Hydrolysis of low-molecular-weight oligosaccharides and oligosaccharide alditols by pig intestinal sucrase/isomaltase and glucosidase/maltase

Article abstract of DOI:10.1016/S0008-6215(00)00008-2

The ability of purified pig intestinal sucrase/isomaltase (SI; EC 3.2.1.10/48) and glucosidase/maltase (GM; EC 3.2.1.20) to hydrolyze di- and oligosaccharides consisting of D-glucose and D-fructose residues and the corresponding alditols was studied. The products, after incubation, reflect different binding patterns at both catalytic sites of SI. The active site of the sucrase subunit cleaves α,β-(1→2) glycosidic bonds, and only two monomer units of the substrates bind with favorable affinity. Oligosaccharides and reduced oligosaccharides containing α-(1→6) and α-(1→1) glycosidic bonds are hydrolyzed by isomaltase, and for the active site of this subunit more than two subsites were postulated. Moreover, different binding sites for various aglycons seem to exist for isomaltase. Oligosaccharide alcohols are cleaved at lower rates if the reduced sugar residue occupies the aglycon binding site. GM also hydrolyzes α-(1→1) linkages, but at a lower rate. The enzyme has the ability to bind compounds containing residues other than D-glucose. There are indications for similarities between GM and the isomaltase subunit of SI in the binding mode of oligosaccharides. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00008-2

www.elsevier.nl/locate/carres
Carbohydrate Research 326 (2000) 264–276
Hydrolysis of low-molecular-weight oligosaccharides and
oligosaccharide alditols by pig intestinal
sucrase/isomaltase and glucosidase/maltase
Sabine Hertel a,*, Fritz Heinz b, Manfred Vogel a
a Zentralabteilung Forschung, Entwicklung und Ser6ices, Su¨dzucker AG Mannheim/Ochsenfurt,
D
-67283 Obrigheim/Pfalz, Germany
b Zentrum Biochemie, Arbeitsbereich Enzymologie OE 4312, Medizinische Hochschule Hanno6er,
-30623 Hanno6er, Germany
Received 21 November 1998; received in revised form 27 October 1999; accepted 20 December 1999
D
Abstract
The ability of purified pig intestinal sucrase/isomaltase (SI; EC 3.2.1.10/48) and glucosidase/maltase (GM; EC
3.2.1.20) to hydrolyze di- and oligosaccharides consisting of -glucose and -fructose residues and the corresponding
D
D
alditols was studied. The products, after incubation, reflect different binding patterns at both catalytic sites of SI. The
active site of the sucrase subunit cleaves a,b-(1“2) glycosidic bonds, and only two monomer units of the substrates
bind with favorable affinity. Oligosaccharides and reduced oligosaccharides containing a-(1“6) and a-(1“1)
glycosidic bonds are hydrolyzed by isomaltase, and for the active site of this subunit more than two subsites were
postulated. Moreover, different binding sites for various aglycons seem to exist for isomaltase. Oligosaccharide
alcohols are cleaved at lower rates if the reduced sugar residue occupies the aglycon binding site. GM also hydrolyzes
a-(1“1) linkages, but at a lower rate. The enzyme has the ability to bind compounds containing residues other than
D
-glucose. There are indications for similarities between GM and the isomaltase subunit of SI in the binding mode
of oligosaccharides. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Pig intestinal sucrase; Isomaltase; Glucosidase; Maltase; Affinity to monomeric units of substrates; Characteristic of
active sites; Enzyme specificity
1. Introduction
drolase that acts at the nonreducing end of
oligosaccharides specifically cleaving a,b-(1“
2), a-(1“4) and a-(1“6) bonds, and GM
exclusively cleaves a-(1“4) glycosidic bonds.
Corresponding to the substrate specificity of
their catalytic sites, the subunits of SI are
called sucrase and isomaltase. In humans [1,2],
rats [3] and rabbits [4], both subunits also
show maltase activity. In monkeys, this activ-
ity is restricted to the isomaltase subunit [5],
and in pigs to the sucrase subunit [6,7]. At
both subunits, the binding site for the glycon
moiety only accepts pyranoside structures,
Sucrase/isomaltase (SI; EC 3.2.1.10/48) and
glucosidase/maltase (GM; EC 3.2.1.20) belong
to the intestinal brush border glycoside hydro-
lases and are responsible for the final steps of
carbohydrate digestion, leading to absorbable
monosaccharides. Both enzyme complexes
consist of two noncovalently linked subunits,
each harboring one active site. SI is an exohy-
* Corresponding author. Tel.: +49-6359-803277; fax:
+49-6359-803331.
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00008-2

