Suitability and limitations of methods for characterisation of activity of malto-oligosaccharide-forming amylases

Article abstract of DOI:10.1016/S0008-6215(00)00151-8

The suitability and limitations of essential methods and reference substrates used for characterisation of activity of amylolytic enzymes is investigated. Saccharogenic, chromogenic and chromatographic methods are included. The results are discussed in relation to the measurement of reaction rates, determination of action mode and product specificity and the impact on identification and nomenclature of malto-oligosaccharide-forming amylases. An accurate determination of reaction rates using the saccharogenic methods strongly depends on the degree of polymerisation (DP) of the standards used and the hydrolysis products formed by the amylase. Particularly the use of glucose as standard can lead to overestimates due to the differences in the reducing potential of glucose and malto-oligosaccharides. The reliability of the chromogenic methods for determination of action mode depends on the DP of the substrate and the specificity of the amylase. For a characterisation of the starch hydrolysis products and the variation in the DP during hydrolysis, high performance anion-exchange chromatography with pulsed amperometric detection provided a fast and reliable method. A literature survey revealed varying and inconsistent use of nomenclature of malto-oligosaccharide forming amylases. Therefore a systematic approach identifying three main classes of activity is suggested using not only the mode of action and the DP of the major product but also the stage of hydrolysis at which this product is formed. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00151-8

www.elsevier.nl/locate/carres
Carbohydrate Research 329 (2000) 109–119
Suitability and limitations of methods for
characterisation of activity of
malto-oligosaccharide-forming amylases
Lene Duedahl-Olesen, Lars Haastrup Pedersen *, Kim Lambertsen Larsen
Department of Life Science, Aalborg Uni6ersity, Sohngaardsholms6ej 49, DK-9000 Aalborg, Denmark
Received 29 November 1999; accepted 8 May 2000
Abstract
The suitability and limitations of essential methods and reference substrates used for characterisation of activity of
amylolytic enzymes is investigated. Saccharogenic, chromogenic and chromatographic methods are included. The
results are discussed in relation to the measurement of reaction rates, determination of action mode and product
specificity and the impact on identification and nomenclature of malto-oligosaccharide-forming amylases. An
accurate determination of reaction rates using the saccharogenic methods strongly depends on the degree of
polymerisation (DP) of the standards used and the hydrolysis products formed by the amylase. Particularly the use
of glucose as standard can lead to overestimates due to the differences in the reducing potential of glucose and
malto-oligosaccharides. The reliability of the chromogenic methods for determination of action mode depends on the
DP of the substrate and the specificity of the amylase. For a characterisation of the starch hydrolysis products and
the variation in the DP during hydrolysis, high performance anion-exchange chromatography with pulsed ampero-
metric detection provided a fast and reliable method. A literature survey revealed varying and inconsistent use of
nomenclature of malto-oligosaccharide forming amylases. Therefore a systematic approach identifying three main
classes of activity is suggested using not only the mode of action and the DP of the major product but also the stage
of hydrolysis at which this product is formed. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Malto-oligosaccharide-forming amylase; Starch hydrolysis products; Saccharogenic and chromogenic methods; High-
performance anion-exchange chromatography with pulsed amperometric detection
a-
D
-(1“4) O-glucosidic linked residues and
-(1“6) O-glucosidic link-
1. Introduction
branch points at a-
D
ages. The linear amylose molecule contains
one reducing end (one free anomeric carbon
atom) and one non-reducing end, while an
amylopectin molecule has one reducing end
and numerous non-reducing ends due to the
branched structure. In order to access the
chemical energy stored in starch, organisms
need starch-hydrolysing enzymes.
Starch is one of the most abundant polysac-
charides in nature where it serves as energy
storage in plant seeds and tubers [1]. Starch is
a homopolymer of glucose consisting of an
essentially linear fraction termed amylose with
glucose linked by a- -(1“4) O-glucosidic
D
linkages and a highly branched fraction
termed amylopectin with linear segments of
Enzymes capable of hydrolysing starch are
denoted amylolytic enzymes and are classified
*
Corresponding author. Fax: +45-98-141808.
as hydrolases which cleave a- -(1“4) O-
D
E-mail address: i5lhp@civil.auc.dk (L. Haastrup Pedersen).
0
008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00151-8

1
10
L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
glucosidic or a-
D
-(1“6) O-glucosidic link-
Closely related to a-amylases are cyclodextrin
ages. Amylolytic enzymes have been divided
into several subclasses according to their ac-
tivities on starch [2,3]. The distinction has
been made according to the configuration of
the anomeric carbon atom in the products, i.e.
a- or b-amylases or action mode on the sub-
strate, i.e., endo- or exo-activity (Table 1). The
a-amylases (E.C. 3.2.1.1) are endo-acting en-
glycosyltransferases (CGTases, E.C. 2.1.4.19).
The action on starch of CGTases results in the
formation of cyclic malto-oligosaccharides of
six or more glucose units [4,9], while a lower
rate of hydrolysis activity is found
concomitantly.
Several bacterial amylases capable of de-
grading starch to malto-oligosaccharides of a
certain DP have been identified. Traditionally
they are termed product-specific amylases and
include amylases forming either maltotriose,
maltotetraose, maltopentaose, or malto-
hexaose as the major product [2–4,10]. The
pure malto-oligosaccharides have mainly been
used as fine chemicals and for diagnostic kits
[10,11] and due to the difficulties in prepara-
tive separation of malto-oligosaccharides a
considerable interest in amylases forming
malto-oligosaccharides of a specific DP has
evolved. As the product size distribution is
developing during the time of hydrolysis some
confusion related to the properties and the
nomenclature of these amylases has resulted.
