The production, purification and characterisation of two novel α-D-mannosidases from Aspergillus phoenicis

Article abstract of DOI:10.1016/j.carres.2005.01.005

1,6-α-D-Mannosidase from Aspergillus phoenicis was purified by anion-exchange chromatography, chromatofocussing and size-exclusion chromatography. The apparent molecular weight was 74 kDa by SDS-PAGE and 81 kDa by native-PAGE. The isoelectric point was 4.6. 1,6-α-d-Mannosidase had a temperature optimum of 60°C, a pH optimum of 4.0-4.5, a K_m of 14 mM with α-d-Manp-(1→6)-d-Manp as substrate. It was strongly inhibited by Mn²⁺ and did not need Ca²⁺ or any other metal cofactor of those tested. The enzyme cleaves specifically (1→6)-linked mannobiose and has no activity towards any other linkages, p-nitrophenyl-α-d- mannopyranoside or baker's yeast mannan. 1,3(1,6)-α-d-Mannosidase from A. phoenicis was purified by anion-exchange chromatography, chromatofocussing and size-exclusion chromatography. The apparent molecular weight was 97 kDa by SDS-PAGE and 110 kDa by native-PAGE. The 1,3(1,6)-α-d-mannosidase enzyme existed as two charge isomers or isoforms. The isoelectric points of these were 4.3 and 4.8 by isoelectric focussing. It cleaves α-d-Manp-(1→3)-d- Manp 10 times faster than α-d-Manp-(1→6)-d-Manp, has very low activity towards p-nitrophenyl-α-d-mannopyranoside and baker's yeast mannan, and no activity towards α-d-Manp-(1→2)-d-Manp. The activity towards (1→3)-linked mannobiose is strongly activated by 1 mM Ca ²⁺ and inhibited by 10 mM EDTA, while (1→6)-activity is unaffected, indicating that the two activities may be associated with different polypeptides. It is also possible that one polypeptide may have two active sites catalysing distinct activities.

Full text of DOI:10.1016/j.carres.2005.01.005

Carbohydrate
RESEARCH
Carbohydrate Research 340 (2005) 609–617
The production, purification and characterisation of two novel
a-D-mannosidases from Aspergillus phoenicis
Vasileios I. Athanasopoulos, Keshavan Niranjan and Robert A. Rastall*
School of Food Biosciences, The University of Reading, PO Box 226, Whiteknights, Reading RG6 6 AP, UK
Received 11 June 2004; received in revised form 30 November 2004; accepted 3 January 2005
Available online 21 January 2005
Abstract—1,6-a-D-Mannosidase from Aspergillus phoenicis was purified by anion-exchange chromatography, chromatofocussing
and size-exclusion chromatography. The apparent molecular weight was 74 kDa by SDS-PAGE and 81 kDa by native-PAGE.
The isoelectric point was 4.6. 1,6-a-D-Mannosidase had a temperature optimum of 60 ꢀC, a pH optimum of 4.0–4.5, a K
m
of
2
+
2+
1
4 mM with a-D-Manp-(1!6)-D-Manp as substrate. It was strongly inhibited by Mn and did not need Ca or any other metal
cofactor of those tested. The enzyme cleaves specifically (1!6)-linked mannobiose and has no activity towards any other linkages,
p-nitrophenyl-a-D-mannopyranoside or bakerÕs yeast mannan. 1,3(1,6)-a-D-Mannosidase from A. phoenicis was purified by anion-
exchange chromatography, chromatofocussing and size-exclusion chromatography. The apparent molecular weight was 97 kDa
by SDS-PAGE and 110 kDa by native-PAGE. The 1,3(1,6)-a-D-mannosidase enzyme existed as two charge isomers or isoforms.
The isoelectric points of these were 4.3 and 4.8 by isoelectric focussing. It cleaves a-D-Manp-(1!3)-D-Manp 10 times faster than
a-D-Manp-(1!6)-D-Manp, has very low activity towards p-nitrophenyl-a-D-mannopyranoside and bakerÕs yeast mannan, and no
2
+
activity towards a-D-Manp-(1!2)-D-Manp. The activity towards (1!3)-linked mannobiose is strongly activated by 1 mM Ca
and inhibited by 10 mM EDTA, while (1!6)-activity is unaffected, indicating that the two activities may be associated with different
polypeptides. It is also possible that one polypeptide may have two active sites catalysing distinct activities.
ꢁ
2005 Elsevier Ltd. All rights reserved.
Keywords: 1,6-a-D-Mannosidase; Enzyme purification; Linkage specificity; Glycosidase
1
. Introduction
on (1!2)-only, EC 3.2.1.114 to those acting on (1!3)-
and (1!6)-linkages and EC 3.2.1.137 to those acting
on (1!2)- and (1!6)-linkages. There are no (1!6)- or
(1!3)-linkage specific mannosidases included yet. There
has, however, been one (1!6)-specific enzyme reported
Alpha-D-mannosidases are assigned the following five
EC numbers by the Nomenclature Committee of the
International Union of Biochemistry and Molecular
Biology: EC 3.2.1.24 to those that do not have strict
linkage specificity, EC 3.2.1.77 to those acting on
1
2
in the literature, and is commercially available, but it
has not been fully characterised. A (1!3)-linkage spe-
cific exo-mannosidase has not been discovered to date.