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
265
whereas the aglycon component can be a fura-
nosyl as well as a pyranosyl ring. Different
kinds of aglycons, however, occupy individual
binding sites [7].
of GM have no affinity for furanoside struc-
tures.
The purpose of this work was to gain new
information about the specificity of the active
sites of SI and GM for certain glycosidic
bonds and the affinity for the different
monomeric parts of the substrates. This sub-
ject is of importance as there are not many
definitive reports on the specificity of these
enzymes on oligosaccharides and their aldi-
tols. Therefore, the hydrolysis of selected
In contrast to SI, the two active sites in
GM cannot be distinguished with respect
to substrate specificity [8]: both cleave
a-(1“4) glycosidic bonds and hydrolyze mal-
tooligosaccharides up to maltoheptaose. As
described for glucoamylases of different
origins [9–12], the active sites of GM are
divided into a series of binding sites [2,5,13],
called subsites, which interact with the
monomer units of the substrates. According
to Heymann [7], the subsites of the active sites
oligosaccharides, consisting of
D
-glucose and
D
-fructose residues linked via a-(1“1), a-
(1“6) or a,b-(1“2) glycosidic bonds (Fig.
1(A)) and of their corresponding sugar alco-
Fig. 1. Structures of (A) oligosaccharides and (B) reduced di- and oligosaccharides (for abbreviations see Table 1), positions of
hydrolysis (indicated by arrows), and rate of cleavage by SI and GM, respectively, after 24 h of incubation (for the hydrolysis of
the oligosaccharide alcohols by SI, the cleavage rate expresses the sum hydrolysis of Glc-ol and Man components).

266
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
Fig. 1. (Continued)
hols (Fig. 1(B)), by the isolated intestinal a-
glucosidases was determined and compared
with the cleavage of maltose.
6G-1,1-GPM/1G-1,6-GPS and 1,1-GPS. In
spite of the maximum cleavages of 62, 27 and
35%, the rates of hydrolysis of these com-
pounds were comparatively low.
With the exception of Isomele, the oligosac-
charides containing a-(1“6) and a-(1“1)
glycosidic bonds were cleaved by SI to an
extent of 92–100% (Fig. 2(A)). However, the
progress curves compared with that of Mal
indicate a lower velocity for the hydrolysis of
the oligosaccharides. After 2 h of reaction, the
rates of hydrolysis were determined to be 11,
27, 45, 60 and 69% for Isomele, 6%G-Tre,
2. Results
The amount of enzyme in the incubation
assay was chosen to give a maltase activity of
0.7 U/mL for SI and GM, leading to 90 and
95% hydrolysis of maltose, respectively, after
2 h of incubation (Fig. 2).
An overview of the percentage rates of hy-
drolysis for the oligosaccharides, reduced dis-
accharides and reduced oligosaccharides,
expressed by the decrease of the substrates
(for abbreviations see Table 1), is shown in
Figs. 2 and 3. SI had the capability of hy-
drolyzing all of the sugars and reduced sugars
tested, whereas GM only cleaved 1G-Pala,
6%G-Pala,
6%G,6%%G-Pala,
and
1G-Pala,
respectively.
For the reduced oligosaccharides, with the
exception of 6G-1,1-GPM/1G-1,6-GPS, the
rates of hydrolysis by SI (Fig. 3(A)) were
similar to those of the corresponding oligosac-

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
267
Fru was released. A very small amount of
Iso-DP2 was detectable. The time course of
the reaction of 6%G-Tre, and the formation of
intermediate products, was qualitatively simi-
lar to the hydrolysis of 6%G-Pala (Fig. 4(C)).
6%G,6%%G-Pala was cleaved by SI leading to the
production of 6%G-Pala and Glc (Fig. 4(D)).
As the intermediate trisaccharide was a sub-
strate itself, it accumulated only for the first 2
h. After this lag phase, the hydrolysis of 6%G-
Pala predominated.
1G-Pala was hydrolyzed by both SI and
GM (Fig. 5(A and B)). The SI hydrolysis
pattern was similar to those of 6%G-Pala and
6%G-Tre. The hydrolysis by GM was slow,
giving increasing amounts of Pala, because
GM had no activity on this disaccharide.
As mentioned above, the action of SI
on 6%G-Isomalt, 6%G,6%%G-Isomalt and 6%G-1-
Isomalt resembled the action on the cor-
responding oligosaccharides. Intermediate di-
saccharide or trisaccharide alcohols were
formed, which, as well as the substrates, were
mixtures of isomers containing Man and Glc-
ol components. The end products were Glc,
Man and Glc-ol. When 6%G-Isomalt was hy-
drolyzed by SI (Fig. 6(A)), 1,1-GPM and 1,6-
GPS appeared in a 2:3 ratio, in accordance
with the composition of the initial substrate.
During incubation, the concentration of the
charides (Fig. 2(A)). The rate of hydrolysis of
6G-1,1-GPM/1G-1,6-GPS was nearly below
50% compared with that of the analogous
1G-Pala. This relationship was also observed
for the action of GM (Fig. 2(B) and Fig.
3(B)).
The reduced disaccharides 1,1-GPM, 1,6-
GPS and 1,1-GPS were hydrolyzed at a slower
rate by SI than the reduced oligosaccharides
and Mal (Fig. 3). The best activity was found
for 1,1-GPS, being more than twice as high as
that of 1,6-GPS. Furthermore, 1,6-GPS was
clearly a better substrate for SI than 1,1-
GPM. GM only hydrolyzed 1,1-GPS.
The reaction mixtures of oligosaccharides
and the oligosaccharide alditols gave varying
intermediates and monomeric end products.
The time courses of the substrates and all
products of the enzymatic hydrolysis are
shown in Figs. 4–7 (see Table 1 for
abbreviations).
The hydrolysis of Isomele by SI (Fig. 4(A))
resulted in the initial formation of Pala and
Glc. After 3 h of incubation, the increase of
the Pala concentration was reduced, and Fru
was released. No sucrose was detected. The
incubation of 6%G-Pala (Fig. 4(B)) also led to
the production of Pala and Glc, but at a much
higher rate than from Isomele. After 3 h, the
concentration of Pala remained constant, and
Fig. 2. Percentage rates of hydrolysis of Isomele, 6%G-Pala, 1G-Pala, 6%G-Tre and 6%G,6%%G-Pala (for abbreviations see Table 1) by
SI (A) and GM (B) compared with the splitting of Mal.

