Calcium ion-dependent increase in thermostability of dextran glucosidase from streptococcus mutans

Article abstract of DOI:10.1271/bbb.110256

Dextran glucosidase from Streptococcus mutans (SmDG), which belongs to glycoside hydrolase family 13 (GH13), hydrolyzes the non-reducing terminal glucosidic linkage of isomaltooligosaccharides and dextran. Thermal deactivation of SmDG did not follow the s

Full text of DOI:10.1271/bbb.110256

Biosci. Biotechnol. Biochem., 75 (8), 1557–1563, 2011
Calcium Ion-Dependent Increase in Thermostability
of Dextran Glucosidase from Streptococcus mutans
*
Momoko KOBAYASHI, Hironori HONDOH,y Haruhide MORI, Wataru SABURI,
Masayuki OKUYAMA, and Atsuo KIMURA
Research Faculty of Agriculture, Hokkaido University, Kita-9 Nishi-9 Kita-ku, Sapporo 060-8589, Japan
Received April 1, 2011; Accepted May 16, 2011; Online Publication, August 7, 2011
[doi:10.1271/bbb.110256]
Dextran glucosidase from Streptococcus mutans
(SmDG), which belongs to glycoside hydrolase family
13 (GH13), hydrolyzes the non-reducing terminal glu-
cosidic linkage of isomaltooligosaccharides and dextran.
Thermal deactivation of SmDG did not follow the single
exponential decay but rather the two-step irreversible
deactivation model, which involves an active intermedi-
ate having 39% specific activity. The presence of a low
concentration of CaCl2 increased the thermostability
of SmDG, mainly due to a marked reduction in the
rate constant of deactivation of the intermediate. The
addition of MgCl2 also enhanced thermostability, while
KCl and NaCl were not effective. Therefore, divalent
cations, particularly Ca2þ, were considered to stabilize
SmDG. On the other hand, CaCl2 had no significant
effect on catalytic reaction. The enhanced stability by
Ca2þ was probably related to calcium binding in the
ꢀ ! ꢁ loop 1 of the ðꢀ=ꢁÞ8 barrel of SmDG. Because
similar structures and sequences are widespread in
GH13, these GH13 enzymes might have been stabilized
by calcium ions.
zymes share a core three-domain structure composed of
domains A, B, and C. Domain A is a ðꢁ=ꢀÞ8 barrel fold,
domain B is a long protruding loop between ꢁ3 and ꢀ3
in the ðꢁ=ꢀÞ8 barrel, and domain C is an antiparallel
ꢁ-sheet at the C-terminus. The catalytic center, subsite
ꢀ1, is structurally well-conserved by seven invariant
amino acid residues, while the other parts are structur-
ally diverse so that the enzymes catalyze different
reactions.4) GH13 enzymes are classified into 36
subfamilies on the basis of amino acid sequence.5)
Dextran glucosidase belongs to subfamily 31
(GH13 31), together with oligo-1,6-glucosidase, ꢀ-glu-
cosidase, sucrose isomerase, isomaltulose synthase, and
trehalulose synthase. The enzymes act on ꢀ-glucoside at
the non-reducing ends of the substrates, and catalyze
hydrolysis, transglucosylation, and isomerization.
Some GH13 enzymes require calcium ions for activity
and stability. In particular, the most predominant ꢀ-
amylases are dependent on calcium ions.6) Many ꢀ-
amylases possess two calcium-binding sites in domain
B, and the bound calcium ions stabilize the protruding
loop structure.7) Other GH13 enzymes, such as cyclo-
dextrin glucanotransferase, cyclodextrinase, and ꢀ-glu-
cosidase, are also calcium dependent,8–11) and possess
Ca2þ in the equivalent positions and/or different parts of
the structures.12–23) Trehalulose synthase, a member of
subfamily GH13 31, has one calcium ion in the ꢁ ! ꢀ
loop 1.13) In cyclodextrin glucanotransferase8) and ꢀ-
amylase 2 (TVAII) of Thermoactinomyces vulgaris,22)
calcium ions function to enhance the stability.
SmDG is one of the enzymes containing Ca2þ in the
solved structures.12) It has a simple three-domain
architecture, and the ꢁ ! ꢀ loops of domain A and
part of domain B are involved in the formation of a
pocket-shaped active site, in which the residues in
subsite ꢀ1 are conserved structurally at the bottom. On
the outside of the active pocket, one calcium ion is
coordinated tightly by atoms of Asp21, Asn23, Asp25,
Ile27, and Asp29 in the ꢁ ! ꢀ loop 1 of domain A
(Fig. 1).12)
In this report, we present the denaturation kinetics of
SmDG. The denaturation of proteins is generally
described by a single-step irreversible denaturation
model, in which denaturation is expressed by a single
rate constant. Another model is the two-step irreversible
Key words: dextran glucosidase; thermostability; two-
step irreversible deactivation; calcium-bind-
ing site; glycoside hydrolase family 13
Dextran glucosidase, found in Streptococcus mutans
(SmDG) as the dexB gene product, hydrolyzes (1 ! 6)-
ꢀ-D-glucosidic linkages at the non-reducing end of
substrates to release glucose.1) The enzyme acts pref-
erably on short-chain isomaltooligosaccharides, but also
hydrolyzes dextran.1,2) SmDG catalyzes transglucosyla-
tion at high concentrations of the substrate, resulting in
the formation of ꢀ-1,6-glucosidic linkages.2)
Glycoside hydrolases are classified into glycoside
hydrolase families on the basis of sequence similarity.