In vivo, mannosidases are involved in processing com-
plex-type oligosaccharides in mammalian cells, and
(
1!2)- and (1!3)-linkages, EC 3.2.1.113 to those acting
3
,4
mannans in yeasts. In vitro, mannosidases are useful
tools for enzymatic analysis of high-mannose oligosac-
Abbreviations: AEC, anion-exchange chromatography; BCA, bicinch-
onic acid; BSA, bovine serum albumin; CHR, chromatofocussing
chromatography; CV, column volumes; EDTA, ethylenediamine tetra
acetic acid; pNPM, p-nitrophenyl-a-manno-D-pyranoside; RSD, rela-
tive standard deviation; SDS, sodium dodecyl sulfate; SEC, size-
exclusion chromatography; U, unit.
5
–7
charide structures,
8–13
sis.
and for oligosaccharide synthe-
The production of a novel 1,6-a-D-mannosidase from
Aspergillus phoenicis ATCC 14332 (syn. Aspergillus
*
Corresponding author. Tel.: +44 0118 378 6726; fax: +44 0118 931
080; e-mail: r.a.rastall@reading.ac.uk
1
saitoi) has already been reported and the enzyme has
4
0
0
008-6215/$ - see front matter ꢁ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2005.01.005

610
V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
been employed in regioselective synthesis of manno-oli-
1
3
gosaccharides. It has not yet been purified and charac-
terised. During the course of this work the effect of
fermentation time on enzyme production was studied
and the enzyme was purified and characterised. Unex-
pectedly, when A. phoenicis was grown on the commer-
ꢂ
cial yeast mannan hydrolysate Bio-Mos for 7 days in
an attempt to maximise 1,6-a-D-mannosidase produc-
tion, hydrolytic activity towards (1!3)-linked mannobi-
ose was detected. It has been reported in the literature
that, when grown on wheat bran, A. phoenicis produces
1
5
a 1,2-a-D-mannosidase and a 1,3(1,6–1,2)-a-D-man-
nosidase, which has high affinity for (1!3)-linkages
1
6
Figure 1. Anion-exchange chromatography of the ultrafiltration con-
centrated culture extract on a 4.6 mm · 100 mm column packed with
POROS 20 HQ media. The column was eluted with a sodium chloride
gradient 0–1 M. Fractions with enzyme activity are indicated by bars.
but also hydrolyses (1!6)- and (1!2)-linkages. Since
there was no 1,2-a-D-mannosidase activity detected
when A. phoenicis was grown on Bio-Mos , and previ-
ꢂ
ously reported mannosidases from A. phoenicis did not
possess (1!3)- or (1!3)/(1!6)-specific activities, we
thought it to be desirable to purify and partially cha-
racterise this enzyme too.
2. Results
ꢂ
When grown on Bio-Mos , the microorganism secreted
a 1,6-a-D-mannosidase into the culture medium. The en-
zyme activity was highest after 7 days fermentation,
although production of biomass reached a plateau on
day 5 (data not presented). When the medium contained
ꢂ
mannose, glucose or sucrose instead of Bio-Mos the
extract contained no detectable mannosidase activity,
but the enzyme was produced in media lacking any sin-
Figure 2. Chromatofocussing chromatography of the 1,6-a-D-man-
nosidase fraction from POROS 20 HQ column, onto a Pharmacia
Biotech Mono P H/R 5/5 column. The column was eluted with
Polybuffer 74, setting a pH gradient of 7.1–4.0. Fractions with enzyme
activity are indicated by the bar.
ꢂ
gle component other than Bio-Mos .
To establish that production of the enzyme was in-
ꢂ
duced by Bio-Mos , the same strain of Aspergillus was
also grown on wheat bran following the method pub-
1
5
lished by Ichishima et al. The microorganism grew as
expected and produced 1,2-a-D-mannosidase activity,
as well as activity towards p-nitrophenyl-a-D-mannopyr-
anoside and bakerÕs yeast mannan. After 7 days ferment-
ation time, which resulted in maximum 1,6-a-D-
mannosidase production, a second a-D-mannosidase
activity was detected towards a-D-Manp-(1!3)-D-
Manp. The extract did not exhibit any hydrolytic activ-
ity towards a-D-Manp-(1!2)-D-Manp.
The elution profiles of 1,6-a-D-mannosidase in each
purification step are shown in Figures 1–3; the purifica-
tion steps used are summarised in Table 1. In the first
step (Fig. 1) fractions were assayed for (1!6)- and
(
1!3)-activities. In subsequent steps (Figs. 2 and 3)
fractions were assayed for (1!6)-activity only. The
apparent molecular weight of the 1,6-a-D-mannosidase
was 74 kDa by SDS-PAGE (Fig. 4) and 81 kDa by
native-PAGE (data not shown). The isoelectric point
was 4.6 (data not shown). The enzyme had hydrolytic
activity towards (1!6)-linked mannobiose only (Fig. 5).
Figure 3. Size-exclusion chromatography of the 1,6-a-D-mannosidase
fraction from Mono P H/R 5/5 column, onto a Pharmacia Biotech
Superdex 200 H/R 10/30 column (fractionation range 10–600 kDa
MW). The column was eluted with 0.01 M sodium acetate–0.15 M
NaCI buffer, pH 4.5. The bar indicates fractions with enzyme activity
that were pooled.

