Bioscience, Biotechnology, and Biochemistry

ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: https://www.tandfonline.com/loi/tbbb20

Purification of Amylases and Other Enzymes by a

Forced-affinity Chromatography Method

Mikihiko Kobayashi, Yutaka Sasaki & Shoichi Kobayashi

To cite this article: Mikihiko Kobayashi, Yutaka Sasaki & Shoichi Kobayashi (1997) Purification
of Amylases and Other Enzymes by a Forced-affinity Chromatography Method, Bioscience,
Biotechnology, and Biochemistry, 61:5, 813-816, DOI: 10.1271/bbb.61.813

To link to this article: https://doi.org/10.1271/bbb.61.813

Published online: 12 Jun 2014.

Submit your article to this journal

Article views: 914

View related articles

Citing articles: 2 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tbbb20
Biosci. Biotech. Biochel11., 61 (5),813-816, 1997

Purification of Amylases and Other Enzymes by a Forced-affinity Chromatography

Method
Mikihiko KOBAYASHI, Yutaka SASAKI, and Shoichi KOBAYASHI
National Food Research Institute, Kannondai, Tsukuba 305, Japan
Received October 22, 1996

An affinity matrix of soluble starch gel was prepared by cross-linking catalyzed by epichlorohydrin.
The elution pattern of Taka-amylase A (TAA) indicated that the amount of enzyme bound to the starch
gel column increased with increases in the ammonium sulfate (AmS) concentration in the equilibrating
buffer. T AA had an affinity for the gels with a starch structure, and desorbed from the column with the
buffer containing no AmS. Bound T AA was also eluted with starch and cyclodextrin solution. The AmS
stimulative effect was partially replaced by polyethylene glycol and surfactants. Besides T AA, various other
amylases bound satisfactorily to the starch gel. Moreover, affinity purifications of dextranase, cellulase,
and pectinase were done by gels with dextran, cellulose, and pectin structures, respectively. By the aid of
forced effects of AmS, various carhohydrases could be purified by the affinity gels of polysaccharide linked
by epichlorohydrin.

Key words: affinity chromatography; amylases; forced-affinity chromatography; glucanases

