Enzyme and Microbial Technology 34 (2004) 278–282

Thermostabilization of Bacillus subtillis CCMI

966 xylanases with trehalose
Study of deactivation kinetics
Paula Sá-Pereira a,∗ , Ana Sofia L. Carvalho b , Maria Costa- Ferreira a ,
Maria Raquel Aires-Barros b
a Department of Biotechnology, Instituto Nacional de Engenharia e Tecnologia Industrial, Estrada do Paço do Lumiar, 22, 1649-038 Lisboa, Portugal
b Centre for Biological and Chemical Engineering, Instituto Superior Técnico, 1049-001 Lisboa, Portugal

Received 1 August 2002; received in revised form 13 October 2003; accepted 6 November 2003

Abstract
The effect of trehalose (0.5 M) on the thermal stability of two endo-1,3(4)-␤-xylanases, alkaline Xyl I (pI > 9.3) and a neutral Xyl II
(pI 7.5), purified from the culture broth of Bacillus subtilis CCMI 966 was studied at 60 and 70 ◦ C. Thermal deactivation kinetics of Xyl I
and Xyl II were analyzed in the absence and in the presence of trehalose. Suitable mechanisms were proposed to describe the deactivation
process. Trehalose not only stabilized both enzymes at 60 and 70 ◦ C, but also increased their initial activity. Addition of trehalose allowed
both isoenzymes to attain a final state with residual activity. In contrast, without trehalose, the final state had no activity. At 60 and 70 ◦ C,
the Xyl I half-lives increased 2.5- and 2-fold, respectively, in the presence of trehalose. The half-life for Xyl II at 60 ◦ C, increased about
eightfold. At 70 ◦ C, the half-life increased about twofold and gave rise to a final state with residual activity of about 35%.
© 2003 Elsevier Inc. All rights reserved.
Keywords: Bacillus subtilis; Xylanase; Trehalose; Thermostability; Enzyme deactivation kinetics; Modelling

1. Introduction Bacillus subtilis CCMI 966 strain produces two extracel-

Xylan, an abundant component of plant biomass, has a
relatively complex structure based on a non-debranched
␤-glycosidically linked xylose backbone. Typically, back-
bone depolymerization is accomplished by the action of
endo-␤-1,4-xylanases (␤-1,4-d-xylan xylanohydrolases,
EC 3.2.1.8), ␤-1,4-xylosidases (␤-1,4-d-xyloside xylohy-
drolases, EC 3.2.1.37) and exoxylanases (␤-1,4-d-xylan
xylohydrolases). Different xylanase isoenzymes differing in
their specificities are involved in the cleavage of glycoside
linkages of heteroxylan backbone [1,2].
The bioconversion of xylan from industrial by-products
can be improved through the selection of xylan degrading
enzymes that are more stable to extreme reaction conditions
or by improving their operational stability [3]. Knowledge
about thermal deactivation kinetics of xylanases deactivation
processes can contribute toward a better understanding of
xylanase stability [4–6].
∗Corresponding author. Tel.: +351-21-7165141; fax: +351-21-7163636.
E-mail address: paula.pereira@ineti.pt (P. Sá-Pereira).
0141-0229/$ – see front matter © 2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.enzmictec.2003.11.014
lular thermostable endo-1,3(4)-␤-xylanases, Xyl I and Xyl
II, with unique properties among bacteria [7–9]. These ex-
tracellular isoenzymes have been purified from culture me-
dia and further characterized as an alkaline Xyl I (pI > 9.3)
and a neutral Xyl II (pI 7.5). They are multimeric enzymes,
which have a modular architecture with cellulose-binding
properties, suggesting the presence of a cellulose-binding
domain [8]. The two isoenzymes differ in their physicochem-
ical properties and in their substrate specificities. Optimal
xylanase production, occurred after 18-h fermentation, using
xylan as main carbon source. In contrast, when trehalose was
used instead of xylan, production decreased about threefold.
However, this extracellular xylanase activity was sustained
throughout the 80-h cultivation period studied. These results
suggested that trehalose may be taken up by Bacilli, acting
simultaneously as a xylanase inducer and a stabilizer [9].
Although protein stabilization by low molecular weight
solutes is a widely used strategy [3,10], few reports on the
effect of trehalose on xylanase stability are referred to in
the literature. George et al. [6] reported that trehalose did
not have a significant effect on the xylanase stability from
Thermomonospora sp.
 P. Sá-Pereira et al. / Enzyme and Microbial Technology 34 (2004) 278–282 279