V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
611
Table 1. Purification of 1,6-a-D-mannosidase from A. phoenicis based on assay with a-D-Manp-(1!6)-D-Manp as substrate
Purification step
Volume (mL)
Total protein (mg)
Total activity (U)
Yield (%)
Specific activity (U/mg)
Fold purification
Culture filtrate
a
UF concentrate
1200
40
6
855
368
14.4
1.1
108
108
51
—
0.13
0.3
—
100
47
2
b
AEC
CHR
3.5
6.4
27
49
c
2
1.2
7
4.8
6.5
4.4
d
SEC
0.10
48
370
a
b
c
UF: ultrafiltration.
AEC: anion-exchange chromatography.
Chromatofocussing.
Size-exclusion chromatography.
d
Figure 5. TLC plates showing the hydrolytic activity of the 1,6-a-D-
mannosidase towards differently linked disaccharides, monitored over
2
4 h incubation time at 30 ꢀC, pH 4.5, using two ascents of butanol–
ethanol–acetic acid–water 9:6:3:1 v/v/v/v. The plates were dipped in
mL of 0.1 M Ce(SO diluted in 100 mL of 15% v/v H SO and the
sugar spots visualised at 100 ꢀC for 15 min. M(1!6)M: a-D-Manp-
1!6)-D-Manp etc., S: mannose 2% w/v standard.
5
4
)
2
2
4
(
Figure 4. Silver-stained SDS-PAGE of the purification steps of the
,6-a-D-mannosidase. Lane numbers correspond to: 1: concentrated
1
extract, 2: extract after anion-exchange chromatography, 3: extract
after chromatofocussing, 4: extract after size-exclusion chromatogra-
phy, M: molecular weight standard markers, Sigma wide-range. On the
right: molecular weights (kDa) of the standards.
It did not hydrolyse the other two mannobioses, pNPM
or bakerÕs yeast mannan, even after prolonged incuba-
tion. It was free from b-galactosidase, b-mannosidase
and b-N-acetylglucosaminidase activities.
The effect of temperature on activity and stability of
,6-a-D-mannosidase is presented in Figures 6 and 7.
The temperature optimum was 60 ꢀC and more than
Figure 6. Effect of temperature on activity of 1,6-a-D-mannosidase.
The enzyme activity towards or a-D-Manp-(1!6)-D-Manp was
measured at different temperatures (20–70 ꢀC). Values are means
of duplicate determinations.
1
9
3
(
5% of activity retained after storage at 4 ꢀC for
months or at ambient temperature (ꢀ23 ꢀC) for 7 days
data not presented). It was also stable at 30 ꢀC for 24 h
The optimum pH of 1,6-a-D-mannosidase was 4.0–4.5
(Fig. 8). Although activity fell sharply at pH values
greater than 5.0, the enzyme retained ꢀ36% of its maxi-
mum activity at pH 3.0. The 1,6-a-D-mannosidase was
but was inactivated rapidly above that temperature.
Freeze and thaw cycles and freeze drying had a detri-
mental effect with ꢀ60% activity lost (data not
presented).
2
+
strongly inhibited by 1 mM Mn and partially inhibited

612
V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
Figure 7. Thermal stability of 1,6-a-D-mannosidase. Aliquots of the
enzyme were incubated for 24 h at different temperatures and the
residual activity determined at 30 ꢀC as described in Section 4.4. Values
are means of duplicate determinations.
Figure 9. Chromatofocussing chromatography of the 1,3(1,6)-a-D-
mannosidase fraction from POROS 20 HQ column, onto a Pharmacia
Biotech Mono P H/R 5/5 column. The column was eluted with
Polybuffer 74, setting a pH gradient of 7.1–4.0. Fractions with enzyme
activity towards a-D-Manp-(1!3)-D-Manp are indicated by the bar.
Table 2. Effect of ethylenediamine tetra acetic acid (EDTA) and
various metal ions in the form of chloride salts on the activity of the
1
,6-a-D-mannosidase
Metal ion (1 mM)
Activity (% of blank)
Blank
EDTA (10 mM)
100
107
100
100
98
2+
Zn
Cu
2
+
2+
Ca
3
+
Fe
Co
97
96
2+
+
Li
Fe
96
94
2
+
Figure 8. Effect of pH on the activity of the 1,6-a-D-mannosidase. The
2
+
enzyme assay was performed at 30 ꢀC in 100 mM of: CH
HCl/CH COONa m, CH COOH/CH COONa and
KH PO
3
COOH +
HPO
Mg
Mn
60
18
2+
3
3
3
d
K
2
4
/
2
4
j. Values are means of duplicate determinations.
2
+
b y Mg but EDTA did not affect activity of the enzyme
even at a concentration of 10 mM (Table 2). The 1,6-a-
D-mannosidase had a K , calculated from a Michaelis–
m
Menten plot by nonlinear regression, of 14 mM with
a-D-Manp-(1!6)-D-Manp as substrate.