Affinity chromatography provides effective purification
of biologically active molecules including enzymes. 1.2) Since
this method is based on biological recognition, introduction
of key compounds such as the enzyme substrates, products,
or cofactors into the gel matrix of affinity supports greatly
affects on the efficiency of chromatography. A large number
of synthetic methods for introduction of ligands to the gel
have been reported.
Interaction between enzyme and substrate has been used
for the batch-wise purification of carbohydrases such as
potato a-glucan phosphorylase 3 ) and cyclodextrin (CD)
glucanotransferase,4) where the enzymes were adsorbed on
raw starch and then recovered by appropriate methods. The
batch-wise purifications would be happened on difficulty in
filtration and washing steps and lead to low recovery of
enzymes.
Although affinity chromatography purification over-
comes these drawbacks of the batch method, preparation
of affinity gels is rather expensive and tedious, and
sometimes necessitates the use of hazardous reagent such
as CNBr. Therefore, in the case of carbohydrase purifica-
tion, use of polysaccharide gels would provide less-expensive
affinity gels in large quantity. Extensive purification of
concanavalin A was done with Sephadex G-IOO gel,5)
which was prepared by the cross-linking of dextran with
epichlorohydrin. Moreover, affinity chromatography of
a-l,3-g1ucanase 6 ) and :x-l)-debranching enzyme on Sepha-
dex G-lS0 7l have been reported. Dextransucrase bound
tightly to the DEAE-Sephadex gel and was desorbed
by the gradient elution with 0-4 M guanidine-HCI after
washing with 0-2 M NaCl. 8) These examples illustrated
that polysaccharide gels served as the affinity matrix and
afforded highly purified enzyme preparations.
In spite of the sufficient preparation of affinity columns,
we often encounter to an unexpected result of no binding
of our target enzyme. To stimulate the binding of a-amylase
to starch granules, addition of AmS (ammonium sulfate),
sodium nitrate, and ethanol were already notified to be
effective. For microbial f3-amylase adsorption on starch 9 )
and affinity purification of plant f3-amylases on a-CD-
Sepharose 6B columns,10) use of AmS gave a favourable
binding of enzyme to the matrix.
Therefore, we focused on the convenient preparation of
affinity gel matrix by cross-linking of starch and some other
polysaccharides with epichlorohydrin. Evaluation of
polysaccharide gels as the affinity matrix was done by
monitoring the elution profiles of various types of
carbohydrases including amylases. In these experiments,
AmS played an important role in stimulation of the bind-
ing of enzymes to the gel, and we call this AmS mediated
affinity system, forced-affinity chromatography. Although
we presented only a limited amount of results on the
carbohydrases, the forded-affinity system could be applica-
ble to a wide variety of enzymes concerning the metabolism
of carbohydrates.
Materials and Methods
Materials. Soluble starch (~1252, Merck), citrus pectin (164-00552),
oc-amylase from Bacillus subtilis (015-03731), J1-amylase from wheat
(012-13751), and pectinase from a mold (Fluka 76290) were purchased
from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). CGTase
(cyclodextrin glucanotransferase), isoamylase, and pullulanase were the
products of Hayashibara Biochemical Inc. Glucoamylase (AF 6), cellulase
A, and cellulase T were the products of Amano Pharmaceutical Co.
Taka-amylase A (TAA) and endodextranase from Chaetomium gracile
were the products of Sankyo Co. Meicelase was donated by Meiji Seika
Kaisha Ltd. :x-Amylases from porcine pancreas (Type VI-B), and B.
amyloliquefacience were the products of Sigma Co. and Seikagaku Kogyo
Co., respectively.
Preparation 4 soluble starch gel. Soluble starch (5 g) was dissolved in
5 M NaOH solution (3 ml) and then diluted with water (3 ml). To this
reaction mixture, epichlorohydrin (2 m!) was added and vigorously mixed
in the water-bath at about 90°C for 10min. The resulting gel block was
kept at room temperature overnight. The gel was homogenized with a
mortar, washed with water, and suction dried. The particulate starch gel
was stored at 4°C. Raw potato starch, carboxymethyl-starch, and guar

Abbreviations: TAA, Taka-amylase A; AmS, ammonium sulfate; CGTase, cyclodextrin glucanotransferase: PEG, polyethylene glycol.