The present investigation is a study of the
suitability and limitations of essential methods
and reference substrates used to characterise
amylolytic activity regarding substrate conver-
sion, action mode and product specificity. The
results are discussed in relation to the identifi-
cation and nomenclature of malto-oligosac-
charide-forming amylases.
zymes that randomly attack the internal a-
D
-
(
1“4) O-glucosidic linkages of starch except
for those adjacent to the ends of the substrate
and those in the vicinity of branch points. The
end products are a-limit dextrins, which are
branched saccharides not prone to further hy-
drolysis and malto-oligosaccharides of varying
degrees of polymerisation (DP), characteristic
of the particular enzyme [2,3]. b-Amylases
(
E.C. 3.2.1.2) are exo-acting amylases which
hydrolyse a-
D
-(1“4) O-glucosidic linkages in
starch from the non-reducing end of the sub-
strate, leading to b-maltose and b-limit dex-
trins. a-Glucosidases (E.C. 3.2.1.20) and
glucoamylases (E.C. 3.2.1.3) are also exo-act-
ing amylases producing a-
glucose, respectively [2–4]. Glucoamylases
also hydrolyse a- -(1“6) O-glucosidic link-
D
-glucose and b-
D
-
D
ages, however at a slower rate than the a-
D
-
(
1“4) linkages. Debranching enzymes, which
can hydrolyse the a-
D-(1“6) O-glucosidic
linkages in amylopectin are pullulanases (E.C.
3
.2.1.41) and isoamylases (E.C. 3.2.1.68).
Table 1
Starch hydrolysing enzymes and their mode of action on starch.
Enzyme name
Mode of
action
Glucosidic
linkages
Main product(s) formed
Ref.
attacked
a-Amylase E.C. 3.2.1.1)
Endo
a-
D
-(1“4)
mixture of malto-oligosaccharides with
a-configuration
b-maltose
a-glucose
b-glucose
[2–4]
b-Amylase (E.C. 3.2.1.2)
a-Glucosidase (E.C. 3.2.1.20)
Glucoamylase (E.C. 3.2.1.3)
Exo
Exo
Exo
a-
a-
a-
D
D
D
-(1“4)
-(1“4)
-(1“4)
[2–4]
[2–4]
[2–4]
and
a-
D
-(1“6)
-(1“6)
Isoamylase (E.C. 3.2.1.68)
Endo
a-D
linear malto-oligosaccharides up to 25 glucose
units without branches
[2–4]
Pullulanase (E.C. 3.2.1.41)
Exo-maltotetrao-hydrolase
Endo
Exo
a-
D
D
-(1“6)
-(1“4)
various linear malto-oligosaccharides
maltotetraose
[2–4]
[5,6]
a-
(
E.C. 3.2.1.60)
Exo-maltohexao-hydrolase
E.C. 3.2.1.98)
Exo
a-D
-(1“4)
maltohexaose
[6–8]
(

L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
111
Table 2
Methods for determination of amylolytic activity, chemical reaction and detection method
Method
Substrate
Reaction principle
Detection
Ref.
Saccharogenic
methods
Ferricyanide
native or
soluble starch
native or
soluble starch arsenomolybdat
native or
alkaline ferricyanide(III) reduced to ferrocyanide(II)
spectrophotometric [14,15]
spectrophotometric [16,17]
spectrophotometric [18]
spectrophotometric [19]
Somogyi
alkaline cupri(II) ions reduced to cupro(I) ions, colour with
-
Nelson
DNS
alkaline 3,5-dinitrosalicylic acid reduced to
soluble starch 3-amino-5-nitrosalicylic acid
native or
CuSO4–
alkaline cupri(II) ions reduced to cupro(I) ions, colour with
biquinoline
soluble starch 4,4%-dicarboxy-2,2%-biquinoline
Chromogenic
methods
Cerealpha™
a
BPNPG7
endo-attack followed by glucoamylase/a-glucosidase action
spectrophotometric [20,21]
spectrophotometric [22]
on p-NP part release p-NP
b
Betamyl™
PNPG5
exo-attack and a-glucosidase action on p-NP part releases
p-NP
Dyed-starch
AZCL
release of blue colour when amylose is hydrolysed
visually or
[12]
spectrophotometric
c
–
amylose
Amyloclastic
methods
Iodometric
−
starch or
amylose
amyloseꢀI3 complex (blue) hydrolysis decreases the complex
spectrophotometric [23,24]
and the blue colour
Viscosity
starch in high hydrolysis results in a decrease of the solution viscosity
concentrations
viscosity
measurement
[4]
a
Blocked p-nitrophenolmaltoheptaose.
p-nitrophenolmaltopentaose.
