Fungalα-arabinofuranosidases of glycosyl hydrolase families 51 and 54 show a dual arabinofuranosyl- and galactofuranosyl-hydrolyzing activity

Article abstract of DOI:10.1515/hsz-2012-0134

Aspergillus niger possesses a galactofuranosidase activity, however, the corresponding enzyme or gene encoding this enzyme has never been identified. As evidence is mounting that enzymes exist with affinity for both arabinofuranose and galactofuranose, we investigated the possibility that α- l -arabinofuranosidases, encoded by the abfA and abfB genes, are responsible for the galactofuranosidase activity of A. niger. Characterization of the recombinant AbfA and AbfB proteins revealed that both enzymes do not only hydrolyze p -nitrophenyl-α- l -arabinofuranoside (pNp-α- Araf) but are also capable of hydrolyzing p -nitrophenyl-β- d -galactofuranoside (pNp-β-Galf ). Molecular modeling of the AbfB protein with pNp-β-Galf confirmed the possibility for AbfB to interact with this substrate, similarly as with pNp-α-Araf. We also show that galactomannan, a cell wall compound of A. niger, containingβ-linked terminal and internal galactofuranosyl moieties, can be degraded by an enzyme activity that is present in the supernatant of inulin- grown A. niger. Interestingly, purified AbfA and AbfB did not show this hydrolyzing activity toward A. niger galactomannan. In summary, our studies demonstrate that AbfA and AbfB,α- l -arabinofuranosidases from different families, both contain a galactofuranose (Galf) -hydrolyzing activity. In addition, our data support the presence of a Galf - hydrolase activity expressed by A. niger that is capable of degrading fungal galactomannan. Copyright by Walter de Gruyter ? Berlin ? Boston.

Full text of DOI:10.1515/hsz-2012-0134

Biol. Chem., Vol. 393, pp. 767–775, August 2012 • Copyright © by Walter de Gruyter • Berlin • Boston. DOI 10.1515/hsz-2012-0134
Fungal α-arabinofuranosidases of glycosyl hydrolase
families 51 and 54 show a dual arabinofuranosyl- and
galactofuranosyl-hydrolyzing activity
Boris Tefsen1,a,*, Ellen L. Lagendijk2,a, Joohae Park2,
hydrolase activity expressed by A. niger that is capable of
degrading fungal galactomannan.
Michiel Akeroyd3, Doreen Schachtschabel2, Robert
Winkler4, Irma van Die1 and Arthur F.J. Ram2,5
Keywords: AbfA; AbfB; Aspergillus niger;
galactofuranose; galactofuranosidase; glycosyl hydrolase.
1 Department of Molecular Cell Biology and Immunology,
VU University Medical Center, van den Boechorststraat 7,
NL-1081 BT Amsterdam, The Netherlands
2 Department of Molecular Microbiology and Biotechnology,
Institute of Biology Leiden, Leiden University, Sylviusweg
Introduction
72, NL-2333 BE Leiden, The Netherlands
Fungi secrete various enzymes to degrade complex molecules
3 DSM Biotechnology Center, Beijerinck Laboratory, PO
in their surroundings in order to provide them with the essen-
Box 1, NL-2600 MA Delft, The Netherlands
tial building blocks of life. Aspergillus spp. are saprophytic
4 Department of Biotechnology and Biochemistry,
and secrete many different glycosyl hydrolases (GHs), which
CINVESTAV Unidad Irapuato, Km 9.6 Libramiento Norte
catalyze the breakdown of polysaccharides. The availability
Carr. Irapuato-León, 36821 Irapuato Gto., Mexico
of genome sequences for aspergilli has allowed a detailed
5 Kluyver Centre for Genomics of Industrial Fermentation,
mining for GHs (Martens-Uzunova et al., 2006; Pel et al.,
PO Box 5057, NL-2600 GA Delft, The Netherlands
2007; Coutinho et al., 2009). For some particular GH activi-
*Corresponding author
e-mail: b.tefsen@vumc.nl
ties, however, the corresponding gene and protein have never
been identified. An example of such an unattributed activity
is the enzymatic degradation of Galf-containing glycoconju-
gates. This activity has been described for Aspergillus niger
(Cousin et al., 1989; Wallis et al., 2001), for several other fila-
Abstract
mentous fungi (Daley and Strobel, 1983; Cousin et al., 1989;
Reyes et al., 1992; Van Bruggen-Van Der Lugt et al., 1992),
and for a prokaryote, Bacillus sp. (Ramli et al., 1995). No pro-
Aspergillus niger possesses a galactofuranosidase activity,
however, the corresponding enzyme or gene encoding this
enzyme has never been identified. As evidence is mounting
tein or gene responsible for this activity has been identified in
that enzymes exist with affinity for both arabinofuranose
any of these cases.
and galactofuranose, we investigated the possibility that
It was recently shown that thiodisaccharides with either
α-l-arabinofuranosidases, encoded by the abfA and abfB
Galf or arabinofuranose (Araf) as terminal units inhibit the
genes, are responsible for the galactofuranosidase activ-
ity of A. niger. Characterization of the recombinant AbfA
and AbfB proteins revealed that both enzymes do not only
hydrolyze p-nitrophenyl-α-l-arabinofuranoside (pNp-α-
activity of an exo-galactofuranosidase from Penicillium fel-
lutanum, a close relative of Aspergillus (Repetto et al., 2009).
In addition, an α-l-arabinofuranosidase (AbfD3) was found in
Thermobacillus xylanilyticus, which displays transglycosyla-
Araf ) but are also capable of hydrolyzing p-nitrophenyl-
tion activity and is able to use pNp-β-Galf as the sugar donor
β-d-galactofuranoside (pNp-β-Galf ). Molecular modeling
(Remond et al., 2005). These data demonstrate that enzymes
exist with affinity for both Araf and Galf and suggest the pos-
sibility that one or more arabinofuranosidases in A. niger play
of the AbfB protein with pNp-β-Galf confirmed the pos-
sibility for AbfB to interact with this substrate, similarly as
with pNp-α-Araf. We also show that galactomannan, a cell
a role in the degradation or synthesis of Galf-containing gly-
wall compound of A. niger, containing β-linked terminal
coconjugates. The filamentous fungus A. niger secretes two
and internal galactofuranosyl moieties, can be degraded by
an enzyme activity that is present in the supernatant of inu-
α-l-arabinofuranosidases, AbfA and AbfB (Flipphi et al.,
1993a,b). It has been described that AbfA, belonging to GH
lin-grown A. niger. Interestingly, purified AbfA and AbfB
family 51, hydrolyzes terminal α-linked Araf from small sub-
did not show this hydrolyzing activity toward A. niger
strates, such as pNp-α-Araf and short-chain arabino-oligo-
galactomannan. In summary, our studies demonstrate that
saccharides. AbfB, belonging to GH family 54, has a similar
AbfA and AbfB, α-l-arabinofuranosidases from different
activity on these substrates but is also able to hydrolyze poly-
families, both contain a galactofuranose (Galf)-hydrolyzing
meric substrates such as branched arabinan and arabinoxylan
activity. In addition, our data support the presence of a Galf-
(Rombouts et al., 1988).