NII-Electronic Library Service

814 M. KOBAYASHI, Y. SASAKI, and S. KOBAYASHI
gum gels were prepared in the same ways.
Preparation of pectin gel was done by a different method. Citrus pectin
(1 g) was dissolved in water (20 ml) and mixed with water-soluble
carbodiimide (EDC, l-ethyl-3-(3-dimethylaminopropyl) carbodiimide
I g/30 ml of water). The pH of reaction mixture was maintained at pH
4.75 with 1M Hel. Hexamethylenediamine (1 g/5ml water) was adjusted
to pH 4.75 with HCl and added to this reaction mixture to maintain the
pH of the reaction mixture. The reaction proceeded at pH 4.75 and room
temperature with stirring for 1 h. A portion of the resulting loose gel
(4 ml) was copolymerized with the polyacrylamide gel stock solutions (4 ml)
for preparing 3.75% acrylamide gel. Then, the hard gel obtained was
homogenized, thoroughly washed with water, and suction dried.
F'orced-affinity chromatography. For a convenient binding measurement,
soluble starch gel (2 ml volume) was packed into a disposable plastic syringe
(3 ml volume) and equilibrated with 40 mM acetate buffer (pH 5.2)
containing IMAmS (ammonium sulfate). Each 200111 of enzyme solution
dissolved in the above acetate buffer-AmS was put on the column and
eluted with the buffer-AmS solution. Fractions (1 ~3ml/tube) were collected
and then the enzyme protein adsorbed on the affinity column was recovered
by elution with plain acetate buffer.
Assay ofen:::yme activity. Amylase activity was measured by incubating
suitably diluted enzyme (25 JlI) with 1% soluble starch in 40 mM acetate
buffer (pH 5.2, 25 Ill) at 30"C for an appropriate period. Reducing sugar
was measured by the Nelson-Somogyi method as described previously.6)
The protein concentration was measured at 280 mn. Assays of puIIulanase,
dextranase, cellulase, and pectinase were done essentially by the same
procedure, using the respective substrate,
Results
Forced-affinity chromatography of T AA
Soluble starch is a good substrate for :x-amylases. T AA
showed high affinity to this substrate, which was
demonstrated by using affinity gel electrophoresis. How-
ever, no binding of TAA to the soluble starch gel, which
was prepared by the cross-linking of soluble starch by
epichlorohydrin, was observed as shown in Fig. la. Addition
of AmS in the buffer markedly stimulated the binding of
TAA to starch gel and the amount of TAA bound to the
gel was increased according to the increase in the AmS
concentration in the buffer (Fig. 1b~d). The height of pro-
tein peaks, eluted with the buffer containing no AmS, were
also increased, demonstrating the higher binding of T AA at
higher AmS concentrations. Most of the TAA activity
was absorbed on the starch gel equilibrated with 3 M AmS
in the buffer.
Specificity of T AA to affinity gels
Although this forced-affinity chromatography system was
~ a b c d
~
80 0.16
2-
>-
.~
> 0
+=i
~ 40 0.08 N
Q)
E Soluble starch
>-
N
c:
w 0
0 5 0 5 0 5 0 5
Tube No. (3 ml)
Fig. 1. Adsorption of T AA on Soluble Starch-gel.
TAA (1 mg, 40 units) dissolved in 0-3\1 AmS40m\1 acetate buffer (pH 5.2, 0.5mI)
was put on column of soluble starch-gel (2 ml), whi<.:h was equilibrated with acetate
buffer (a). IMAmS -buffer (b). 2 M AmS buffer and 3 M AmS-buffer (d),
respectIvely, 'rhe column was eluted with thIS butfer then eluted WIth the butTer
containing no AmS (indicated by arrov,,).
mediated by AmS, recognition of substrate by T AA was
essential for the affinity binding. As shown in Table I,
binding of T AA was closely related to the structure of gel
matrixes. Commercial gels of Toyopearl HW-55F and
Bio-Gel P-2 showed less than 20% of the binding of T AA
although 1 M AmS was included in the buffer. Similar re-
sults were obtained with Sephadex G-IO and Sepharose 6B.
Binding ofTAA to these commercial gels having non-starch
structure might be caused by a non-specific, hydrophobic
interaction as described later in Discussion. All of these
gels were composed of non-starch matrixes and T AA
seemed to distinguish these gels from starch gels. Actual-
ly, raw-potato starch gel and the carboxymethyl derivative
of starch gel gave much better binding ofTAA than those of
non-starch gels described above. Results from carboxy-
methyl starch, however, suggested that substitution of
hydroxyl groups of the starch molecule by carboxymethyl
groups reduced the binding affinity of T AA to the matrix.
Effects of reagents on TAA adsorption
T AA binding to soluble starch gel was further examined
in the presence of various reagents. Among the 11 reagents
tested, PEG (polyethylene glycol) gave rather satisfactory
results, where 31 ~67% of T AA subjected to the column
were retained (Table II). In spite of the higher binding being
attained with higher concentrations of PEG, lower recovery
of the total amount of T AA was noticed. This may suggest
that part of T AA was still retained in the column and/or
T AA activity was inhibited by the high concentration of
PEG. SMC (sucrose monocaprate) 50mM, 5% each of
Triton X-IOO and Tween 80, and 1 M urea gave more than
20% ofTAA binding with the recovery of an ordinary level.
One M Na 2 S0 4 and 33% glycerol gave only 18% binding,
and ethanol, acetone, and dioxane caused no noticeable
stimulation of T AA binding.
Effects of ligands on TAA desorption
T AA bound to the starch gel in the presence of AmS
could be readily recovered by elution with buffer contain-
ing no AmS as shown in Fig. 1. In the typical case of
affinity chromatography, the bound enzyme was eluted
with the solution containing some ligands of substrate
analogs. Therefore, the effects of ligands on the desorption
of bound T AA were examined. One-tenth M !X-cyclodextrin,
maltitol, IX-methyl glucoside, or sorbitol dissolved in 1 M
AmS were used instead of buffer elution, although (X-
Table I. Adsorption of TAA to Various Gel Matrixes"
~ TAA activity (%)
Gel matrix
eX)
Non-adsorbed Adsorbed
<t
24 76
Raw-potato starch 29 71
Carboxymethyl-starch 48 52
Toyopearl HW-55F 81 19
Bio-Gel P-2 91 9
a TAA (7.4mg, 444 units) dissolved in 1 M AmS-40mM acetate buffer
(pH 5.2) was put on a column containing 2 ml volume of each gel.
Amounts of TAA activity recovered at the break-through fractions
(non-adsorbed T AA) and subsequent buffer-eluted fractions
(adsorbed T AA) were measured.
NII-Electronic Library Service
 Affinity Purification of Amylases and Other Enzymes 815

