Journal of Biotechnology 63 (1998) 199 – 210

Softwood hemicellulose-degrading enzymes from Aspergillus

niger: Purification and properties of a b-mannanase

Pia Ademark a, Arthur Varga a, József Medve a, Vesa Harjunpää b,

Torbjörn Drakenberg b, Folke Tjerneld a, Henrik Stålbrand a,*
a
Department of Biochemistry, Center for Chemistry and Chemical Engineering, Lund Uni6ersity, PO Box 124,
S-22100 Lund, Sweden
b
VTT Chemical Technology, FIN-02044 VTT (Espoo), Finland

Received 19 December 1997; received in revised form 5 June 1998; accepted 12 June 1998

Abstract

The enzymes needed for galactomannan hydrolysis, i.e. b-mannanase, a-galactosidase and b-mannosidase, were
produced by the filamentous fungus Aspergillus niger. The b-mannanase was purified to electrophoretic homogeneity
in three steps using ammonium sulfate precipitation, anion-exchange chromatography and gel filtration. The purified
enzyme had an isoelectric point of 3.7 and a molecular mass of 40 kDa. Ivory nut mannan was degraded mainly to
mannobiose and mannotriose when incubated with the b-mannanase. Analysis by 1H NMR spectroscopy during
hydrolysis of mannopentaose showed that the enzyme acts by the retaining mechanism. The N-terminus of the
purified A. niger b-mannanase was sequenced by Edman degradation, and comparison with Aspergillus aculeatus
b-mannanase indicated high identity. The enzyme most probably lacks a cellulose binding domain since it was unable
to adsorb on cellulose. © 1998 Elsevier Science B.V. All rights reserved.

Keywords: Aspergillus niger; b-Mannanase; a-Galactosidase; Hemicellulase

1. Introduction cellulose, contains b-1,4-linked D-mannopyranose

Hemicelluloses are complex polysaccharides
which are abundant in higher plant cell walls.
Galactoglucomannan, the major softwood hemi-
* Corresponding author. Fax: +46 46 2224534; e-mail:
Henrik.Stalbrand@biokem.lu.se
and D-glucopyranose units. The residues in the
main chain are partially substituted by a-1,6-
linked D-galactosyl side groups. The complete en-
zymatic degradation of hemicelluloses involves
several specific activities.
Endo-1,4-b-D-mannanase (EC 3.2.1.78) cata-
lyzes the random cleavage of b-D-1,4-mannopyra-