268
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
Table 1
Nomenclature for di- and oligosaccharides and their reduced forms and abbreviations used in this paper
Systematic name Abbreviation
D
D
D
D
-Glucose
-Fructose
-Mannitol
-Glucitol
Glc
Fru
Man
Glc-ol
6-O-a-
6-O-a-
1-O-a-
4-O-a-
D
D
D
D
-Glucopyranosyl-
-Glucopyranosyl-
-Glucopyranosyl-
-Glucopyranosyl-
D
D
D
D
-fructofuranose (isomaltulose, palatinose)
-glucopyranose (isomaltose)
-fructopyranose (trehalulose)
-glucopyranose (maltose)
Pala
Iso-DP2
Tre
Mal
a-D
-Glucopyranosyl-(1“6)-a-
D
-glucopyranosyl-(1“6)-
D
-glucopyranose (isomaltotriose)
Iso-DP3
1,1-GPM
1,6-GPS
1,1-GPS
Isomele
6%G-Pala
1G-Pala
6%G-Tre
6%G,6%%G-Pala
1-O-a-
6-O-a-
1-O-a-
D
D
D
-Glucopyranosyl-
-Glucopyranosyl-
-Glucopyranosyl-
D
D
D
-mannitol
-glucitol
-glucitol
a-D
-Glucopyranosyl-(1“6)-b-fructofuranosyl-a-
-Glucopyranosyl-(1“6)-a- -glucopyranosyl-(1“6)-
-glucopyranosyl- -fructofuranose
-glucopyranosyl-(1“1)-
D-glucopyranoside (isomelezitose)
a-
D
D
D
-fructofuranose
1,6-Di-O-a-
D
D
a-
a-
D
D
D
D
-Glucopyranosyl-(1“6)-a-
-Glucopyranosyl-(1“6)-a-
-fructofuranose
D
D
-fructopyranose
D
-glucopyranosyl-(1“6)-a-D-gluco-pyranosyl-(1“6)-
a-
-Glucopyranosyl-(1“6)-a-
D
-glucopyranosyl-(1“1)-
-glucopyranosyl-(1“6)-
-mannitol and 1,6-di-O-a-
-glucopyranosyl-(1“1)-
-glucopyranosyl-(1“1)-
-glucopyranosyl-(1“6)-a-
-glucopyranosyl-(1“6)-a-glucopyranosyl-(1“6)-a-
D
-mannitol and
6%G-Isomalt
a-
D
-glucopyranosyl-(1“6)-a-
D
D-glucitol
1,6-Di-O-a-
D
-glucopyranosyl-
-Glucopyranosyl-(1“6)-a-
a- -glucopyranosyl-(1“6)-a-
-Glucopyranosyl-(1“6)-a-
and a-
D
D
-glucopyranosyl-
D
-glucitol
6G-1,1-GPM/1G-1,6-GPS
6%G-1-Isomalt
a-D
D
D-mannitol and
D
D
D
-glucitol
-gluco-pyranosyl-(1“1)-
-glucopyranosyl-(1“6)-D
a-D
D
D
D
-mannitol 6%G,6%%G-Isomalt
-glucitol
D
D
disaccharide alcohols increased for 6 h, fol-
lowed by a stagnation period. At the same
time, the production of Man and Glc-ol was
observed, but compared with the release of
Fru from the corresponding oligosaccharide,
the hexitols were released at a much slower
rate. 6%G,6%%G-Isomalt also consisted of Man
and Glc-ol components in a ratio of 2:3. The
same ratio was observed for the intermediates
of SI hydrolysis, the sum of the components
being given in Fig. 6(C). The first hydrolytic
reaction produced 6%G-Isomalt, which itself is
a substrate of SI (see Fig. 6(A)). Initially, the
trisaccharide accumulated, but after 3 h, hy-
drolysis of the intermediate to Isomalt and
Glc predominated. Early in the reaction of
6%G-1-Isomalt with SI, the rates of release of
1,1-GPM and 1,1-GPS essentially corre-
sponded to the Glc release (Fig. 6(B)), indicat-
ing that Man and Glc-ol components of the
substrate were cleaved to nearly the same
extent, as was observed for 6%G-Isomalt and
6%G,6%%G-Isomalt. In the later periods of the
reaction, the concentrations of the intermedi-
ates 1,1-GPM and 1,1-GPS diverged, whereas
1,1-GPM and 1,6-GPS ran parallel in the case
of 6%G-Isomalt (Fig. 6(A)). Thus, a different
activity of SI with regard to the intermediate
disaccharide alcohols was indicated, agreeing
with the results reported in Fig. 3.
As mentioned above for the analogous
trisaccharide, 6G-1,1-GPM/1G-1,6-GPS was
hydrolyzed by both SI and GM. The Glc-ol
and Man components of this substrate were
not hydrolyzed to the same extent: SI favored
1G-1,6-GPS, as indicated by higher concentra-
tions of the intermediate 1,6-GPS compared
with 1,1-GPM. Also, GM hydrolyzed only the
Glc-ol component of the substrate.
3. Discussion
The rates of hydrolysis of disaccharide alco-
hols by SI and GM were calculated from the
D-glucose released, and it was found to be in
agreement with the consumption of the sub-
strate. However, for oligosaccharides and