In the present work, the effect of trehalose on the thermal

stability of the two purified extracellular iso-endo-1,3(4)-␤-xylanases,
Xyl I and Xyl II, from B. subtilis was studied. Thermal
Scheme 1.
deactivation kinetics of Xyl I and Xyl II was analyzed
using suitable models. Mechanisms were proposed for the
deactivation process.
2.4. Data analysis

2. Materials and methods Thermal deactivation kinetics of Xyl I and Xyl II were
studied based on a single step first-order deactivation scheme
2.1. Bacterial strain, cultivation conditions and xylanase and sequential deactivation schemes involving one and two
purification intermediate states, according to Sadana [15]. The deacti-
vation schemes used to modulate the experimental deacti-
A Bacillus subtilis strain with xylanolytic properties [7] vation data obtained for Xyl I and Xyl II, without and with
was isolated during a screening of water samples from trehalose, are described in this section.
a hot-spring in Azores (Portugal). For long term storage, Series deactivation involving two active intermediates and
1 ml aliquots of Bacillus grown in the liquid medium having a final state Ed , that is totally inactive, can be repre-
Luria-Bertani, described in Sá-Pereira et al. [11], were kept sented by Scheme 1, where E, E1 , E2 and Ed are the different
in 25% glycerol at −70 ◦ C under sterile conditions. enzyme states; k1 , k2 and k3 are the first-order deactivation
Enzymatic extracts were produced in 1-l shake-flasks with rate coefficients; β1 , and β2 are the ratio of specific activities
10% inocula (v/v) in 250 ml of culture media DXM [9]. The δ1 /δ and δ2 /δ1 , respectively, being δ, δ1 and δ2 the specific
pH was adjusted to 6.0 with HCl. The bacterial culture was activities of the E, E1 and E2 enzyme states, respectively.
grown at 50 ◦ C, in a rotatory shaker at 200 rpm. Xyl I and Assuming that at time t equal to zero, the concentration
Xyl II were purified by ammonium sulfate precipitation and of E1 , E2 and Ed enzyme states is equal to zero, the activity
anionic exchange chromatography. a, expressed as a weighted-average function of the different
enzyme states, is given by:
2.2. Xylanase activity
δ1 [E] + δ2 [E1 ] + δ3 [E2 ]
a=
The xylanolytic activity was measured by the DNS δ1 [E0 ]
(3,5-dinitrosalicylic acid) assay [12]. The amount of re-

β1 k1 β 2 k1 k 2
ducing sugars liberated from oat spelt xylan (OSX, 5 g l−1 , = 1+ + exp(−k1 t)
Sigma), were determined by solubilization of the substrate k2 − k 1 (k2 − k1 )(k3 − k1 )


in 0.05 M phosphate buffer, at pH 6.0 (450 ␮l), according β1 k1 β2 k1 k2
to Bailey et al. [13], with addition of 50 ␮l of Xyl I or Xyl + + exp(−k2 t)
k1 − k 2 (k1 − k2 )(k3 − k2 )
II. The buffer contains orthophosphate ions, since it was
β2 k1 k2
observed that in some Bacillus xylanases this ion signifi- + exp(−k3 t) (1)
cantly improves, not only the stability but also the xylanase (k1 − k3 )(k2 − k3 )
activity [14]. Considering a series enzyme deactivation involving a single
The mixture containing the substrate xylan and the isoen- active intermediate and a final state, E2 , with some activity,
zyme was incubated for 10 min at 60 ◦ C, and the reaction the series deactivation scheme described above is simplified
was stopped by addition of 750 ␮l of DNS. The substrate to Scheme 2 and the weighted-average activity expression
with trehalose and without enzyme and the substrate with is:
trehalose and with the inactivated enzyme were used as con-
trols. The xylanase activity was determined by using a cali-
bration curve for d-xylose (Sigma). The activity of xylanase
was expressed in International Units (IU). One IU is defined Scheme 2.
as the amount of enzyme that releases 1 ␮mol xylose min−1 ,
at 30 ◦ C.  
β 2 k1 β 2 k2
a = β2 + 1 + − exp(−k1 t)
2.3. Thermal stability k2 − k 1 k2 − k 1
 
β 1 k1 β 2 k2
The thermostability of Xyl I and Xyl II was studied at 60 − − exp(−k2 t) (2)
k2 − k 1 k2 − k 1
and 70 ◦ C, during 3–24 h at pH 6.0. Studies were performed
by pre-incubating the xylanase in phosphate buffer, 0.05 M, When δ2 = 0, than β2 = 0 and the enzyme state E2 is
pH 6.0, (235 mU ml−1 of Xyl I or 151 mU ml−1 of Xyl II) completely deactivated and equal to Ed . Thus, when the final
without and with 0.5 M trehalose (Sigma). The activity of enzyme state has no activity, Scheme 3 is proposed and the
each sample was then quantified, using the DNS assay. weighted-average activity expression is:
280 P. Sá-Pereira et al. / Enzyme and Microbial Technology 34 (2004) 278–282