The 1,3(1,6)-a-D-mannosidase enzyme was also puri-
fied and characterised. The elution profiles in each puri-
fication step are presented in Figures 1, 9 and 10, and
the steps of purification are summarised in Table 3. In
the first step (Fig. 1) fractions were assayed for (1!3)-
and (1!6)-activity. In subsequent steps (Figs. 9 and
1
0) fractions were assayed for (1!3)-activity only. The
apparent molecular weight of the 1,3(1,6)-a-D-mannosi-
dase was 97 kDa by SDS-PAGE (Fig. 11) and 110 kDa
by native-PAGE (data not shown). The enzyme appears
to exist as two charge isomers (isoforms) as evidenced by
the appearance of two bands on an isoelectric focussing
gel with pI values of 4.3 and 4.8 (data not shown). The
enzyme hydrolysed (1!3)- and (1!6)-linked mannobi-
oses only, although at a different rate (Fig. 12). It also
hydrolysed pNPM but at a very slow rate (0.7 U/mg of
the purified enzyme), and bakerÕs yeast mannan only
Figure 10. Size-exclusion chromatography of the 1,3(1,6)-a-D-man-
nosidase fraction from Mono P H/R 5/5 column, onto a Pharmacia
Biotech Superdex 200 H/R 10/30 column (fractionation range 10–
6
0
00 kDa MW). The column was eluted with 0.01 M sodium acetate–
.15 M NaCI buffer, pH 4.5. Fractions with enzyme activity towards
a-D-Manp-(1!3)-D-Manp are indicated by the bar.
after prolonged incubation (the activity was not quanti-
fied). It was free from b-galactosidase, b-mannosidase
and b-N-acetylglucosaminidase activities.

V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
613
Table 3. Purification of 1,3(1,6)-a-D-mannosidase from A. phoenicis based on assay with a-D-Manp-(1!3)-D-Manp as substrate
Purification step
Volume (mL)
Total protein (mg)
Total activity (U)
Yield (%)
Specific activity (U/mg)
Fold purification
Culture filtrate
UF conc.
AEC
1200
40
6
855
368
2.9
—
—
Very low
0.011
1.1
—
—
4.1
3.3
1.9
1.3
100
81
46
32
100
545
CHR
SEC
2
1.2
0.31
0.07
6.0
19.7
1770
A single determination was carried out for each purification step.
Time-course assays of the two activities of 1,3(1,6)-a-
D-mannosidase, in the absence as well as in the presence
2
+
of calcium ion showed that Ca stimulated (1!3)-
activity by 70% but (1!6)-activity was enhanced by less
than 4%, a very low value which is within the RSD of
2
+
the assay. The effects of EDTA and Ca on each activ-
ity are presented in Table 4. (1!3)-Activity is stimulated
2
+
b y Ca and inhibited by EDTA, while 1,6 activity is
unaffected. The rate of hydrolysis of a-D-Manp-(1!3)-
D-Manp was more than 11 times higher than that of a-
2
+
D-Manp-(1!6)-D-Manp when 1 mM Ca was present.
3. Discussion
Various carbon sources supported growth but none of
them could induce production of 1,6-a-D-mannosidase
except Bio-Mos . The result was the same when the min-
ꢂ
eral solution composition was altered. Additionally, the
same strain of Aspergillus produced 1,2-a-D-mannosi-
dase when grown on wheat bran according to the method
1
5
of Ichishima et al. These results indicate that we are not
dealing with a different strain of microorganism and that
Bio-Mos induces production of the enzymes.
A fraction of 1,6-a-D-mannosidase free of (1!3)-
activity could be obtained by anion-exchange chroma-
tography (AEC). However, (1!3)-activity co-eluted
Figure 11. Silver-stained SDS-PAGE of the purification steps of the
ꢂ
1,3(1,6)-a-D-mannosidase. Lane numbers correspond to: 1: anion-
exchange chromatography, 2: chromatofocussing, 3: size-exclusion
chromatography, M: molecular weight standard markers, Sigma wide-
range. On the right: molecular weights (kDa) of the standards.
Figure 12. TLC plates showing the hydrolytic activity of the 1,3(1,6)-a-D-mannosidase towards differently linked disaccharides, monitored over 24 h
incubation time at 30 ꢀC, pH 5.0, using two ascents of butanol–ethanol–acetic acid–water 9:6:3:1 v/v/v/v. The plates were dipped in 5 mL of 0.1 M
Ce(SO
4
)
S: mannose 2% w/v standard.
2
diluted in 100 mL of 15% v/v H
2
SO
4
and the sugar spots visualised at 100 ꢀC for 15 min. M(1!6)M: a-D-Manp-(1!6)-D-Manp etc.,

614
V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
Table 4. Effect of ethylenediamine tetra acetic acid (EDTA) and
calcium chloride on the activity of the 1,3(1,6)-a-D-mannosidase
towards a-D-Manp-(1!3)-D-Manp and a-D-Manp-(1!6)-D-Manp
(1!3)-linkage compared to (1!6)-linkage, but it is re-
ported to cleave (1!2)-linkages too, and is completely
inactive towards p-nitrophenyl-a-D-mannopyranoside.