0168-1656/98/$ - see front matter © 1998 Elsevier Science B.V. All rights reserved.
PII S0168-1656(98)00086-8
200 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
nosyl linkages within the main chain of galac-
tomannan, glucomannan, galactoglucomannan
and mannan. Mannanases from a variety of dif-
ferent organisms have been studied, including
bacteria, fungi, higher plants and animals (re-
viewed by Dekker and Richards, 1976 and Viikari
et al., 1993). The interest in b-mannanase and
other hemicellulose-degrading enzymes has re-
cently increased, partly because of their potential
applicability in the food and paper and pulp
industries (Viikari et al., 1993, 1994).
The degradation of galactomannan and galac-
toglucomannan by b-mannanase is greatly af-
fected by the extent and pattern of substitution of
the mannan backbone. The interference of D-
galactosyl side groups with hydrolysis has been
carefully analyzed using b-mannanases from A.
niger (McCleary and Matheson, 1983) and Tri-
choderma reesei (Tenkanen et al., 1997). The com-
plete conversion of galactomannan into
D-galactose and D-mannose requires the presence
of two additional enzymes, a-galactosidase (EC
3.2.1.22) and b-mannosidase (EC 3.2.1.25). These
enzymes catalyze the cleavage of terminal a-1,6-
linked D-galactosyl and b-1,4-linked D-mannopy-
ranosyl residues, respectively. A considerable
variation of the ability to hydrolyze galactosyl
side groups from polymeric substrates exists
among the a-galactosidases.
A. niger, a filamentous fungus, is one of the
most used organisms in the industrial production
of fermented foods, organic acids, and enzymes
(Barbesgaard, 1977, Bennett, 1985). The purifica-
tion of b-mannanase from A. niger has been
described earlier (Eriksson and Winell, 1968, Tsu-
jisaka et al., 1972, Yamazaki et al., 1976, Mc-
Cleary, 1979, 1988), but it has not been clear if
more than one b-mannanase is secreted. A multi-
plicity of extracellular b-mannanases appears to
be common among fungi and has been noticed,
for example, in T. reesei (Stålbrand et al., 1993,
Stålbrand, 1995) and Trichoderma harzianum
(Torrie et al., 1990). In the present study the aim
was to separate analytically the major galactoglu-
comannan-degrading enzymes from A. niger and
to purify and characterize biochemically the b-
mannanase.
2. Materials and methods
2.1. Culture
A. niger ATCC-46890 was obtained from the
QM Culture Collection, Department of Botany,
University of Massachusetts, Amherst. The cul-
ture was maintained on potato dextrose agar
slants.
2.2. Screening of growth media
The effect of different culture media on b-man-
nanase production was investigated. A. niger
ATCC-46890 was grown at 28°C in shake-flasks
containing one of the following carbon sources:
0.5–2% (w/v) locust bean gum galactomannan
(Sigma), 0.5% Solka Floc cellulose, or 1% glucose.
All media were supplemented with 4% Vogel’s
medium N (Vogel, 1964), 0.1% peptone and 0.1%
citric acid monohydrate. Samples were collected
and assayed for b-mannanase activity at various
times during the cultivation.
2.3. Enzyme production
The fungus was cultivated in 1-liter shake-flasks
on a medium containing 0.2% peptone, 2% locust
bean gum, 4% Vogel’s medium N and 0.015%
Tween 80; 100 ml of medium was inoculated with
5 ml of mycelia from a 3-day-old culture. After 7
days of growth at 28°C and 250 rpm, mycelia
were removed and the culture fluid was filtered
through a sterile 0.45-mm pore size membrane
filter.
2.4. Acti6ity assays
b-Mannanase activity was assayed as described
by Stålbrand et al. (1993) using locust bean gum
galactomannan (Sigma G-0753) as substrate. a-
Galactosidase and b-mannosidase activities were
assayed with p-nitrophenyl-a-D-galactopyra-
noside (Sigma N-0877) at pH 4.5 and p-nitro-
phenyl-b-D-mannopyranoside (Sigma N-1268) at
pH 5.3, respectively, as described by Rättö and
Poutanen (1988). All enzyme activities are ex-
pressed in SI units (katals).
 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
2.5. Protein assay
Protein concentrations were measured by the
Micro BCA Protein Assay (Pierce, Rockford, IL),
using bovine serum albumin as standard. All
chromatographic runs were monitored for protein
by absorbance at 280 nm.
2.6. Gel electrophoresis and zymogram analysis
Sodium dodecyl sulfate – polyacrylamide gel
electrophoresis (SDS-PAGE) and isoelectric fo-
cusing (IEF) were performed using the PhastSys-
tem™ (Pharmacia Biotech, Uppsala, Sweden).
Proteins were detected with silver staining as de-
scribed in the Pharmacia PhastSystem™ instruc-
tions. Isoelectric focusing was carried out in the
pH range 3–9 using gels of the type PhastGel®
IEF 3-9 (Pharmacia). Marker proteins in the pI
range 3.5–9.3 (Broad pI Kit, Pharmacia) were
used as standards. SDS-PAGE was carried out on
12.5% polyacrylamide gels (PhastGel® Homoge-
neous 12.5, Pharmacia) using the Low Molecular
Weight Calibration Kit (Pharmacia) as standard.
b-Mannanase activity was detected by using an
IEF zymogram technique with locust bean gum as
substrate and staining with Congo red (Stålbrand
et al., 1993). a-Galactosidase activity was detected
by using 5-bromo-4-chloro-3-indolyl-a-D-galac-
topyranoside (xagal, Boehringer Mannheim) as
substrate. The zymogram was prepared as de-
scribed by den Herder et al. (1992), with the
exception that the incubation was made at pH 5.3
instead of 4.5. a-Galactosidase activity could be
seen as light blue bands on the transparent gel.
2.7. Purification of the b-mannanase
Culture filtrate (100 ml) was brought to 80%
saturation by the addition of 56.1 g of solid
ammonium sulfate. The precipitate was collected
by centrifugation and then dissolved in 20 ml of
citrate buffer, pH 5.3; 3 ml of this solution was
desalted and equilibrated with 20 mM ammonium