The largest group in the CAZy database (http:
//www.cazy.org)3) is glycoside hydrolase family 13
(GH13), which contains ꢀ-amylase, cyclodextrin gluca-
notransferase, ꢀ-glucosidase, and many different activ-
ities in addition to dextran glucosidase. Most GH13
enzymes act on ꢀ-glycosides, such as starch, glycogen,
and related oligo- and polysaccharides, and catalyze
predominantly hydrolysis and transglycosylation with
retention of the ꢀ-anomeric configuration. GH13 en-
y
*
To whom correspondence should be addressed. Tel/Fax: +81-11-706-2808; E-mail: kimura@abs.agr.hokudai.ac.jp
Present address: Graduate School of Biosphere Science, Hiroshima University, 1-4-4 Kagamiyama, Higashi-Hiroshima 739-8528, Japan
Abbreviations: BSA, bovine serum albumin; CGTase, cyclodextrin glucanotransferase; GH, glycoside hydrolase; pNP, p-nitrophenol; pNPG,
p-nitrophenyl ꢀ-glucopyranoside; SmDG, dextran glucosidase from Streptococcus mutans

1558
M. KOBAYASHI et al.
The reaction mixture (50 mL), composed of 2 mM pNPG, 40 mM
sodium acetate buffer (pH 6.0), 0.02% bovine serum albumin (BSA),
0.02% sodium azide, and an adequate concentration of SmDG, was
incubated at 37 ꢁC for 10 min. The enzymatic reaction was stopped by
the addition of 100 mL of 1 M Na2CO3, and liberated pNP was
calculated from A400 of the mixture using "1 mM, 400 nm ¼ 16:7, which
was experimentally obtained. To determine the catalytic parameters,
the enzymatic reaction mixture (250 mL) was formulated to contain
0.2–12 mM pNPG and 1.2 mM EDTA in addition to 40 mM sodium
acetate buffer (pH 6.0), 0.02% BSA, 0.02% sodium azide, and SmDG.
To quantitate liberated pNP, 100 mL of the mixture was withdrawn and
mixed with 200 mL of 1 M Na2CO3. To measure glucose concen-
trations, the enzyme reaction (100 mL) was terminated by the addition
of 50 mL of 4 M Tris–HCl buffer (pH 7.0), and glucose was quantified
by the glucose oxidase-peroxidase method28) using Glucose CII Test
Wako (Wako Pure Chemical Industries, Osaka, Japan).
model, in which an irreversible active intermediate
is postulated before the enzyme is completely inacti-
vated.24) This model has been used to describe the
thermal denaturation of Bacillus licheniformis ꢀ-amy-
lase,24) and has been employed for other enzymes.25,26)
We applied the two-step irreversible denaturation model
to evaluate the thermostability of SmDG and to inves-
tigate the effect of calcium ions on thermostability. The
thermal denaturation was described well by the model,
and the presence of Ca2þ enhanced thermostability.
Increased stability with the calcium ion is a novel
characteristic of GH13 31 enzymes. The possibility that
many enzymes belonging to several GH13 subfamilies
have the similar properties is discussed.
Kinetic parameters of catalytic reaction. The reaction scheme
shown in Fig. 2 gave rise to equations of reaction velocities of
aglycone-release (vag), hydrolysis (vh), and transglucosylation (vtg)
expressed as a function of the substrate concentration ([S]):29,30)
Materials and Methods
Enzyme. Recombinant His6-tagged SmDG was prepared as described
previously.2) It was produced in Escherichia coli BL21(DE3) Codon-
Plus RIL (Stratagene, La Jolla, CA, USA) transformant harboring the
expression plasmid derived from pET23d (Novagen, Darmstadt,
2
2
vag ¼ ðkcat2½Sꢂ þ kcat1Km2½SꢂÞ=ð½Sꢂ
þ Km2½Sꢂ þ Km1Km2
Þ
ð1Þ
ð2Þ
ð3Þ
Germany) and purified to homogeneity through
a Ni-chelating
2
vh ¼ kcat1Km2½Sꢂ=ð½Sꢂ þ Km2½Sꢂ þ Km1Km2
Þ
Sepharose column (Chelating Sepharose FastFlow; GE Healthcare,
Little Chalfont, UK). The purified enzyme was dialyzed thoroughly
against 20 mM sodium acetate buffer (pH 6.0) and 0.02% sodium azide.