It is strongly (13 times) activated by addition of 1 mM
Reagent and
concentration
Activity on
Activity on
a-D-Manp-(1!3)-D-Manp a-D-Manp-(1!6)-
2+
Ca . Whether calcium has the same effect on the other
(% of blank)
D-Manp (% of blank)
two activities has not been investigated. Since the above
enzyme has been purified from an extract of a culture
that also produces 1,2-a-D-mannosidase, it can be sur-
mised that the low activity towards (1!2)-linkages
was contaminating activity and the enzyme is actually
the same as the one we report here. Unfortunately, no
other detail is available to compare with our findings.
Regardless, we did not observe as strong activation by
Ca as reported in the above publication. Additionally,
the 1,3(1,6)-a-D-mannosidase reported here, cleaves
p-nitrophenyl-a-D-mannopyranoside. It is therefore
concluded that this is a different enzyme.
Blank
100
170
77
100
103
110
2+
Ca 1 mM
EDTA 10 mM
with a part of the (1!6)-activity in this and every sub-
sequent purification step. It is not clear whether the
two activities are properties of the same polypeptide or
are present on two distinct polypeptides.
2
+
The 1,6-a-D-mannosidase, purified 370-fold over the
culture crude extract, was obtained in 4.4% yield with
a specific activity of 48 U/mg of protein. The final yield
is low, partly due to a significant yield sacrifice in the
AEC step (in order to avoid (1!3)-activity contamina-
tion), and partly to low recovery in the chromatofocuss-
ing step. The apparent molecular weight of the 1,6-a-D-
mannosidase under denaturing and nondenaturing con-
ditions was 74 and 81 kDa, respectively, indicating that
the enzyme is a monomer. The isoelectric point deter-
mined by IEF, was 4.6 which is in agreement with the
elution position of the enzyme in the chromatofocussing
purification step.
The 1,3(1,6)-a-D-mannosidase, purified 1770-fold over
the culture crude extract, was obtained in a reasonably
high yield (32%) with a specific activity of 20 U/mg of
protein. The 1,3(1,6)-a-D-mannosidase appears to be a
monomer with molecular weight 97 kDa by SDS-PAGE
and 110 kDa by native-PAGE. Two distinct bands ap-
peared in the IEF at 4.3 and 4.8 indicating the presence
of charge isomers or isoforms. This is consistent with
two peaks of (1!3)-activity being seen in the chroma-
tofocussing step.
The pH optimum of the 1,6-a-D-mannosidase is 4.0–
4.5, which is quite low, like the Aspergillus niger
1
7,18
1
the Vigna umbellata and the watermelon manno-
9
20
sidases. All other mannosidases have optimum pH
7
,15,16,20–23
Like some mannosidases
this enzyme retained most of its
between 5.0 and 7.0.
1
6,19
reported earlier,
activity even at a pH of 3.5.
The thermal stability of 1,6-a-D-mannosidase was
studied after 24 h incubation. Most published reports
determine thermal stability over a very short incubation
7
,15,21
time, usually 10 min,
but a longer duration stabi-
lity study gives more meaningful results, especially useful
when lengthy concentration/purification steps have to be
employed (ultrafiltration, gel filtration), or in cases
where the enzymes have to be transported. The 1,6-a-
D-mannosidase was found to be very stable at ambient
temperature. It also exhibited a high temperature opti-
mum, ꢀ60 ꢀC. Other microbial mannosidases have
1
5,16
20,21
optima ranging from 30 ꢀC
to 50–55 ꢀC
but a
1
9
plant mannosidase has an optimum of 60 ꢀC.
2+
A. phoenicis is reported to produce a specific 1,2-a-D-
16
mannosidase and a nonspecific a-D-mannosidase.
The 1,6-a-D-mannosidase has not been reported previ-
ously from this species although it is the second enzyme
with such specificity to de described, the first produced
The 1,6-a-D-mannosidase did not need Ca or any
other metal ion for activity. The 1,6-a-D-mannosidase
from X. manihotis is also Ca independent and so
1
5
2
+
2
appear to be all linkage specific 1,2-a-D-man-
1
7,21,23,24
nosidases.
In a very recent study it was found
2
by Xanthomonas manihotis.
that A. phoenicis 1,2-a-mannosidase has a calcium-bind-
ing consensus sequence and binds Ca (1 mol/mol of
2
+
The second enzyme reported here cleaves only (1!3)-
and (1!6)-mannobiose linkages. The rate of (1!3)-
cleavage is 11 times higher than the rate of (1!6)-cleav-
age. It is activated by calcium (increase 2-fold compared
to EDTA treated sample), but this is observed only with
2
5
enzyme). Calcium, when bound, increases thermal sta-
bility of the enzyme but is not required for enzymatic
activity. The only exception reported is the 1,2-a-D-man-
nosidase from Trichoderma reesei, which is inactivated
2
+
(
1!3)-activity. (1!6)-Activity was not affected by
by EDTA and its activity restored by Ca , although
6
2
either reagent. This may possibly indicate that the two
activities are associated with different polypeptide
chains. It is also possible that a single polypeptide may
have two distinct active sites, the conformation and/or
activity of one being influenced by the presence of cal-
cium ions. A similar enzyme from A. phoenicis, reported
no calcium-binding consensus sequence was identified.
2
+
The 1,3(1,6)-a-D-mannosidase was activated by Ca ,
2
,16,22,27
18,20
like most,
mannosidases.
but not all
nonlinkage specific
The 1,3(1,6)-a-D-mannosidase belongs to the EC
3.2.1.114 mannosidases. A new EC number is needed
for the 1,6-a-D-mannosidase.