acetate buffer, pH 7.5, using a Fast Desalting
Column (Pharmacia). The sample was then di-
rectly loaded on a Resource™ Q 1-ml anion-ex-
change column (Pharmacia) equilibrated with the
201
same buffer. Proteins were eluted with a linear
gradient (0–60%) of 1 M ammonium acetate, pH
7.5. Fractions of 0.25 ml were collected and as-
sayed for b-mannanase, a-galactosidase and b-
mannosidase activities.
The ten fractions most active in b-mannanase
were pooled and further purified on a HiLoad™
16/60 gel filtration column pre-packed with Su-
perdex 200 prep grade (Pharmacia). Elution was
achieved in 4 h using 200 mM NaCl in 100 mM
phosphate buffer, pH 6.5. The flow rate was 0.5
ml min − 1. Fractions of 1 ml were collected and
assayed for b-mannanase and a-galactosidase ac-
tivities. Both ion-exchange chromatography and
gel filtration were performed using the FPLC®
System (Pharmacia).
2.8. pH optimum and stability
The purified b-mannanase was incubated at
50°C at different pH values (pH 2.6–7: 50 mM
citrate–phosphate buffer; pH 7.5–8: 50 mM
phosphate buffer) at a concentration of 1.7 mg
ml − 1. Bovine serum albumin (BSA, Sigma) was
added at a concentration of 100 mg ml − 1. Sam-
ples were withdrawn at zero time and 24 h and
assayed under standard conditions. The pH opti-
mum was determined by using substrates with
different pH values.
2.9. Temperature stability
The purified b-mannanase was incubated (3.4
mg ml − 1) in citrate–phosphate buffer (50 mM,
pH 5.3) including BSA (100 mg ml − 1) at different
temperatures for 24 h. Following incubation, the
remaining activities were measured as described
above.
2.10. N-terminal sequence determination
The N-terminal amino acid sequence of the
purified b-mannanase was determined by auto-
mated Edman degradation, using an ABI 477A
protein sequencer connected to an ABI 120A
analyzer (Applied Biosystems, USA). The analysis
was performed at the Biomolecular Resource Fa-
cility at Lund University.
202 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
2.11. Adsorption studies
The T. reesei mannanase (pI 5.4) was purified as
described by Stålbrand et al. (1993). Microcrys-
talline cellulose (Avicel, M 2331) was purchased
from Merck (Darmstadt, Germany). The adsorp-
tion experiments were performed in 1.8-ml plastic
tubes (Nunc, Denmark) at pH 5.3 in 50 mM
sodium citrate buffer. The enzyme/substrate ratio
was varied in the range 0.2 – 80 mmol g − 1 Avicel
and the incubation was carried out at 4°C with
continuous mixing by inversion of the tubes. After
incubation for 60 min, the Avicel with bound
enzyme was removed by filtration through a small
syringe filter (Millex-GV4, pore size 0.22 mm;
Millipore). The residual mannanase activity was
measured and the adsorption was expressed as
percent of the enzyme added. The time needed to
achieve maximal adsorption at the given condi-
tions was predetermined.
2.12. Hydrolysis of i6ory nut mannan
Ivory nut (Phytelephas macrocarpa) mannan, an
unbranched b-1,4-linked mannan polymer, was
purchased from Megazyme, Australia. The poly-
mer was suspended (2.5 g l − 1 in 20 mM sodium
acetate buffer, pH 4.5) at 80°C with continuous
stirring. After addition of the purified b-man-
nanase (2000 nkat g − 1 substrate), the solution was
incubated at 40°C for 48 h with continuous mix-
ing. Samples were removed at various times,
heated at 100°C for 5 min and then stored at
−20°C. The products formed during hydrolysis
were analyzed on a Pharmacia HPLC system
equipped with an Erma ERC-7515A refractive
index detector (Erma, Tokyo, Japan); 20-ml sam-
ples were applied on an Aminex HPX-87H column
(Bio-Rad, Richmond, CA). Elution was carried
out at 65°C with 5 mM H2SO4 at a flow rate of 0.6
ml min − 1. Mannobiose and mannotriose
(Megazyme, Australia) were used as standards.
2.13. Hydrolysis of locust bean gum
Locust bean gum was prepared at a concentra-
tion of 0.25% (w/v) in 20 mM sodium acetate
buffer, pH 4.5. Culture filtrate was added at a
concentration corresponding to a mannanase ac-
tivity of 2000 nkat g − 1 substrate. The solution was
then incubated at 40°C for 48 h with continuous
mixing. Aliquots were removed at various times,
heated at 100°C for 5 min, and then stored at
− 20°C. The hydrolysis products were quantified
by HPLC, using an Aminex HPX-87P column
(Bio-Rad, Richmond, CA). Millipore water was
used as eluant at 85°C at a flow rate of 0.6 ml
min − 1. Mannose and galactose (Sigma) were used
as standards.
2.14. Analysis of mannopentaose degradation by
nuclear magnetic resonance (NMR)
The hydrolysis reaction was carried out directly
in the NMR tube. Mannopentaose was dissolved
in 50 mM deuterated acetate buffer in 2H2O (pD
4.5) to a concentration of 3 mM. After accumula-
tion of an initial spectrum, b-mannanase was
added to the substrate solution in the NMR tube.
The tube was quickly transferred to the spectrom-
eter at 55°C and a new spectrum accumulation
started as soon as possible. New spectra were
thereafter started every 10 min. The 1H NMR
spectra were recorded at 599.94 MHz using a
Varian Unity 600 spectrometer. Each spectrum
was the sum of 26 accumulations taken over 10
min. For details see Harjunpää et al. (1995).
2.15. Acid hydrolysis
Acid hydrolysis of locust bean gum and ivory
nut mannan was carried out. A volume of 5 ml of
0.25 M H2SO4 including 0.25% of the polymer was
autoclaved for 2 h at 120°C. More H2SO4 was
added to final concentration of 0.4 M and the
solution was again heated for 2 h at 120°C.
Samples of 100 ml were then analyzed by HPLC
using Aminex HPX-87P and HPX-87H columns.
3. Results
3.1. Screening of growth media
b-Mannanase activity was assayed in the A.
niger cultures containing either galactomannan,
 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210 203