Azo-colour crosslinked amylose.
b
c
2
. Results and discussion
Saccharogenic methods.—It is evident, that
during hydrolysis of a-(1“4) glucans the
number of reducing ends increases, except for
the activity of CGTases, which mainly result
in cyclic products. Kinetic studies of amylase-
catalysed reactions is therefore obtained by
saccharogenic methods where the number (or
molar concentration) of glucosidic linkages
hydrolysed in a certain time is determined
[4,13]. An increase in the amount of reducing
ends released is determined by oxidation of
the anomeric carbon atom at the reducing end
of the hydrolysis product. A corresponding
reduction of a metal ion or a compound im-
parts colour either in the reduced state itself
or as a complex with a ligand (Table 2). The
colourimetric measurement of the formation
of reducing ends has been accomplished at
alkaline conditions primarily by the use of
three different oxidising reagents: copper
Determination of enzyme acti6ity.—The
most commonly used methods for activity de-
termination of amylases can be grouped into
saccharogenic, chromogenic and amyloclastic
methods [12]. Amyloclastic methods measure
the degradation of starch by viscometric, tur-
bidimetric, iodometric or nephelometric pro-
cedures [13]. The saccharogenic methods are a
direct measure of hydrolysis which determine
the number of reducing ends formed [14–19]
and the chromogenic methods, which include
the use of coloured substrates [12,20,21]. The
mechanism of the different assays is sum-
marised in Table 2. These methods do not
qualitatively determine the actual products
formed and therefore care should be taken
when the activity of malto-oligosaccharide-
forming amylases is to be measured.

1
12
L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
[
16,17,19], ferricyanide [14,15] and 3,5-dini-
reducing power detected by the two assays.
An overestimation of the amylase activity
would occur when glucose instead of maltose
was used as reference. Particularly for the
DNS method the use of maltotriose as a refer-
ence resulted in a curve different from the
trosalicylic acid [18,25].
For quantification of the number of reduc-
ing ends formed the most common reference
used is maltose, particularly if malto-oligosac-
charides or dextrins are the products formed
maltose standard. For the CuSO /biquinoline
[14]. The standard curves obtained with glu-
4
assay a curve similar to the one for maltose
was obtained with maltotriose, maltotetraose
and maltoheptaose. Investigations of the Som-
ogyi–Nelson method [16,17] showed that glu-
cose and maltose as references did not give
rise to different curves (data not shown).
These results were confirmed by the findings
that the copper procedure [16] gives equimolar
reducing equivalents for equimolar reducing
ends of malto-oligosaccharides.
Studies of the ferricyanide method also re-
vealed a higher reducing potential of glucose
than malto-oligosaccharides [14]. The method
was modified by Hizukuri et al. [15] to obtain
the same reducing effect of glucose and malto-
oligosaccharides up to DP 40. The dinitrosali-
cylic acid reagent gives a decreasing reducing
potential with an increase in the number of
cose, maltose and maltotriose by the use of
the CuSO /biquinoline assay [26] and the
4
DNS method [18] are illustrated in Fig. 1. The
CuSO /biquinoline assay was found to be sen-
4
sitive to small concentrations of products
formed in the range 0.025–0.25 mM (Fig.
1
(A)). The DNS method could be used at
concentrations of 0.5–3.0 mM (Fig. 1(B)) and
the curves were still found to be linear up to
1
0 mM. Identical concentrations of glucose
and maltose did not contribute equally to the
D
-glucose units in the oligosaccharide chain
[4,27,28].
Malto-oligosaccharide-forming
amylases
have been found to form either maltotriose,
maltotetraose, maltopentaose or maltohexaose
as the dominating product. As
D-glucose and
malto-oligosaccharides do not have the same
effect on the oxidising reagents of most sac-
charogenic assays, we found that
D
-glucose is
inappropriate for use as a reference. A correct
determination of the activity would be possi-
ble only by the use of the predominant hydro-
lysis product as reference for the saccharo-
genic assay. Even a mixture of products can
result in a misleading activity determination.
Chromogenic methods.—For the chro-
mogenic methods, substrates such as insoluble
substrate-dye complexes, which liberate solu-
ble dye fragments upon hydrolysis [12] or
p-nitrophenol (p-NP) derivatives of malto-
oligosaccharides, which release the p-NP upon
hydrolysis [22], are used. p-NP derivatised
malto-oligosaccharides (maltopentaose, mal-
tohexaose or maltoheptaose) are hydrolysed
by the amylases in conjunction with a-glucosi-
dase or glucoamylase leading to the liberation
Fig. 1. Standard curves for determination of reducing sugar
equivalents by saccharogenic assays using glucose and malto-
oligosaccharides. Assay: (A) CuSO /biquinoline assay, (B)
4
2
DNS assay, the linear correlation coefficient (R ) are seen. ꢀ
glucose, ꢁ maltose, ꢂ maltotriose, X maltotetraose, and ꢃ
maltoheptaose.

L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
113
of p-NP and measured spectrophotometrically
as the yellow coloured nitrophenoxide ion at
weakly alkaline conditions [20–22]. Due to
the need of a-glucosidase for the assay of
p-NP-derivatised malto-oligosaccharides the
procedure is limited to a pH range of 5.0–6.0.
For the endo-acting amylases the determina-
tion of activity can be achieved by the use of
a p-NP-maltoheptaoside blocked with a 4,6-
O-benzylidene at the non-reducing end ac-
cording, e.g., to the Cerealpha method [21]
method is performed by addition of an I ꢀKI
2
solution to time course samples withdrawn
from starch or amylose solutions added to the
actual amylase [23]. The amylose helix forms
an inclusion complex with iodine and triiodide
−
3
(I and I ) which results in a blue colour
2
[24,29]. When the starch or amylose is cleaved
by amylases the substrate decreases in size,
which also decreases the size of the complex
and the intensity of blue colour. John et al.
[24] found an acrolimit at which no colour
was detected for linear chains of 20 glucose
units or less. Moreover, helices in amylopectin
do not contribute to the blue colour obtained
for 39 glucose units or more [24]. The stability
of the formed iodine–starch complex depends
on the temperature and pH [30]. In the present
investigation analysis of the intensity of
starch–iodine blue colour during hydrolysis
was performed at room temperature and con-
stant pH.