Here, we describe the cloning of the abfA and abfB genes
of A. niger in Pichia pastoris and prove the capacity of the
aThese authors contributed equally to this work.
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

768 B. Tefsen et al.
recombinant AbfA and AbfB proteins to hydrolyze pNp-
β-Galf in addition to pNp-α-Araf. Molecular modeling of
AbfB, using the crystal structure of AbfB from A. kawachii
(Miyanaga et al., 2004), supports that both the pNp-β-Galf
and the pNp-α-Araf substrates may fit within the active site
of the enzyme. However, purified P. pastoris expressed AbfA
and AbfB do not have the capacity to release Galf from the
biologically relevant polysaccharide galactomannan. Finally,
we demonstrate that supernatant from both wild-type A. niger
and ΔabfA mutant grown on inulin has a hydrolyzing activity
that releases Galf from the biologically relevant polysaccha-
ride galactomannan, suggesting the presence of a dedicated
β-galactofuranosidase expressed by A. niger N402 that has
yet to be identified.
of pNp-β-d-xyloside, pNp-N-acetyl-β-d-glucosaminide, pNp-
β-l-arabinopyranoside,pNp-β-d-galactopyranoside,pNp-α-d-
galactopyranoside, pNp-α-d-mannoside, pNp-β-d-mannoside,
pNp-α-d-glucoside, pNp-β-d-glucoside, pNp-α-l-fucoside, or
pNp-α-d-N-acetylneuraminic acid by these enzymes (data not
shown). These data showed that α-arabinofuranosidases from
different GH families possess a distinctive β-Galf-hydrolyzing
activity. To exclude any possible effects of differential glyco-
sylation on the activity of these hydrolases, we also expressed
a His-tagged AbfA in A. niger. This protein displayed similar
GH activity as the His-tagged AbfA expressed in P. pastoris
(data not shown).
Characterization of the β-galactofuranosidase activity
of AbfA and AbfB
Results
Acomparativekineticstudywasperformedtocharacterizeboth
the α-arabinofuranosidase and the α-galactofuranosidase
activity of AbfA and AbfB. Michaelis-Menten diagrams can
be found in Supplementary Figure S2. The Km of the activ-
ity toward pNp-α-Araf was an order of magnitude lower
(23-fold for AbfA and 26-fold for AbfB) than the activity
toward pNp-β-Galf for both enzymes (Table 1), whereas
the Vmax was higher for pNp-α-Araf than for pNp-β-Galf
for both enzymes. Thus, both enzymes display a higher
hydrolyzing activity toward pNp-α-Araf than to pNp-β-
Galf. AbfB seems to have a lower affinity (higher Km) but
a higher efficiency (higher Vmax) for both substrates com-
pared with AbfA (Table 1). To characterize both enzymes
in more detail, the influence of pH (Figure 2) and tempera-
ture (Figure 3) on their activity toward both substrates was
examined. The optimum for either glycosyl-hydrolyzing
activity of both enzymes appears to be slightly above pH
3. At increasing temperatures, a higher activity of AbfA
toward pNp-β-Galf was measured, whereas the hydrolyz-
ing activity toward pNp-α-Araf was quite stable across a
wide temperature range (Figure 3A). In contrast to AbfA,
the β-galactofuranosidase activity of AbfB did not decrease
at lower temperatures. Both the α-arabinofuranosidase and
the β-galactofuranosidase activities of AbfB were optimal
between 30°C and 35°C (Figure 3B).
α-Arabinofuranosidases from GH families 51 and 54
possess both α-arabinofuranosidase and
β-galactofuranosidase activity
To investigate a putative Galf-hydrolyzing activity of the
α-arabinofuranosidases AbfA and AbfB, cDNAs coding for
the A. niger genes abfA and abfb were amplified and cloned
into pGAPZαB and expressed in P. pastoris. This resulted in
the secretion of His-tagged AbfA and AbfB into the growth
medium, from which the recombinant proteins were subse-
quently isolated and purified (Supplementary Figure S1).
AbfA and AbfB were analyzed for their glycosyl-hydrolyzing
capacity using a panel of pNp-sugar substrates. Two substrates
from this panel were hydrolyzed by both enzymes, pNp-α-
Araf and pNp-β-Galf (Figure 1), indicating a dual hydrolyzing
activity of AbfA and AbfB. We did not find any hydrolysis
AbfA
AbfB
A
B
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
3D modeling of AbfB with its proposed substrates
The availability of the crystal structure of the A. kawachii
AbfB protein (Miyanaga et al., 2004) allowed the modeling
of both the pNp-α-Araf and pNp-β-Galf into the correspond-
ing substrate binding site. As both the N- and C-terminus
(blue) are far away from the catalytic pocket or the gly-
cosylation site (light blue), the addition of the tag is not
expected to affect the catalytic domain or the glycosylation
of the AbfB protein (Figure 4A). The A. niger AbfB pro-
tein is 97% identical to the amino acid sequence of the A.
kawachii AbfB protein and the catalytic residues involved in
catalysis and substrate binding residues (Cys176, Cys177,
Met195, Trp206, Asp219, Glu221, Asn222, Leu224, and
Asp297) are conserved. In A. kawachii AbfB, the carboxyl
groups of Glu221 and Asp297 are located on either site of
Galf
Araf
Galf
Araf
Figure 1 GH activity of AbfA and AbfB toward different pNp
substrates.
(A) pNp-β-Galf (Galf) and pNp-α-Araf (Araf) (2 mm) were incubated
for 3 h with 1 μg purified AbfA. Liberation of freed pNp molecules
was measured at OD 405 nm. The average values of three samples
were normalized for hydrolysis of pNp-α-Araf (0.3 nmol/μg/min)
and depicted as relative hydrolysis rate. Error bars depict normalized
standard deviations. (B) pNp-β-Galf (Galf) and pNp-α-Araf (Araf)
(2 mm) were incubated for 1 h with 0.1 μg purified AbfB. Liberation
of free pNp molecules was measured at OD 405 nm. The average
values of three samples were normalized for hydrolysis of pNp-
α-Araf (9.2 nmol/μg/min) and depicted as relative hydrolysis rate.
Error bars depict normalized standard deviations.
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

α-Arabinofuranosidases show galactofuranosidase activity 769
Table 1 Apparent kinetic parameters for GH activities of AbfA
and AbfB.