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
269
oligosaccharide alcohols, a complete descrip-
tion of the degradation requires a time-course
study including the substrates as well as all
intermediates and products. Furthermore,
analysis of the various intermediates gives in-
sight into the specific breakdown of the differ-
ent glycosidic bonds.
SI has the capacity to hydrolyze all the
substrates tested, but the rates of hydrolysis
differ. Comparing the rate of hydrolysis for
Isomele with those for 6%G-Pala, 1G-Pala,
6%G-Tre and 6%G,6%%G-Pala, Isomele is the
slowest to be hydrolyzed (Fig. 2). Isomele
contains two nonreducing ends, and it could
be expected that both the a,b-(1“2) and the
a-(1“6) glycosidic bond would be hy-
drolyzed. In fact, only the a,b-(1“2) bond
was hydrolyzed, since Pala was the only disac-
charide intermediate to be observed (Figs. 1
and 3).
bond of Isomele was hydrolyzed at the active
site of sucrase, and that the cleavage of the
other oligosaccharides as well as of the oligo-
and disaccharide alcohols tested, all saccha-
rides containing a-(1“6) and/or a-(1“1) gly-
cosidic linkages, occurred at the active site of
the isomaltase subunit. The different binding
sites and the low rate of Isomele hydrolysis
indicate a different binding mode for the
oligomeric structures at both active sites of the
SI complex. The active site of the sucrase
subunit contains only two subsites and, thus,
can bind only two monomeric parts of a sub-
strate with positive affinity, providing a possi-
ble explanation for the reduced activity on
Isomele compared with sucrose. The rate of
hydrolysis was 29% for Isomele (Fig. 3) com-
pared with 86% for sucrose under the same
conditions (data not shown). Contrary to the
sucrase activity, the active site of isomaltase
seems to contain other binding sites in addi-
tion to the binding sites for the monomeric
glycon and aglycon moieties, as the tri- and
Considering the different substrate specific-
ities of the active sites of the SI complex (see
Section 1), it was assumed that the a,b-(1“2)
Fig. 3. Percentage rates of hydrolysis of 1,1-GPM, 1,6-GPS, 1,1-GPS, 6%G-Isomalt, 6G-1,1-GPM/1G-1,6-GPS, 6%G-1-Isomalt
6G%,6%%G-Isomalt (for abbreviations see Table 1) by SI (A) and GM (B).

270
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
Fig. 4. Time course of substrates (continuous lines), intermediates (long dashed lines) and end products (short dashed lines) for
tests of (A) Isomele, (B) 6%G-Pala, (C)6%G-Tre and (D) 6%G,6%%G-Pala (for abbreviations see Table 1) with SI over a period of 24
h.
Our results for pig intestinal enzymes indi-
cate that the subsite model is also valid for
certain oligosaccharide substrates containing
tetrasaccharides tested can be efficiently hy-
drolyzed by this subunit. Assuming a mini-
mum of four subsites at the active site of
isomaltase [9,11], this hypothesis would ex-
plain the higher rate of hydrolysis for
6%G,6%%G-Pala compared with 6%G-Pala. Hey-
mann et al. [2] thus suggested that the active
site of isomaltase of human small intestinal SI
could contain four subsites in contrast to only
two subsites for the active site of sucrase.
D
-glucose,
D
-fructose,
D
-glucitol and
D
-man-
nitol residues. Intermediate Pala in the experi-
ments with 6%G-Pala and 6%G,6%%G-Pala was
hydrolyzed at only a slow rate, because the
affinity of subsite 2 for the fructosyl residue is
reduced compared with the glucosyl residue.
Pala released from Isomele first has to leave