2
2
vtg ¼ kcat2½Sꢂ =ð½Sꢂ þ Km2½Sꢂ þ Km1Km2
Þ
where kinetic parameters, kcat1, kcat2, Km1, and Km2, were defined as
follows by the rate constants shown in Fig. 2:
Biochemical assay. The protein concentration was determined by
quantification of the various amino acids by the ninhydrin colorimetric
method using an amino acid analyzer (JEOL model JLC-500/V,
Tokyo, Japan) after hydrolysis of 26 mg of SmDG in 6 M HCl at 110 ꢁC
for 24 h.27)
Km1 ¼ k3ðkꢀ1 þ k2Þðkꢀ4 þ k5Þ=fk4k5ðkꢀ1 þ k2Þ
þ k1ðkꢀ4 þ k5Þðk2 þ k3Þg
Km2 ¼ fk4k5ðkꢀ1 þ k2Þ þ k1ðkꢀ4 þ k5Þðk2 þ k3Þg
=k1k4ðk2 þ k5Þ
kcat1 ¼ k1k2k3ðkꢀ4 þ k5Þ=fk4k5ðkꢀ1 þ k2Þ
þ k1ðkꢀ4 þ k5Þðk2 þ k3Þg
Enzyme activity (1 U) was defined as the amount of enzyme that
catalyzes the release of p-nitrophenol (pNP) at an initial rate of
1 mmol sꢀ1 from 2 mM p-nitrophenyl ꢀ-glucopyranoside (pNPG;
Nacalai Tesque, Kyoto, Japan) under standard reaction conditions.
kcat2 ¼ k2k5=ðk2 þ k5Þ
The transglucosylation ratio was defined as follows, and was
expressed as a function of the substrate concentration using a single
parameter:
Asp25
Asp25
Asn23
Asn23
Transglucosylation ratio (%) ¼ vtg=vag ꢃ 100
Ile27
Ile27
Ca2+
Ca2+
Asp21
Asp21
¼ ½Sꢂ=ðKTG þ ½SꢂÞ ꢃ 100
ð4Þ
where KTG was defined as follows:
Asp29
Asp29
KTG ¼ kcat1Km2=kcat2
The KTG constant is the substrate concentration that gives a
transglucosylation ratio of 50%.
Thermal denaturation of the enzyme. An enzyme solution (750 mL),
composed of 1.4 nM SmDG, 0.033% BSA, 2 mM EDTA, 0–30 mM
CaCl2, and 67 mM sodium acetate buffer (pH 6.0), was incubated at
319 K (46 ꢁC). An aliquot of 40 mL was withdrawn at indicated time
and cooled on ice. The residual pNPG-hydrolyzing activity of the
aliquot was measured by the standard assay.
β
1
α
1
β
α1
1
Fig. 1. A Calcium Ion Situated in ꢁ ! ꢀ Loop 1 in the ðꢁ=ꢀÞ8 Barrel
Structure of SmDG12) (Stereo View).
A calcium ion interacts with five atoms of side chains (Asp21
OD1, Asn23 OD1, Asp25 OD1 and OD2, and Asp29 OD2), one
carbonyl oxygen of the main chain (Ile27), and two water molecules
(not shown). The strand with ꢁ1 and the helix with ꢀ1 indicate
ꢁ-strand 1 and ꢀ-helix 1 in the catalytic ðꢁ=ꢀÞ8 barrel, respectively.
pH stability. An enzyme solution (100 mL), composed of 2.1 nM
SmDG, 0.01% BSA, 0.6 mM EDTA (or 5 mM CaCl2), and 90 mM
sodium acetate buffer (pH 3.6–6.2) or 90 mM Hepes buffer (pH 6.5–
pNP
k1
k3
SmDG pNPG
(ES complex)
SmDG + pNPG
SmDG-Glc
pNPG
SmDG + Glc
k2
k-1
k-4
k4
k5
SmDG-Glc pNPG
(ES complex)
SmDG + pNPG2
Fig. 2. Kinetic Scheme for the Reaction Catalyzed by SmDG.
SmDG-Glc represents the glucosyl-enzyme intermediate. pNPG2 is p-nitrophenyl isomaltoside.

Calcium-Dependent Thermostability of Dextran Glucosidase
1559
6
A
B
100
50
0
5
4
3
2
1
0
0
20
40
60
0
20
40
60
Incubation time (min)
Incubation time (min)
C
D
100
80
60
40
20
0
100
80
60
40
20
0
0
20
40
60
0
20
40
60
Incubation time (min)
Incubation time (min)
Fig. 3. Time-Dependent Heat Inactivation of SmDG at 46 ꢁC.
A, SmDG of 1.4 nM was incubated at 46 ꢁC in 2 mM EDTA without the addition of CaCl2 and cooled on ice at the indicated times, and the
residual activity was measured. The plots of residual activity versus incubation time did not follow the theoretical line based on the single-step
denaturation model (Eq. (5), broken line), but fitted the two-step irreversible denaturation model (Eq. (6), solid line). B, Semi-logarithmic plots
of residual activity (sꢀ1) versus incubation time. C, Effects of various concentrations of CaCl2: 0 mM ( ), 1 mM ( ), 2 mM ( ), 3 mM ( ), 5 mM
(circled dot), 15 mM ( ), and 30 mM ( ). D, Effects of various kinds of salt at 15 mM: KCl ( ), NaCl ( ), MgCl2 ( ), and CaCl2 ( ), and the
absence of any salt as control ( ). Residual activities were measured and plotted against incubation times. The fit is to the expression for the
two-step irreversible model (Eq. (6)).
