures that
In fact, ␣-l-arabinofuranosidases are exo-type enzymes
require the sequential and synergistic action of
ber of different enzymes for their complete hydrolysis [1].
Arabinoxylans are composed of backbone of (1,4)-linked
-d-xylopyranosyl residues where single ␣-arabinofuranosyl sub-
stituents are attached to the C(O)-2, C(O)-3 or to both C(O)-2,3 of
the xylose residues [2]. The xylan backbone can also be substituted
with ␣-d-glucopyranosyl uronic acid, or its 4-O-methyl derivative
and acetyl groups [3]. In addition, ferulic acid and pcoumaric acid
may be covalently linked to arabinoxylan via esterification at the
C5 position of some of the arabinosyl units [4].
Because of this heterogeneous composition of arabinoxylans,
their enzymatic degradation requires the action of a battery of
debranching and depolymerizing activities. Debranching activities
mainly include ␣-l-arabinofuranosidase (EC 3.2.1.55) and sub-
sequently, feruloyl esterases (EC 3.1.1.73), ␣-glucuronidases (EC
a
num-
that generally catalyse the cleavage of the terminal ␣-l-
arabinofuranosyl residues of arabinoxylan, arabinan and
arabinogalactan [7]. This class of enzymes is important since
removal of the arabinose residues from the xylan backbone
may have a synergistic effect with endoxylanases, presumably
because endo-xylanase hydrolysis of glycosidic bonds is inhibited
by arabinose-substituted xylose residues [8]. In general, these
enzymes are produced by bacteria [9], actinomycetes [10] and
several fungi [11–13] but only a few ␣-l-arabinofuranosidases
from thermophilic fungus had been reported [14]. Depending on
their amino acid sequences, ␣-l-arabinofuranosidases have been
classified into four of glycoside hydrolase families (families 43, 51,
54 and 62) (http://afmb.cnrs-mrs.fr/CAZY/). This classification is
based on the amino acid sequence of their active sites rather than
their substrates specificity [15].
a
␤
3
.2.1.139) and/or acetyl xylan esterases (EC 3.1.1.72), whereas
Researchers have recently realized that effective ␣-l-
arabinofuranosidase production is important for a wide range
of applications. It is, for instance, of particular importance for
the effective conversion of hemicellulosic biomass into fuels
and chemicals, the efficient delignification of pulp, the proper
∗ Corresponding author. Tel.: +216 74 875 818x1090; fax: +216 871 816.
E-mail address: hafeth.belghith@cbs.rnrt.tn (H. Belghith).
1
381-1177/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2010.11.003

M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
193
utilization of plant materials for animal feed [14] and the improved
hydrolysis of grape monoterpenyl glycosides during wine fermen-
tation [16]. ␣-l-arabinofuranosidase was used for polysaccharides
hydrolysis to liberate l-arabinose, which is used commercially as a
low-calorie sweetener. l-Arabinose inhibits sucrase and prevents
elevation of blood glucose levels induced by sucrose intake [17].
As a result, it can potentially inhibit obesity and prevent or treat
diseases associated with hyperglycemia [18].
In view of this, there is a need to develop suitable ␣-l-
arabinofuranosidases for use in the conversion of hemicelluloses
to fermentable sugars which could be used for the subsequent
production of bio-ethanol, and other value-added chemicals [6].
In this context, we have isolated a new Talaromyces thermophilus
strain named AX4 that proved efficiency in the production of large
amounts of thermostable xylanolytic enzymes, including an endo-
xylanase [19], a ␤-xylosidase [20], an ␣-l-arabinofuranosidase and
a ␤-d-mannosidase [21].
an oligoelements solution (MnSO , 1.6 g/l; ZnSO , 1.4 g/l; FeSO ,
4 4 4
5 g/l and CoCl , 2 g/l). The medium was sterilized by autoclaving
2
◦
(120 C, 20 min); appropriate antibiotics (ampicillin 50 ␮g/ml and
tetracycline 20 ␮g/ml) were added and subsequently inoculated by
8
2 ml suspension of spores (5.10 spores/ml). Ten milliliters broth
samples was taken regularly during the course of fermentation
and examined for growth, contamination, and hemicellulase activ-
ities. At the end of the fermentation (5 days), the mycelium was
harvested by centrifugation at 6000 rpm for 10 min, and the super-
natant was used to purify the ␣-l-arabinofuranosidase enzyme.
2.4. Enzyme essays
␣-l-Arabinofuranosidase, ␤-xylosidase and ␤-mannosidase
activities were determined by measuring the release
of
p-nitrophenol
from,
respectively,
p-nitrophenyl-␣-l-
arabinofuranoside (pNPA), p-nitrophenol-␤-d-xylopyranoside
(pNPX) and p-nitrophenol-␤-d-mannopyranoside (pNPM). The
assay mixture contained 200 ␮l of each substrate (2 mM) in 50 mM
potassium phosphate buffer (pH 7.0), and 200 ␮l of enzyme solu-
We describe here the purification and catalytic properties of the
␣-l-arabinofuranosidase produced by this fungus. We also studied
the mode of action of this enzyme and its synergistic effect with
immobilized endo-␤-1,4-xylanase for the breakdown of alkali-
extracted wheat arabinoxylan.
◦
tion. After incubation for 10 min at 50 C, the reaction was stopped
by the addition of 1.6 ml of Na2CO3 (1 M) [22]. The absorbance
(nm) due to the release of p-nitrophenol was measured at 405 nm.
2
. Materials and methods
One unit (IU) of enzyme activity was defined as the amount of
enzyme that released 1 ␮mole of p-nitrophenol/min in the reaction
mixture under the assay conditions.
2.1. Chemicals
The xylanase activity was assayed by measuring the reducing
groups released from Birchwood xylan [23]. The reaction mixture
consisted of 500 ␮l of 1% solution, 400 ␮l of 50 mM phosphate
buffer, pH 7.0, and 100 ␮l of enzyme solution. After incubation at
Q-Sepharose (Big beads), Sephacryl S-200 (high resolu-
tion) and Standard proteins (LMW gel filtration calibration
kit) were from Amersham Biosciences (Uppsala, Sweden).