Fig. 1. Isoelectric focusing of the culture filtrate, silver stained (A) and activity stained for b-mannanase (B), and of the purified
b-mannanase with silver staining (C). Isoelectric focusing with a-galactosidase activity staining of the a-galactosidase peak
eluting adjacent to b-mannanase in ion-exchange chromatography (D), the culture filtrate (E) and the a-galactosidase peak
eluting far ahead of b-mannanase in ion-exchange chromatography (F). Standards are marked at the sides. For experimental
details see text.

cellulose or glucose, as described in Section 2. 3.3. Purification of the b-mannanase

The growth was interrupted after 7 days when
the b-mannanase production had reached a
maximum in all flasks. A. niger produced b-
mannanase activity (32 – 56 nkat ml − 1) when
grown on media containing locust bean gum
galactomannan (results not shown). The highest
activity (56 nkat ml − 1) was observed in the
medium supplied with 2% locust bean gum. No
b-mannanase activity could be detected when
glucose and cellulose were used as sole carbon
sources.
3.2. Characterization of the culture filtrate en-
zymes
The following enzyme activities, necessary for
complete hydrolysis of galactomannan, were de-
tected in the culture filtrate of A. niger at the
time for harvest: b-mannanase 90 nkat ml − 1,
a-galactosidase 47 nkat ml − 1, b-mannosidase
8.1 nkat ml − 1. Isoelectric focusing and zy-
mogram analysis of the culture filtrate showed
several poorly separated a-galactosidase bands in
the pI interval 4.2 – 6.6, with two major areas
around pI 4.4 and 5.9 (Fig. 1E). A single clear
single band of b-mannanase activity was de-
tected at pI 3.7 (Fig. 1B).
Ammonium sulfate precipitation was used
mainly as a concentrating step before applying the
sample to the Resource Q column; 97% of the
b-mannanase activity was recovered in the precip-
itate. Analysis of the collected fractions from an-
ion-exchange chromatography (Fig. 2) showed a
single peak of b-mannanase activity. Two peaks
of a-galactosidase activity, one eluting adjacent to
the b-mannanase, and a single b-mannosidase
peak were also resolved. Analysis of the b-man-
nanase fractions by SDS-PAGE (Fig. 3A) showed
two major proteins, one of 40 kDa (b-man-
nanase) and one of 76 kDa. Further purification
by gel filtration yielded a b-mannanase prepara-
tion that was pure as judged from IEF (Fig. 1C)
and SDS-PAGE (Fig. 3B). The specific activity
was determined to be 3860 nkat mg − 1 protein.
No a-galactosidase activity could be detected in
the mannanase peak after an assay time of 90
min. An overall recovery of 46% and a 46-fold
purification of the b-mannanase were obtained.
The purification is summarized in Table 1.
3.4. Enzyme properties
The isoelectric point of the purified b-man-
nanase was 3.7 (Fig. 1C) and the apparent
 204

Table 1
Purification of b-mannanase from A. niger culture filtrate

Purification step Volume Activity Total activity Total protein (mg) Specific activity Yield Fold purification
(ml) (nkat ml−1) (nkat) (nkat mg−1) (%)

Culture filtrate 15 90 1350 16 84 (100) (1)

(NH4)2SO4 precipitation 3.0 436 1308 6.7 195 97 2.3
Ion-exchange chro- 2.5 350 875 0.70 1250 65 15
matography
Gel filtration 6.0 103 618 0.16 3860 46 46
P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210 205

Fig. 2. Anion-exchange chromatography of the ammonium sulfate precipitated culture filtrate on a Resource™ Q 1-ml column. For
experimental details see text. ——, absorbance at 280 nm; - - -, gradient; “, b-mannanase ×0.1; , a-galactosidase; ,
b-mannosidase.

molecular weight 40 kDa (Fig. 3B). Zymogram
analysis of the a-galactosidase peak eluting close
to b-mannanase in ion-exchange chromatography
showed a single band at pI 6.6 (Fig. 1D). The
a-galactosidase peak eluting early in the chro-
matogram appeared as a broad, diffuse band at
around pI 4.4 (Fig. 1F). Optimal pH for b-man-
nanase activity was 3.5 (Fig. 4). The enzyme
proved to be stable in the pH range 3.5 – 7 (Fig. 4)
and at temperatures of 50°C and below (data not
shown). The N-terminal sequence of the b-man-
nanase is shown in Fig. 5 together with the gene
deduced amino acid sequences of T. reesei
(Stålbrand et al., 1995) and A. aculeatus (Christ-
gau et al., 1994).