Glycosyl-hydrolyzing activity of A. niger strains
grown on glucose and inulin
Substrate
Km (mm)
Vmax (nmol/μg/min)
A. niger has been shown previously to secrete
a
β-galactofuranosidase when grown on inulin (Wallis et al.,
2001). To investigate whether AbfA and/or AbfB may be
responsible for the hydrolysis of pNp-β-Galf under these con-
ditions, the hydrolyzing activity toward different substrates of
A. niger strain N402, grown on glucose and inulin, respectively,
was determined. Culture filtrates were collected and subjected
to GH assays using a panel of substrates. A. niger grown on inu-
lin showed a 2.8-fold higher hydrolyzing activity toward pNp-
β-Galf than toward pNp-α-Araf, whereas by contrast, growth
on glucose resulted in a 4.7-fold higher hydrolyzing activity
toward pNp-α-Araf than to pNp-β-Galf (Figure 5). If AbfA or
AbfB would be responsible for the Galf-hydrolyzing activity of
A. niger grown under these conditions, we would expect to see
a similar rise in capacity to hydrolyze pNp-α-Araf compared
with activity toward pNp-β-Galf, resulting in a ratio of both
activities comparable to the results shown for the activity of
AbfAandAbfB in Figure 1. Therefore, these results suggest that
A. niger expresses another Galf-hydrolyzing activity, in addi-
tion to AbfA or AbfB. Interestingly, although outside the scope
of this study, we observed a huge increase in hydrolysis activity
toward pNp-β-glucose and pNp-β-N-acetylglucosamine when
A. niger was grown on inulin (Figure 5).
To provide further evidence for the possibility that AbfA is
not responsible for the β-galactofuranosidase activity, an abfA
deletion strain (ΔabfA) was constructed and verified by Southern
blot analysis (Supplementary Figure S3). The ΔabfA mutant
showed a phenotype similar to wild-type strain N402 when
grown on glucose, inulin, maltose, or xylose (data not shown).
The culture filtrate of the ΔabfA mutant grown on inulin was
tested for hydrolysis of pNp-α-Araf and pNp-β-Galf. The ΔabfA
mutant displayed a 5.3-fold higher activity toward pNp-β-Galf
AbfA
AbfB
AbfA
AbfB
pNp-α-Araf
pNp-β-Galf
0 .7
17
1 .1
28
1 .0
0.3
6 4
42
The apparent kinetic parameters Km and Vmax were obtained by
fitting the data for hydrolysis of these substrates at variable con-
centrations to the Michaelis-Menten equation using nonlinear
regression analysis (SPSS 15.0 program; SPSS).
the anomeric C1 of α-l-Araf indicating that they form the
catalytic subunit (Miyanaga et al., 2004). For the α-l-Araf
substrate, three H-bonds are postulated, which is confirmed
by our model as the co-crystallized substrate displays three
H-bonds of 2.88, 2.92, and 3.14 Å length (Figure 4B). Both
substrates that are hydrolyzed by AbfB, pNp-α-Araf and
pNp-β-Galf, were fitted into the modeled A. niger AbfB
protein structure. As shown in Figure 4C (pNp-α-Araf) and
Figure 4D (pNp-β-Galf), both substrates fit into the pocket
at the catalytic site. The residues Asp297 and Glu221 are
located on either side of the anomeric C1 of both substrates,
indicating they are most likely responsible for catalysis,
whereas Gly296 and Asn223 seem mainly needed for sub-
strate coordination. These pNp substrates form two H-bonds
each in the manual docking, with lengths of the H-bonds
being 2.59 and 2.80 Å for pNp-Galf and 2.49 and 2.51 Å for
pNp-Araf. As both substrates fit into the substrate pocket in
proximity to the catalytic residues, the modeling supports
the biochemical data that AbfB has hydrolyzing activity
toward both pNp-α-Araf and pNp-β-Galf.
AbfA
AbfB
A
B
1
1
0.8
0.6
0.4
0.2
0
pNp-α-Araf
pNp-β-Galf
pNp-α-Araf
pNp-β-Galf
0.8
0.6
0.4
0.2
0
2
3
4
5
6
7
8
9
2
3
4
5
6
7
8
9
pH
pH
Figure 2 pH dependence of AbfA- and AbfB-hydrolyzing activities on pNp-α-Araf and pNp-β-Galf.
(A) Purified AbfA was incubated in duplicate with 0.5 mm pNp-α-Araf (solid lines; 0.5 μg enzyme) and pNp-β-Galf (dotted lines; 0.5 μg
enzyme) at 37°C at the indicated pH. (B) Purified AbfB was incubated in duplicate with 0.5 mm pNp-α-Araf (solid lines; 10 ng enzyme) and
pNp-β-Galf (dotted lines; 1 μg enzyme) at 37°C at the indicated pH. Absolute hydrolysis of both substrates was calculated from liberated pNp
molecules measured at OD405. These data were normalized against the highest measured value for each substrate (for AbfA, 0.6 nmol pNp-
α-Araf/μg/min and 9 pmol pNp-β-Galf/μg/min; for AbfB, 9.7 nmol pNp-α-Araf/μg/min and 3.3 nmol pNp-β-Galf/μg/min) and depicted as
relative hydrolysis rate with error bars representing normalized SEM.
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

770 B. Tefsen et al.
AbfA
AbfB
A
B
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
pNp-α-Araf
pNp-β-Galf
pNp-α-Araf
pNp-β-Galf
25 30 35 40 45 50 55 60
Temperature (°C)
25 30 35 40 45 50 55 60
Temperature (°C)
Figure 3 Temperature dependence of AbfA- and AbfB-hydrolyzing activities on pNp-α-Araf and pNp-β-Galf.
(A) Purified AbfA was incubated in duplicate with 0.5 mm pNp-α-Araf (solid lines, 150 ng enzyme) and pNp-β-Galf (dotted lines, 1.5 μg
enzyme) at pH 4.5 at the indicated temperatures. (B) Purified AbfB was incubated in duplicate with 0.5 mm pNp-α-Araf (solid lines, 8 ng
enzyme) and pNp-β-Galf (dotted lines, 40 ng enzyme) at pH 4.5 at the indicated temperatures. Absolute hydrolysis of both substrates was cal-
culated from liberated pNp molecules measured at OD405. The average values were normalized against the highest measured value for each
substrate (for AbfA, 0.2 nmol pNp-α-Araf/μg/min and 20 pmol pNp-β-Galf/μg/min; for AbfB, 6.4 nmol pNp-α-Araf/μg/min and 0.5 nmol
pNp-β-Galf/μg/min) and depicted as relative hydrolysis rate. Error bars indicate normalized SEM.
than to pNp-α-Araf (data not shown), indicating that AbfA is
not essential for galactofuranosidase activity in Aspergillus.
inulin-grown N402, purified AbfA, or purified AbfB. Resulting
samples were subsequently analyzed for remaining galacto-
mannan in a sandwich ELISA using the anti-galactofuranose
antibody EB-A2. Both AbfA and AbfB were unable to hydro-
lyze galactomannan (Figure 6A). In contrast, enzymes in the
medium of inulin-grown A. niger were able to hydrolyze Galf
residues from galactomannan. The breakdown of galactoman-
nan by supernatant of A. niger was greatly diminished when it
was heat-inactivated prior to incubation with galactomannan,
indicating that the enzymatic activity was destroyed by this
Galactomannan is hydrolyzed by supernatant of
inulin-grown A. niger, but not by AbfA or AbfB
We next tested the ability of recombinant AbfA and AbfB to
hydrolyze Galf moieties in galactomannan, a polysaccharide
present in the cell wall of A. niger. Galactomannan was isolated
from wild-type strain N402 and incubated with medium of
A
B
GLU221
ASP297
GLY296
C
ASN223
N
C
D
GLU221
GLU221
ASP297
GLY296
ASP297
GLY296
ASN223
ASN223
Figure 4 Manual 3D modeling of A. niger AbfB protein with three substrates.