◦
Xylan
p-nitrophenyl-␣-l-arabinopyranoside,
-d-xylopyranoside, p-nitrophenyl-␤-d-cellobioside,
(Oat
spelt),
p-nitropenyl-␣-l-arabinofuranoside,
50 C for 10 min, the liberated reducing sugars were determined by
p-nitropenyl-
DNS method [24]. One unit was defined as the amount of enzyme
that releases 1 ␮mol xylose equivalents/min under the assay con-
ditions.
␤
p-nitrophenol-␤-d-mannopyranoside, methylumbelliferyl-4,7-␤-
d-arabinofuranoside, and 3,5-dinitrosalicylic acid were purchased
from Sigma–Aldrich (Steinhiem, Germany). Polyethylene gly-
col (PEG) was from Fluka (Steinhiem, Germany). Column for
high-pressure liquid chromatography (HPLC, Bio-Sil SEC-125,
2.5. ˛-l-Arabinofuranosidase purification
The production of ␣-l-arabinofuranosidase by T. thermophilus
strain in submerged batch fermentation was examined. The strain
was grown in optimized nutrient medium containing wheat bran
7
.8 mm × 300 mm) was from Bio-Rad (Montpellier, France).
◦
2
.2. Micro-organism
as carbon source. Cultivation was performed at 50 C for 5 days.
The extracellular proteins were recovered by centrifugation and
the supernatant (1 l) was treated with ammonium sulfate (60%
saturation). The precipitate was collected by centrifugation at
9000 rpm for 15 min, dissolved in 20 mM phosphate buffer, pH
7.0, and then dialyzed overnight against the same buffer. The
dialyzed enzyme solution was loaded on a Q-sepharose column
(2.5 cm × 22 cm) pre-equilibrated with 20 mM phosphate buffer,
pH 8. The column was extensively washed with the same buffer.
␣-l-Arabinofuranosidase activity was eluted with a gradient of
The thermotolerant fungal strain was identified as T. ther-
mophilus stolk by CBS (Centraalbureau voor schimmelculturen,
Holland) Code reference: detail 274-2003. The deposit number of T.
thermophilus in the national strain bank of Tunisia is: Tunisian Col-
lection of Micro-organisms CTM10.103 (Centre of Biotechnology of
Sfax, Tunisia).
2
.3. Cultivation in bioreactor
0
–0.5 M NaCl in the same buffer, at a flow rate of 30 ml/h. The
The cultivation was carried out in a 3.6 l stirred tank bioreactor
Infors, AG GH-4103 Bottmingen, Switzerland) with 1.6 l working
volume. The bioreactor (dished bottom glass jacketed reactor) was
equipped with instrumentation for measurement and/or control
of agitation, temperature, pH and dissolved oxygen concentration.
The agitator was equipped with two 6-bladed Rushton impellers of
active fractions were pooled, concentrated using PEG 6000 and
dialyzed overnight against 20 mM phosphate buffer, pH 7.0. The
␣-l-arabinofuranosidase obtained from the ion exchanger was
further purified by gel filtration Sephacryl S-200 (high resolu-
tion) (1.5 cm × 96 cm) equilibrated and eluted by 25 mM phosphate
buffer, pH 7.0 at a flow rate of 0.5 ml/min. Fractions showing ␣-
l-arabinofuranosidase from the last purification step were then
pooled, concentrated by Centricon 10 (Amicon) and analyzed by
FPLC using Mono Q column anion exchange chromatography (Q12
Bio-Rad, 15 × 68 mm) equilibrated with 20 mM phosphate buffer,
pH 8. The column was operated at a flow rate of 5 ml/min and active
fractions (3 ml each) were eluted with a gradient of 0-0.5 M NaCl
in the same buffer. Purified ␣-l-arabinofuranosidase was dialyzed
overnight in 20 mM phosphate buffer, pH 7.0 and used for further
studies.
(
4
5
5 mm diameter. The cultivation temperature was maintained at
◦
0 C and dissolved oxygen was kept at above 20% of the saturation
for the medium by varying the agitation rate and flow rate of air. The
pH was controlled at 6.7 ± 0.1 by automatic addition of ammonium
hydroxide (4 M) and phosphoric acid (2 M). The T. thermophilus
was cultivated in the optimized liquid medium [21]: KH PO , 1 g;
2
4
K HPO , 2.5 g; MgSO , 1.2 g; CaCl , 0.3 g; yeast extract, 1 g; Tween
2
4
4
2
8
0, 1 ml; water, 1 l and 3.7% wheat bran as sole carbon source. The
pH of the medium was 6.7 and was supplemented with 1 ml/l of

1
94
M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
kinetic parameters, Km (mM) and Kcat (s−1) were calculated from
Lineweaver–Burk plot of Michaelis–Menten equation.
2.6. Molecular mass determination
The molecular mass of the purified ␣-l-arabinofuranosidase
was estimated using SDS-PAGE electrophoresis and gel filtration
HPLC). Sodium dodecyl sulfate-polyacrylamide gel electrophore-
2.10. Endo-xylanase immobilization
(
sis (SDS-PAGE) [25] was carried out to determine the purity
and the molecular weight of the enzyme. The gel was stained
with Coomassie brilliant blue R-250 and destained in 10%
T. thermophilus endo-xylanase retained 100% of its activity after
h of incubation at 100 C. This high thermostability makes diffi-
◦
1
cult the total elimination of the endo-xylanase by just boiling [19]
consequently we exploited its immobilization by covalent bind-
ing in chitosan support for subsequently subtracted from mixture
reaction by centrifugation or filtration. For this purpose (0.5 g) of
chitosan was dissolved in 50 ml of 0.1 M HCl containing 1.5% (v/v)
acetic acid and 20% of absolute ethanol.