The starch–iodine blue colour method can
be used for the determination of the relative
activity of an amylase [13]. To investigate the
hydrolysis action of three different amylases,
the progress in starch hydrolysis was followed
by simultaneous starch–iodine blue colour
formation and liberation of reducing ends (see
Fig. 2). A steep slope is the result of a fast
reduction of the starch–iodine blue colour
due to random endo-activity by the a-amylase
(BAN, slope, a= −26). The maltogenase and
the extracellular enzyme preparation from
Pseudomonas stutzeri NRRL B 3389 showed
evidence of a prevalent exo-activity indicated
by a flat slope of a= −2 and a= −12, re-
spectively (see Fig. 2).
Direct comparison of reaction rates of a
maltose and a malto-oligosaccharide-forming
(DP\2) exo-amylase by the use of a saccha-
rogenic assay may be difficult since the ex-
pected result would be fewer reducing ends
released by the latter during the progress of
the reaction. The hydrolysis activity of CGT-
ases cannot be detected by any of these assays
since they only produce low amounts of re-
ducing ends. Furthermore, the cyclodextrins
formed by CGTases complex with iodine and
interfere with the starch–iodine method [26].
The information obtained from the saccharo-
genic assays, the chromogenic assays, or the
(see Table 2). The blocking of the non-reduc-
ing end prevents activity of exo-amylases. For
the determination of exo-activity a non-
blocked mixture of p-NP-maltoheptaoside, p-
NP-maltohexaoside and a-glucosidase is used
as, e.g., in the Betamyl method [22] (see Table
2
). Strict exo-activity on a starch substrate
leading to the formation of maltoheptaose
DP7) as the end product would result in
(
limited or no cleavage at all of the p-NP-mal-
topentaose substrate depending on the subsite
affinities.
Enzyme activity on short substrates (e.g.
maltopentaose) may not lead to results use-
able to predict the activity on polymeric sub-
strates as native starch which could vary in
structure and be branched. The substrate can
indeed be too short for detection of the actual
activity, which is important to consider when
using p-NP-derivatised malto-oligosaccharide
substrates. Furthermore, the activity of the
enzyme can vary depending on the length of
the substrate as known from a-glucosidases,
which show a lower activity on long substrates
than on short substrates [2].
Amylolytic activity can be determined using
insoluble substrates with covalently bound
dyes. The insoluble part can be either starch,
amylose or amylopectin [22] e.g. the azo-
colour crosslinked amylose termed AZCL–
amylose (Table 2). The dye is water soluble
and is liberated as a result of hydrolysis. How-
ever, the size of the starch hydrolysis product
or the cleavage mechanism cannot be deter-
mined by this method.
Amyloclastic methods.—The amyloclastic
methods include the iodometric method (the
so-called starch–iodine blue colour method)
and the measurement of the changes of the
viscosity of a starch solution. The iodometric

1
14
L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
Fig. 2. Determination of endo- and exo-action mode of some enzyme preparations. The formation of starch–iodine blue colour
is plotted against the production of reducing sugar equivalents. The slope of the linear curve (a) indicates the prevalence of endo-
or exo-action mode.
Table 3
Selected properties of the TLC, HPLC and HPAEC–PAD methods for detection of starch hydrolysis products formed by amylase
a
activity
Method
Pretreatment
Reproducibility
Analysis time
Separation of
products
Cleaning
procedure
Use of a
gradient
DP-separated
TLC
HPLC
HPAE–PAD
++
++
+++
++
+++
+++
+++
+++
+++
++
++
+++
+++
++
+++
−
+
+++
10
20
50
a
−
, impossible; +, hardly possible or not easy; ++, can be done, fair; +++, easy, good.
starch–iodine blue colour method, cannot re-
veal which starch hydrolysis product is pre-
dominantly formed.
Analysis of the starch hydrolysis products.—
For determination of the starch hydrolysis
products formed by the malto-oligosaccha-
TLC-plate and also temperature, and the con-
ditions for colour detection and development
including distribution of the reagent (see
Table 3).
High pressure liquid chromatography
(HPLC) with refractive index (RI) detection
does not easily provide a stable baseline and is
sensitive to the choice of eluent and sample
matrix components [37]. Silica-based columns
must be operated at a pHB7.5 and therefore,
saccharides with high DP, which are soluble
only in alkaline solution cannot be analysed
on these types of column. Separation of
malto-oligosaccharides up to DP 20 by ad-
sorption chromatography on chemically
bonded phases [38] has been achieved (see
Table 3).
ride-forming
amylases,
chromatographic
methods using different separation techniques
have been applied. These methods show dif-
ferent advantages and limitations for identifi-
cation of the starch hydrolysis products.
Thin-layer chromatography (TLC) is a low
cost method and still commonly used as an
analytical tool for the detection of starch hy-
drolysis products because of its simplicity and
relatively high sensitivity [5,7,31–36]. How-
ever, the variation on a single analysis as well
as between analyses can occur as a result of
variation in application of samples to the
The development of high performance an-
ion-exchange chromatography (HPAE) com-

L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
115
bined with pulsed amperometric detection
PAD) has provided a highly improved sepa-
ration of malto-oligosaccharides [39,40] and
increased the sensitivity of detection signifi-
cantly. For a homologous malto-oligosaccha-
ride series, it can be generalised that the larger
DP, the longer the retention time (see Fig.
starch hydrolysis products and malto-
oligosaccharide-forming amylases give rise to
major amounts of one single product. The
initial starch hydrolysis products obtained
with amyloglucosidase from Aspergillus niger
(
(E.C. 3.2.1.3), maltogenase from Bacillus
stearothermophilus (E.C. 3.2.1.-), a-amylase
from Bacillus amyloliquefaciens (BAN, E.C.