7.9), was incubated for 15 min at 37 ꢁC and cooled on ice. After quick
neutralization, residual activity toward 2 mM pNPG was assayed using
10 mL of the enzyme solution under the standard reaction conditions
except that 100 mM sodium acetate buffer (pH 6.0) and 0.002% BSA
were used. The pH range in which SmDG retained more than 90% of
the original activity was considered to be the stable range.
where kB, h, R, and T represent the Boltzmann constant (1:381 ꢃ
10ꢀ23 JꢄKꢀ1), the Planck constant (6:626 ꢃ 10ꢀ34 Jꢄs), the gas constant
(8.3145 J Kꢀ1 molꢀ1), and the absolute temperature 319 K (46 ꢁC),
respectively.
Results
Thermal denaturation kinetics. When a fully-active native enzyme
(N) undergoes the single-step irreversible thermal deactivation process
and is converted directly into a completely inactive form (D),
Thermal denaturation process for SmDG
SmDG of 1.4 nM was incubated at 46 ꢁC for 60 min.
CaCl2 was not added, but 2 mM EDTA was added to
exclude the possible effect of contaminated bivalent
cations. Residual activity during the heat denaturation
process was monitored (Fig. 3A, B). The time-depend-
ent decrease in activity was not well fitted to the
theoretical line based on the simple one-step denatura-
tion model (Eq. (5), Fig. 3A). The plots of ln A versus
incubation time t were distinctly nonlinear (Fig. 3B),
indicating that the denaturation process was not correctly
described by the simple one-step model. The theoretical
line based on the two-step irreversible denaturation
model, however, was well fitted to the plots (Eq. (6),
Fig. 3A). Hence, the heat denaturation of SmDG was
expressed by the two-step irreversible denaturation
model, in which fully active SmDG (N) was inactivated
and converted to an inactive form (D) through an
irreversible active intermediate (X). By applying the
two-step model equation (Eq. (6)), the specific activities
A1 and A2 of the fully active native form and the
irreversible intermediate, respectively, and rate con-
N ! D
the residual activity (A) is expressed as follows:
A ¼ A1 expðꢀkdtÞ
ð5Þ
where A1 is the original activity of N, kd is the rate constant of N ! D,
and t is the time of the thermal process. When an enzyme undergoes
the two-step irreversible thermal deactivation process, which includes
one transitional active intermediate X,
N ! X ! D
the residual activity A is expressed by the following equation:24)
A ¼ fA1 ꢀ A2kd1=ðkd1 ꢀ kd2Þg expðꢀkd1tÞ
þ A2kd1=ðkd1 ꢀ kd2Þ expðꢀkd2tÞ
ð6Þ
where A1 and A2 are the specific activities of N and X, respectively,
and kd1 and kd2 are the rate constants at N ! X and X ! D,
respectively. The parameters were obtained by the non-linear least
squares method (the Levenberg-Marquardt algorithm) with Origin 8.1
(OriginLab, Northampton, MA, USA) in the t-A plots. Based on the
z
Eyring transition state theory, activation Gibbs energy ꢀG was
calculated by the following equation:
z
k ¼ ðkBT=hÞ expfꢀꢀG =ðRTÞg
ð7Þ

1560
M. KOBAYASHI et al.
z
Table 1. Specific Activities, Rate Constants, and ꢀG in the Irreversible Two-Step Heat Denaturation Process and Effects of the Presence of Salts
a
b
c
d
z
e
z
ꢀG
2
f
A1
A2
(mmol sꢀ1 mmolꢀ1
kd1
kd2
ꢀG
1
)
(sꢀ1
)
(kJ molꢀ1
)
no saltg
187 ꢅ 1:6
181 ꢅ 1:6
189 ꢅ 1:4
199 ꢅ 1:1
192 ꢅ 1:9
191 ꢅ 2:1
72:6 ꢅ 5:8
159 ꢅ 1:2
155 ꢅ 0:99
165 ꢅ 1:8
50:2 ꢅ 8:3
60:7 ꢅ 10
0:343 ꢅ 0:028
0:403 ꢅ 0:075
0:407 ꢅ 0:042
0:555 ꢅ 0:066
0:291 ꢅ 0:026
0:310 ꢅ 0:037
0:0627 ꢅ 0:0044
0:0015 ꢅ 0:00023
81.1
80.7
80.7
79.9
81.6
81.4
85.7
95.6
94.7
88.1
85.5
85.3
5 mM CaCl2
15 mM CaCl2
15 mM MgCl2
15 mM KCl
15 mM NaCl
0:00204 ꢅ 0:00021
0:0249 ꢅ 0:00032
0:0669 ꢅ 0:0093
0:0722 ꢅ 0:0096
aA1, pNP-releasing activity from 2 mM pNPG of the native form (N) of SmDG.
bA2, the same as A1, but of the intermediate (X) of SmDG.
c,d
k
and kd2, the rate constant of irreversible denaturation step from N to X and from X to the inactive denatured form (D) of SmDG, respectively.
d1
e,f ꢀG and ꢀG 2, Gibbs’s free energy required for steps of N to X, and X to D, respectively.
gHeat denaturation was done in the presence of 2 mM EDTA.
z
z
1
stants kd1 and kd2 of the irreversible change from N to X
and X to D, respectively, were obtained (Table 1).