weight calibration kit for SDS electrophoresis (Fermentas, France):
-galactosidase (116 kDa), Bovine serum albumin (66.0 kDa), oval-
bumin (45.0 kDa), lactate dehydrogenase (35.0 kDa), REase Bsp981
25.0 kDa), ␤-lactoglobulin (18.4 kDa), lysozyme (14.4 kDa) was
A low molecular
␤
(
◦
glutaraldehyde (GA) at 30 C for 2 h. The solublized chitosan was
used as molecular weight standard.
precipitated by the addition of 1 ml NaOH (1.0 mol/l). The precip-
itate was separated by centrifugation (10 min at 6000 rpm) and
washed with distilled water to remove excess of GA. The wet
chitosan was mixed with 2.0 ml (8.5 U/ml) of the endo-xylanase
The purified ␣-l-arabinofuranosidase was also injected on a
high performance liquid chromatography (HPLC; KNAUER advance
scientific instruments, Germany) gel filtration column (Bio-Sil SEC-
1
25, 7.8 mm × 300 mm, BioRad, France). The elution was realized
◦
solution and stirred at 4 C for 24 h. The unbound enzyme was
with 50 mM phosphate buffer pH 7.0 at a flow rate of a 0.8 ml/min.
Molecular weight standard from Amersham used to calibrate
the column were ribonuclease A (13.7 kDa), chymotripsinogen A
removed by washing with phosphate buffer 20 mM until no protein
or activity was detected [19]. We shall recall that the immobilized
endo-xylanase presented 81% and 72% of activity and immobiliza-
tion yield, respectively.
(
25 kDa), ovalbumin (43 kDa) and albumin (67 kDa).
.7. Zymogram analysis
-l-Arabinofuranosidase activity was also revealed by zymo-
2
2
.11. Chemical arabinoxylan extraction and monosaccharide
composition
␣
gram analysis. For this purpose, proteins are mixed with the same
loading buffer as in SDS-PAGE, but they are not boiled. After elec-
trophoresis, the gel is incubated for 2 h in 50 mM phosphate buffer
pH 7.0 to get rid of SDS and allowing the renaturation of proteins.
Then, the gel was superposed against an overlay of 1% agar con-
taining 5 mM of MUA or 2 mM of pNPA. After a suitable period of
Arabinoxylan fraction was chemically extracted from wheat
bran after two steps. The first step led to the delignification of
wheat bran and the second to the solubilization of the arabinoxy-
lan fraction. The delignification step adapted from [27], involves the
oxidation of sodium hypochlorite and allows not only the removal
of lignin but also the defibrillation of the polysaccharide network
due to the swelling of cellulose fibers. Therefore, it facilitates the
extraction of hemicellulose [27]. 50 g of destarched wheat bran
◦
incubation (15 min) at 50 C, activity was observed under UV light
(
MUA) or directly following the appearance of a yellow spot (pNPA).
.8. Influence of temperature, pH, and chemicals
-l-Arabinofuranosidase activity was determined under stan-
(
particle size less than 0.5 mm diameter) was incubated in 0.3%
2
◦
of H SO4 and 40% of NaClO2 at 70 C during 2 h [27]. The mix-
2
ture was centrifuged (15 min at 7000 rpm), the supernatant was
eliminated and the pellet was recovered and washed by distilled
water and lyophilized. The arabinoxylan extraction step was advo-
␣
dard conditions except different temperatures were assayed within
◦
the range 30–70 C as indicated in the results. The enzyme activ-
◦
cated by the incubation of delignified wheat bran at 60 C for 2 h
ity was then determined at various pH values ranging from 4.0 to
in the alkaline solution (KOH 1 M). The heat treatment in alkaline
medium, results in the saponification of the ester bonds linking fer-
ulic acid to arabinose, releasing individual arabinoxylan molecules
from cell wall structure [28]. After this step, the pH of the mix-
ture was neutralized by adding HCl solution 12 M and centrifuged
for 20 min at 7000 rpm. The supernatant was recovered and con-
1
1.0 at the optimal temperature. The thermostability of the enzyme
was evaluated by measuring the residual activity after each hours of
pre-incubation of the enzyme in the absence of substrate at various
temperatures between 50 and 70 C. The pH stability was studied
◦
◦
by pre-incubating the enzyme at pH ranging from 5.0 to 11.0 at 4 C
for 24 h and the residual enzymatic activity was determined under
standard conditions. The effects of metal ions and some chemicals
were assessed by pre-incubating the enzyme with the test com-
pound at 4 C for 15 min and then assaying the residual enzymatic
activity under standard conditions. Apo-␣-l-arabinofuranosidase
centrated. The heteroxylans were finally precipitated by ethanol
◦
(
2 Vol.; for 18 h; at 4 C), recovered by filtration (Whatman GF/A
◦
glass microfiber filters), washed with ethanol then acetone, and
◦
dried in an oven at 40 C [28]. The yield of arabinoxylan by this
extraction process was approximately 45% (Data not shown). To
determine the monosaccharide composition and content, 10% of
was made by dialyzing the enzyme against 5 mM EDTA dissolved in
2
0 mM phosphate buffer pH 7.0 for 20 h. The EDTA was removed by
◦
arabinoxylan was hydrolyzed using 0.4 M HCl for 2 h at 100 C.
dialyzing apo-enzyme intensively against 15 l of bi-distilled deion-
ized water in 24 h [26]. Activity of apo-␣-l-arabinofuranosidase
was determined in the presence of different ions (Ca , Mn , Co
and Na ). The enzyme was incubated for 2 h at 4 C in the presence
of each metal before measuring the activity.
After hydrolysis, the sample was cooled and subsequently neutral-
ized by addition of NaOH. Each sample was then filtered and the
monosaccharide composition and content (g/l) were determined
by HPLC.
2+
2+
2+
+
◦
2.9. Determination of kinetic parameters
2.12. High-performance liquid chromatography
The kinetic parameters of the purified enzyme were determined
The hydrolysis products were monitored by high-performance
liquid chromatography (HPLC; Bio-Rad Aminex 87-C, column
7.8 mm × 300 mm). A solution of simple sugar and disaccharide
(glucose, xylose, arabinose and xylobiose), at 1 g/l each, were used
as a standards.
for pNPA as substrate and were estimated using a specific pro-
gram of enzymology (Hyper 32 exe program, version 1.0.0, 2003).