Table n. Effects of Various Reagents on T AA Adsorption to Soluble

Starch Gel il
0.6
T AA activity
Reagent 0.4
Non-adsorbed Adsorbed Total
(%) (OA)) {unitst
- --- ---~-~-----
~------~-- 0.2
1M AmS 24 76 404 ~
5% PEG 69 31 372 0 0
IX)
II (Yo PEG 54 46 228 0.6 N
22% PEG 40 60 160 c(

33% PEG 33 67 108 0.4

80
1M Na 2 S0 4 82 18 260
50 mM Sucrose monocapratc 76 24 344
1M Urea 75 25 400 40 0.2
5% Triton X-I 00 80 20 276
5% Tween 80 63 37 304
33% Glycerol 82 18 368 505050505
10% Acetone 100 344 Tube No. (2 mI)
10% Dioxane 100 236
10% Ethanol 93 7 348 Fig. 2. Adsorption of Various Amylases on Soluble Starch-gel.
Each amylase was dissolved in 1M AmS40mM acetate buffer (pH 5.2) in 200,u1
II
Details were shown in the footnote to Table r. volume and put on a column of soluble starch gel (2ml) equilibrated with the above
/J buffer. Elution was done as described in the legend to Fig. I. (a), ex-Amylase from
Sum of the activity recovered in the non-adsorbed and adsorbed porcine pancreas, 193mg. 12 J units; (b). :x-amylase from B. amyloliquefacience, 4mg,
fractions. 255 units; (c), cellulase A preparation assayed for amylase activity, 58 mg, 98 units;
(d), cellulase T preparation assayed for amylase activity, 44 mg, 300 units; (e),
:x-amylase from B. suhtilis, 60 mg, 232 units; (f) CGTase, 60 mg, 68 units; (g),
cyc10dextrin gave better recovery of bound TAA (26%) glucoamylase AF-6, 60mg, 141 units; (h) isoamylase, 60mg, 159 units; (i) {:i-amylase
from wheat, 14 mg, 122 units; (j) pullulanase, 8 mg, 114 units.
among the 4 reagents tested, which was about one-third
of the case of buffer elution. Elution with 0.2% soluble
starch in 1 M AmS gave about 67 % recovery of bound TAA. a b c d
These results indicated that low-molecular weight sugars "i'60 0.12
x
had rather poor ability to desorb the bound T AA. E
Pllrtfication of various amylases
"-.e-
::::J