Without any additional salts, the specific activity of the
native form (A1 ¼ 187 sꢀ1) was almost identical to the
specific activity (188 ꢅ 2:2 sꢀ1) of intact SmDG. The
specific activity of the intermediate (A2 ¼ 72:6 sꢀ1) was
39% of A1. The rate constant of kd1 was approximately 5
times higher than kd2. Changes in Gibbs’s free energy,
100
80
60
40
20
0
z
z
ꢀG
and ꢀG
for inactivation processes N ! X
1
2
and X ! D, respectively, were calculated to be
81.1 kJ molꢀ1 and 85.7 kJ molꢀ1 at 319 K (46 ꢁC) by
Eq. (7) (Table 1).
3
4
5
6
7
8
pH
Fig. 4. Effect of CaCl2 on the pH Stability of SmDG.
Effects of Ca2þ on the stability of SmDG
SmDG of 2.1 nM was incubated at various pH values at 37 ꢁC for
15 min in the absence and the presence of 5 mM CaCl2, and residual
activity was measured. SmDG was stable at pH 5.8–6.7 without
Next, the thermostability of SmDG was assayed at the
same temperature, 46 ꢁC, but in the presence of CaCl2 at
0–30 mM (Fig. 3C). The activity decreased during
incubation at all CaCl2 concentrations, but the rates of
denaturation were drastically different depending on the
CaCl2 concentration. The activity dropped to 50% in
5 min at 0–1 mM CaCl2, but the presence of 2 mM CaCl2
prolonged this to 30 min, and 3–30 mM CaCl2 stabilized
SmDG so that it retained 70% activity or higher even
after 60 min of incubation. The most positive effect on
stability was obtained at 5 mM CaCl2. All of the time-
courses of decrease in residual activity were expressed
well by the two-step denaturation model (Eq. (6),
Fig. 3C). By fitting the data to Eq. (6), the specific
activities, A1 and A2, and the rate constants of
denaturation, kd1 and kd2, were obtained (Table 1).
Regardless of the CaCl2 concentration in the thermal
process, the specific activity of N (A1) was in a good
agreement with that of intact SmDG. The rate constant
from N to X (kd1) was not significantly affected by the
addition of 5 or 15 mM CaCl2 either (kd1 ¼ 0:403 and
0.407 sꢀ1, respectively). In contrast, the specific activity
of the irreversible intermediate (A2) was increased 2-
fold (A2 ¼ 159 and 155 sꢀ1, respectively) and the rate
constant toward the completely denatured form, kd2, was
decreased 42-fold by the addition of 5 mM CaCl2
CaCl2
( ) and at pH 5.5–7.0 with 5 mM CaCl2 ( ).
The presence of CaCl2 also affected the pH stability
of SmDG (Fig. 4). Without CaCl2, SmDG retained its
activity in a narrow pH range (pH 5.8–6.7) after 15 min
of incubation at 37 ꢁC. The range was, however,
broadened by addition of 5 mM CaCl2, to pH 5.5–7.0.
Effects of other cations on the thermostability of
SmDG
The thermal denaturation kinetics at 46 ꢁC of SmDG
were analyzed in the presence of other salts: 15 mM KCl,
NaCl, and MgCl2 (Fig. 3D). KCl and NaCl did not
significantly affect the time-dependent decrease in
activity, and hence the kinetic parameters were similar
to those without salt (Table 1). The presence of MgCl2,
however, decreased the rate of inactivation of SmDG,
although the positive effects on the thermostability of
SmDG was lower than those due to 15 mM CaCl2
(Fig. 3D, Table 1). In the presence of 15 mM MgCl2, kd1
even increased slightly, but kd2 dropped to 40%, and A2
rose to 230% (Table 1).
Kinetics of the catalytic reaction
(kd2 ¼ 0:0015 sꢀ1). The effect of the presence of
SmDG catalyzes hydrolysis and transglucosylation
simultaneously. The catalytic reaction of SmDG for
pNPG was analyzed kinetically. The initial velocities of
liberation of pNP (vpNP) and glucose (vGlc) from 0.2–
12 mM pNPG were measured. Since aglycone (pNP) is
produced by both hydrolysis and transglucosylation,
vpNP must be the sum of the velocities of hydrolysis and
transglucosylation. vGlc represents hydrolysis, and the
z
15 mM CaCl2 on ꢀG at the first denaturation step
1
z
was very limited, but ꢀG at the second step was
2
increased by approximately 10 kJ molꢀ1. The results
indicate that the intermediate was protected by CaCl2
from irreversible denaturation toward the completely
inactive form, and the intermediate itself possessed
higher activity.

Calcium-Dependent Thermostability of Dextran Glucosidase
1561
250
250
200
150
100
50
100
80
60
40
20
0
A
B
C
200
150
100
50
0
0
0
2
4
6
8
10 12
0
2
4
6
8
10 12
0
2
4
6
8
10 12
pNPG (mM)
pNPG (mM)
pNPG (mM)
Fig. 5. Kinetics of Catalytic Reaction and the Effect of CaCl2 on Activity.