The enzyme was incubated with pNPA at various concentrations
(
ranging from 0.5 to 20 mM) at optimal conditions. The enzyme

M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
195
1
1
,4
,2
1
A
0
0
0
0
,8
,6
,4
,2
0
0
50
100
150
Time (hours)
2
5
0
5
0
5
0
6
5
4
3
2
1
0
B
2
Fig. 2. SDS–PAGE of purified ␣-l-arabinofuranosidase from T. thermophilus (A) and
the overlay zymogram gel corresponding (B). The enzyme was electrophoresed at
pH 8.6 on a 10% acrylamide gel and stained with Coomassie Brillant Blue R-250.
M: Molecular weight standards; P: Purified ␣-l-arabinofuranosidase (∼10 ␮g); Z1:
1
1
␣-l-arabinofuranosidase activity was detected with 5 mM 4-methylumbelliferyl-␣-
l-arabinofuranoside (MUA); Z2: ␣-l-arabinofuranosidase activity was detected with
mM pNPA.
2
3.2. ˛-l-Arabinofuranosidase purification
0
24
34
48
56
72
96
120 144
Time (hours)
◦
Following growth of T. thermophilus for 120 h at 50 C on a
wheat bran-containing medium, the culture supernatant was con-
centrated with 60% of ammonium sulfate and directly adsorbed to
a Q-sepharose column. The ␣-l-arabinofuranosidase activity was
eluted as a sharp peak between 0.2 and 0.25 M NaCl. Following con-
centration and desalting of the active fractions, the specific activity
increased with 58% recovery of the activity (Table 1). This material
was applied directly to a Sephacryl S-200 column and the active
fractions were collected as a single sharp peak, which was applied
on to a Mono-Q column and a pure ␣-l-arabinofuranosidase was
finally eluted in a single peak. A summary of the purification pro-
cedures is presented in Table 1. The purified enzyme behaves as a
single band on SDS polyacrylamide gel electrophoresis (SDS–PAGE;
Fig. 2), coinciding with the activity band, as revealed on zymogram
with MUA or pNPA (Fig. 2B). The final purification step resulted in
a yield of 8% of the activity and a 26.7-fold increase in the specific
activity.
Fig. 1. Time course profiles of hemicellulases production by T. thermophilus in batch
fermentation using wheat bran as only carbon source. (A) Kinetic production of ␣-
l-arabinofuranosidase (ꢀ), ␤-xylosidase (ꢀ) and ␤-d-mannosidase (ꢁ). (B) Kinetic
production of the endo-xylanase (ꢀ) and reducing sugars (ꢂ). Enzyme and reducing
sugars essays were detailed in Section 2.
3
. Results
3.1. Hemicellulases production by T. thermophilus in batch
fermentation
The production of hemicellulases by T. thermophilus strain was
studied in a 3.6-l bioreactor by using previously optimized cul-
ture medium [21]. When grown on wheat bran as the only carbon
source, this filamentous fungus produces a wide spectrum of hemi-
cellulases (Fig. 1). In contrast, no cellulase activity was detected
in culture filtrate. Compared with the culture of T. thermophilus
in Erlenmeyer flask, the batch fermentation led to an improve-
ment in the rate of production of all xylanolytic hydrolases. The
xylanase activity is predominant with a production rate of 22 U/ml
instead of 5 U/ml after 120 h of culture [19]. Wheat bran induced
high titres of ␤-xylosidase (1.2 U/ml) and ␣-l-arabinofuranosidase
3.3. Enzyme properties
The molecular mass of the ␣-l-arabinofuranosidase was deter-
mined to be about 35 kDa, either by gel filtration (data not
shown) or following SDS-PAGE analysis, indicating that it exists
as monomeric protein.
(
0.85 U/ml), furthermore, it does not appear to be an effective
inducer of ␤-mannosidase activity (0.4 U/ml). The time course of ␣-
l-arabinofuranosidase production begins after 24 h of culture and
reached a maximum after 100 h corresponding to 2.25 U/mg of spe-
cific activity. At the end of the culture (120 h) a decrease of reducing
sugars (0.6 mg/ml) was observed in the growth medium.
The ␣-l-arabinofuranosidase displayed maximal activity at
◦
5
5 C (Fig. 3A) and between pH 6.0 and 7.0 (Fig. 4). The activ-
◦
◦
ity was stable over 7 h at 50 C and was slightly affected at 55 C
◦
with a half-life of 4 h. At 70 C, the enzyme was rapidly inactivated
and had a half-life of 30 min (Fig. 3B). After incubation for 24 h
Table 1
Summary of the purification strategy for ␣-l-arabinofuranosidase from T. thermophilus.
Step
Total protein (mg)
Total activity (U)
Specific activity (U/mg protein)
Recovery (%)
Purification fold
Culture supernatant
Ammonium Sulfate (60%)
Q-Sepharose
S-200 gel filtration
Mono-Q Sepharose
328
190
99
22
1.2
742
521
436
271
59
1.8
2.7
4.5
12.3
49
100
70
58
36
8
1
1.5
2.4
6.7
26.7

1
96
M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
A 1
1
20
00
B
1
1
20
00
8
6
4
2
0
0
0
0
0
8
6
4
2
0
0
0
0
0
20
40
60
80
0
1
2
3
4
5
6
7
8
Temperature (°C)
Time (hours)
Fig. 3. Effects of temperature on activity (A) and stability (B) of T. thermophilus ␣-l-arabinofuranosidase. Enzyme activity was assayed in (50 mM) phosphate buffer pH 7 at
◦
◦
◦
◦
◦
different temperatures. For thermal stability, the pure enzyme was incubated in the same buffer at 50 C (ꢀ), 55 C (ꢂ), 60 C (ꢃ), 65 C (ꢁ) and 70 C (ꢀ). For different time
intervals, residual activity was determined by standard method.