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
271
the active site of the sucrose subunit and move
to the isomaltase subunit, leading to an even
lower rate of hydrolysis.
Due to the lack of specificity for a-(1“6)
glycosidic bonds, GM hydrolyzed only 1G-
Pala, and none of the other oligosaccharides
(Fig. 2(B)). When the extent of hydrolysis had
reached 60%, contamination by SI was ex-
cluded as a cause of this conversion. The
accumulation of Pala and negligible concen-
trations of monosaccharides (Glc and Fru;
Fig. 5) reflect that the ‘Tre part’ of the trisac-
charide was exclusively hydrolyzed. Agreeing
with an earlier subsite model for the active
sites of GM and glucoamylases of different
origins [5,7,9–13] based on the interaction
with linear maltooligosaccharides, our results
for 1G-Pala indicate that the subsite model is
not limited to those simple a-(1“4)-linked
oligosaccharides. Furthermore, the hydrolysis
of 1G-Pala by GM shows activity towards
a-(1“1) glycosidic linkages as well as affinity
to the enzyme for fructosyl residues. Although
the activity on a-(1“1) glycosidic bonds was
clearly less than on a-(1“4) linkages, further
evidence for this specificity was provided by
the results found for 6G-1,1-GPM/1G-1,1-
GPS and 1,1-GPS. Affinity for fructosyl units
agrees with our recent results [14] on the
competitive inhibition of maltose hydrolysis
A comparison of the structures of 6%G-Pala,
1G-Pala and 6%G-Tre (Fig. 1) shows the influ-
ence of the position of the
D-fructose residue
on the activity of SI (Fig. 2(A)). The enzyme
hydrolyzed 1G-Pala faster than the two other
trisaccharides, though it consists of a ‘Pala
part’ and a ‘Tre part’, which were hydrolyzed
by a negligible amount only in the experi-
ments with 6%G-Pala and 6%G-Tre (see Fig. 4(B
and C)). Furthermore, the beginning of the
‘aglycon’ moiety of 1G-Pala is a fructofura-
nosyl ring, whereas it is a glucopyranosyl ring
in 6%G-Pala and 6%G-Tre. During hydrolysis of
glucosyloligosaccharides, the aglycon part oc-
cupies subsite 2, the subsite with the highest
affinity for
quently, a decrease of the activity should re-
sult when -fructose is accommodated at this
D-glucose residues [2]. Conse-
D
subsite. So, the highest activity found on 1G-
Pala is clearly in contradiction with the occu-
pation of subsite 2 by the
D
-fructose residue.
Apparently, different binding sites exist for
glucopyranoside and fructofuranoside aglycon
moieties of oligosaccharides, as described by
Heymann [7] for the SI-catalyzed hydrolysis of
dimeric substrates.
by
L
-ribulose, which indicated a low affinity of
GM for furanoside structures.
Fig. 5. Time course of substrates (continuous lines), intermediates (long dashed lines) and end products (short dashed lines) for
tests of 1G-Pala (for abbreviations see Table 1) with SI (A) and GM (B) over a period of 24 h.

272
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
Fig. 6. Time course of substrates (continuous lines), intermediates (roughly dashed lines) and end products (finely dashed lines)
for tests of (A) 6%G-Isomalt, (B) 6%G-1-Isomalt and (C) 6G%, 6%%G-Isomalt (for abbreviations see Table 1) with SI over a period of
24 h.
The activity of SI towards 6%G-Isomalt,
6%G,6%%G-Isomalt and 6%G-1-Isomalt and the
corresponding oligosaccharides was at the
same level (Figs. 2 and 3), indicating that
hydrogenation of the reducing end of
oligosaccharides having a minimum of three
monomeric units has no influence on hydroly-
sis. The hydrolytic capacity of SI, however,
was decreased for 6G-1,1-GPM/1G-1,6-GPS
and the disaccharide alcohols, having the hex-
itol unit corresponding to the aglycon binding
site of the enzyme. In agreement with the high
affinity of subsite 2 for glucopyranosydic
structures (see above), hydrolysis was found to
be favored for substrates with cyclic sugar
residues that interact with the aglycon binding
site.
In steady-state kinetic experiments, carried
out with crude homogenates of human intesti-
nal mucosa, the KM values of 1,1-GPM and
1,6-GPS were found to be much higher than
the value for maltose (Mal: 2.7 mM; 1,1-
GPM: 11.4 mM; 1,6-GPS: 7.7 mM), and the
relative rates of hydrolysis for Mal, 1,1-GPM
and 1,6-GPS are ‘100:4:2’ [15]. This agrees
with the reduced capacity for hydrolysis com-
pared with Mal found in our experiments. For
the isomer 1,1-GPS, no kinetic data were re-

