Examination of the active sites of human salivary α-amylase (HSA)

Article abstract of DOI:10.1016/S0008-6215(00)00221-4

The action pattern of human salivary amylase (HSA) was examined by utilising as model substrates 2-chloro-4-nitrophenyl (CNP) β-glycosides of maltooligosaccharides of dp 4-8 and some 4-nitrophenyl (NP) derivatives modified at the nonreducing end with a 4,6-O-benzylidene (Bnl) group. The product pattern and cleavage frequency were investigated by product analysis using HPLC. The results revealed that the binding region in HSA is longer than five subsites usually considered in the literature and suggested the presence of at least six subsites; four glycone binding sites (-4, -3, -2, -1) and two aglycone binding sites (+1, +2). In the ideal arrangement, the six subsites are filled by a glucosyl unit and the release of maltotetraose (G₄) from the nonreducing end is dominant. The benzylidene group was also recognisable by subsites (-3) and (-4). The binding modes of the benzylidene derivatives indicated a favourable interaction between the Bnl group and subsite (-3) and an unfavourable one with subsite (-4). Thus, subsite (-4) must be more hydrophylic than hydrophobic. As compared with the action of porcine pancreatic α-amylase (PPA) on the same substrates, the results showed differences in the three-dimensional structure of active sites of HSA and PPA. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00221-4

www.elsevier.nl/locate/carres
Carbohydrate Research 329 (2000) 579–585
Examination of the active sites of human salivary
a-amylase (HSA)
Lili Kandra*, Gyo¨ngyi Gye´ma´nt
Uni6ersity of Debrecen, Institute of Biochemistry, PO Box 55, H-4010 Debrecen, Hungary
Received 9 June 2000; accepted 24 July 2000
Dedicated to Professor Andra´s Lipta´k on the occasion of his 65th birthday
Abstract
The action pattern of human salivary amylase (HSA) was examined by utilising as model substrates 2-chloro-4-ni-
trophenyl (CNP) b-glycosides of maltooligosaccharides of dp 4–8 and some 4-nitrophenyl (NP) derivatives modified
at the nonreducing end with a 4,6-O-benzylidene (Bnl) group. The product pattern and cleavage frequency were
investigated by product analysis using HPLC. The results revealed that the binding region in HSA is longer than five
subsites usually considered in the literature and suggested the presence of at least six subsites; four glycone binding
sites (−4, −3, −2, −1) and two aglycone binding sites (+1, +2). In the ideal arrangement, the six subsites are
filled by a glucosyl unit and the release of maltotetraose (G4) from the nonreducing end is dominant. The benzylidene
group was also recognisable by subsites (−3) and (−4). The binding modes of the benzylidene derivatives indicated
a favourable interaction between the Bnl group and subsite (−3) and an unfavourable one with subsite (−4). Thus,
subsite (−4) must be more hydrophylic than hydrophobic. As compared with the action of porcine pancreatic
a-amylase (PPA) on the same substrates, the results showed differences in the three-dimensional structure of active
sites of HSA and PPA. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Human salivary a-amylase; Substrate specificity; Maltooligosides; HPLC separation
1. Introduction
Clinical importance.—a-Amylase
the tissue of origin. The pancreas and salivary
glands have amylase concentrations that are
several orders of magnitude greater than that
of other normal tissue, and these two organs
probably account for almost all of the serum
amylase activity in normal persons. Measuring
a-amylase activity in serum, urine, saliva and
other biological fluids is a useful diagnostic
tool in evaluating diseases of the pancreas and
salivary glands (e.g., acute pancreatitis, paroti-
tis) [1,2].
(EC
3.2.1.1) is one of the major secretory products
of the pancreas and salivary glands in hu-
mans, playing a role in digestion of starch and
glycogen. Human a-amylases of both salivary
and pancreatic origin (HSA and HPA, respec-
tively) have been extensively studied from the
view point of clinical chemistry because they
are important as indicators of dysfunction of
A variety of methods are now available that
are based on different principles using such
substrates as limit dextrins [3], modified starch
[4], maltooligosaccharides [5,6], or mal-
tooligosaccharide derivatives containing 4-
* Corresponding author. Tel.: +36-52-512900/2256; fax:
+36-52-512913.
E-mail address: kandra@tigris.klte.hu (L. Kandra).