A 3D model of AbfB was constructed with three manual substrate dockings. (A) Overview of the complete structure of AbfB. Two
N-acetylglucosamines shown to be present on ASN202 are indicated by an arrow and the N- and C-termini are indicated by an N and C, respec-
tively. The boxed area left of the glycosylation is the catalytic site that is shown in more detail in the other three panels. (B) A docking model of
the catalytic site with the substrate α-l-Araf. (C) A docking model of the catalytic site with the substrate pNp-α-l-Araf. (D) A docking model
of the catalytic site with the substrate pNp-β-d-Galf. The amino acids in the catalytic site are indicated. ASP297 and GLU221 are important for
catalysis, whereas GLY296 and ASN223 are important for substrate coordination. The substrates are depicted in the middle of each panel with
the pNp-moiety sticking out to the back, when present. The proposed H-bonds are represented by dotted lines.
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

α-Arabinofuranosidases show galactofuranosidase activity 771
1
0.8
0.6
0.4
0.2
0
Discussion
Glucose
Inulin
In this study, we prove for the first time that fungal
α-arabinofuranosidases belonging to GH51 (AbfA) and GH54
(AbfB) also have galactofuranosidase activity. Although the
α-arabinofuranosidase activity of AbfA has been described
and characterized in several Aspergillus spp. including
A. niger, A. oryzae, A. kawachii, and A. awamori (Rombouts
et al., 1988; Flipphi et al., 1993a,b; Koseki et al., 2003;
Matsumura et al., 2004), the pNp-β-Galf substrate was never
included as a potential substrate. The recent observation that
the α-arabinofuranosidase from P. fellutanum was inhibited
by thiodisaccharides with either Galf or Araf as terminal units
(Repetto et al., 2009) suggested that the enzyme has affinity
for both Galf and Araf residues. Representative members of
GH51 and GH54, A. niger AbfA and A. niger AbfB, respec-
tively, were expressed in P. pastoris, which does not secrete
GHs with activities that could interfere with these studies, and
their enzyme activity was determined. For all these proteins,
we detected a dual activity on pNp-β-Galf besides pNp-α-
Araf. The enzymes were only active against the furanose form
of the sugars and not active against the pyranose forms in
agreement with earlier studies (Rombouts et al., 1988; Flipphi
et al., 1993a,b; Koseki et al., 2003; Matsumura et al., 2004).
In our analysis, we consistently found that that the activity of
the enzymes toward pNp-α-Araf was higher compared with
the activity toward pNp-β-Galf. Interestingly, we found that
AbfAhad different temperature optima for both substrates and
a different shape of the temperature curve. We do not have
an explanation for this phenomenon. Earlier studies on α-l-
arabinofuranosidases showed temperature optima between
50°C and 60°C (Rombouts et al., 1988; Kaneko et al., 1998;
De Ioannes et al., 2000; Matsuo et al., 2000; Koseki et al.,
2003), but we could not find any reports on different tempera-
Figure 5 Galactofuranosidase activity increases when A. niger is
grown on inulin.
The indicated pNp substrates (0.5 mm) were incubated in triplicate for
2 h at 37°C with 5 μl supernatant collected from wild-type A. niger
strain N402 grown on glucose (white bars) or inulin (black bars).
Liberated pNp molecules were measured at OD 405 nm. Values were
normalized against the highest measured value (1.3 μmol pNp-α-
Galp/ml/h for glucose-grown and 2.2 μmol pNp-α-Galp/ml/h for
inulin-grown) and depicted as relative hydrolysis rate. Error bars
depict standard deviation. Xyl=pNp-β-d-xyloside; α-Galp=pNp-α-
d-galactopyranoside; β-Galp=pNp-β-d-galactopyranoside; Galf=pNp-
β-d-galactofuranoside; Araf=pNp-α-l-arabinofuranoside; Arap=pNp-
β-l-arabinopyranoside; α-Man=pNp-α-d-mannoside; β-Man=pNp-
β-d-mannoside; α-Glc=pNp-α-d-glucoside; β-Glc=pNp-β-d-gluco-
side; Fuc=pNp-α-l-fucoside; GlcNAc=pNp-N-acetyl-β-d-gluco-
saminide.
treatment (Figure 6B). Supernatant of an inulin-grown ΔabfA
mutant was also tested for its capacity to hydrolyze galacto-
mannan. Clearly, the galactomannan is also hydrolyzed by an
activity within this supernatant (Figure 6B), indicating that
other enzymes than AbfA are responsible for this activity in A.
niger. To our knowledge, this is the first report of enzymatic
breakdown of A. niger galactomannan by its own enzymes.
3.5
2.5
Supernatant
A
B
Galactomannan+supernatant
Galactomannan+enzyme(s)
3.0
Galactomannan+HI-supernatant
2.0
2.5
2.0
1.5
1.0
0.5
0
1.5
1.0
0.5
0
MM
AbfA
AbfB
N402
MM
N402
ΔabfA
Figure 6 Galactomannan is hydrolyzed by supernatant of inulin-grown A. niger, but not by AbfA or AbfB.
(A) Minimal medium (MM), purified AbfA (1.1 μg), purified AbfB (2.1 μg), or 10.5 μl supernatant of inulin-grown wild-type A. niger strain
N402 was incubated with 7.5 ng (gray bars) or without (white bars) galactomannan for 20 h at 37°C at pH 4.5. (B) Supernatant of inulin-grown
N402 and ΔAbfA mutant was diluted 1:20 and heat-inactivated (HI-supernatant, black bars) or not (gray bars). Subsequently, 20 μl of these
treated supernatants were incubated with 5 ng galactomannan for 24 h at 37°C at pH 4.5. Samples were analyzed for remaining galactomannan
after incubation in an ELISA sandwich using antibody EB-A2 at an OD of 450 nm. These are typical experiments that have been performed at
least three times independently.