1
1
20
00
8
6
4
2
0
0
0
0
0
2
3
4
5
6
7
pH
8
9
10
11
12
Fig. 5. Effect of metals on the reactivation of apo-␣-l-arabinofuranosidase from T.
thermophilus. Error bars represent the standard deviation.
Fig. 4. Effects of pH on activity (ꢂ) and stability (ꢀ) of T. thermophilus ␣-l-
arabinofuranosidase. For pH optimum, enzyme activity was assayed by the standard
method after changing the buffer to obtain the desired pH. The buffers used were cit-
rate (pH 4.0–5.0), phosphate (pH 6.0–7.0), AMPSO (pH 8–9), and glycine-hydroxide
was significantly inhibited by Hg2+, Cu2+, and Zn2+ cations.
Inhibition was also observed in the presence of SDS and ethylenedi-
aminetetraacetic acid (EDTA). The enzymatic activity was slightly
(
4
pH 10–11). pH stability was determined by pre-incubating the enzyme for 24 h at
C at these same different pH and the residual activity was assayed by the standard
◦
2+
2+
2+
and Na+ ions.
stimulated in the presence of Mn , Ca , Co
method.
The removal of metals from ␣-l-arabinofuranosidase resulted
into almost complete loss of activity and apoenzyme of ␣-l-
arabinofuranosidase displayed only about 7.6% residual activity.
◦
at 4 C at various pH (ranging from 5.0 to 11.0) (Fig. 4), the ␣-l-
arabinofuranosidase was shown to be very stable between pH 6.0
and 8.0. We even found that the enzyme, after incubation at pH 6.0,
was slightly activated relatively to its starting activity, measured
under standard conditions. At pH 9.0, ␣-l-arabinofuranosidase
retains up to 70% of its initial activity.
2+
2+
2+
and Na+
The effect of different metal ions like Mn , Ca , Co
on the reactivation of apo-enzyme was determined (Fig. 5). The
activity of apo-␣-l-arabinofuranosidase was recovered (48% of
2+
reactivation) in the presence of 1 mM Mn , other metals exhibited
a low reactivation rate.
The influence of various reagents on the purified ␣-l-
arabinofuranosidase activity was studied (Table 2). The activity
3.4. Substrate specificity to pNP glycosides
Table 2
Relative rate of hydrolysis of various p-nitrophenyl gly-
Effect of various reagents on the activity of the purified T. thermophilus ␣-l-
arabinofuranosidase. The data presented are an averages and standard errors of
two independent experiments.
cosides by the purified ␣-l-arabinofuranosidase were exam-
ined (Table 3). The enzyme showed the highest activity
towards p-nitrophenyl-␣-l-arabinofuranoside (100%) but was
not able to hydrolyse p-nitrophenyl-␣-l-arabinopyranoside. ␣-
l-Arabinofuranosidase was less active (52%) on p-nitrophenyl-
Reagents
Concentration (mM)
Relative activity (%)
Control
CaCl2
BaCl2
CoCl2
MnSO4
MgSO4
ZnCl2
FeSO4
CuSO4
HgCl2
NaCl
–
5
5
5
5
5
5
5
5
100
133 ± 6
99 ± 4
123 ± 2
135 ± 8
114 ± 3
11 ± 4
58 ± 4
7 ± 2
␤
-d-xylopyranoside as substrate. Only 3.4% of that activity was
found with p-nitrophenyl-␤-d-mannopyranoside and no activity
Table 3
Substrate specificity of purified ␣-l-arabinofuranosidase.
Synthetic substrate
Relative activity (%)
5
5
2 ± 0
119 ± 5
13 ± 6
118 ± 2
120 ± 1
100 ± 2
30 ± 6
p-Nitrophenyl-␣-l-arabinofuranoside
p-Nitrophenyl-␣-l-arabinopyranoside
p-Nitrophenyl-␤-d-xylopyranoside
p-Nitrophenyl-␤-d-glucopyranoside
p-Nitrophenyl-␤-d-mananoside
p-Nitrophenyl-␤-d-cellobioside
100
0
EDTA
Urea
10
10
10
10
10
52 ± 4
2
-Mercaptoethanol
0
DTT
SDS
3.4 ± 1.5
0

M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
197
Fig. 6. Synergistic action of ␣-l-arabinofuranosidase and immobilized endo-␤-1,4-xylanase. Plots show the release of reducing sugars (A), xylose (B), arabinose (C), and
xylobiose (D) from 5% (w/v) wheat arabinoxylan. This substrate was treated for 4 h with 0.65 U/ml ␣-l-arabinofuranosidase alone (Ab); with immobilized family 11 endo-␤-
1
,4-xylanase alone (X), with immobilized endo-␤-1,4-xylanase for 4 h, with subsequent elimination by centrifugation, followed by ␣-l-arabinofuranosidase for 4 h (X/Ab);
with ␣-l-arabinofuranosidase for 4 h with subsequent inactivation by boiling, followed by endo-␤-1,4-xylanase for 4 h (Ab/X) or with both enzymes simultaneously (X + Ab).
could be detected with p-nitrophenyl-␤-d-glucopyranoside and p-
nitrophenyl-␤-d-cellobioside.
of liberated xylose and arabinose. On the other hand the endo-
xylanase was capable of releasing almost 30 mg/g of reducing
sugars (Fig. 6A), containing 1.26 mg/g, 0.4 mg/g and 19 mg/g of
xylose, arabinose and xylobiose, respectively (Fig. 6A–C). Fam-
ily 11 xylanases are known to be able to cleave arabinose from
polymeric substrates [29], although no activity against pNPA
could be detected. When both enzymes acted simultaneously,
we noticed a clear improvement in the rate of reducing sugar
release (>50 mg/g, superior to each enzyme acting individually).
Sequential treatment of the substrate with endo-xylanase fol-
lowed by ␣-l-arabinofuranosidase increases the rate of arabinose
released from 0.44 mg/g for the endo-xylanase alone to 6.2 mg/g
The rate dependence of the enzymatic reaction on the pNPA con-
◦
centration at pH 7.0 and 55 C followed Michaelis–Menten kinetics.