3
(C)). However, the response for similar mo-
3
.2.1.1), CGTase from Paenibacillus sp. F8
lar concentrations of different malto-oligosac-
charides decreases with increasing molecular
size until DP6 (Fig. 4).
(
E.C. 2.4.1.19), and exo-maltotetraohydrolase
from P. stutzeri NRRL B 3389 (E.C. 3.2.1.60)
are shown in Fig. 3. It is clearly indicated that
malto-oligosaccharide-forming amylases can
be detected by the sole hydrolysis product
formed (Fig. 3(B, E)). Even CGTases give rise
to a characteristic chromatogram indicated by
the irregular peak shapes of b-CD and g-CD
in Fig. 3(D). Furthermore, cyclodextrins dis-
play approximately 11 times lower response
on a molar basis, than malto-oligosaccharides
Different amylases give rise to different
[
41].
The HPAE-PAD is particularly well suited
for the analysis of the time course of an
enzymatic reaction, since formed products can
be qualitatively detected. The formation of
initial and end products by starch hydrolysis
can be easily determined by comparison of
chromatograms obtained at different time in-
tervals. The elution conditions can be adjusted
and results can be obtained within a short
time. With the HPAE–PAD direct quantifica-
tion of non-derivatised carbohydrates at low
picomole values with minimal sample prepara-
tion can be performed. The detection is sensi-
tive and consumes only small amounts of
sample. Compared with the TLC technique,
which is a cheap method requiring few materi-
als and HPLC with RI-detection, HPAE–
PAD is superior for the identification of the
starch hydrolysis products formed by amylase
activity. The three methods are listed accord-
ing to reproducibility and separation of prod-
ucts according to their DP in Table 3.
Conclusions.—For activity determinations
care should be taken in the choice of the
saccharogenic assay and the use of saccharide
reference. Comparison of the amylase activi-
ties published in the literature is difficult due
to the use of different methods and saccharide
references for determination of reducing
sugar. A significant problem was shown using
Fig. 3. HPAE–PAD chromatograms obtained after starch
hydrolysis with different amylolytic enzymes. (A) amyloglu-
cosidase, (B) maltogenase, (C) a-amylase (BAN), (D) Paeni-
bacillus sp F8 CGTase, (E) P. stutzeri maltotetraohydrolase.
The detected peaks represent DP1=glucose, DP2=maltose,
DP3=maltotriose, DP4=maltotetraose, DP5=maltopen-
taose, DP6=maltohexaose, DP7=maltoheptaose, a-CD, b-
CD, g-CD=a-, b-, g-cyclodextrin, respectively.

1
16
L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
Fig. 4. Standard curves for HPAE–PAD for glucose and malto-oligosaccharides. ꢀ Glucose, ꢁ Maltose, ꢂ Maltotriose, X
Maltotetraose, ꢄ Maltopentaose, ꢅ Maltohexaose, ꢃ Maltoheptaose.
the DNS and CuSO /biquinoline methods as
Malto - oligosaccharide - forming amylases
have been named according to a predominant
hydrolysis product formed at the beginning of
the hydrolysis [32,34,35,43] or a hydrolysis
product dominating at the steady state of the
hydrolysis [44–46]. The steady state is ob-
tained when the formed products are not
changing by further hydrolysis [47] also de-
scribed as end products. Hayashi et al. [34,35]
described an amylase as a maltohexaose-form-
ing amylase based on the initial product for-
mation. An amylase of similar activity was
described by Kim et al. [44] as a malto-
tetraose-forming amylase based on the end
product formation. An indication of the stage
of hydrolysis at which the specific product is
produced could minimise the misleading first
impression of very different enzymes.
4
they both gave different responses to glucose
and maltose. For the DNS method [18] the
actual product formed should be used as the
reference, while other methods mainly need
attention by the use of glucose as reference
resulting in an overestimation of activity com-
pared with malto-oligosaccharides. Only com-
parison between investigations using the same
methods and even the same saccharide refer-
ence is recommended.
The identification of the starch hydrolysis
products produced by the malto-oligosaccha-
ride-forming amylases is an essential step in
the determination of the activity and classifi-
cation of the amylase. Chromatography meth-
ods, preferably HPAE–PAD can be used for
the detection of these products. For detection
of amylolytic activity on pure substrates of
low DP the HPLC and the TLC techniques
are appropriate. An improved separation of
hydrolysis products can be obtained by
HPAE–PAD within short analysis times. Ini-
tial products of high DP (up to DP 50) can
also be detected by this method. Due to the
high sensitivity of detection, this method is an
excellent tool for following the starch hydroly-
sis over time. The present investigation
showed that the response for similar molar
concentrations of different malto-oligosaccha-
rides decreases with increasing molecular size
until DP6.
Some malto-oligosaccharide-forming amy-
lases have been found to form maltotriose
[31], maltotetraose [48], maltopentaose [32],
and maltohexaose [49] by an endo-type of
action mode. However, the main part of the
malto-oligosaccharide-forming amylases has
been found to form starch hydrolysis products
of a specific DP by an exo-type of action
mode (Table 4). Exo-maltotetraohydrolase
(E.C. 3.2.1.60) from P. stutzeri NRRL B 3389
[5] and exo-maltohexaohydrolase (E.C.