A and B, The dependence of initial velocities of aglycone release ( ), hydrolysis ( ), and transglucosylation (hollow triangle) on the pNPG
concentration was analyzed in the absence (A) and the presence (B) of 15 mM CaCl2. C, Plots of the transglucosylation ratio versus the pNPG
concentration in the absence of CaCl2 ( ) and the presence of 15 mM CaCl2 ( ). The data fitted the theoretical lines well.
Table 2. Kinetic Parameters of SmDG in the Catalytic Reaction on pNPG in the Absence and the Presence of CaCl2
kcat1
(sꢀ1
kcat2
(sꢀ1
Km1
Km2
KTG
CaCl2
)
)
(mM)
(mM)
(mM)
0 mM
15 mM
173 ꢅ 20
175 ꢅ 13
255 ꢅ 3:8
205 ꢅ 4:4
0:288 ꢅ 0:066
0:366 ꢅ 0:054
1:65 ꢅ 0:12
1:27 ꢅ 0:16
1:03 ꢅ 0:065
1:26 ꢅ 0:087
difference (vpNP-vGlc) indicates the transglucosylation
velocity. By the reaction scheme shown in Fig. 2 and
steady-state kinetics, the initial velocities of hydrolysis,
transglucosylation, and the both reactions are expressed
by Eq. (2), Eq. (3), and Eq. (1), respectively. In Fig. 5A,
the velocities in the absence of CaCl2 are plotted as a
function of the pNPG concentration. The velocities of
aglycone-release and transglucosylation increased with
the pNPG concentration, but the hydrolysis rates
reached a maximum at 1 mM pNPG, and decreased with
increase in the pNPG concentration. The data fitted the
theoretical curves well (Fig. 5A), and the kinetic
parameters, kcat1, kcat2, Km1, and Km2, were determined
from the theoretical equations (Table 2). The plots of
transglucosylation ratios, defined by Eq. (4), versus the
pNPG concentrations fitted the theoretical saturation
curve as well (Fig. 5C, solid squares). The transgluco-
sylation parameter, KTG, the substrate concentration
giving a transglucosylation ratio of 50%, was deter-
mined to be 1.0 mM (Table 2). The addition of 15 mM
CaCl2 did not have a strong impact on the kinetic
behavior of the reaction on pNPG (Fig. 5B, C). None of
the parameters was significantly different from those in
the absence of CaCl2, but the maximum aglycone-
releasing velocity, kcat2, decreased 20%, and trans-
glucosylation fell with 20% increased KTG (Table 2).
calcium chloride, and the two calcium ions were not
predicted to be held in the protein molecule in a solution
with a low concentration of CaCl2, but were predicted to
be released. On the other hand, the other calcium ion on
the ꢁ ! ꢀ loop 1 was predicted to be held in the protein
due to tight binding.
Our biochemical analyses of SmDG revealed the
function of the calcium ion in enhancing the stability of
SmDG. In the absence and the presence of CaCl2, the
denaturation kinetics of SmDG were not described by a
single exponential decay, but by the two-step irrever-
sible denaturation model, which was used to describe
the deactivation of B. licheniformis ꢀ-amylase24) and
several other enzymes.25,26) The model postulates that a
fully active native enzyme (N) is transformed into a
completely inactive form (D) via an active intermediate
(X) through two irreversible steps. The deactivation
intermediate of SmDG in the absence of any additional
salt retained 39% of the activity of intact SmDG. In
terms of rate constants, the change from N to X was 5.5-
fold faster than the second step, from X to D, in the
deactivation of SmDG. The addition of divalent cations
in the thermal process did not cause a significant
difference in the specific activity of N, but enhanced the
thermostability of SmDG through an increase in the
specific activity of the intermediate and the reduction of
kd2, while kd1 increased slightly. In the case of 5 mM
CaCl2, the specific activity of X increased 2.2-fold, and
kd2 fell 42-fold. Therefore, according to the model of
two-step irreversible denaturation, the effect of calcium
ions on the stability of SmDG during the thermal process
is achieved by increased specific activity and enhanced
stability of the deactivation intermediate.
Discussion
SmDG is a GH13 31 enzyme that has a calcium ion
bound to the ꢁ ! ꢀ loop 1 of the core ðꢁ=ꢀÞ8 barrel in
the solved three-dimensional structures (Fig. 1).12) The
calcium ion is coordinated in the site by eight atoms
from five amino acid residues and two water molecules.
The crystal structures of SmDG hold two other calcium
ions between symmetry-related enzyme molecules. The
crystal used in structural analysis was, however, grown
in the presence of a high concentration (200 mM) of
The thermostability of SmDG was increased by
MgCl2 as well, but not by NaCl or KCl, which rather
decreased the residual activity (Fig. 3D). Therefore, the
enhanced stability was due to the influence of the
divalent cations, particularly the calcium ion. On the

1562
M. KOBAYASHI et al.
Table 3. Amino Acid Sequences of the ꢁ ! ꢀ Loop 1 Calcium-Binding Site of Structure-Solved GH13 Enzymes
Subfamily
Enzyme
Organism
Calcium-binding sitea
Ref.