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00221-4

580
L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
nitrophenyl (NP) [7–10], or 2-chloro-4-nitro-
phenyl (CNP) [11–13] as aglycon groups.
Some blocked substrates carrying a modified
residue at the nonreducing end have also been
reported, and it has been shown that such
groups as ethylidene [14], 3-oxobutylidene
[15], benzylidene [16], benzyl [17], and galacto-
pyranosyl [18] protect the substrates from hy-
drolysis by exo-glucosidases, which are used
as coupling enzymes in the a-amylase kits.
The method for a-amylase, with 2-chloro-4-ni-
trophenyl-a-maltotrioside as substrate, re-
quires no helper enzymes and is suitable for
routine assay of clinical specimens [19].
envisaged as good candidates for further stud-
ies of the action pattern of HSA. Compared
with other substrate series so far reported, for
example, maltooligosaccharides [22] or a-NP-
maltooligosaccharides [23], the CNP- and Bnl-
NP-maltooligosaccharides are b-glycosides.
Their preparation and their use in the map-
ping of the active centre of PPA was first
reported by our laboratory [24]. The b linkage
is stable and is not hydrolysed by a-amylases.
In addition, the 4,6-O-benzylidene-modified
oligomers are very stable towards hydrolysis
by exo-glucosidases and are useful for moni-
toring the digestion products modified at the
nonreducing end.
Present studies were aimed at determining
the relationship between the action pattern of
human salivary amylase (HSA) and the mode
of binding of the CNP and Bnl-NP glycoside
series to HSA by product analysis using an
HPLC method. In addition, the effect of agly-
cons and the role of the benzylidene group at
the nonreducing end on the action pattern of
HSA were examined. These results are dis-
cussed on the basis of the subsite concept,
introduced by Phillips and his co-workers in
the study of lysozyme and Schechter and
Berger in their work on proteinase papain
[25].
In the course of our studies of convenient
substrates for a-amylase, 2-chloro-4-nitro-
phenyl (CNP) and 4,6-O-benzylidene-modified
4-nitrophenyl (Bnl-NP) b-maltooligosaccha-
rides (dp 4–8) were synthesised from cy-
clodextrins using a chemical procedure [20].
For the preparation of short chain length
CNP-maltooligosides in the range dp 3–6, a
new chemoenzymatic procedure was devel-
oped using rabbit skeletal muscle glycogen
phosphorylase b [21]. Fig. 1 shows the struc-
ture of this substrate series.
Enzymological
significance.—The
ho-
mologous maltooligosaccharide substrates are
indispensable tools in the investigation of the
binding site and the action of different depoly-
merising enzymes. In these studies well
defined, high purity, low-molecular-weight
substrates are preferred because purity and
their reaction patterns can be exactly deter-
mined. Therefore, our substrate series were
2. Results
Action pattern and clea6age frequencies of
HSA on CNP-maltooligosaccharides.—A se-
ries of CNP-maltooligosaccharides (dp 4–9)
was used in the HSA reaction to determine
unambiguously the exact glycosidic linkage
being cleaved, as well as the cleavage fre-
quency tested by high-performance liquid
chromatography (HPLC) on an aminopropyl-
silica (NH2) column. Only the chromogenic
reaction products could be detected with a
diode array detector (DAD) at 302 nm, and
therefore these compounds were quantified.
The concentrations of glycosidic fragments
produced by amylase reaction showed linear-
ity with the reaction time, and the distribution
of products was calculated for a given sub-
strate. Reproducible values were obtained at
four incubation times; and the mean values
are given in Table 1.
Fig. 1. Structure of the substrates of HSA.

L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
581
Table 1
Yields a of products from the hydrolysis of CNP-maltooligosides by HSA
Substrate
Products (mol/mol of 2-chloro-4-nitrophenyl products) b
G1-CNP
G2-CNP
G3-CNP
G4-CNP
G5-CNP
G6-CNP
G4-CNP
G5-CNP
G6-CNP
G7-CNP
G8-CNP
G9-CNP
0.10
0.02
0.85
0.86
0.51
0.18
0.16
0.17
0.05
0.12
0.44
0.50
0.27
0.19
0.05
0.32
0.41
0.26
0.16
0.30
0.08
a The yields given are an average of at least four determinations.
b Experimental error is 95% of the values given.