3.2.1.98) from Klebsiella pneumonia [7] have
been classified and given systematic names as

L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
117
suggested by studies of their hydrolytic action
6,53]. However, the nomenclature is not used
can be illustrated by the exo-maltotetraose
forming amylase described by Kim et al.
[44] initially forming maltohexaose which
is further degraded to the end product
maltotetraose.
[
systematically for malto-oligosaccharide-form-
ing amylases throughout the literature. The
maltohexaose-forming amylase from Bacillus
caldo6elox was found to cleave starch by an
exo-type of action mode [52], still it was
classified as an a-amylase which is not consis-
tent with the exo-activity. Clearly, the action
mode is of importance for classification of the
malto-oligosaccharide-forming amylases.
We therefore suggest a nomenclature in-
cluding the action mode of the enzyme, the
predominant starch hydrolysis product
formed and the stage of hydrolysis at which
this product is formed. This way three differ-
ent classes of amylases forming one predomi-
nant malto-oligosaccharide can be identified:
3
. Malto-oligosaccharide-forming a-amylase
are amylases predominantly forming a
malto-oligosaccharide of a specific DP by
endo-action mode. This term should not
include a-amylases forming a broad distri-
bution in the DP of products as e.g. the B.
amyloliquefaciens amylase.
Since glucose is not a malto-oligosaccharide
glucose formers are excluded, while b-amylase
still can be included as an exo-maltose hydro-
lase. Recently, this more precise terminology
has been used [34,35,44,51,52], however with
some inconsistency as indicated above.
1
. Exo - malto - oligosaccharide hydrolase
should include amylases, which predomi-
nantly form a malto-oligosaccharide of a
specific DP at the beginning and through-
out the entire period of hydrolysis by an
exo-action mode, as, e.g., the exo-maltotet-
raohydrolase found by Robyt and Acker-
man [5].
3
. Materials and methods
Materials.—Soluble starch, glucose, potas-
sium–sodium tartrate, and CuSO ·5H O were
4
2
obtained from E. Merck KGa, Darmstadt,
Germany. Maltose, maltotriose, maltotetra-
ose, maltopentaose, maltohexaose and malto-
heptaose, 3,5-dinitrosalicylic acid (DNS), and
4,4%-dicarboxy-2,2%-biquinoline were obtained
from Sigma Chemical Co., St. Louis, USA. a-,
b- and g-cyclodextrin were obtained from
Wacher Chemie GmbH, Burghausen, Ger-
many. Maltotetraose was prepared as de-
2
. Exo-malto-oligosaccharide-forming amy-
lase for amylases which form a malto-
oligosaccharide of specific DP as end
product by an exo-action mode. This amy-
lase would form an initial product which is
further hydrolysed to the predominant end
product determining the classification. This
Table 4
Bacterial malto-oligosaccharide forming amylases of endo- and exo-action mode
a
Origin
Malto-oligosaccharide
Mode of action
Ref.
Streptomyces griseus NA-468
Bacillus subtilis
Bacillus sp. MG-4
maltotriose (DP3)
maltotriose (DP3)
maltotriose (DP3)
exo
endo
–
[50]
[31]
[53]
P. stutzeriNRRL B-3389
Pseudomonas sp.IMD353
Bacillus sp. GM8901
Bacillus licheniformis 584
Isolate 163-26 (DSMZ 5853)
Aerobacter aerogenes (K. pneumonia)
Bacillus circulans F-2
B. circulans G-6
maltotetraose (DP4)
maltotetraose (DP4)
maltotetraose (DP4)
maltopentaose (DP5)
maltopentaose (DP5)
maltohexaose (DP6)
maltohexaose (DP6)
maltohexaose (DP6)
maltohexaose (DP6)
maltohexaose (DP6)
exo
endo
exo
endo
exo
exo
exo
endo
exo
exo
[5]
[48,51]
[44]
[32]
[42]
[7,8]
[33,43]
[49]
[34,35]
[52]
Bacillus sp. H.167
B. caldo6elox
a
Examples are listed with origin, mode of action, formed products, and references.

1
18
L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
scribed previously [54]. All other chemicals
were unless stated otherwise of analytical
grade. Amyloglucosidase (AMG, E.C.
added and mixed. Deionised water (4.0 mL)
was added and mixing was repeated. The blue
colour developed after 10 min at ambient
temperature was measured spectrophotometri-
cally at 600 nm.
Analysis of hydrolysis products.—Hydroly-
sis products of starch and reference malto-
oligosaccharides (25–200 mM) were detected
using high performance anion-exchange chro-
matography with pulsed amperometric detec-
tion (HPAE-PAD). A CarboPac PA-1 column
3.2.1.3), a glucose producing amylase from A.
niger, maltogenase (E.C. 3.2.1.-), a maltose
producing amylase from B. stearothermophilus
and a bacterial a-amylase (BAN, E.C. 3.2.1.1)
from B. amyloliquefaciens, were all obtained
from Novo Nordisk, Bagsværd, Denmark.
Purified Paenibacillus sp. F8 CGTase was ob-
tained as previously described [41]. Crude exo-
maltotetraohydrolase (E.C. 3.2.1.60) was
prepared as described by Robyt and Acker-
man [5] and used without further purification.