31
31
Dextran glucosidase (SmDG)
Trehalulose synthase (MutB)
ꢀ-Amylase A
ꢀ-Glucosidase (GSJ)
ꢀ-Amylase (SusG)b
Streptococcus mutans ATCC 25175
Pseudomonas mesoacidophila MX-45
Halothermothrix orenii
21
49
44
21
73
13
30
DtNgDgIgD
DtNgDgIgD
DsDgDgIgD
DaNgDgIgD
DsDgDgYgD
DgNlDgVgD
DsNdDgWgD
12
13
14
15
16
17
18
36
unclassified
unclassified
unclassified
unclassified
Geobacillus sp. HTA-462
Bacteroides thetaiotaomicron VPI-5482
Thermotoga maritima MSB8
Saccharomyces cerevisiae S288C
4-ꢀ-Glucanotransferase (TM0364)
Oligo-ꢀ-1,6-glucosidase
2
ꢁ-Cyclodextrin glucanotransferase
Maltogenic ꢀ-amylase
Neopullulanase
Bacillus circulans 251
54
54
DgNpaNN-(17)-GgD
DgDttNN-(20)-GgD
NgNpsiS-(18)-GgD
NgDpsND-(19)-GgD
NgDpsND-(18)-GgD
NgDssND-(35)-GgD
19
20
21
22
23
22
2
20
Geobacillus stearothermophilus C599
Geobacillus stearothermophilus TRS40
Thermoactinomyces vulgaris R-47
Flavobacterium sp. 92
147
143
137
174
20
unclassified
unclassified
ꢀ-Amylase 2 (TVAII)
Cyclomaltodextrinase
ꢀ-Amylase 1 (TVA I)
Thermoactinomyces vulgaris R-47
aCapital letters indicate residues involved in Ca2þ binding. Parenthetic numbers in the sequence show numbers between residues.
bMg2þ is a ligand.
other hand, chloride ions and the ionic strength of the
solution might reduce the thermostability of SmDG. The
slight decrease in thermostability with CaCl2 at higher
than 5 mM might have been caused in the same way.
It can be assumed that the enhancement of thermo-
stability due to the addition of calcium ions at a low
concentration is related to the tight binding of a calcium
ion in the ꢁ ! ꢀ loop 1. The position of calcium
binding in the SmDG structure is distinctly different
from that commonly found in ꢀ-amylases, which often
possess Ca2þ in domain B (a long protruding loop
connecting ꢁ3 and ꢀ3 in the ðꢁ=ꢀÞ8 barrel) maintaining
its structural integrity.7) The predominant ꢀ-amylases
have no calcium ion in the equivalent position in ꢁ ! ꢀ
loop 1, as observed for SmDG, but calcium binding to
the site in loop 1 has been found in many GH13
enzymes (representative examples are listed in Table 3).
Even so, little has been known about functions of the
calcium ion in that position to date. In cyclodextrin
glucanotransferase and cyclodextrinase, which contain
a calcium ion in loop 1 in the solved structures, the
addition of a chelating reagent such as EDTA or
ethyleneglycol bis(2-aminoethylether) tetraacetic acid
reduced protein stability,8,10,22) but it was unclear
whether that effect was due to removal of the equivalent
calcium ion in loop 1 or another, because the enzymes
held two or more calcium ions in a protein molecule. A
cyclodextrinase, so-called ꢀ-amylase 2 of T. vulgaris, is
the sole evidence that the calcium ion in the loop 1 is
related to increases in the thermostability of the enzyme,
because the enzyme holds just one equivalent calcium
ion per protein molecule, and chelation with ethyl-
eneglycol bis(2-aminoethylether) tetraacetic acid resulted
in a reduction in the thermostability by approximately
5 ꢁC under the experimental conditions.22) Our results
for SmDG provide even more precise evidence as to
what positive effects the calcium ion has on the stability
of the protein.
consensus sequences are widely distributed in many
GH13 enzymes. The first consensus sequence of the
short-loop group is shared by GH13 enzymes belonging
to subfamilies 16, 17, 23, 29, 30, 31, 35, and 36. The
second long-loop group is comprised of GH13 enzymes
of subfamilies 2, 19, 20, and 22. Exceptions are found
only in subfamily 20. The both sequences are also
shared by other GH13 enzymes not yet classified into
any subfamily. It has not been known yet whether an
equivalent calcium ion is held in the loop 1 site in many
of the enzymes. Some solved three-dimensional struc-
tures, for instance, oligo-1,6-glucosidase from Bacillus
cereus,31) possess no calcium ion as ligand in the loop 1
site in spite of a shared consensus sequence, but it might
happen because of the crystal conditions. There is
possibility that by the addition of calcium ions to some
concentration to the protein solution, calcium ions can
be introduced into the calcium-binding site in ꢁ ! ꢀ
loop 1, resulting in an increase in the stability of the
enzyme, as observed for SmDG.