Fig. 4. Effect of pH on b-mannanase activity (“) and stability

Fig. 3. SDS-PAGE with silver staining of the b-mannanase
fractions after ion-exchange chromatography (A) and after gel
filtration (B). Standards are marked at the sides. For experi-
mental details see text.
(). The pH optimum, expressed in percent of maximum, was
determined by measuring the activity under standard condi-
tions using buffers of different pH. The pH stability was
determined by incubating the purified enzyme at different pH
values for 24 h at 50°C. The residual activity is shown as
percent of the original activity.
206 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
Fig. 5. N-terminal amino acid sequence comparison of b-man-
nanases from Aspergillus niger, Aspergillus aculeatus and Tri-
choderma reesei. Identical amino acid residues are boxed. The
A. niger b-mannanase sequence was determined by Edman
degradation. The other two sequences are gene deduced. Ref-
erences: 1, Christgau et al. (1994); 2, Stålbrand et al. (1995).
3.5. Adsorption on cellulose
The adsorption of A. niger b-mannanase on
cellulose was compared with that of a b-man-
nanase from T. reesei which has a cellulose
binding domain (CBD) (Stålbrand et al., 1995)
(Fig. 6). The A. niger b-mannanase was unable
to bind to cellulose at all the enzyme/substrate
ratios tested. As expected, the T. reesei man-
nanase was readily adsorbed. About 80% of the
enzyme was bound at a protein concentration of
80 mmol g − 1 Avicel, and it approached 100%
bound when the concentration decreased below
10 mmol g − 1 Avicel. The amount of adsorbed
enzyme reached a maximum after 30 min incu-
bation; no further adsorption was detected even
after 4 h (data not shown).
Fig. 6. Adsorption of Aspergillus niger and Trichoderma reesei
b-mannanases on cellulose after 60 min of incubation at 4°C.
The enzyme/substrate ratio was varied in the range 0.2–80
mmol g − 1 Avicel. , T. reesei b-mannanase; “, A. niger
b-mannanase.
3.6. Hydrolysis experiments
Locust bean gum was incubated with H2SO4
as described in Section 2. An aliquot of 234 mg
locust bean gum yielded 132 mg mannose and 32
mg galactose after acid hydrolysis. The mannose/
galactose ratio was determined as 4.1 on a mo-
lar basis, which is very close to previously
reported values (Rol, 1973). Acid hydrolysis of
234 mg of ivory nut mannan yielded 194 mg of
mannose.
Incubation of locust bean gum with the cul-
ture filtrate yielded mainly mannose and galac-
tose (Fig. 7A). The molar ratio of
mannose/galactose produced by enzymatic hy-
drolysis for 48 h was 3.3 and the amount of
released monomers 66% compared to the value
obtained by total acid hydrolysis. Ivory nut
mannan was degraded mainly to mannobiose
and mannotriose when incubated with the
purified b-mannanase for 48 h (Fig. 7B). The
molar ratio of released mannobiose/mannotriose
was 1.2.
The hydrolysis of mannopentaose was ana-
lyzed by 1H NMR spectroscopy. The initial
products were equal amounts of mannobiose
and mannotriose (Fig. 8B), showing that the two
central glycosidic bonds of the oligosaccharide
are cleaved. Prolonged hydrolysis also results in
the formation of mannose from the secondary
hydrolysis of mannotriose. The progress curves
in Fig. 8B show convincingly that the b-man-
nanase acts by the retaining mechanism (Sinnott,
1990) since there is an initial fast increase in the
signal intensity from b-anomeric protons.
The 1H NMR signal from the internal
anomeric proton in mannotriose (shown in Fig.
8A) can be used to deduce the degradation pat-
tern. If the hydrolysis occurs at the second gly-
cosidic bond from the non-reducing end only,
the mannotriose formed would have the equi-
librium a/b ratio and the two resonances ob-
served for the anomeric proton from the internal
ring would have a time-independent ratio. This
is clearly not the case. On the other hand, if
only the third bond is hydrolyzed, the formed
mannotriose would be completely in the b-
 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210 207