40 0.08
~

Wide applicability of the current forced-affinity system

.s; 2
N

was illustrated by various amylases besides TAA. In the ~ c(

case of a-amylase preparations from porcine pancreas (Fig.
2a), B. amyloliquefacience (Fig. 2b), and B. subtilis (Fig.
2e), more than 6OCYo of activity were bound to the column.
The amounts of protein eluted at the breakthrough fraction
were varied according to the extent of purity of the enzyme
preparations. Since two cellulase preparations containcd
high amylase activity, crude enzymes were fractionated in
the same way and assayed for the amylase activity (Fig. 2c,
d). Both amylase were recovered from the bound fraction
and contained rather small amount of proteins. CGTase
(cyc1odextrin glucanotransferase) and pullulanase seemed
totally bound to the column (Fig. 2[ j). Moreover,
g]ucoamylase, isoamylase, and I)-amylase from wheat
showed also efficient binding to the starch gel column (Fig.
2g, h, i). These results indicated that various types of
enzymes using starch as the substrate could readily bind to
the starch gel in the presence of 1 M AmS. The column used
in these experiments contained 3 mI of gel, and about 60 mg
of enzyme powder was pui on the column.
Purification of enzymes other than amylase
Owing to the convenient preparation of starch gel by the
epichlorohydrin method, affinity purification of various
types of amylases could be attained as described above.
Besides the handmade preparation of epichlorohydrin gels,
there are many commercial gels available for the purification
of various carbohydrases.
Endodextranase from C. gracile did not bind to the
«I
Q)
E 20 0.04
>-
~
W
5 o 5 o 5 o 5
Tube No. ( 3 ml )
Fig. 3. Adsorption of Dextranase, Meicelase, and Pectinase on the
Respective Substrate-gel.
(a) Dextranase from C. gracile (1 mg, 69 units) dissolved in 40 mM acetate buffer (pH
5.2) was put on a Sephadex G-IO column (2 ml) and eluted with the buffer. (b)
Dextranase was chromatographed With 1 M AmS~acetate buffer system as in Fig. 1.
(c) Meicelase (1 mg, 98 units) was subjected to the AmS~acetate buffer system.
Phospho-cellulose column was used as the amnity matrix. (d) Pectinase (2 mg, 62 units)
was subjected to the AmS~acetate buffer system. Pectin gel described in the text was
used as the affinity matrix.
Sephadex G-IO column, though more than two-thirds of
the activity was adsorbed on the column in the presence
of 1 M AmS (Fig. 3a, b). However, use of larger pore-size
gels such as G-IOO should be cautious because of the hy-
drolysis of gel by the dextranase action.
Cellulase was adsorbed on the P-cellulose column as
shown in Fig. 3c. In this case, the ion-exchange ability of
phospho-cellulose might be negligible because of the high-
ionic strength of 1 M AmS. Other types of cellulose
resin such as DEAE- and CM-types might be expected to
work as the case of P-cellulose.
Affinity purification of pectinase required somewhat
different approaches because of the difficulty of pectin gel