Acknowledgment
We thank Mr. Tomohiro Hirose of the Instrumental
Analysis Division, Equipment Management Center,
Creative Research Institution, Hokkaido University, for
amino acid analysis.
References
1) Russell RR and Ferretti JJ, J. Gen. Microbiol., 136, 803–810
(1990).
2) Saburi W, Mori H, Saito S, Okuyama M, and Kimura A,
Biochim. Biophys. Acta, 1764, 688–698 (2006).
3) Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V,
and Henrissat B, Nucleic Acids Res., 37, D233–D238 (2009).
4) MacGregor EA, Janecek S, and Svensson B, Biochim. Biophys.
Acta, 1546, 1–20 (2001).
5) Stam MR, Danchin EG, Rancurel C, Coutinho PM, and
Henrissat B, Protein Eng. Des. Sel., 19, 555–562 (2006).
6) Vallee BL, Stein EA, Sumerwell WN, and Fischer EH, J. Biol.
Chem., 234, 2901–2905 (1959).
As shown in Table 3, the calcium ion-holding loop 1
structures fall into two groups: a short-loop group sharing
consensus sequence Dx(N/D/T/S)xDg(I/X)gD, and a
long-loop group sharing (D/N)g(D/N)xx(N/x)(N/D/S)
and GgD connected with long toes of loop 1, where the
capital letters indicate residues involved in direct
interaction with the calcium ion, and x denotes any
residue. SmDG belongs to the first short-loop group. The
7) Machius M, Declerck N, Huber R, and Wiegand G, Structure, 6,
281–292 (1998).
8) Thiemann V, Donges C, Prowe SG, Sterner R, and Antranikian
¨
G, Arch. Microbiol., 182, 226–235 (2004).
9) Kelly RM, Dijkhuizen L, and Leemhuis H, Appl. Microbiol.
Biotechnol., 84, 119–133 (2009).
10) Kim TJ, Shin JH, Oh JH, Kim MJ, Lee SB, Ryu S, Kwon K,

Calcium-Dependent Thermostability of Dextran Glucosidase
1563
Kim JW, Choi EH, Robyt JF, and Park KH, Arch. Biochem.
Biophys., 353, 221–227 (1998).
TV, Beier L, Wilson KS, and Davies GJ, Biochemistry, 38,
8385–8392 (1999).
11) Plant AR, Parratt S, Daniel RM, and Morgan HW, Biochem. J.,
1, 865–868 (1988).
21) Hondoh H, Kuriki T, and Matsuura Y, J. Mol. Biol., 326, 177–
188 (2003).
12) Hondoh H, Saburi W, Mori H, Okuyama M, Nakada T,
Matsuura Y, and Kimura A, J. Mol. Biol., 378, 913–922
(2008).
22) Kamitori S, Abe A, Ohtaki A, Kaji A, Tonozuka T, and Sakano
Y, J. Mol. Biol., 318, 443–453 (2002).
23) Fritzsche HB, Schwede T, and Schulz GE, Eur. J. Biochem.,
270, 2332–2341 (2003).
13) Ravaud S, Robert X, Watzlawick H, Haser R, Mattes R, and
Aghajari N, J. Biol. Chem., 282, 28126–28136 (2007).
14) Sivakumar N, Li N, Tang JW, Patel BK, and Swaminathan K,
FEBS Lett., 580, 2646–2652 (2006).
24) Violet M and Meunier JC, Biochem. J., 263, 665–670 (1989).
25) Henley JP and Sadana A, Enzyme Microb. Technol., 7, 50–60
(1985).
15) Shirai T, Hung VS, Morinaka K, Kobayashi T, and Ito S,
Proteins, 73, 126–133 (2008).
26) Nury S, Meunier JC, and Mouranche A, Eur. J. Biochem., 180,
161–166 (1989).
16) Koropatkin NM and Smith TJ, Structure, 18, 200–215 (2010).
17) Roujeinikova A, Raasch C, Sedelnikova S, Liebl W, and Rice
DW, J. Mol. Biol., 321, 149–162 (2002).
27) Moore S and Stein WH, J. Biol. Chem., 176, 367–388 (1948).
28) Miwa I, Okudo J, Maeda K, and Okuda G, Clin. Chim. Acta, 37,
538–540 (1972).
18) Yamamoto K, Miyake H, Kusunoki M, and Osaki S, FEBS J.,
277, 4205–4214 (2010).
29) Brumer H III, Sims PF, and Sinnott ML, Biochem. J., 339, 43–
53 (1999).
19) Lawson CL, van Montfort R, Strokopytov B, Rozeboom HJ,
Kalk KH, de Vries GE, Penninga D, Dijkhuizen L, and Dijkstra
BW, J. Mol. Biol., 236, 590–600 (1994).
30) Kawai R, Igarashi K, Kitaoka M, Ishii T, and Samejima M,
Carbohydr. Res., 339, 2851–2857 (2004).
31) Watanabe K, Hata Y, Kizaki H, Katsube Y, and Suzuki Y,
J. Mol. Biol., 269, 142–153 (1997).
20) Dauter Z, Dauter M, Brzozowski AM, Christensen S, Borchert