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
273
ported. Our results demonstrate that 1,1-GPS
was hydrolyzed by SI to a much lower extent
than maltose, but with a notably higher rate
than both of the other disaccharide alcohols.
Similarly, only 1,1-GPS was hydrolyzed by
GM.
The disaccharide alcohols differ from mal-
tose by the glycosidic linkage and by the struc-
ture of the aglycon part. The a-(1“6) and
a-(1“1) glycosidic bonds of 1,6-GPS, 1,1-
GPM and 1,1-GPS correspond to each other,
because in both cases the glycon is linked with
a terminal carbon atom of the hexitol chain.
Therefore, it was concluded that the different
activities of SI and GM are influenced by the
conformation of the alditol chains. 1,1-GPM
total of 10 atoms from the terminal O-6 to C-
2% of the -glucose part [16,17]. In 1,6-GPS,
the nearly planar chain of the -hexitol com-
prises only eight atoms (from O-2% to C-2).
Due to unfavorable 1,3-interactions, the
molecule rotates about the C-2ꢀC-3 bond
through 120° and attaches the terminal hy-
D
D
droxymethyl group of the
D
-glucitol in a 67°
bend [17–19]. The -hexitol chains of 1,1-
D
GPM and 1,6-GPS thus meet the ‘alditol
rules’ of Jeffrey and Kim [20]: ‘the carbon
chain of an alditol adopts the extended, planar
zigzag conformation when the configurations
at alternate carbon centers are different, and is
bent and non-planar when they are the same’.
Also based on these rules, the conformation of
consists of a
D
-mannitol part adopting a
the
D
-glucitol part of 1,1-GPS results from a
nearly planar zigzag conformation, extending
rotation about the C-2ꢀC-3 bond through
the a-(1“6) glycosidic bond and comprising a
120° [19], producing two extended-chain por-
Fig. 7. Time course of substrates (continuous lines), intermediates (long dashed lines) and end products (short dashed lines) for
tests of 6G-1,1-GPM/1G-1,6-GPS (for abbreviations see Table 1) with SI (A) and GM (B) over a period of 24 h.
Scheme 1. Conformation of the alditol chains in 1,1-GPM, 1,6-GPS and 1,1-GPS (for abbreviations see Table 1) according to
Immel and Lichtenthaler [17], Munir et al. [18], and Lichtenthaler and Lindner [19].

274
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
the tri- and tetrasaccharides tested) and by
isomaltase (that dominates the hydrolysis of
all other oligosaccharides) may be explained if
a similar binding mechanism is valid for GM
and SI from pig intestine. The parallels found
for the hydrolysis of 1,1-GPS and 6G-1,1-
GPM/1G-1,6-GPS by GM and isomaltase of-
fer further evidence for this conclusion.
tions crossing each other (from C-1 to O-3
and from O-2 to O-6). With these conforma-
tions (Scheme 1), the aglycon parts of 1,6-GPS
and 1,1-GPS show increasing deviations from
the planar zigzag chain of the
D-mannitol
moiety of 1,1-GPM. So, it can be concluded
that the rate of hydrolysis by SI, and in part
by GM, increases with the deviation of the
alditol chain from the extended linear confor-
mation. Since SI and GM bind the aglycon
part of their substrates via hydrogen bonds
[21], these interactions will decrease with vari-
ation in the configurations of the hydroxyl
groups of the relevant residue from those of
the structure of the natural substrate.
Amongst the disaccharide alcohols, the varia-
tion from a cyclic structure of the aglycon is
highest for 1,1-GPM. The compound cannot
be hydrolyzed by GM and is hydrolyzed only
4. Experimental
Isomele was isolated from the product solu-
tion of sucrose isomerization to isomaltulose,
using immobilized cells of Protaminobacter ru-
brum (CBS 574.77). 6G-Pala, 6%G,6%%G-Pala,
and 6%G-Tre were obtained by incubating su-
crose as the glucosyl donor with isomaltulose
or trehalulose, in the presence of dextransu-
crase. 1G-Pala was obtained by reacting su-
crose and isomaltulose in the presence of
inulinase (SP230). After isolation from the
reaction solutions by gel permeation chro-
matography on Fractogel HW40S (TOSO-
HAAS) and rechromatography on the same
column, the tri- and tetrasaccharides were
purified to minimum purity of 95%. The
oligosaccharide alcohols were prepared by re-
ducing the purified oligosaccharides (redis-
solved in a 1:1 mixture of water and MeOH)
with NaBH4 (at 35 °C for 3 h; afterwards pH
adjustment to 6.5 by adding 0.1 M HCl).
After neutralization, the oligosaccharide aldi-
tols were isolated from the reaction by Frac-
togel HW40S column chromatography.
6%G-Isomalt, 6G-1,1-GPM/1G-1,6-GPS and
6%G,6%%G-Isomalt were isomeric mixtures of
Man and Glc-ol components in a ratio of 2:3;
the composition of 6%G-1-Isomalt concerning
Man and Glc-ol components was not known.
1,1-GPM and 1,6-GPS were obtained by frac-
tional crystallization of isomalt followed by
fractional recrystallization. For 1,1-GPS, the
starting material was trehalulose, which was
first catalytically hydrogenated with an H2–Pt
catalyst. 1,1-GPM was removed from the
product mixture by fractional crystallization,
and afterwards, 1,1-GPS was purified by
to a low extent by SI. The bend in the -gluci-
D
tol chain of 1,6-GPS seems to improve the
interaction of the hydroxyl groups with the
active site of isomaltase, whereas the bend in
the middle of the hexitol chain, as realized for
1,1-GPS, is most favorable for this kind of
interaction for both SI and GM.
The findings for the experiments with 6G-
1,1-GPM/1G-1,6-GPS (SI favoring 1G-1,6-
GPS or GM selecting the Glc-ol component
only, respectively) can be explained by the fact
that the bonds cleaved correspond to those
linkages of 1,1-GPM and 1,1-GPS (see Fig. 1).
Referring to the subsite model described for
GM and for the isomaltase subunit of SI,
6G-1,1-GPM/1G-1,6-GPS should be hy-
drolyzed to a higher extent than 1,1-GPS. The
results given in Fig. 3, however, show identical
rates of hydrolysis for both substrates, sug-
gesting that subsite 3 was not occupied during
hydrolysis of 6G-1,1-GPM/1G-1,6-GPS, pre-
sumably because the bend in the
D-glucitol
chain creates a distance between subsite 3 and
the relevant residue.
Based on their experiments with linear iso-
malto- and maltooligosaccharides, Heymann
et al. [2] postulated a similar binding mode of
linear glucosyloligosaccharides to human
small intestinal GM and to the isomaltase
subunit of SI. Our results, which demonstrate
the analogy between the hydrolysis of 1G-Pala
by GM (representing the exception amongst
cation-exchange chromatograpy in the Ca2+
-
form resin. In all cases, the purity of the
polyols was over 95%.