Fig. 7. (a) HPLC analysis with the Aminex HPX-87P column of the products formed after hydrolysis of locust bean gum at 40°C
for 48 h. Incubation of the substrate with culture filtrate (A) and without enzyme addition (B). Gal and Man indicate galactose and
mannose, respectively. The large peak eluting at approximately 29 ml is acetate from the buffer. (B) HPLC analysis with the Aminex
HPX-87H column of the products formed during hydrolysis of ivory nut mannan at 40°C for 48 h. Incubation of the substrate with
purified b-mannanase (A) and without enzyme addition (B). M2 and M3 indicate mannobiose and mannotriose, respectively. The
large peak eluting at approximately 18 ml is acetate from the buffer.

anomeric form with the a-form obtained through 4. Discussion

mutarotation. An equal probability for cleavage
of the two central glycosidic bonds would result in
the formation of 68% b and 32% a-mannotriose.
It is clear that more of the b-form is formed than
of the a-form; however, it is not straightforward
to obtain this excess since the mutarotation can
not be totally isolated from the hydrolysis. Model
calculations, not shown, indicate a weak prefer-
ence for cleavage of the third glycosidic bond over
the second one.
The b-mannanase gene sequences of two fungal
strains, A. aculeatus (Christgau et al., 1994) and
T. reesei (Stålbrand et al., 1995), have been pub-
lished. The b-mannanase produced by A. aculea-
tus is 62 amino acids shorter and lacks the
C-terminal CBD sequence present in the T. reesei
enzyme. Alignment of the N-terminus of the
purified A. niger b-mannanase with that of the A.
aculeatus b-mannanase shows a high sequence
208 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
identity (eight identical residues out of 12) (Fig.
5). The resemblance to the T. reesei mannanase is
somewhat lower: five residues out of 12 are identi-
cal. The A. niger and A. aculeatus b-mannanases
are both smaller than the one from T. reesei and
Fig. 8. (A) 1H NMR spectra of the region with the resonance
from the anomeric proton from the internal ring in man-
notriose. (B) Progress curves for the hydrolysis of mannopen-
taose by A. niger mannanase showing the intensities of various
1
H NMR resonances from the anomeric protons as a function
of time. Conditions: 3 mM mannopentaose and 3.6 mM man-
nanase at 5°C. Symbols used in the figure are: “, internal;
, b-mannanase has proved to be stable in the pH
terminal; , reducing b; ", reducing a.
have very similar molecular weights (40000 and
45000, respectively). Consequently, judged from
its size, it would not be surprising if the A. niger
mannanase also lacks a CBD. This speculation is
strongly supported by the adsorption data (Fig.
6). The T. reesei mannanase bound efficiently on
cellulose whereas no binding could be observed
for the A. niger mannanase. The strong adsorp-
tion of T. reesei mannanase on cellulose has previ-
ously been reported by Tenkanen et al. (1995).
The addition of galactomannan (locust bean
gum) to the culture media induced formation of
b-mannanase, a-galactosidase and b-mannosi-
dase, whereas no detectable b-mannanase activity
was produced when the fungus was grown on
cellulose or glucose as sole carbon source. It is
worth mentioning that cellulose induces higher
b-mannanase activity than galactomannan in T.
reesei (Rättö and Poutanen, 1988, Arisan-Atac et
al., 1993) and that the enzyme also contains a
CBD (Stålbrand et al., 1995). However, the A.
niger b-mannanase in this study was not induced
by cellulose, which is an interesting observation
considering that it probably does not have a
CBD.
Purification of the A. niger mannanase was
achieved using only three steps. The described
method is very simple and yields a high recovery
of mannanase activity (46%), suggesting that it
might be suitable for upscaling trials. The specific
activity using locust bean gum as substrate was
3860 nkat mg − 1 protein, which is six times the
value obtained by McCleary (1988). The isoelec-
tric point of the single b-mannanase detected was
3.7 and the molecular mass 40 kDa. In earlier
investigations the pI has been determined to the
following values: 3.2 (Tsujisaka et al., 1972), 4.0
(McCleary, 1979, 1988) and 4.1 (Yamazaki et al.,