Reciprocal plot analysis showed an apparent Km value of 0.77 mM.
The turnover number (Kcat) and catalytic efficiency (Kcat/Km) of the
−1
−1 −1
s , respectively,
enzyme were found to be 14.3 s and 1.8 104 M
for pNPA.
3.5. Synergistic action with immobilized endo-xylanase
Analysis of hydrolysis products of arabinoxylan by HPLC shows
that it was composed of 38%, 23% and 6% of xylose (X), arabinose
A) and glucose, respectively. The A/X ratio of the arabinoxylan was
(
Fig. 6C). The ␣-l-arabinofuranosidase was active only when the
arabinoxylan was pre-treated initially with an endo-xylanase.
-l-Arabinofuranosidase was unable to release xylose from poly-
(
␣
then calculated to be 0.6. This extracted arabinoxylan fraction was
used to test the combinatory effects of ␣-l-arabinofuranosidase
and the immobilized family 11 endo-␤-1,4-xylanase (both purified
from culture filtrate of T. thermophilus) on the release of ara-
binose, xylose and xylobiose. Various combination of 0.65 U/ml
meric xylan but the simultaneous action of both enzymes released
more xylose than endo-xylanase acting alone (Fig. 6B). Treat-
ment of the arabinoxylan by the endo-xylanase produced 19 mg/g
of xylobiose. Simultaneous action with ␣-l-arabinofuranosidase
(Ab + X) can improve this rate to reach 26 mg/g. However, ␣-l-
␣
-l-arabinofuranosidase and 1 g of immobilized family 11 endo-
arabinofuranosidase was not able to release additional xylobiose
from wheat arabinoxylan after pre-treatment with endo-xylanase
(Fig. 6D). The treatment with ␣-l-arabinofuranosidase followed by
the endo-xylanase (Ab/X) did not improve hydrolysis efficiency of
arabinoxylan, due to the inability of the ␣-l-arabinofuranosidase
to change the texture of the polymer substrate.
xylanase (14 U/g of support) were incubated with 5% (w/v) wheat
bran arabinoxylan in 50 mM phosphate buffer. Reactions were
allowed to proceed at 55 C for 4 h. The hydrolysis reaction prod-
◦
ucts were analyzed by HPLC (xylose, arabinose and xylobiose were
used as standards). Substrates were incubated with each enzyme
individually, with both enzymes simultaneously, or sequentially
as follow: condition (a), first incubation of the substrate with the
immobilized endo-␤-1,4-xylanase for 4 h followed by its subse-
quent elimination by centrifugation (10 min at 7000 rpm), then
by incubation with ␣-l-arabinofuranosidase for a second cycle
for 4 h. Condition (b), consist in the incubation of the substrate
4. Discussion
The generation of fermentable sugars from lignocellulosic
resources is an important step in the biotransformation of renew-
able biomass into useful products such fuels, chemicals and
polymers [30]. It is clear that a complete xylanolytic system of
filamentous fungus including debranching and depolymerizing
activities is necessary to achieve a complete hydrolysis of hemicel-
lulose [31]. When cultivated on wheat bran as only carbon source,
T. thermophilus produces a wide spectrum of hemicellulases (Fig. 1).
This hemicellulase complex has been previously reported as a good
◦
with the ␣-l-arabinofuranosidase for 4 h at 50 C, followed by
◦
heat inactivation (15 min at 100 C) and finally a last cycle of
4
h beginning by the addition of the immobilized endo-xylanase.
Fig. 6 illustrates, that ␣-l-arabinofuranosidase was unable to act
alone in the presence of a complex substrate such as arabinoxy-
lan. In this condition, the rate of reducing sugars release was
very low, as confirmed by HPLC analysis, showing the absence

1
98
M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
alternative for lignocellulosic plant material saccharification and
particularly for arabinose and xylose production [21]. No cellulase
activity was detected in culture filtrate after growth on wheat bran.
In this context, the use of cellulase-free xylanases has a wide range
of biotechnological applications, especially in the pulp and paper
industries. They are used primarily as bleaching agents to reduce
the amount of chlorine required to achieve desirable levels of paper
brightness [32].
The pentose composition of the extracted arabinoxylan indi-
cated a 0.6 ratio of arabinose/xylose. This result was quite in
good agreement with previously reported data [44] and may indi-
cate the quality of the arabinoxylan fraction. The evidence from
synergy experiments (Fig. 6) suggests that T. thermophilus ␣-l-
arabinofuranosidase is unable to hydrolyze ␣-l-arabinofuranose
bound to internal xylosyl units of alkali-extracted arabinoxy-
lan. This inability was circumvented when the enzyme was
acting in concert with family 11 endo-xylanases. Clearly the
low-molecular-mass substituted xylo-oligosaccharide products
generated by endo-xylanase action are better substrates for the ␣-
l-arabinofuranosidase. Similar result had been reported previously
for Aspergillus awamori ␣-l-arabinofuranosidase [31], suggesting
that T. thermophilus ␣-l-arabinofuranosidase may belongs to family
51 of glycosyl hydrolases. This group of ␣-l-arabinofuranosidases
degrades only low molecular mass l-arabinofuranose-containing
xylooligosaccharides and synthetic substrates [15]. Several factors
can affect the binding of ␣-l-arabinofuranosidase to the arabinoxy-
lan, such as, the steric hindrance caused by the substituents on
the xylose backbone of the arabinoxylan molecule and the con-
formational changes of the polysaccharide chains especially after
hydrolysis of ester linkages by alkali treatment [31]. Treatment
of arabinoxylan with the simultaneous action of endo-xylanase
and ␣-l-arabinofuranosidase can, therefore, be considered as a
potential alternative for the eventual production of xylobiose. This
compound is a selective growth stimulant to intestinal Bifidobac-
terium, which is beneficial for the maintenance of healthy intestinal
microflora. The selective stimulative effect of xylobiose on Bifi-
dobacterium was greater than that of other oligosaccharides [45].