NII-Electronic Library Service

816 M. KOBAYASHI, Y. SASAKI, and S. KOBAYASHI
preparation by epichlorohydrin. Therefore, citrus pectin
was first activated with water-soluble carbodiimide and
cross-linked with hexamethylenediamine. The resulting
rather sol-like loose gel was then converted into a tight gel
by the aid of polyacrylamide. The elution profile of pectinase
from this doubly cross-linked pectin gel showed less than
half the amount of enzyme was recovered at the break-
through fraction (Fig. 3d).
Guar gum is known to contain about 20% of C-6
branched galactose residues. Unlike the citrus pectin, guar
gum was readily converted into a tight gel by the epichlo-
rohydrin method. An affinity column of pectinase with
guar gum gel gave similar results to that from pectin gel.
Since guar gum is composed of mannose and galactose,
mannanase might have a potent affinity to this gel.
Discussion
A convenient preparation of affinity matrix of poly-
saccharide gels by cross-linking with epichlorohydrin
seemed to outdo the ordinary affinity gel matrix when the
purification of various enzymes related to the polysaccharide
metabolism. A large amount of polysaccharide gels could
be prepared by the use of epichlorohydrin in economical,
safe, simple, and rapid ways. Soluble starch, dextran,
cellulose, and guar gum gels were prepared in the same way,
while pectin gave no rigid gel with the epichlorohydrin
reaction. Although this may be ascribed to the occurrence
of carboxyl groups, no definite evidence was obtained.
A marked stimulation of binding of enzymes to the gels
was obtained by the addition of AmS to the equilibrating
buffer. Increase in AmS concentration led to the thorough
binding of applied enzyme (Fig. 1), although hydrophobic
binding of other proteins may be stimulated by a higher
AmS concentration. An apparent decrease in specific activi-
ty of bound TAA (Fig. 1b-d) was partly ascribed to the
inhibitory effects of AmS on the activity assay system
with the Nelson-Somogyi method. Moreover, this method
required no ligand elution, where omission of AmS in the
eluant caused immediate desorption of enzyme from the
affinity gel. Tn spite of the simple and economical system
of this forced-affinity chromatography method, recognition
of substrate by enzyme defined the specificity of affinity
column. In Table I, TAA binding to the gels was demon-
strated for only the strach-based gels. Replacement of
AmS by other reagents led to rather poor binding of
T AA to the gel (Table II). Therefore, at present, AmS is
the best reagent to enforce the binding of enzyme to the
affinity gel. The low recovery of T AA activity shown in
Table II might be ascribed to the following two reasons.
(1) Part of T AA was not eluted and still remained in the
column. (2) Inactivation of T AA was done by the column
conditions such as high concentration of organic solvents
or salt.
Recently, many glucanases were shown to have a
substrate binding site besides the catalytic site. ex-Glucan
phosphorylase from rabbit muscle was the first example and
a glycogen storage (binding) site was deduced from the
X-ray crystallography. I 1) Isolation of glucan binding sites
was reported for glucosyltransferase from cariogenic bac-
teria,12) glucoamylase,13) cellulase, 14) and chitinase. 15 ) In
the case of T AA, a substrate binding peptide fragments
was isolated by the forced-affinity column method, where
T AA was digested into peptides by the acid protease and
starch binding fragments were collected (unpublished
results). In this study, SDS-gel electrophoresis showed the
occurrence of several minor protein bands in the affinity
purified amylase preparations, presumably ascribed to
protein fragments binding to the starch gel.
Acknmvledgments. This work was supported in part by a Grant-in-Aid
from the Ministry of Agriculture, Forestry, and Fisheries of Japan through
the National Project ofGlyco-Technology. We also thank to Ms. Setsuko
Ohya for her expert technical assistance.
References
1) M. Wilchek, T. Miron, and J. Kohn, Methods En:::ymol., 104, 3-55
(1984).
2) 1. Turkova, in "Affinity Chromatography," Elsevier, Amsterdam,
1978, pp. 1-223.
3) A. Kamogawa, T. Fukui, and Z. Nikuni, J. Biochem., 63, 361-369
(J 968).
4) S. Kobayashi, K. Kainuma, and S. Suzuki, Carbohydr. Res., 61,
229-238 (1978).
5) B. B. L. Agrawal and I. J. Goldstein, Biochem. J., 96, 23C (1965).
6) K. Imai, T. Kikuta, M. Kobayashi, and K. Matsuda, Agric. BioI.
Chem., 41, 1339-1346 (1977).
7) Y. Mitsuishi, M. Kobayashi, and K. Matsuda, Agric. BioI. Chem.,
43, 2283-2290 (1979).
8) K. Funane, M. Yamada, M. Shiraiwa, H. Takahara, N. Yamamoto,
E. Ichishima, and M. Kobayashi, Biosci. Biotech. Biochem., 59,
776-780 (1995).
9) M. Hoshino, Y. Hirose, K. Sano, and K. Mitsugi, Agric. Bioi. Chem.,
39,2415-2416 (1975).
10) A. Totsuka and C. Fukazawa, Prot. Express. Purijil., 4, 333-336
(1993).
11) R. 1. Fletterik and N. B. Madsen, Ann. Rev. Biochem., 49, 31-61
(1980).
12) G. Mooser and C. Wong, Infect. Immun., 56, 880-884 (1988).
13) S. Hayashida, K. Nakahara, W. Kanlayakrit, T. Hara, and Y.
Teramoto, Agric. BioI. Chem., 53, 143~149 (1989).
14) N. B. Gilkes, D. G. Kilburn, R. C. Miller, Jr., and A. J. Warren, J.
BioI. Chern., 264, 17802-17808 (1989).
15) T. Yamagami and G. Funatsu, Biosci. Biotech. Biochem., 59,
1076 -I 081 (1995).
NII-Electronic Library Service