1976). The pI and the molecular mass for a b-
mannanase purified from a commercial enzyme
preparation produced by a fungus from the As-
pergillus niger-oryzae group have been determined
to 3.95 (Ahlgren et al., 1967) and 42 kDa
(Eriksson and Winell, 1968), respectively. The
molecular weight has also been determined earlier
to 45000 (McCleary, 1979, 1988). The A. niger
range 3–8 and below 70°C (McCleary, 1979,
 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
1988) and the pH optimum has been determined
to 3.6 (Yamazaki et al., 1976) and 3.0 (McCleary,
1979). The pI and molecular weight data in the
literature agree well with the results of the present
investigation, and accordingly A. niger seems to
secrete only one b-mannanase under the circum-
stances used. It can, however, not be ruled out
that A. niger secretes additional b-mannanases
which are present at very low levels or are induced
if using other growth conditions. At least two
enzymes with a-galactosidase activity, in the pI
range 4.2–6.6, were detected in the A. niger cul-
ture filtrate. The production of multiple forms of
a-galactosidase by A. niger is also confirmed by
earlier investigations (Lee and Wacek, 1970, den
Herder et al., 1992).
The formation of mannobiose and mannotriose
after hydrolysis of ivory nut mannan (Fig. 7B)
indicates that the purified b-mannanase is an
endoenzyme (Reese and Shibata, 1965). Similar
results were obtained with b-mannanases purified
from T. reesei (Stålbrand et al., 1993), Aspergillus
tamarii (Civas et al., 1984) and Aspergillus gigan-
teus (Reese and Shibata, 1965). The ability of A.
niger b-mannanase to attack manno-oligosaccha-
rides with a degree of polymerization of three and
higher was shown by Eriksson and Winell (1968).
The purified mannanase partly hydrolyzed guar
galactomannan to mannobiose and mannotriose,
but showed no activity against mannobiose. Ya-
mazaki et al. (1976) reported that the A. niger
mannanase was able to degrade mannotriose and
mannotetraose, but not mannobiose. Mannotriose
is however hydrolyzed very slowly; a degree of
polymerization of at least four is required for a
significant hydrolysis rate (McCleary and
Matheson, 1983).
The results in Fig. 8 show that the A. niger
b-mannanase has an activity towards mannopen-
taose that is very similar to that of the two major
b-mannanases from T. reesei (Harjunpää et al.,
1995). Like a T. reesei mannanase it acts accord-
ing to the retaining mechanism and is therefore
also able to perform transglycosylation. Further-
more, also like the T. reesei mannanase, there is
no strong preference for hydrolysis of either the
second or third glycosidic linkage. The transglyco-
sylation capacity of A. niger b-mannanase was
209
demonstrated by McCleary and Matheson (1983).
A transient synthesis of higher oligosaccharides
was detected during the early stages of reaction
with mannopentaose and mannotetraose, and the
final products after further hydrolysis were mainly
mannobiose and mannotriose.
Locust bean gum galactomannan was degraded
to galactose and mannose when incubated with
the culture filtrate enzymes (Fig. 7A). The com-
plete enzymatic hydrolysis of this polymer re-
quires the action of b-mannanase, a-galactosidase
and b-mannosidase, which are all produced by A.
niger. The detailed co-operative action of these
enzymes on galactomannan and galactoglucoman-
nan, however, remains to be investigated.
Acknowledgements
This study was supported by the Swedish Na-
tional Board for Industrial and Technical Devel-
opment (NUTEK).
References
Ahlgren, E., Eriksson, K.-E., Vesterberg, O., 1967. Character-
ization of cellulases and related enzymes by isoelectric
focusing, gel filtration and zone electrophoresis. Acta
Chem. Scand. 21, 937 – 944.
Arisan-Atac, I., Hodits, R., Kristufek, D., Kubicek, C.P.,
1993. Purification, and characterization of a b-mannanase
of Trichoderma reesei C-30. Appl. Microbiol. Biotechnol.
39, 58 – 62.
Barbesgaard, P., 1977. In: Smith, J.E., Pateman, J.A. (Eds.),
Genetics and Physiology of Aspergillus (British Mycology
Society Symposium Series No. 1). Academic Press, Lon-
don, pp. 391 – 404.
Bennett, J.W., 1985. In: Timberlake, W.E. (Ed.), Molecular