Treatment of the substrate with either only endo-xylanase or
sequential reactions of Ab/X should theoretically give the same
results but two main causes can affect the result. The first one is the
nature of alkali-extracted arabinoxylan (heterogeneity due to its
partial purification), the second one concerns experimental errors.
The synergistic interaction of the ␣-l-arabinofuranosidase and
endo-xylanase of T. thermophilus is obviously essential for the
extensive hydrolysis of complex arabinoxylan polymers.
To better exploit the T. thermophilus xylanolytic system, the
purification and characterization of the ␣-l-arabinofuranosidase
activity were investigated. Except ␣-l-arabinofuranosidase belong-
ing to family 62 that cannot hydrolyse effectively the pNPA as
substrate [33], the results observed after purification suggest
that the enzyme preparation from T. thermophilus contains only
one ␣-l-arabinofuranosidase enzyme. This was also confirmed by
overlay zymography method (data not shown). In contrast, multi-
ple forms of ␣-l-arabinofuranosidase have been detected in the
culture broths of Aspergillus niger [11] and Penicillium capsula-
tum [12]. Moreover, the analysis by gel filtration showed that
the ␣-l-arabinofuranosidase from T. thermophilus is a monomer
with an apparent molecular mass ∼35 kDa. Several fungi pos-
sess ␣-l-arabinofuranosidase with a low molecular mass, such as
Aspergillus sojae (34.4 kDa) [34], Aspergillus terreus (39 kDa) [35],
and Aspergillus nidulans (36 kDa) [36]. The molecular mass of fun-
gal ␣-l-arabinofuranosidase is ranging from 31 kDa to 105 kDa [6],
this indicates considerable diversity in ␣-l-arabinofuranosidase
molecular weight that influences biochemical characteristics. The
maximal activity of the purified ␣-l-arabinofuranosidase from T.
◦
thermophilus was observed at 55 C and pH 7.0. This enzyme was
very stable over a considerable temperature range, as observed
for several ␣-l-arabinofuranosidase of fungal strains [6]. Stabil-
ity and activity at high temperatures are desirable properties in
this class of enzymes, considering the fact that the most indus-
trial processes using xylanolytic enzymes are carried out at high
temperatures [37]. According to their physicochemical properties,
some fungal ␣-l-arabinofuranosidases were active and stable only
at acidic pH [6]. However, most bacterial ␣-l-arabinofuranosidases
that are purified until now, have the optimum pH around 7.0,
such as Bacillus stearothermophilus [9] and Ruminococcus albus
5. Conclusion
2+
2+
2+
[
38]. The addition of metal ions such as Zn , Cu and Hg
inhibited the ␣-l-arabinofuranosidase activity significantly, sug-
gesting that ␣-l-arabinofuranosidase is a thiol-sensitive enzyme
because these heavy metal ions bind free mercapto groups (–SH)
in cysteine residues. Moreover, sensitivity to Hg2 indicates that
The results presented in this paper indicate that ␣-l-
arabinofuranosidase of T. thermophilus play an important role in
the assimilation of arabinose moieties from arabinose-containing
xylooligosaccharides generated by endo-xylanase. For this reason,
+
␣
-l-arabinofuranosidase has an active-site thiol group [39]. This
␣
-l-arabinofuranosidase can be suitable for practical applications
effect was also observed in the ␣-l-arabinofuranosidase of Strep-
tomyces sp. PC22 [40], Aspergillus nidulans [36] and two Penicillium
chrysogenum ␣-l-arabinofuranosidases [13]. Apo-forme of enzyme
was reactivated in the presence of 1 mM of Mn2 , probably this
in food processing, such as efficient degradation of hemicel-
lulosic biomass for production of biofuels or l-arabinose and
its derivatives. In the future, accessory enzymes such as ␣-l-
arabinofuranosidases combining with xylanase might be widely
applied in animal feed to convert xylan to simple sugars effectively
and enhance feed digestibility. Studies on this novel enzyme, such
as its gene cloning are now in progress.
+
␣
-l-arabinofuranosidase is mostly dependant on this metal ion to
maintain an active structure. This property made it novel for this
group of enzyme.
The ␣-l-arabinofuranosidases belonging to family 3 and 43
of glycosyl hydrolases can hydrolyze the pNPX following a non-
specific reaction [41]. This can be explained by the fact that
the d-Xylopyranose and ␣-l-arabinofuranose are spatially sim-
ilar such that the glycosidic bonds and hydroxyl groups can
be overlaid, leading to bifunctional enzymes [42]. Other notable
property of this enzyme was its apparent high affinity for the
pNPA (Km: 0.77 mM). Lower Km values for homogeneous ␣-l-
arabinofuranosidase have been previously reported for a few other
enzymes such as ␣-l-arabinofuranosidase of A. niger (0.6 mM) [11]
and that of Streptomyces sp. PC22 (0.23 mM) [40]. The T. ther-
mophilus ␣-l-arabinofuranosidase hydrolyses pNPA with a catalytic
efficiency 3.5 fold higher than ␣-l-arabinofuranosidase of Penicil-
lium purpurogenum [43].
Acknowledgment
This work received financial support from “Ministère de
l’Enseignement Supérieur de la Recherche Sientifique et de la
Technologie, Tunisia” granted to the “Laboratoire de Génétique
Moléculaire des Eucaryotes” (Centre des Biotechnologies de Sfax,
Tunisia). We also thank Mr. Khalifa ben khadhra (CBS, Sfax) for his
help in the HPLC analysis.
References
[
1] J. Puls, K. Poutanen, in: M.P. Coughlan (Ed.), Enzyme Systems for Lignocellulose
Degradation, Elsevier Applied Science, Barking, UK, 1989, pp. 151–166.

M. Guerfali et al. / Journal of Molecular Catalysis B: Enzymatic 68 (2011) 192–199
199
[
2] G. Cleemput, S.P. Roels, M. Van Oort, P.J. Grobet, J.A. Delcour, Cereal Chem. 70
[26] F.W. Wagner, in: J.F. Riordan, B.L. Vallee (Eds.), Methods in Enzymology, Part –
A, Academic Press, New York, 1988, p. 158.