S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
275
hydrolysis were detected by HPAEC (column:
Carbopac PA1, 4 mm×250 mm, anion ex-
changer; two columns in series in the case of
reduced oligosaccharides; pre-column: Carbo-
pac PA1, 4 mm×50 mm, anion exchanger;
eluent: 0.22 M NaOH with a 0–20% gradient
of 0.5 M sodium acetate in 15 min for Isomele
and all non-reduced forms of oligosaccharides
or 0.12 M NaOH with a 1–33% gradient of
0.5 sodium acetate in 25 min for all reduced
oligosaccharides; flow-rate: 1 mL/h; sample:
0.020 mL; temperature: 25 °C; detection:
pulsed amperometric detector, gold electrode,
potentials E and pulse time settings t of the
detector are E1=0.05 V and t1=480 ms,
E2 =0.60 V and t2=300 ms, and E3 = −0.60
V and t3=240 ms).
Hexokinase (yeast), glucose-6-phosphate de-
hydrogenase (Leuconostoc mesenteroides), pa-
pain (Carica papaya), ATP, NAD and
triethanolamine hydrochloride were from
Boehringer Mannheim. Ammonium sulfate
was obtained from Serva. All other biochemi-
cals were purchased from E. Merck. DEAE-
cellulose (DE52) was from Whatman, and a
prepacked HiLoad 16/60 Superdex 200
column was purchased from Pharmacia
Biotech.
SI and GM were isolated from pig intestinal
mucosa by treatment with papain, and
purified by a procedure containing fractiona-
tion with ammonium sulfate, followed by ion-
exchange chromatography on DEAE-cellulose
and gel filtration on Superdex 200 [6,7,22–24].
The enzyme preparations achieved yields of
45% for SI and 31% for GM, respectively,
related to the activities after papain solubiliza-
tion. Expressed by sucrase activity for SI and
by maltase activity for GM, the specific activi-
ties were 15.3 and 8.2 U/mg. For SI, the
maltase to sucrase quotient was 0.7. For GM,
the sucrase activity could be reduced to 1.3%
of the sum of maltase and sucrase activities.
Protein concentrations were estimated ac-
cording to Bradford [25]. The purity of the
enzyme preparations was demonstrated by
SDS gel electrophoresis according to Davis
[26] and Laemmli [27].
Acknowledgements
The authors thank A. Pantke and H.
Dattge for their technical assistance with the
HPAEC-analyses.
References
[1] G. Galand, Comp. Biochem. Physiol., B94 (1989) 1–11.
[2] H. Heymann, D. Breitmeier, S. Gu¨nther, Biol. Chem.
Hoppe-Seyler, 376 (1995) 249–253.
[3] E.H. van Beers, H.A. Bu¨ller, R.J. Grand, A.W.C. Einer-
handet, J. Dekker, Crit. Re6. Biochem. Mol. Biol., 30
(1995) 197–262.
Oligosaccharides and oligosaccharide aldi-
tols were incubated with the enzymes for 24 h
at 37 °C (substrate concentration: 15–20 mM
in 100 mM triethanolamine buffer, pH 7, in a
final assay volume of 1.25 mL; enzyme
dosage: 0.7 U/mL maltase activity for SI and
GM, determined by detecting the rate of liber-
ated Glc using the hexokinase/Glc-6-P dehy-
drogenase method [28]; one unit of activity is
defined as the rate of hydrolysis of 1 mmol of
substrate per minute at 37 °C at saturating
substrate concentration. Samples of the reac-
tion mixture (0.12 mL) were removed after 0,
1, 2, 3, 4, 6, 8, 16 and 24 h, and the reaction
was stopped by heating at 95 °C for 2 min. To
determine the rate of hydrolysis, the concen-
tration of liberated Glc was measured at room
temperature by the hexokinase/Glc-6-P dehy-
drogenase endpoint method [28]. Further-
more, substrates and all products of enzymatic
[4] A. Quaroni, E. Gershon-Quaroni, G. Semenza, Eur. J.
Biochem., 52 (1975) 481–486.
[5] H.A. Kapadia, S. Sivakami, Ind. J. Biochem. Biophys., 27
(1990) 93–97.
[6] I.R. Rodriguez, F.R. Taravel, W.J. Whelan, Eur. J.
Biochem., 143 (1984) 575–582.
[7] H. Heymann, Isolierung und Charakterisierung 6on Di-
und Oligosacchari-dasen aus Du¨nndarmmukosa, Disserta-
tion, Universita¨t Hannover, 1991.
[8] O. Noren, H. Sjo¨stro¨m, G.M. Cowell, J. Tranum-Jensen,
O.C. Hansen, K.G. Wehlider, J. Biol. Chem., 261 (1986)
12306–12309.
[9] J.A. Thoma, G.V.K. Rao, C. Brothers, J. Spradlin, J.
Biol. Chem., 246 (1971) 5621–5635.
[10] K. Hiromi, Y. Nitta, C. Numata, S. Ono, Biochim.
Biophys. Acta, 302 (1973) 362–375.
[11] K. Hiromi, M. Ohnishi, A. Tanaka, Mol. Cell. Biochem.,
51 (1983) 79–95.
[12] R. Fagerstro¨m, J. Gen. Microbiol., 137 (1991) 1001–
1008.
[13] H. Heymann, S. Gu¨nther, Biol. Chem. Hoppe-Seyler, 375
(1994) 451–455.
[14] S. Hertel, F. Heinz, Food Ingredients Europe. Conference
Proceedings, Miller Freemann Plc, Maarsen, The
Netherlands, 1997, pp. 89–93.