HSA exhibits an interesting pattern of ac-
tion on CNP maltooligosaccharides by cleav-
ing the maltose (85%) and maltotriose (86%)
units as the main products from the nonreduc-
ing ends of CNP-G4 and CNP-G5, respec-
tively, and leaving CNP-G2 glycosides. In the
case of the hexamer glycoside, the maximum
frequency of attack shifts from the nonreduc-
ing end to the inside of the chain and the
maltotetraose (G4) becomes the major
product. The cleavage of the maltotetraose
from the nonreducing ends of the oligomers
remained favourable in the case of the hydrol-
ysis of the heptamer, octamer and nonamer
glycosides also, resulting in 50, 41 and 30% of
maltotetraose, respectively.
Patterns of action and clea6age frequencies
of HSA on the 4,6-O-benzylidene-blocked NP-
maltooligosaccharides (Bnl-NP).—A series of
4-nitrophenyl-b-maltooligosaccharides (dp 4–
8) blocked at their nonreducing end with a
4,6-O-benzylidene group was hydrolysed with
HSA. Only the 4-nitrophenyl glycosides
formed were quantified, and the results are
given in Table 2 as an average of at least four
determinations. The results show that the
product patterns of CNP-glycosides and the
These interesting and unexpected results
will be discussed later in Section 3.
3. Discussion
Human a-amylases are divided into two
groups by their site of biosynthesis. One group
is the pancreatic a-amylase and the other is
the salivary a-amylase. Their amino acid se-
quences have been deduced from the nucle-
otide sequences of cDNAs [26], and that of
HSA was also determined from the cloned
gene [27]. These sequences are quite similar,
with 97% homology between the two a-amy-
lases. Each contains several isozymes, which
have been separated by isoelectric focusing
and polyacrylamide gel electrophoresis [28].
Human saliva contains more than five a-amy-
lase isozymes [29]. There are some differences
in the physico-chemical properties of these
isozymes, but their active sites may be
identical.
Table 2
Yields of products from the hydrolysis of 4,6-O-benzylidene-
protected NP-maltooligosides by HSA
Substrate
Products
benzylidene modified NP-glycosides differ
significantly.
(mol/mol of 4-nitrophenyl products) a
Only one prominent glycosidic fragment,
G2-NP was released from the tetramer gly-
coside. From the pentamer two (G2-NP and
G3-NP) and from the hexamer three (G2-NP,
G3-NP and G4-NP) glycosidic products were
formed. Four cleavages occur in the longer
benzylidene blocked substrates, yielding G2-
NP, G3-NP, G4-NP and G5-NP glycosides.
G2-NP
G3-NP
G4-NP
G5-NP
BnlG4-NP
BnlG5-NP
BnlG6-NP
BnlG7-NP
BnlG8-NP
1.00
0.17
0.62
0.67
0.32
0.83
0.13
0.12
0.27
0.25
0.11
0.22
0.10
0.19
a Experimental error is 95% of the values given.

582
L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
binding sites [33], there is no third subsite for
the aglycone binding site [34].
We have studied the active site of HSA by
examination of its modes of action on CNP
b-maltooligosides (Fig. 2) and modified NP
b-maltooligosides, which had 4,6-O-benzyli-
dene groups at their nonreducing ends (Fig.
3).
It has been found that all CNP b-mal-
tooligosaccharide substrates have more than
one bond cleaved; three cleavages occur in the
case of CNP-G4, CNP-G5, CNP-G6 and CNP-
G7 oligomers and four glycosidic bonds were
hydrolysed in CNP-G8. For the shorter sub-
strates (G4 and G5) the CNP moiety would be
equivalent to a glucose residue, as was found
for PPA in our previous work [21]. In these
cases the aglycons (CNP), which were in b-
glycosidic linkages, could interact with subsite
(+2), but less favourably than a glucose
residue. In the ideal arrangement subsite (+2)
was filled by a glycopyranosyl unit and the
hydrolysis resulted in the prominent forma-
tion of CNP-G2 (85% from CNP-G4 and 86%
Fig. 2. Schematic representation of CNP-glycosides binding
to subsites of HSA. Bond cleavage frequencies are expressed
as percentages of total cleavage events calculated from the
reaction products. G, glucosyl residues; CNP, 2-chloro-4-ni-
trophenyl groups (connected to the reducing end in b form);
(— ) linkages; (¡) catalytic site.