Determination of reducing sugars.—For de-
termination of reducing sugars standard
curves of glucose and malto-oligosaccharides
were prepared. All measurements were per-
(
Dionex Corporation, Sunnyvale, CA, USA)
was used with a gradient of 1.0 M NaOAc
from 0 to 60% over 30 min in 100 mM NaOH
at a flow rate of 1.0 mL/min on a Dionex
DX-300 system (Dionex Corporation, Sunny-
vale, CA, USA). Starch hydrolysis products
were identified by comparison of their reten-
tion times with glucose, maltose, maltotriose,
maltotetraose, maltopentaose, maltohexaose,
maltoheptaose, a-, b- and g-CDs. Since the
retention times of homologous linear malto-
oligosaccharides increases with the DP, linear
malto-oligosaccharides of intermediate DP
could easily be identified [40].
formed in triplicates. For the CuSO /biquino-
4
line assay [26] standard curves of glucose,
maltose, maltotriose, maltotetraose and malto-
heptaose, respectively, in the concentration
range 0.025–0.3 mM were prepared.
For the DNS method of Bernfeld [18] stan-
dard curves of glucose, maltose and mal-
totriose were prepared, respectively, in the
concentration range 0.1–3.0, 5.0 and 10.0
mM. Determinations were done spectrophoto-
metrically at 540 nm.
Acknowledgements
Enzyme assay.—2.0 mL soluble starch solu-
tion (1.25% (w/v)) in buffer (acetate buffer pH
This work was a part of a Ph.D. pro-
gramme supported by Aalborg University.
Mette Steengaard and Tanja Nielsen are
thanked for technical assistance on perfor-
mance of the saccharogenic methods.
5.0, phosphate buffer pH 7.0) and 0.5 mL
enzyme solution (diluted to give a proper level
of activity) was mixed to give a final concen-
tration of 1% starch. The reaction was incu-
bated at optimal temperature and stopped by
boiling the mixture for 10 min. The formation
of reducing sugars was determined in dupli-
References
cates with the CuSO /biquinoline assay [26].
4
[
[
1] J.J.M. Swinkels, Starch/St a¨ rke, 37 (1985) 1–5.
2] W.M. Fogarty, C.T. Kelly, Starch-degrading enzymes of
microbial origin, in M.J. Bull (Ed.), Progress in Indus-
trial Microbiology, Vol. 15, Elsevier Scientific, Amster-
dam, 1979, pp. 87–150.
Maltose was used as a reference except for
AMG where glucose was used.
Determination of starch–iodine blue colour
formation.—An I ꢀKI solution (0.20% (w/v)
2
[
3] W.M. Fogarty, Microbial amylases, in W.M. Fogarty
(Ed.), Microbial Enzymes and Biotechnology, Applied
Science, London, 1983, pp. 1–92.
KI and 0.02% (w/v) I ) was prepared accord-
2
ing to Fuwa [23]. The starch–iodine blue
colour formation was measured in duplicates
with the following modifications. A sample of
[4] J.F. Robyt, Enzymes in the hydrolysis and synthesis of
starch, in R.L. Whistler, J.N. Bemiller, E.F. Paschall
(
Eds.), Starch: Chemistry and Technology, second ed.,
5
00 mL was withdrawn from the enzymatic
Academic, San Diego, 1984, pp. 87–123.
5] J.F. Robyt, R.J. Ackerman, Arch. Biochem. Biophys.,
[
hydrolysis mixture of soluble starch at differ-
1
45 (1971) 105–114.
ent time intervals. pH was adjusted by 250 mL
[
6] T. Nakakuki, K. Azuma, K. Kainuma, Carbohydr. Res.,
1
.0 M HCl and 250 mL I ꢀKI solution were
128 (1984) 297–310.
2

L. Duedahl-Olesen et al. / Carbohydrate Research 329 (2000) 109–119
119
[
[
[
7] K. Kainuma, S. Kobayashi, T. Ito, S. Suzuki, FEBS
[31] Y. Takasaki, Agric. Biol. Chem., 49 (1985) 1091–1097.
[32] N. Saito, Archi6es. Biochem. Biophys., 155 (1973) 290–
298.
[33] H. Taniguchi, C.M. Jae, N. Yoshigi, Y. Maruyama,
Agric. Biol. Chem., 47 (1983) 511–519.
[34] T. Hayashi, T. Akiba, K. Horikoshi, Agric. Biol. Chem.,
52 (1988) 443–448.
[35] T. Hayashi, T. Akiba, K. Horikoshi, Appl. Microbiol.
Biotechnol., 28 (1988) 281–285.
Lett., 26 (1972) 281–285.
8] N. Monma, T. Nakakuki, K. Kainuma, Agric. Biol.
Chem., 47 (1983) 1769–1774.
9] S. Kitahata, Cyclodextrin glucanotransferase, in The
Amylase Research Society of Japan (Ed.), Handbook of
Amylases and Related Enzymes, their sources, isolation
methods, properties and applications. Pergamon, Oxford,
1988, pp 154–164.
[
10] K. Kainuma, a-Amylases which produce specific
oligosaccharides, in The Amylase Research Society of
Japan (Ed.) Handbook of Amylases and Related En-
zymes, Their Sources, Isolation Methods, Properties and
Applications, Pergamon, Oxford, 1984, pp 50–62.
[
36] H. Okemoto, S. Kobayashi, M. Monma, H. Hashimoto,
K. Hara, K. Kainuma, Appl. Microbiol. Biotechnol., 25
(
1986) 137–142.
[
37] V.M.B. Cabalda, J.F. Kennedy, K. Jumel, Analysis of
amylodextrins, in R.B. Friedman, (Ed.) Biotechnology of
Amylodextrin Oligosaccharides, ACS Symp. Ser. 458,
American Chemical Society, Washington DC, 1991, pp
[
11] T. Kimura, T. Nakakuki, Starch/St a¨ rke, 42 (1990) 151–
157.