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

772 B. Tefsen et al.
ture optima for different substrates of the same enzyme, like
we see for AbfA.
cell walls (Martens-Uzunova et al., 2006). The increase
of β-galactofuranosidase activity upon carbon starvation
has led others to suggest that the β-galactofuranosidase
is involved in the recycling of fungal cell wall material
during starvation (Mennink-Kersten et al., 2006) as Galf
residues are present in the galactomannan fractions and
on glycoproteins present in the cell wall (Tefsen et al.,
2011). Genome-wide expression analysis in A. niger malt-
ose-starved mycelium, a condition that has been reported
to induce β-galactofuranosidase activity (Wallis et al.,
2001), did not result in the induction of abfA and abfB (B.
Nitsche and A. Ram, unpublished results). This lack of
induction of abfA and abfB during starvation supports our
conclusions that these enzymes are not responsible for the
β-galactofuranosidase activity displayed by A. niger.
The genome of A. niger contains three other genes encod-
ing arabinofuranosidases, AbfC (An08g01710), AbfD
(An09g00880), AbfE (Martens-Uzunova and Schaap,
2009), and an arabinoxylan arabinofuranohydrolase, AxhA
(An03g00960) (Gielkens et al., 1997) that is also able to
hydrolyze Araf residues. Similar to AbfA and AbfB, the
genes encoding AbfC, AbfD, and AxhA are not induced upon
starvation. Expression data are not available for the abfE
gene, as this gene is not present on the Affymetrix micro-
array (Martens-Uzunova and Schaap, 2009), but our prelimi-
nary data in which the A. niger AbfE protein is expressed in
P. pastoris indicated that AbfE also has a higher activity
toward pNp-α-Araf compared with pNp-β-Galf. Altogether,
this suggests that the β-galactofuranosidase activity in starved
cultures is not encoded by the known members of the abf gene
families.
The availability of a crystal structure of the AbfB protein
from A. kawacchi, allowed the 3D modeling of the A. niger
AbfB protein with both the pNp-α-Araf and the pNp-β-Galf
as substrates. The modeling strongly supports the possibility
that both substrates fit in a similar way in the catalytic bind-
ing pocket of the enzyme, allowing a similar mechanism of
hydrolysis of the substrate. Interestingly, pNp-Galf seems to
fit less well into the substrate cavity than pNp-Araf, causing
steric limitations that might cause less efficient H-bonding
that is required for the substrate coordination in the catalytic
action.
To investigate whether α-arabinofuranosidases could be
responsible for the β-galactofuranosidase activity observed
by Wallis et al. (2001), we cultured A. niger strain N402
in growth medium with inulin as a carbon source, and we
could confirm the presence of a high β-galactofuranosidase
activity. Interestingly, the relative hydrolysis activity of the
culture filtrate was higher toward pNp-β-Galf than to pNp-
α-Araf (Figure 5), in contrast to the ratio shown for purified
AbfA and AbfB. We observed a shift in the ratio between
the hydrolyzing activity toward pNp-β-Galf and the activity
toward pNp-α-Araf in the medium of A. niger grown on inu-
lin when compared with A. niger grown on glucose (Figure
5). This indicates that another enzyme with a preference for
pNp-Galf likely exists. To prove that the AbfA protein is not
responsible for the β-galactofuranosidase activity in the cul-
ture filtrate of inulin-grown cells, we deleted the abfA gene by
constructing a knockout strain. The knockout strain showed
a similar β-galactofuranosidase activity as the wild-type
strain, indicating that AbfA is not essential for the observed
β-galactofuranosidase activity in A. niger.
In addition, we showed that the culture filtrate of inulin-
grown N402 cells contains an enzyme activity that can degrade
cell wall galactomannan (Figure 6). As can be concluded from
Figure 5, the culture filtrate contained no activity against pNp-
α-mannose but a high activity against pNp-β-Galf, suggesting
that the breakdown of the galactomannan is due to the pres-
ence of a β-galactofuranosidase. Interestingly, recombinant
AbfA and AbfB did not show β-galactofuranosidase activity
toward galactomannan isolated from the cell wall of A. niger,
indicating that both AbfA and AbfB are not responsible for
this specific β-galactofuranosidase activity. However, we
cannot exclude that these enzymes are capable of degrading
Galf-containing lipids or proteins, which contain other glyco-
sidic linkages (Tefsen et al., 2011).
To attribute a putative biological role for the dedicated
Galf-hydrolase is challenging. On the one hand, Galf has
never been found in plant polysaccharides, which argues
against the possibility that these organisms need such a
Galf-hydrolase to provide them with Galf residues destined
as an energy source from plant debris. Araf residues, on
the other hand, are abundantly present in the plant poly-
saccharides hemicellulose and pectin. The genes encod-
ing AbfA and AbfB are strongly induced by hemicellulose
and pectin, suggesting that the enzymes are primarily
involved in the hydrolysis of Araf residues present in plant
In summary, in this study, we demonstrated that AbfA and
AbfB, α-l-arabinofuranosidases from GH families 51 and 54,
both contain a Galf-hydrolyzing activity and we have char-
acterized these activities. Molecular modeling supported the
experimental data. Furthermore, we demonstrated the pres-
ence of an enzyme activity expressed by A. niger that is capa-
ble of degrading fungal galactomannan.
Materials and methods
Strains and media
E. coli DH5α was grown in Luria broth and was used as clon-
ing host. Pichia pastoris X33 (Invitrogen) was grown in yeast
extract/peptone/dextrose (YPD) medium. Ampicillin (10 μg/ml) and
Zeocine (20 and 100 μg/ml) were added when necessary. A. niger
strains N402 and ΔabfA were grown for 7 days at 28°C on minimal
Aspergillus medium (MAM), while shaking at 180 rpm (Bennett and
Lasure, 1991) containing 1% inulin as a carbon source. The abfA
deletion strain was constructed in MA169.4, a ku70, pyrG derivative
of N402 (Carvalho et al., 2010).
Cloning of AbfA and AbfB
The abfA and abfB coding sequences were amplified from a cDNA
library of A. niger constructed from RNA obtained after grown on
xylose (kind gift from P.A. Vankuyk) using Phusion Hot Start High-
Fidelity DNA Polymerase (Finnzymes, Woburn, MA, USA), and
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

α-Arabinofuranosidases show galactofuranosidase activity 773
Table 2 Primers used in this study.