276
S. Hertel et al. / Carbohydrate Research 326 (2000) 264–276
[15] U. Grupp, G. Siebert, Res. Exp. Med., 173 (1978) 261–
[22] A. Cogoli, H. Mosiman, C. Vock, A.-K. von Baltharzar,
G. Semenza, Eur. J. Biochem., 30 (1972) 7–14.
[23] J. Kolinska, J. Kraml, Biochim. Biophys. Acta, 284
(1972) 235–247.
278.
[16] H.J. Lindner, F.W. Lichtenthaler, Carbohydr. Res., 93
(1981) 135–140.
[17] S. Immel, F.W. Lichtenthaler, Justus Liebigs Ann. Chem.,
(1995) 1925–1937.
[18] M. Munir, B. Schneider, H. Schiweck, Carbohydr. Res.,
164 (1987) 477–485.
[24] L.M.J. Lee, A.K. Salvatore, P.R. Flanagan, G.G.
Forstner, Biochem. J., 187 (1980) 437–446.
[25] M.M. Bradford, Anal. Biochem., 72 (1976) 245–
248.
[19] F.W. Lichtenthaler, H.J. Lindner, Justus Liebigs Ann.
Chem., (1981) 2371–2382.
[20] G.A. Jeffrey, H.S. Kim, Carbohydr. Res., 14 (1970) 207–
216.
[21] A. Cogoli, G. Semenza, J. Biol. Chem., 250 (1975) 7802–
7809.
[26] B.J. Davis, Ann. NY Acad. Sci., 121 (1964) 404–427.
[27] U.K. Laemmli, Nature, 227 (1970) 680–685.
[28] A. Kunst, B. Draeger, J. Ziegenhorn, in H.U.
Bergmeyer, J. Bergmeyer, M. Graßl (Eds.), Methods of
Enzymatic Analysis, Vol. VI, third ed., VCH, Weinheim,
Germany, 1984, pp. 163–172.
.