Genetics of Filamentous Fungi. Alan R. Liss, New York,
pp. 345 – 366.
Christgau, S., Kauppinen, S., Vind, J., Kofod, L.V., Dalboge,
H., 1994. Expression cloning, purification and characteri-
zation of a b-1,4-mannanase from Aspergillus aculeatus.
Biochem. Mol. Biol. Int. 33, 917 – 925.
Civas, A., Eberhard, R., Le Dizet, P., Petek, F., 1984. Glycosi-
dases induced in Aspergillus tamarii. Secreted a-D-galac-
tosidase and b-D-mannanase. Biochem. J. 857 – 863.
Dekker, R.F.H., Richards, G.N., 1976. Hemicellulases: their
occurrence, purification, properties, and mode of action.
Adv. Carbohydr. Chem. Biochem. 32, 277 – 352.
210 P. Ademark et al. / Journal of Biotechnology 63 (1998) 199–210
den Herder, I.F., Mateo Rosell, A.M., van Zuilen, C.M., Punt,
P.J., van den Hondel, C.A.M.J.J., 1992. Cloning and ex-
pression of a member of the Aspergillus niger gene family
encoding a-galactosidase. Mol. Gen. Genet. 233, 404–410.
Eriksson, K.-E., Winell, M., 1968. Purification and characteri-
zation of a fungal b-mannanase. Acta Chem. Scand. 22,
1924 – 1934.
Harjunpää, V., Teleman, A., Drakenberg, T., 1995. Kinetic
and stereochemical studies of manno-oligosaccharide hy-
drolysis catalysed by b-mannanases from Trichoderma ree-
sei. Eur. J. Biochem. 234, 278–283.
Lee, Y.C., Wacek, V., 1970. Galactosidases from Aspergillus
niger. Arch. Biochem. Biophys. 138, 264–271.
McCleary, B.V., 1979. Modes of action of b-mannanase en-
zymes of diverse origin on legume seed galactomannans.
Phytochemistry 18, 757–763.
McCleary, B.V., 1988. b-D-Mannanase. Methods Enzymol.
160, 596 – 610.
McCleary, B.V., Matheson, N.K., 1983. Action patterns and
substrate-binding requirements of b-D-mannanase with
mannosaccharides and mannan-type polysaccharides. Car-
bohydr. Res. 119, 191–219.
Rättö, M., Poutanen, K., 1988. Production of mannan-degrad-
ing enzymes. Biotechnol. Lett. 10 (9), 661–664.
Reese, E.T., Shibata, Y., 1965. b-Mannanases of fungi. Can.
J. Microbiol. 11, 167 –183.
Rol, F., 1973. Locust bean gum. In: Whistler, R.L. (Ed.),
Industrial Gums, 2nd ed. Academic Press, New York, pp.
323 – 337.
Sinnott, M.L., 1990. Catalytic mechanisms of enzymic glycosyl
transfer. Chem. Rev. 90, 1171–1202.
Stålbrand, H., 1995. Hemicellulose-degrading enzymes from
fungi. Characterization of b-mannanase and the man1
gene of Trichoderma reesei. Doctoral Thesis, Lund
University.
Stålbrand, H., Siika-aho, M., Tenkanen, M., Viikari, L., 1993.
Purification and characterization of two b-mannanases
from Trichoderma reesei. J. Biotechnol. 29, 229 – 242.
Stålbrand, H., Saloheimo, A., Vehmaanperä, J., Henrissat, B.,
Penttilä, M., 1995. Cloning and expression in Saccha-
romyces cere6isiae of a Trichoderma reesei b-mannanase
gene containing a cellulose binding domain. Appl. Env.
Microbiol. 61, 1090 – 1097.
Tenkanen, M., Buchert, J., Viikari, L., 1995. Binding of
hemicellulases on isolated polysaccharide substrates. En-
zyme Microb. Technol. 17, 499 – 505.
Tenkanen, M., Makkonen, M., Perttula, M., Viikari, L., Tele-
man, A., 1997. Action of Trichoderma reesei mannanase on
galactoglucomannan in pine kraft pulp. J. Biotechnol. 57,
191 – 204.
Torrie, J.P., Senior, D.J., Saddler, J.N., 1990. Production of
b-mannanase by Trichoderma harzianum E58. Appl. Mi-
crobiol. Biotechnol. 34, 303 – 307.
Tsujisaka, Y., Hiyama, K., Takenishi, S., Fukumoto, J., 1972.
Studies on the hemicellulases part III. Purification and
some properties of mannanase from Aspergillus niger van
Tieghem sp. Nippon Nogei Kagaku Kaishi 46, 155 – 161.
Viikari, L., Tenkanen, M., Buchert, J., Rättö, M., Bailey, M.,
Siika-Aho, M., Linko, M., 1993. Hemicellulases for indus-
trial applications. In: Saddler, J.N. (Ed.), Biotechnology in
Agriculture, No. 9. Bioconversion of Forest and Agricul-
tural Plant Residues. C.A.B. International, Wallingford,
pp. 131 – 182.
Viikari, L., Kantelinen, A., Sundquist, J., Linko, M., 1994.
Xylanases in bleaching: from an idea to the industry.
FEMS Microbiol. Rev. 13, 335 – 350.
Vogel, H.J., 1964. Distribution of lysine pathways among
fungi: evolutionary implications. Am. Nat. 98, 435 – 446.
Yamazaki, N., Sinner, M., Dietrichs, H.H., 1976. Isolierung
und eigenschaften einer b-1,4-Mannanase aus Aspergillus
niger. Holzforschung 30, 101 – 109.