The active sites of depolymerases and espe-
cially of such endoglycanases as HSA are
estimated to be composed of tandem subsites
geometrically complementary to several glu-
cose residues. To examine the active site of an
enzyme, X-ray crystallographic analysis of the
complex of an enzyme and its substrate ana-
logue is a powerful method. Many structural
studies of PPA have been described, both free
and in complex with oligosaccharides [30]. In
addition there are structures of the more rele-
vant HPA and its complexes with analogues
[31]. However, there is only an X-ray analysis
of the human salivary amylase [32] performed
in the absence of oligosaccharides. The use of
modified, low-molecular-weight substrates
should therefore be an effective way to eluci-
date the number of subsites in the active site
area of HSA.
Fig. 3. Schematic representation of benzylidene-modified NP-
glycosides binding to subsites of HSA and PPA. Bond cleav-
age frequencies are expressed as percentages of total cleavage
events calculated from the reaction products. G, glucosyl
residues; NP, 4-nitrophenyl groups (connected to the reducing
end in b form); Bnl, benzylidene groups; (—) linkages; (¡)
catalytic site.
From enzymological studies, human sali-
vary amylase is believed to have five subsites,
three glycone binding sites and two aglycone

L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
583
from CNP-G5). This favourable release of
subsite (−4).
CNP-G2 was also observed for the hexamer
glycoside (CNP-G6), where four glycosyl
residues from the nonreducing end were
bound in subsites (−4, −3, −2, −1) and
led to the formation of maltotetraose in a 51%
yield. In the case of the longer substrates
(CNP-G7, CNP-G8 and CNP-G9) release of
maltotetraose from the nonreducing ends of
the substrates remains predominant resulting
in 50, 41 and 30% of CNP-G3, CNP-G4 and
CNP-G5, respectively.
In an ideal arrangement two glucose
molecules are fixed in the aglycone binding
sites (subsites +1 and +2) of the active
center and the BnlG2 residue is fixed in the
glycone binding sites (subsites −3, −2 and
−1). However, the action patterns of the
non-modified CNP-glycosides suggest that
there is a fourth subsite on the glycone bind-
ing site (subsite −4), which has a higher
affinity for the binding of a glycopyranosyl
unit than that of benzylidene group. There-
fore, the subsite (−4) must be rather more
hydrophylic than hydrophobic.
Our results provide evidence that the bind-
ing region in the HSA molecule is longer than
five contiguous subsites usually considered in
the literature. It is suggested from our studies
with HSA on modified maltooligosaccharides
that HSA has at least six subsites (−4, −3,
−2, −1, +1, +2) like PPA [30] and that
the glycosidic bond split by this enzyme is
between subsite (−1) and subsite (+1). On
the other hand, our results are in general
agreement with those obtained for the human
pancreatic amylase [31], since a total of five
subsites and an additional glycone subsite was
proposed in that case. However, the binding
modes of the benzylidene derivatives to the
active site of salivary a-amylase were found to
be different from those of porcine pancreatic
a-amylase (Fig. 3). This can be explained by
the slight difference in their primary and crys-
tal structure. Namely, the amino acid 163
located in the vicinity of subsite (−3) is valine
in the case of porcine pancreatic amylase [30],
but it is serine for human salivary amylase
[32]. This change in polarity might result in
differences in the action pattern of substrates,
especially if the benzylidene group is located
at position (−4).
Our results show that the full amylase activ-
ity could be obtained when the six subsites
(−4, −3, −2, −1, +1, +2) were occupied
and these experiments are consistent with a
six-subsite model, which had been proposed
for PPA by Gilles and co-workers in 1996 [30].
The number of binding modes of 4,6-O-
benzylidene protected oligomers was different
from that of the unmodified ones. It was
envisaged that the 4,6-O-benzylidene group
would not mimic a glycopyranosyl unit, but
we found that it was recognisable by subsites
(−3) and (−4). The hydrolysis product of
BnlG4-NP was 100% BnlG2, and that from
BnlG5-NP was 83%. These suggest a very
favourable interaction between the benzyli-
dene residue and subsite (−3). It is important
to mention that the same observations were
made by Nagamine and co-workers [35,36] on
modified phenyl a-maltooligosides protected
with iodo or 2-pyridylamino groups at their
nonreducing ends.