1
6
by Amano and Kobata, has equally high affinity for

V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
615
4
. Experimental
100 lL substrate (5.5 mM pNPM in 0.01 M sodium ace-
tate buffer pH 5.0) was incubated with 20 lL of the en-
zyme solution at 30 ꢀC for 2 h. The enzyme reaction was
stopped by the addition of 1.1 mL of 0.1 M sodium car-
bonate, and the absorbance was measured at 420 nm
against the blank. The concentration of p-nitrophenol
released was determined from a standard curve. One
unit of enzyme is defined as the amount of enzyme
which releases 1 lmol p-nitrophenol per minute at
30 ꢀC. Other glycosidase activities (b-galactosidase,
b-mannosidase and b-N-acetylglucosaminidase) were
determined likewise on the corresponding p-nitro-
phenyl-glycosides.
The activity on bakerÕs yeast mannan was determined
by the Nelson–Somogyi reducing sugar assay by incu-
bating a volume of a 2% w/v yeast mannan solution in
0.01 M sodium acetate buffer pH 5.0 with an equal vol-
ume of the purified enzyme preparations at 30 ꢀC for
2 h. The presence or absence of enzyme activity on this
substrate was merely checked qualitatively.
4
.1. Materials
Mannose was purchased from Sigma (Poole, Dorset,
UK). Mannose containing disaccharides linked a-D-
Manp-(1!2)-D-Manp, a-D-Manp-(1!3)-D-Manp and
a-D-Manp-(1!6)-D-Manp were purchased from Dextra
Laboratories Ltd., Reading, UK. a-D-Manp-(1!6)-D-
Manp in large quantities was prepared according to
1
3
the procedure described by Athanasopoulos et al.
ꢂ
Bio-Mos was a gift from Alltech (UK) Ltd., Stamford,
UK. All other chemicals used were purchased from
Sigma or BDH-Merck. The TLC plates used were
1
0 cm Silica Gel 60 by Merck, Darmstadt, Germany.
4
.2. HPLC
Mannose released from the enzyme action on disaccha-
rides was quantified by HPLC as described by Athana-
1
3
sopoulos et al.
.3. Protein assay
For determination of specific activities, protein content
4.5. Fermentation, extraction and concentration of
mannosidase
4
The procedure employed to produce the enzymes was
essentially the same as the one described by Athana-
2
8
was determined by the BCA assay using bovine serum
albumin (BSA) as a standard.
1
3
sopoulos et al. except that the fermentation time
was extended to 7 days (instead of three), and the
concentration by ultrafiltration was 24-fold (instead
of 10-fold).
4
.4. Enzyme activity assays
A qualitative enzyme detection in the fractions collected
after the chromatographic separations was performed as
follows. A volume (2 lL) of each fraction was incubated
with 2 lL of 2% (w/w) a-D-Manp-(1!6)-D-Manp or
a-D-Manp-(1!3)-D-Manp, at pH 4.5, 30 ꢀC for 2–24 h
4.6. Purification of enzymes
The following chromatographic separations were
carried out using a PE Biosystems BioCAD/SPRINT
Perfusion Chromatography System with a semi-
(
depending on the expected enzyme dilution after each
chromatographic technique), and the presence of mono-
saccharide was detected on Silica Gel 60 TLC plates
as described by Athanasopoulos et al.
Enzyme activity was quantified as follows. An appro-
priate enzyme dilution was incubated with 1% (w/w)
a-D-Manp-(1!6)-D-Manp, total volume 240 lL, at
pH 4.5, 30 ꢀC for 20 min. The reaction was stopped by
immersion in boiling water for 2 min followed by cool-
ing in ice bath. The mannose released was determined
by HPLC. The progress of the reactions was always
preparative flow cell (PE Biosystems, Birchwood Science
Park North, Warrington, Cheshire, UK). All separation
procedures were carried out at ambient temperature
(ꢀ23 ꢀC) and the pooled fractions were concentrated
and stored at 4 ꢀC prior to each subsequent purification
step. Before pooling, the fractions were assayed for
enzyme activity qualitatively, using mannobioses, as
described above.
Anion-exchange chromatography (AEC) was carried
out using 4.6 mm · 100 mm (1.66 mL) column packed
with POROS 20 HQ media (surface functional group
was quaternised polyethyleneimine, 10 lm particle size).
The buffers used were, A: 0.025 M bis-Tris (pH 6.3 with
HCl) and B: 0.025 M bis-Tris–1 M NaCI (pH 6.3 with
HCl). A 5 mL volume of the sample was injected into
the column, which was eluted for 5 column volumes
(CV) with buffer A followed by a linear gradient over
40 CV to 40% buffer B, followed by a linear gradient
over 20 CV to 100% buffer B. The flow rate was 4 mL/
min and the column effluent was monitored at 280 nm.
1
3
<15% substrate conversion. One unit is defined as the
amount of enzyme that releases 2 lmol of mannose
per minute at 30 ꢀC. Determinations were carried out
in triplicate and relative standard deviations (RSD) were
calculated. A similar procedure was followed using a-D-
Manp-(1!3)-D-Manp but the incubation time was
6
0 min at pH 5.0 and single determinations were carried
out.