Primer name
Sequence (5′– 3 ′)a
ggg aca act ttg tat aga aaa gtt gTC TTT TGA GAA GAC TGA AAT
ggg gac tgc ttt ttt gta caa act tgG TTG CCG GCT GTC TGG AGG A
ggg gac agc ttt ctt gta caa agt ggC GTT GAT TGG GGC GAG CTC G
ggg gac aac ttt gta taa taa agt tgG GCC GGG TGG GGG TAT TTC T
ggg gac aag ttt gta caa aaa agc agg ctA GGA TTT CGG CAC GGC TAC
ggg gac cac ttt gta caa gaa agc tgg gtT GTG GAG TGG GCG CTT ACA C
GTT CAC GTG ATC TCC TTG AAG GTC TCC ACCf
5-Flank forwardb
5-Flank reverseb
3-Flank forwardc
3-Flank reversec
HygroB forwardd
HygroB reversed
AbfA forwarde
AbfA reversee
AbfB forwardh
AbfB reverseh
AbfA-NheI_Fi
AbfA_BsrGI_Ri
CGT CTA GAA AGT TCG CCG CCA GGA CAG CCA Cg
GTT CAC GTG GGC CCC TGT GAC ATC TAC GAA Gf
CGT CTA GAA ACG AAG CAA ACG CCG TCT CAA TCg
CCT AGC TAG CAT GGT GGC CTT TTC AGC TCTj
TTG CTG TAC AGT TCG CCG CCA GGA CAG CCAk
aRecombination sites used for Multisite Gateway recombination are in small letters.
bFor amplification of the 5′-flanking region of abfA.
cFor amplification of the 3′-flanking region of abfA.
dFor amplification of the hygromycin marker.
eFor amplification of the abfA gene (expression in P. pastoris)
fPmlI site underlined.
gXbaI site underlined.
hFor amplification of the abfB gene (expression in P. pastoris).
iFor amplification of the abfA gene (expression in A. niger).
jNheI site underlined, kBsrGI site underlined.
steps with 25 mm imidazole wash buffer, the enzyme was eluted with
the primers are indicated in Table 2. The primers to amplify both
genes were selected in such a way that the predicted signal seq-
uences of AbfA and AbfB were not included. The PCR products
were ligated in pJET1.2 (CloneJET™ PCR Cloning Kit Fermentas)
and sequenced. Subsequently, a PmlI-XbaI fragment was cloned
into P. pastoris expression vector pGAPZαB, introducing the cod-
ing sequences in frame with the N-terminal α factor signal sequence
and the C-terminal myc epitope and polyhistidine tag (Invitrogen,
Carlsbad, CA, USA). This vector contains the constitutive glyceral-
dehyde-3-phosphate dehydrogenase (GAP) to drive expression and
the Saccharomyces cerevisiae α-factor signal sequence for translo-
cation into the ER. The constructs were transformed by electropo-
ration to P. pastoris X33 (Invitrogen), and zeocin-resistant colonies
(100 μg/ml) were selected for further analysis. For the expression
and purification of a His-tagged AbfA protein in A. niger, the abfA
gene of A. niger was PCR amplified with primers listed in Table 2
using A. niger genomic DNA. The fragment was cloned in pJET1.2
and excised with NheI and BsrGI. The fragment containing the abfA
gene was subsequently cloned in NheI and BsrGI digested pARAn19
(Roth and Dersch, 2010) to give pARAn19-AbfA. The final expres-
sion plasmid was sequenced.
250 mm imidazole. The purification was carried out at 4°C. Enzyme
purity was estimated by SDS-PAGE (Supplementary Figure S1). To
remove imidazole, the enzyme fractions were dialyzed with cellulose
dialyzing tube (MWCO is 2 kDa) (Spectrum Laboratories, Rancho
Dominquez, CA, USA) against distilled water at 4°C. Protein con-
centration was 0.1 mg/ml for AbfA and 0.2 mg/ml for AbfB, mea-
sured with a BCA assay (Pierce). The pARAn19-AbfA plasmid was
transformed to A. niger strain AB1.13 (Mattern et al., 1992) and a
strain containing three copies of the plasmid (JP18.1) was selected
to purify AbfA-His. JP18.1 was grown for 5 days in 25 ml MAM at
30°C. The AbfA-His-tagged protein was purified from A. niger as
described above for His-tagged AbfA P. pichia.
GH assays and enzyme kinetics
GH activity was assayed routinely as follows, unless indicated oth-
erwise. Purified enzyme or medium from A. niger was incubated in
100 mm sodium acetate buffer (pH 4.5) with 0.5 mm pNp-conjugated
glycoside (Sigma) in an end-volume of 60 μl in a flat-bottom poly-
styrene 96-wells plate (Greiner) at 37°C. The reaction was termi-
nated by adding 240 μl of 250 mm NaOH, and absorbance was
measured at 405 nm. The amount of liberated pNp was calculated
using 4-nitrophenol as a standard and was used to determine the hy-
drolase activity of the pNp substrate in nanomoles per microgram per
minute (nmol/μg/min). The experiments to determine the influence
of temperature on the activity were performed in Eppendorf tubes in
heatblocks.
For kinetic studies, different concentrations of substrate were in-
cubated for 1 h at 37°C. The apparent kinetic parameters Vmax and
Km were obtained by fitting the data to the Michaelis-Menten equa-
tion using nonlinear regression analysis (SPSS 15 program; SPSS,
Chicago, IL, USA). Hydrolyzing enzymes in the supernatant were
diluted 1:20 and heat-inactivated by incubation at 99°C for 10 min
for galactomannan breakdown experiments.
Expression and purification of AbfA and AbfB
To determine levels of protein production, P. pastoris transformants
were grown at 30°C for 72 h in YPD at 250 rpm. Each day, 1 ml
of the supernatant was collected. The culture filtrates were analyzed
by dot-blots on nitrocellulose membrane (Schleicher & Schuell)
using anti-His (C-term)-HRP antibody (Invitrogen). Transformants
producing high amounts of His-tagged protein were grown in 50 ml
YPD in a 300-ml flask in an orbital shaker (250 rpm) at 30°C for
48 h. The cultures were centrifuged (5 min, 1500 rpm), and the super-
natant was mixed with 10 mm imidazole (final concentration 5 mm)
before loading at an equilibrated HisPur Ni-NTA 3 ml spin column
(Pierce Biotechnology, Rockford, IL, USA). After three washing
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

774 B. Tefsen et al.
Isolation of galactomannan
containing an abfA deletion cassette (designated pJH8.1). This vector
was linearized with DraI and then transformed to the MA169.4 strain
as described (Meyer et al., 2010). Hygromycin-resistant clones were
analyzed by Southern blot (see Supplementary Figure S3).