As the chain length increases, the concen-
tration of the main product (BnlG2) decreases;
25, 10 and 0% for the hexamer, heptamer and
octamer glycosides, respectively. Surprisingly,
there was a low affinity between the 4,6-O-
benzylidene group and subsite (−4). Forma-
tion of BnlG3 shows a low cleavage frequency
for this hydrolysis product; 0, 17, 23, 11 and
19% for the BnlG4-NP, BnlG5-NP, BnlG6-NP,
BnlG7-NP and BnlG8-NP, respectively. On the
other hand, formation of G2-NP from the
reducing ends of glycosides shows a high
cleavage frequency for this hydrolysis product;
100, 17, 62, 67 and 32% for BnlG4-NP, BnlG5-
NP, BnlG6-NP, BnlG7-NP, respectively. The
relatively low release of G2-NP (17%) from the
pentamer glycoside (BnlG5-NP) corresponds
to a low affinity between the Bnl group and
Benzylidene derivatives may also elucidate
the differences between the active sites of HSA
and HPA rather than other modified
substrates.
4. Experimental
Substrates.—The
tooligomer substrate series 1, 2 (Fig. 1) (dp
homologous
mal-

584
L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
Fig. 4. Separation of CNP-oligomers in the range dp 2–8.
4–8) were synthesised from cyclodextrins by
the method of Lipta´k et al. [20]. The shorter
chain length of CNP-maltooligosides in the
range dp 3–6 were prepared by a chemoenzy-
matic procedure, as well as using rabbit skele-
tal muscle glycogen phosphorylase b [24].
Enzyme.—a-Amylase (EC 3.2.1.1) from hu-
man saliva (Type IX A) was purchased from
Sigma (Chem. Co. Ltd.) and analysed without
separation of the isozymes.
group at 302 nm and the products of the
hydrolysis were identified by using relevant
standards. The quality of the MeCN was gra-
dient grade. The purified water was obtained
from
a
laboratory purification system
equipped with both ion-exchange and carbon
filters (Millipore, Bedford, MA, USA). Fig. 4
shows the separation of CNP-oligomers in the
range dp 2–8.
Hydrolysis of the maltooligosides.—Incuba-
tions in 25 mM glycerophosphate buffer (pH
7.0) containing 5 mM CaCl2 and 50 mM NaCl
were carried out at 37 °C for 5, 7 and 10 min.
The reactions were initiated by the addition of
10 mM of enzyme to the digest containing 1.7
mM of substrate. Samples were taken at the
indicated time intervals and the reaction was
stopped by the injection of the samples into
the chromatographic column. In these studies
we have taken care to exclude the secondary
attacks on the substrates. The product ratios
were always obtained from the early stages of
hydrolysis (conversionB10%), before any sec-
ondary attack could be detected.
Chromatographic analysis.—For HPLC a
Hewlett–Packard 1090 Series II liquid chro-
matograph equipped with a diode array detec-
tor, automatic sampler, and ChemStation was
used. The samples were separated on Supelco
NH2 5 mm column (20×0.46 cm) with differ-
ent ratios of MeCN–water as the mobile
phase flowing at a rate of 1 mL/min at 40 °C.
The effluent was monitored for the NP/CNP
Acknowledgements
This work was supported by a grant from
the Hungarian Scientific Research Fund
(OTKA T 032005).
References
[1] N.J. Greenberger, P.P. Toskes, Harrison’s Principles of
Internal Medicine, Vol. 2, 12th ed., McGraw-Hill, New
York, 1991, p. 1369.
[2] W.B. Salt, S. Schenker, Medicine, 55 (1976) 269–289.
[3] P.T. Nix, R.D. Goldfarb, S. Morgenstern, L.J. Strong,
R.C. Trivedi, Clin. Chem., 24 (1978) 1000.
[4] J.J. Marshall, A.P. Iodice, W.J. Whelan, Clin. Chim.
Acta, 76 (1977) 277–283.
[5] K.J. Whitelow, N. Gochman, R.L. Forrester, L.J. Wataji,
Clin. Chem., 25 (1979) 481–483.
[6] E.O. Haegele, E. Schaich, E. Rauscher, P. Lehmann, N.
Grassl, J. Chromatogr., 223 (1981) 69–84.
[7] R.A. Kaufman, L.J. Dunka, L.M. Hall, Clin. Chem., 26
(1980) 1018.
[8] R. McCroskey, T. Chang, H. David, E. Winn, Clin.
Chem., 28 (1982) 1787–1791.