The activity of a-mannosidase on p-nitrophenyl-a-D-
mannopyranoside (pNPM) was determined as follows:

616
V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
Chromatofocussing chromatography (CHR) was car-
The isoelectric point (pI) was determined by agarose
ꢂ
ried out using a Pharmacia Biotech Mono P H/R 5/5
column. One millilitre of the sample was injected into
the column. A 5 min elution with 0.025 M bis-Tris buffer
pH 7.1 with iminodiacetic acid) was followed by a
0 min elution with Polybuffer 74 (diluted 1/10, pH 4.0
with iminodiacetic acid), setting a pH gradient of 7.1–
.0 for 20 min. The flow rate was 1 mL/min and the col-
isoelectric focussing (IEF) on IsoGel plates, pH 3–10,
(Cambrex Bioscience, Berkshire, UK) according to the
manufacturerÕs instructions, using IEF-Mix 3.6–9.3
markers from Sigma.
(
2
4.9. Effects of metal ions on activity
4
umn effluentÕs pH and absorbance at 280 nm was
monitored.
The effect of EDTA and various metal ions in the form
of chloride salts on the activity of the 1,6-a-D-mannosi-
dase was studied by incubating the enzyme with 1 mM
of each metal ion (10 mM for EDTA) and estimating
the activity as described in Materials and methods.
The results were compared with a reagent-blank enzyme
assay.
Size-exclusion chromatography (SEC) was carried out
using a Pharmacia Biotech Superdex 200 H/R 10/30 col-
umn (fractionation range 10–600 kDa MW). The col-
umn was eluted with 0.01 M sodium acetate–0.15 M
NaCI buffer (pH 4.5 with acetic acid). A 0.3 mL sample
volume was injected into the column. The flow rate was
set at 0.3 mL/min and the effluent monitored at 280 nm.
The sample was eluted with the above sample buffer
for 80 min.
4.10. Temperature optimum and thermal stability
The 1,6-a-D-mannosidase activity towards a-D-Manp-
(
1!6)-D-Manp was measured at different temperatures
4
.7. Characterisation of hydrolytic specificity
(20–70 ꢀC in 0.01 M sodium acetate buffer, pH 4.5) to
determine the temperature optimum. In order to
determine the thermal stability of 1,6-a-D-mannosidase,
aliquots of the enzyme were incubated for 24 h
at different temperatures and the residual activity
determined at 30 ꢀC as described in Section 4.4.
The hydrolytic specificity was determined by incubating
lL of the purified enzyme preparation with 5 lL of 2%
5
w/w of a-D-Manp-(1!2)-D-Manp, a-D-Manp-(1!3)-D-
Manp or a-D-Manp-(1!6)-D-Manp, at pH 4.5 in
0
.01 M sodium acetate buffer at 30 ꢀC for up to 24 h.
Samples (1 lL) were removed at regular time intervals
and applied to TLC plates to monitor the hydrolysis
of the differently linked disaccharides. The TLC plates
4.11. pH Optimum
The effect of pH on the activity of the 1,6-a-D-manno-
sidase was determined by assaying the enzyme activity
in the pH range: 2.5–7.0, in 100 mM buffers
(CH COOH + HCl/CH COONa, CH COOH/CH COO-
1
3
were run and developed as described by.
4
.8. Estimation of molecular weights and isoelectric points
3
3
3
3
Na and K HPO /KH PO ) at 30 ꢀC.
2
4
2
4
The purity and apparent molecular weight of the puri-
fied enzymes were estimated by a discontinuous SDS-
4.12. Kinetic studies
2
9
PAGE system on a 10 · 7 cm slabgel (3.5% stacking
gel and 10% resolving gel), using molecular weight
markers (Sigma wide-range Marker Kit, Sigma, Dorset,
UK). Proteins in the gel were silver stained using the
method of Chrambach. The gel images were acquired
and analysed using the SynGene system of Synoptics
Ltd. and the Gene-Tools software.
The native molecular weight of the purified
enzymes was estimated by native-PAGE. The standards
used were: bovine serum albumin (BSA) monomer
The 1,6-a-D-mannosidase was assayed in 0.1 M sodium
acetate buffer, pH 4.5, at 30 ꢀC with a range of concen-
trations of a-D-Manp-(1!6)-D-Manp (1, 4, 8, 10, 20, 40
and 50 mM) in a final volume of 240 lL for 10 min with
appropriate blanks. The mannose released was deter-
mined by HPLC and the Michaelis–Menten kinetic
parameters were estimated by nonlinear regression using
BioChem LabAssistant software.
3
1
(
66 kDa), BSA dimer (128 kDa) and Jack bean urease
4.13. 1,3(1,6)-a-D-Mannosidase: ratio of activities and
effect of Ca and EDTA
2þ
trimer (272 kDa), all contained in Sigma nondenatured
protein molecular weight marker kit. The standards
and the purified enzymes were electrophoresed on four
different separating gel acrylamide concentrations
The ratio of the two activities was determined by assay-
ing the purified enzyme preparation with the two sub-
strates using a time-course assay. The mannose
concentrations in two reaction mixtures were deter-
mined and the values were plotted against the incub-
ation time to indicate the susceptibility of each linkage
to the enzyme action in the absence as well as in the
(
7%, 8%, 9% and 10%), with a stacking gel acrylamide
concentration of 3.5% and Bromophenol Blue as a
tracking dye. The molecular weights of the two
enzymes determined according to Hedrick and co-
3
0,31
workers.