[
12] K. Okuda, K-I. Naka, N. Shimojo, Clinical assay of
a-amylase activity in body fluids, in The Amylase Re-
search Society of Japan (Ed.), Handbook of Amylases
and Related Enzymes, Their Sources, Isolation Methods,
Properties and Applications, Pergamon, Oxford, 1988, pp
140–170.
[
38] C.A. White, P.H. Corran, J.F. Kennedy, Carbohydr.
Res., 87 (1980) 165–173.
[39] Y.C. Lee, Anal. Biochem., 189 (1990) 151–162.
[40] Y.C. Lee, J. Chromatogr. A, 720 (1996) 137–149.
[41] K.L. Larsen, L. Duedahl-Olesen, H.J.S. Christensen, F.
Mathiesen, L.H. Pedersen, W. Zimmermann, Carbohydr.
Res., 310 (1998) 211–219.
178–185.
[
13] M. Ohnishi, K. Hiromi, General considerations for con-
ditions and methods of amylase assay, in The Amylase
Research Society of Japan (Ed.), Handbook of Amylases
and Related Enzymes, Their Sources, Isolation Methods,
Properties and Applications, Pergamon, Oxford, 1988, pp
[
42] G. Schmid, A. Candussio, A. Bock, Biologically pure
culture of maltopentaose forming amylase producing
akalophilic bacteria, 1994, US Patent Number 5 304 723.
43] H. Taniguchi, Maltohexaose-producing amylase of
Bacillus circulans F-2, in R.B. Friedman (Ed.), Biotech-
nology of Amylodextrin Oligosaccharides. ACS Symp.
Ser. 458, American Chemical Society, Washington DC,
1
0–12.
14] J.F. Robyt, R.J. Ackerman, J.G. Keng, Anal. Biochem.,
5 (1972) 517–524.
[
[
[
4
15] S. Hizukuri, Y. Takeda, M. Yasuda, A. Susuki, Carbo-
hydr. Res., 94 (1981) 205–213.
16] M.J. Somogyi, Biol. Chem., 195 (1952) 19–23.
17] N.J. Nelson, Biol. Chem., 153 (1944) 375–380.
18] P. Bernfeld, Methods Enzymol., 1 (1955) 149–158.
19] M. Sinner, J. Puls, J. Chromatogr., 156 (1978) 197–204.
20] H. Sheehan, B.V. McCleary, Biotechnol. Techniq., 2
1991, pp 111–124.
[
[
[
[
[
[
[
[
[
[
44] T.U. Kim, B.U. Gu, J.Y. Jeong, S.M. Byun, Y.C. Shin,
Appl. En6iron. Microbiol., 61 (1995) 3105–3112.
45] N. Yoshigi, T. Chikano, M. Kamimura, Agric. Biol.
Chem., 49 (1985) 3369–3376.
46] J.F. Kennedy, C.A. White, Starch/St a¨ rke, 31 (1979)
(
1988) 289–292.
21] B.V. McCleary, H. Sheehan, J. Cereal Sci., 6 (1987)
37–251.
93–99.
[
[
47] B.S. Collins, C.T. Kelly, W.M. Fogarty, E.M. Doyle,
Appl. Microbiol. Biotechnol., 39 (1993) 31–35.
48] A.C.M. Bourke, C.T. Kelly, W.M. Fogarty, Biochem.
Soc. Trans., 19 (1991) 16S.
2
22] B.M. Brena, C. Pazos, L. Franco-Fraguas, F. Batista-
Viera, J. Chromatogr. B, 684 (1996) 217–237.
23] H. Fuwa, J. Biochem., 41 (1954) 583–603.
[
[
[
[
49] Y. Takasaki, Agric. Biol. Chem., 46 (1982) 1539–1547.
50] K. Wako, S. Hashimoto, S. Kubomura, A. Yokota, K.
Aikawa, J. Kamaeda, J. Jpn. Soc. Starch. Sci., 26 (1979)
175–181.
24] M. John, J. Schmidt, H. Kneifel, Carbohydr. Res., 119
(
1983) 254–257.
[
[
25] G.L. Miller, Anal. Chem., 31 (1959) 426–428.
26] L.H. Pedersen, H.J.S. Christensen, F. Mathiesen, K.L.
Larsen, W. Zimmermann, Starch/St a¨ rke, 49 (1997) 250–
[51] W.M. Fogarty, C.T. Kelly, A.C. Bourke, E.M. Doyle,
Biotechnol. Lett., 16 (1994) 473–478.
2
53.
27] F. Hostetter, E. Borel, H. Deuel, Hel6. Chim. Acta, 34
1951) 2132–2139.
28] D.J. Bell, D.J. Manners, A. Palmer, J. Chem. Soc. (1952)
760–3763.
29] H.F. Zobel, Starch/St a¨ rke, 40 (1988) 44–50.
30] J. Chrastil, Carbohydr. Res., 159 (1987) 154–158.
[
[
[52] W.M. Fogarty, F. Bealin-Kelly, C.T. Kelly, E.M. Doyle,
Appl. Microbiol. Biotechnol., 36 (1991) 184–189.
[53] Y. Takasaki, H. Shinohara, M. Tsuruhisa, S. Hayashi,
K. Imada, Agric. Biol. Chem., 55 (1991) 1715–1720.
[54] L. Duedahl-Olesen, W. Zimmermann, J.A. Delcour, Ce-
real Chem., 76 (1999) 227–230.
(
3
[
[
.