L. Kandra, G. Gye´ma´nt / Carbohydrate Research 329 (2000) 579–585
585
[23] E.H. Ajandouz, G.J. Marchis-Mouren, Carbohydr. Res.,
268 (1995) 267–277.
[24] L. Kandra, G. Gye´ma´nt, A. Lipta´k, Carbohydr. Res., 298
(1997) 237–242.
[25] (a) D.C. Phillips, Sci. Am., 215 (1966) 78–90. (b) I.
Schechter, A. Berger, Biochem. Biophys. Res. Commun.,
27 (1967) 157–162.
[26] Y. Nakamura, M. Ogawa, T. Nishide, M. Emi, G.
Kosaki, S. Himeno, K. Matsubara, Gene, 28 (1984) 263–
270.
[27] N. Nishide, Y. Nakamura, M. Emi, T. Yamamoto, M.
Ogawa, T. Mori, K. Matsubara, Gene, 41 (1986) 299–
304.
[28] P.J. Keller, D.L. Kauffman, B.J. Allan, B.L. Williams,
Biochemistry, 10 (1971) 4867–4874.
[29] Y. Muus, J.M. Vnenchak, Nature, 204 (1964) 283–285.
[30] C. Gilles, J.P. Astier, G. Marchis-Mouren, C. Cambillau,
F. Payan, Eur. J. Biochem., 238 (1996) 561–569.
[31] G.D. Brayer, G. Sidhu, R. Maurus, E.H. Rydberg, C.
Braun, Y. Wang, N.T. Nguyen, C.M. Overall, S.G. With-
ers, Biochemistry, 39 (2000) 4778–4791.
[32] N. Ramasubbu, V. Paloth, Y.G. Luo, G.D. Brayer, M.J.
Levine, Acta Crystallogr., Sect. D, 52, 435–446.
[33] K. Omichi, S. Hase, T. Ikenaka, J. Biochem., 109 (3)
(1991) 410–415.
[9] E.O. Haegele, E. Schaich, E. Rauscher, P. Lehmann, H.
Bu¨rk, A.W. Wahlefeld, Clin. Chem., 28 (1982) 2201–
2205.
[10] A.W. MacGregor, J.E. Morgan, E.A. MacGregor, Car-
bohydr. Res., 227 (1992) 301–313.
[11] E. Henkel, S. Morich, R. Henkel, J. Clin. Chem. Clin.
Biochem., 22 (1984) 489–495.
[12] J.C.M. Hafkenscheid, M. Hessels, J. Clin. Chem. Clin.
Biochem., 23 (1985) 529–533.
[13] S. Teshima, N. Mitsuhida, M. Ando, Clin. Chim. Acta,
150 (1985) 165–174.
[14] E.O. Haegele, M. Kratzer, E. Schaich, E. Rauscher, Clin.
Chem., 35 (1989) 188–199.
[15] S. Teshima, Y. Hayashi, S. Emi, K. Ishimaru, Clin. Chim.
Acta, 199 (1991) 23–32.
[16] G. Dupuy, G. Hilaire, C. Aubry, Clin. Chem., 33/4 (1987)
524–528.
[17] S. Satomura, Y. Sakata, K. Omichi, T. Ikenaka, Clin.
Chim. Acta, 174 (1988) 315–324.
[18] K. Majima, S. Teshima, Y. Hamada, T. Kikuchi, Y.
Kawamura, S. Kitahata, Clin. Chim. Acta, 234 (1995)
177–179.
[19] E.S. Winn-Deen, H. David, G. Sigler, R. Chavez, Clin.
Chem., 34/10 (1988) 2005–2008.
[20] E. Farkas, L. Ja´nossy, J. Harangi, L. Kandra, A. Lipta´k,
Carbohydr. Res., 303 (1997) 407–415.
[34] K. Omichi, S. Hase, T. Ikenaka, J. Biochem., 111 (1992)
4–7.
[35] Y. Nagamine, K. Omichi, T. Ikenaka, J. Biochem., 104
(1988) 667–670.
[21] L. Kandra, G. Gye´ma´nt, A. Lipta´k, Carbohydr. Res., 315
(1999) 180–186.
[22] J.F. Robyt, D. French, J. Biol. Chem., 245 (1970) 3917–
3918.
[36] K. Omichi, T. Ikenaka, J. Biochem., 99 (1986) 1245–
1252.
.