V. I. Athanasopoulos et al. / Carbohydrate Research 340 (2005) 609–617
617
presence of calcium ion. The fixed time assay was also
used but with longer incubation times, to determine
the effect of EDTA and calcium chloride on the two
activities, and the results were expressed as percentages
of the blanks.
12. Maitin, V.; Athanasopoulos, V.; Rastall, R. A. Appl.
Microbiol. Biotechnol. 2004, 63, 666–671.
1
1
1
3. Athanasopoulos, V. I.; Niranjan, K.; Rastall, R. A.
J. Mol. Catal. B, Enzym. 2004, 27, 215–219.
4. Raper, K. B.; Fennel, D. I. The Genus Aspergillus; The
Williams and Wilkins Company: Baltimore, USA, 1965.
5. Ichishima, E.; Arai, M.; Shigematsu, Y.; Kumagai, H.;
Sumida-Tanaka, R. Biochim. Biophys. Acta 1981, 658, 45–
5
3.
Acknowledgements
1
6. Amano, J.; Kobata, A. J. Biochem. 1986, 99, 1645–1654.
7. Swaminathan, N.; Matta, K. L.; Donoso, L. A.; Bahl, O.
P. J. Biol. Chem. 1972, 247, 1775–1779.
1
V. I. Athanasopoulos would like to thank The State
Scholarships Foundation of Greece (IKY) for financial
support.
1
1
2
8. Matta, K. L.; Bahl, O. P. J. Biol. Chem. 1972, 247, 1780–
1787.
9. Wongvithoonyaporn, P.; Bucke, C.; Svasti, J. Biosci.
Biotechnol. Biochem. 1998, 62, 613–621.
0. Gaikwad, S. M.; Keskar, S. S.; Khan, M. I. Biochim.
Biophys. Acta 1995, 1250, 144–148.
References
1
. NC-IUBMB. In Enzyme Nomenclature: Recommendations
of the Nomenclature Committee of the International Union
of Biochemistry and Molecular Biology; Webb, E. C., Ed.;
Academic: San Diego, CA, 1992.
21. Yoshida, T.; Inoue, T.; Ichishima, E. Biochem. J. 1993,
290, 349–354.
22. Jones, G. H.; Ballou, C. E. J. Biol. Chem. 1969, 244, 1043–
1051.
2
3
4
. Wong-Madden, S. T.; Landry, D. Glycobiology 1995, 5,
23. Yamamoto, K.; Hitomi, L.; Kobatake, K.; Yamaguchi, H.
J. Biochem. 1982, 91, 1971–1979.
24. Ichikawa, Y.; Look, G. C.; Wong, C. H. Anal. Biochem.
1992, 202, 215–238.
25. Tatara, Y.; Lee, B. R.; Yoshida, T.; Takahashi, K.;
Ichishima, E. J. Biol. Chem. 2003, 278, 25289–25294.
26. Maras, M.; Callewaert, N.; Piens, K.; Claeyssens, M.;
Martinet, W.; Dewaele, S.; Contreras, H.; Dewerte, I.;
Penttila, M.; Contreras, R. J. Biotechnol. 2000, 77, 255–
263.
27. Takegawa, K.; Miki, S.; Jikibara, T.; Iwahara, S. Biochim.
Biophys. Acta 1989, 991, 431–437.
28. Smith, P. K.; Krohn, G. T.; Hermanson, A. K.; Mallia,
K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.;
Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem.
1985, 150, 76–85.
29. Laemmli, U. K. Nature 1970, 227, 680–685.
30. Hedrick, J. L.; Smith, A. J. Arch. Biochem. Biophys. 1968,
126, 155–164.
1
9–28.
. Kornfeld, R.; Kornfeld, S. Annu. Rev. Biochem. 1985, 54,
31–664.
6
. Moremen, K. W.; Trimble, R. K.; Herscovics, A. Glyco-
biology 1985, 4, 113–125.
5
6
. Kobata, A. Anal. Biochem. 1979, 100, 1–14.
. Jacob, G. S.; Scudder, P. Methods Enzymol. 1994, 230,
2
. Maruyama, Y.; NakaJima, T.; Ichishima, E. Carbohydr.
Res. 1994, 251, 89–98.
. Johansson, E.; Hedbys, L.; Mosbach, K.; Larsson, P.
Enzyme Microb. Technol. 1989, 11, 347–352.
80–299.
7
8
9
. Ajisaka, K.; Matsuo, I.; Isomura, M.; Fujimoto, H.;
Shirakabe, M.; Okawa, M. Carbohydr. Res. 1995, 270,
1
23–130.
0. Suwasono, S.; Rastall, R. A. Biotechnol. Lett. 1996, 18,
51–856.
1. Singh, S.; Scigelova, M.; Crout, D. H. G. Tetrahedron:
Asymmetry 2000, 11, 223–229.
1
1
8
31. Chrambach, A.; Rodbard, D. Science 1971, 172, 440–451.