(
1993) 324–329.
[
[
[
[
[
[
3] M.M. Smith, R.D. Hartley, Carbohydr. Res. 118 (1983) 65–80.
4] T. Ishii, Plant Sci. 127 (1997) 111–127.
5] P. Biely, Trends Biotechnol. 3 (1985) 286–290.
6] B.C. Saha, Biotechnol. Adv. 18 (2000) 403–423.
7] A. Kaji, Adv. Carbohydr. Chem. Biochem. 42 (1984) 382–394.
8] A.K.M.S. Rahman, N. Sugitani, M. Hatsu, K. Takamizawa, Can. J. Microbiol. 49
[27] M. Bataillon, P. Mathaly, N.A.P. Carolinali, F. Duchiron, Indus. Crop Prod. 8 (1998)
37–43.
[28] E. Chanliaud, L. Saulnier, J.F. Thibault, J. Cereal Sci. 21 (1995) 195–203.
[29] K.K.Y. Wong, L.U.L. Tan, J.N. Saddler, Microbiol. Rev. 52 (1988) 305–317.
[30] Y. Lin, S. Tanaka, Appl. Microbiol. Biotechnol. 69 (2006) 627–642.
[31] T.M. Wood, S.I. McCrae, Appl. Microbiol. Biotechnol. 45 (1996) 538–545.
[32] A. Manimaran, K.S. Kumar, K. Permaul, S. Singh, Appl. Microbiol. Biotechnol. 81
(2009) 887–893.
(
2003) 58–64.
[
9] L. Bezalel, Y. Shoham, E. Rosenberg, Appl. Environ. Microbiol. 40 (1993) 57–62.
[
[
10] W. Zimmermann, B. Winter, P. Broda, FEMS Microbiol. Lett. 55 (1988) 181–
86.
11] F.M. Rombouts, A.G.J. Voragen, M.F. Searle-van Leeuwen, C.C.J.M. Geraerds, H.A.
Schools, W. Pilnik, Carbohydr. Polym. 9 (1988) 25–47.
[33] P. Vincent, F. Shareck, C. Dupont, R. Morosoli, D. Kluepfel, Biochem. J. 322 (1997)
845–852.
[34] I. Kimura, H. Sasahara, S. Tajima, J. Ferment. Bioeng. 80 (1995) 334–339.
[35] E. Luonteri, M. Siika-aho, M. Tenkanen, L. Vikari, J. Biotechnol. 38 (1995)
279–291.
[36] M.T. Fernandez-Espinar, J.L. Pena, F. Pinaga, S. Valles, FEMS Microbiol. Lett. 115
(1994) 107–112.
[37] K.K.Y. Wong, J.N. Saddler, in: M.P. Coughlan, G.P. Hazlewood (Eds.), Hemicellu-
lose and Hemicellulases, Portland, London, 1993, pp. 127–143.
[38] L.C. Greve, J.M. Labavitch, R.E. Hungate, Appl. Environ. Microbiol. 47 (1994)
1135–1140.
[39] S.S. Keskar, M.C. Srinivasan, V.V. Desphande, Biochem. J. 261 (1989) 49–55.
[40] P. Raweesri, P. Riangrungrojana, P. Pinphanichakarn, Bioresour. Technol. 99
(2008) 8981–8986.
[41] K. Wagschal, C. Heng, C.C. Lee, D.W.S. Wong, Appl. Microbiol. Biotechnol. 81
(2009) 855–863.
[42] D.B. Jordan, X.L. Li, C.A. Dunlap, T.R. Whitehead, M.A. Cotta, Appl. Biochem.
Biotechnol. 141 (2007) 51–76.
[43] P. De Ioannes, A. Peirano, J. Steiner, J. Ezyaguirre, J. Biotechnol. 76 (2000)
253–258.
1
[
[
[
[
[
12] E.X.F. Filho, J. Puls, M.P. Coughlan, Appl. Environ. Microbiol. 62 (1996) 168–173.
13] T. Sakamoto, H. Kawasaki, Biochim. Biophys. Acta 1621 (2003) 204–210.
14] M.T.H. Numan, N.B. Bhosle, J. Ind. Microbiol. Biotechnol. 33 (2006) 247–260.
15] B. Henrissat, A. Bairoch, Biochem. J. 316 (1996) 695–696.
16] Z. Gunata, J.M. Brillouet, S. Voirin, R. Baumes, R. Cordonnier, J. Agric. Food Chem.
3
8 (1990) 772–776.
17] S. Osaki, T. Kimura, T. Sugimoto, S. Hizukuri, N. Iritani, J. Nutr. 131 (2001)
96–799.
18] K. Seri, K. Sanai, N. Matsuo, K. Kawakubo, C. Xue, S. Inoue, Metabolism 45 (1996)
368–1374.
19] A.I. Maalej, M. Guerfali, A. Gargouri, H. Belghith, J. Mol. Catal. B: Enzym. 59
2009) 145–152.
20] M. Guerfali, A.I. Maalej, A. Gargouri, H. Belghith, J. Mol. Catal. B: Enzym. 57
2009) 242–249.
21] M. Guerfali, M. Chaabouni, A. Gargouri, H. Belghith, Appl. Microbiol. Biotechnol.
5 (2010) 1361–1372.
[
[
[
[
[
7
1
(
(
8
[
[
[
[
22] T. Yanai, M. Sato, Biosci. Biotechnol. Biochem. 64 (2000) 1181–1188.
23] M.J. Bailey, P. Biely, K. Poutanen, J. Biotechnol. 23 (1992) 257–270.
24] G.L. Miller, Anal. Chem. 31 (1959) 426–428.
[44] H.R. Sørensen, S. Pedersen, A. Viksø-Nielsen, A.S. Meyer, Enzym. Microbial.
Technol. 36 (2005) 773–784.
[45] K.R.J. Palframan, G.R. Gibson, R.A. Rastall, Curr. Issues Intest. Microbiol. 4 (2003)
71–75.
25] U.K. Laemmli, M. Favre, J. Mol. Biol. 80 (1973) 575–592.