Cell walls were isolated from A. niger strain N402 after growth
in complete medium (CM) (Carvalho et al., 2010) for 24 h at
30°C at 250 rpm. Spores (1×109) were inoculated in 1 l of CM in
a 2-l Erlenmeyer flask. The mycelia were isolated by filtering over
Myracloth and yielded around 16 g (wet weight). The mycelia were
grinded in liquid nitrogen using a pestle and mortar, and the bro-
ken mycelia were washed three times with 1 m NaCl and three times
with MilliQ at 4°C by centrifugation (3600 rpm, 10 min). Successful
breakage of the mycelia was confirmed by microscopy. Isolated
cell walls were lyophilized over night yielding 2.1 g of cell walls
(dry weight). The galactomannan fraction was isolated according to
Bardalaye and Nordin (1977). The yield of purified galactomannan
was 3.0 mg/g of freeze dried cell walls. The purity of the galactoman-
nan fraction was confirmed by HPAEC monosaccharide analysis on
a PA10 column (Dionex) (Salvador et al., 2000) after TFA hydrolysis
(data not shown).
Acknowledgments
We thank Annika Pettersson, Dorien van het Oever, and Patricia
vanKuyk for sharing materials and Benjamin Nitsche for unpub-
lished data. We thank Prof. Dr. Petra Dersch (Technical University
Braunschweig, Germany) for the gift of the A. niger expression vec-
tors. This work was financially supported by the Dutch Technology
Foundation (S.T.W., grant 10622) and the Kluyver Centre for
Genomics of Industrial Fermentation.
References
Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006). The
SWISS-MODEL workspace: a web-based environment for pro-
tein structure homology modelling. Bioinformatics 22, 195–201.
Bardalaye, P.C. and Nordin, J.H. (1977). Chemical structure of the
galactomannan from the cell wall of Aspergillus niger. J. Biol.
Chem. 252, 2584–2591.
Sandwich ELISA to monitor hydrolysis of
galactomannan
Supernatant of A. niger strains, purified AbfA, purified AbfB, or
minimal medium was incubated with or without 5 to 7.5 ng galacto-
mannan for 20 to 24 h at 37°C at pH 4.5 in a volume of 20 to 40 μl.
After addition of water to a final volume of 50 μl, these samples were
added to microtiter plate wells containing coated antibody EB-A2
(recognizing Galf moieties on galactomannan; Stynen et al., 1995)
and 50 μl EB-A2 conjugated to HRP (Platelia Aspergillus EIA kit;
BioRad). The wells were incubated for 90 min at 37°C and subse-
quently washed five times. After incubation for 30 min at ambient
temperature with 200 μl TMB detection mixture, the coloring reac-
tion was stopped by addition of 100 μl 1.5 N H2SO4. Remaining Galf
on galactomannan was measured at an OD of 450 nm.
Bennett, J.W. and Lasure, L. (1991). More Gene Manipulations in
Fungi (San Diego, CA: Academic Press).
Carvalho, N.D., Arentshorst, M., Jin Kwon, M., Meyer, V., and Ram,
A.F. (2010). Expanding the ku70 toolbox for filamentous fungi:
establishment of complementation vectors and recipient strains
for advanced gene analyses. Appl. Microbiol. Biotechnol. 87,
1463–1473.
Cousin, M.A., Notermans, S., Hoogerhout, P., and Van Boom, J.H.
(1989). Detection of beta-galactofuranosidase production by
Penicillium and Aspergillus species using 4-nitrophenyl beta-D-
galactofuranoside. J. Appl. Bacteriol. 66, 311–317.
Coutinho, P.M., Andersen, M.R., Kolenova, K., vanKuyk, P.A.,
Benoit, I., Gruben, B.S., Trejo-Aguilar, B., Visser, H., van
Solingen, P., Pakula, T., et al. (2009). Post-genomic insights into
the plant polysaccharide degradation potential of Aspergillus
nidulans and comparison to Aspergillus niger and Aspergillus
oryzae. Fungal Genet. Biol. 46(Suppl 1), S161–S169.
Daley, L.S. and Strobel, G.A. (1983). [beta]-Galactofuranosidase
activity in Helminthosporium sacchari and its relationship to the
production of helminthosporoside. Plant Sci. Lett. 30, 145–154.
De Ioannes, P., Peirano, A., Steiner, J., and Eyzaguirre, J. (2000).
An alpha-L-arabinofuranosidase from Penicillium purpuroge-
num: production, purification and properties. J. Biotechnol. 76,
253–258.
AbfB modeling and substrate docking
The suggested sequence of the processed A. niger AbfB protein was
submitted to the SWISS-MODEL protein structure homology mod-
eling server (Peitsch, 1995; Arnold et al., 2006; Kiefer et al., 2009),
specifying the crystal structure of A. kawachii IFO4308 AbfB (PDB
entry 1WD4) as template (Martens-Uzunova et al., 2006). The se-
quence homology between the template and the submitted sequence
is 98.3%. Three-dimensional *.mol structure files of the substrates
where obtained from the ChemSpider database (http://www.chem-
spider.com/). The corresponding ChemSpider IDs are 133468 for
pNp-α-l-Araf and 113251 for pNp-β-d-Galf, respectively. The sub-
strates were docked manually into the substrate pocket, which is
occupied by the α-l-Araf molecule 1AHR in the protein-substrate
PDB structure 1WD4. For molecular modeling and generation of the
video frames, the PyMOL Molecular Graphics System (version 1.3,
Schrödinger, LLC, version 0.99rc6) was used.
Flipphi, M.J., van Heuvel, M., van der Veen, P., Visser, J., and de
Graaff, L.H. (1993a). Cloning and characterization of the abfB
gene coding for the major α-L-arabinofuranosidase (ABF B) of
Aspergillus niger. Curr. Genet. 24, 525–532.
Flipphi, M.J., Visser, J., van der Veen, P., and de Graaff, L.H.
(1993b). Cloning of the Aspergillus niger gene encoding
α-L-arabinofuranosidase A. Appl. Microbiol. Biotechnol. 39,
335–340.
Gielkens, M.M., Visser, J., and de Graaff, L.H. (1997). Arabinoxylan
degradation by fungi: characterization of the arabinoxylan-arab-
inofuranohydrolase encoding genes from Aspergillus niger and
Aspergillus tubingensis. Curr. Genet. 31, 22–29.
Construction of an A. niger ΔabfA mutant
The 5′- and 3′-flanking regions of AbfA (An01g00330) were ampli-
fied by PCR using the primers listed in Table 2. The resulting frag-
ments were cloned into two destination Gateway vectors. A 3.1-kb
hygromycin (HygroB) cassette was amplified by PCR using HygroB
primers (Table 2) and pAN7.1 (Z32698.1) as template DNA and sub-
sequently cloned in a third Gateway vector. Using Multisite Gateway
recombination, the three fragments were combined to create a vector
Kaneko, S., Ishii, T., Kobayashi, H., and Kusakabe, I. (1998).
Substrate specificities of α-L-arabinofuranosidases produced by
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM

α-Arabinofuranosidases show galactofuranosidase activity 775
two species of Aspergillus niger. Biosci. Biotechnol. Biochem.
62, 695–699.
Kiefer, F., Arnold, K., Kunzli, M., Bordoli, L., and Schwede,
T. (2009). The SWISS-MODEL Repository and associated
resources. Nucleic Acids Res. 37, D387–392.
Ramli, N., Fujinaga, M., Tabuchi, M., Takegawa, K., and Iwahara,
S. (1995). Isolation and characterization of a novel endo-
β-galactofuranosidase from Bacillus sp. Biosci. Biotechnol.
Biochem. 59, 1856–1860.
Remond, C., Plantier-Royon, R., Aubry, N., and O’Donohue, M.J.
(2005). An original chemoenzymatic route for the synthesis of
beta-D-galactofuranosides using an α-L-arabinofuranosidase.
Carbohydr. Res. 340, 637–644.
Repetto, E., Marino, C., Uhrig, M.L., and Varela, O. (2009).
Thiodisaccharides with galactofuranose or arabinofuranose as
terminal units: synthesis and inhibitory activity of an exo β-D-
galactofuranosidase. Bioorg. Med. Chem. 17, 2703–2711.
Reyes, F., Alfonso, C., Martinez, M.J., Prieto, A., Santamaria, F.,
and Leal, J.A. (1992). Purification of a new galactanase from
Penicillium oxalicum catalysing the hydrolysis of β-(1→5)-
galactofuran linkages. Biochem. J. 281, 657–660.
Rombouts, F.M., Voragen, A.G., Searle-van Leeuwen, M.J.F.,
Geraeds, C.C.J.M., Schols, H.A., and Pilnik, W. (1988). The ara-
binases of Aspergillus niger – purification and characterisation of
two α-L-arabinofuranosidases and an endo-1,5-α-L-arabinase.
Carbohydr. Polym. 9, 25–47.
Roth, A.H. and Dersch, P. (2010). A novel expression system for
intracellular production and purification of recombinant affinity-
tagged proteins in Aspergillus niger.Appl. Microbiol. Biotechnol.
86, 659–670.
Salvador, L.D., Suganuma, T., Kitahara, K., Tanoue, H., and Ichiki,
M. (2000). Monosaccharide composition of sweetpotato fiber and
cell wall polysaccharides from sweetpotato, cassava, and potato
analyzed by the high-performance anion exchange chromatogra-
phy with pulsed amperometric detection method. J. Agric. Food
Chem. 48, 3448–3454.
Koseki, T., Okuda, M., Sudoh, S., Kizaki, Y., Iwano, K., Aramaki, I.,
and Matsuzawa, H. (2003). Role of two α-L-arabinofuranosidases
in arabinoxylan degradation and characteristics of the encod-
ing genes from shochu koji molds, Aspergillus kawachii and
Aspergillus awamori. J. Biosci. Bioeng. 96, 232–241.
Martens-Uzunova, E.S. and Schaap, P.J. (2009). Assessment of the
pectin degrading enzyme network of Aspergillus niger by func-
tional genomics. Fungal Genet. Biol. 46(Suppl 1), S170–S179.
Martens-Uzunova, E.S., Zandleven, J.S., Benen, J.A., Awad, H.,
Kools, H.J., Beldman, G., Voragen, A.G., Van den Berg, J.A.,
and Schaap, P.J. (2006). A new group of exo-acting family 28
glycoside hydrolases of Aspergillus niger that are involved in
pectin degradation. Biochem. J. 400, 43–52.
Matsumura, K., Obata, H., Hata, Y., Kawato, A., Abe, Y., and Akita,
O. (2004). Isolation and characterization of a novel gene encod-
ing α-L-arabinofuranosidase from Aspergillus oryzae. J. Biosci.
Bioeng. 98, 77–84.
Matsuo, N., Kaneko, S., Kuno, A., Kobayashi, H., and Kusakabe,
I. (2000). Purification, characterization and gene cloning of two
α-L-arabinofuranosidases from streptomyces chartreusis GS901.
Biochem. J. 346, 9–15.
Mattern, I.E., van Noort, J.M., van den Berg, P., Archer, D.B.,
Roberts, I.N., and van den Hondel, C.A. (1992). Isolation and
characterization of mutants of Aspergillus niger deficient in
extracellular proteases. Mol. Gen. Genet. 234, 332–336.
Mennink-Kersten, M.A., Ruegebrink, D., Wasei, N., Melchers, W.J.,
and Verweij, P.E. (2006). In vitro release by Aspergillus fumiga-
tus of galactofuranose antigens, 1,3-β-D-glucan, and DNA, sur-
rogate markers used for diagnosis of invasive aspergillosis. J.
Clin. Microbiol. 44, 1711–1718.
Stynen, D., Goris, A., Sarfati, J., and Latge, J.P. (1995). A new sen-
sitive sandwich enzyme-linked immunosorbent assay to detect
galactofuran in patients with invasive aspergillosis. J. Clin.
Microbiol. 33, 497–500.
Meyer, V., Ram, A.F., and Punt, P.J. (2010). Genetics, genetic manipu-
lation, and approaches to strain improvement of filamentous fungi.
In: Manual of Industrial Microbiology and Biotechnology, J.E.
Davies and A.L. Demain, eds. (New York: Wiley), pp. 318–329.
Miyanaga, A., Koseki, T., Matsuzawa, H., Wakagi, T., Shoun,
H., and Fushinobu, S. (2004). Crystal structure of a family
54 α-L-arabinofuranosidase reveals a novel carbohydrate-
binding module that can bind arabinose. J. Biol. Chem. 279,
44907–44914.
Tefsen, B., Ram, A.F., van Die, I., and Routier, F.H. (2011).
Galactofuranose in eukaryotes: aspects of biosynthesis and func-
tional impact. Glycobiology 22, 456–469.
Van Bruggen-Van Der Lugt, A.W., Kamphuis, H.J., De Ruiter, G.A.,
Mischnick, P., Van Boom, J.H., and Rombouts, F.M. (1992). New
structural features of the antigenic extracellular polysaccharides
of Penicillium and Aspergillus species revealed with exo-β-D-
galactofuranosidase. J. Bacteriol. 174, 6096–6102.
Wallis, G.L., Hemming, F.W., and Peberdy, J.F. (2001). An extracel-
lular β-galactofuranosidase from Aspergillus niger and its use as
a tool for glycoconjugate analysis. Biochim. Biophys. Acta 1525,
19–28.
Peitsch, M. (1995). Protein modeling by e-mail. Nat. Biotechnol. 13,
658–660.
Pel, H.J., de Winde, J.H., Archer, D.B., Dyer, P.S., Hofmann, G.,
Schaap, P.J., Turner, G., de Vries, R.P., Albang, R., Albermann,
K., et al. (2007). Genome sequencing and analysis of the versa-
tile cell factory Aspergillus niger CBS 513.88. Nat. Biotechnol.
25, 221–231.
Received February 27, 2012; accepted July 2, 2012
Brought to you by | Imperial College London
Authenticated | 155.198.30.43
Download Date | 6/5/14 9:13 PM