1
with 0.5M trehalose

Scheme 3.
without additive
Activity a

δ1 [E] + δ2 [E1 ] 0.1

a=
δ1 [E0 ]
 
β1 k1 k1 β 1
= 1+ exp(−k1 t) − exp(−k2 t) (3)
k2 − k 1 k2 − k 1
60ºC
The specific activity of the intermediate state E1 , may 0.01
be less, equal to, or greater than the specific activity of the 0 0.5 1 1.5 2 2.5 3 3.5
initial state E. Thus β1 , can have values lower, equal or
higher than one. If β1 < 1, the intermediate state has a lower 10
specific activity than the initial enzyme state; when β1 = 1,
the intermediate state is as active as the initial enzyme state; 1 with 0.5M trehalose

and when β1 > 1, the intermediate state as a higher specific

0.1
Activity a

activity than the initial enzyme state. Similar considerations

can be assumed for β2 . In Eq. (2), and since E2 is the final 0.01
state of the enzyme, the residual activity of the enzyme is
also equal to β2 . 0.001 without additive
When β1 equals zero, Eq. (3) gives
0.0001
a = exp(−k1 t) (4)
70ºC
0.00001
the first-order deactivation equation. 0 0.5 1 1.5 2 2.5 3 3.5
Incubation time (h)

3. Results and discussion
3.1. Xyl I and Xyl II thermostability
The stability of purified Xyl I and Xyl II was studied in
the absence and presence of trehalose. Figs. 1 and 2 present
the thermostability of Xyl I and Xyl II, respectively, at
60 and 70 ◦ C, without and with 0.5 M of trehalose. Both
isoenzymes showed a decrease in the stability as a func-
tion of time, with increasing temperature from 60 to 70 ◦ C.
In the absence of trehalose, and after 3 h of incubation at
60 ◦ C, Xyl I retained 20% of its initial activity but at 70 ◦ C
no activity was observed. In contrast, Xyl II was com-
pletely inactivated above 60 ◦ C. In the presence of 0.5 M
of trehalose, the thermal stability of both isoenzymes was
greatly improved. For Xyl I, and after 3-h incubation with
trehalose, a residual activity of 50 and 36%, was obtained,
at 60 and 70 ◦ C, respectively. Similar residual activities
were obtained for Xyl II. The addition of trehalose not only
stabilized both enzymes at 60 and 70 ◦ C, but also increased
their initial activities. This effect is more pronounced for
the Xyl II isoenzyme where the presence of trehalose in-
creased the initial activities about 2- and 2.5-fold, at 60 and
70 ◦ C, respectively. After 24-h incubation with trehalose,
a residual activity of 40% was obtained for both enzymes,
showing the stabilization effect of this additive.
Trehalose has been described to act as a stabilizer of the
structure and function of several macromolecules. Due to its
Fig. 1. Deactivation kinetics of Xyl I at 60 and 70 ◦ C, in the absence and
in the presence of trehalose (experimental and predicted data—markers
and lines, respectively). Studies were performed by pre-incubating Xyl I
(0.7 ␮g ml−1 ) in 0.05 M phosphate buffer pH 6.0, without and with 0.5 M
of trehalose, at 60 and 70 ◦ C. Deactivation profiles were obtained through
the application of the proposed models of Sadana [15].
larger hydrated volume comparatively to other sugars (about
2.5 larger), trehalose can substitute more water molecules
in solution thereby maintaining the folded state of proteins
[16]. This higher exclusion effect is responsible for its effi-
ciency in the protection enzymes against thermal inactiva-
tion [16,17]. This structure-stabilizing molecule enables Xyl
I and Xyl II to maintain their activity at high temperatures,
at which are normally inactive.
3.2. Xyl I and Xyl II deactivation kinetics
In order to modulate the experimental deactivation data
obtained for Xyl I and Xyl II at 60 and 70 ◦ C, series or
sequential deactivation schemes described by Sadana, [15]
were applied. The experimental deactivation data of Xyl I
as function of time, at 60 and 70 ◦ C, with and without 0.5 M
of trehalose, were adjusted using several deactivation steps
(Fig. 1).
Sequential deactivation in three steps was proposed for
Xyl I deactivation at 60 ◦ C, with and without trehalose as
well as at 70 ◦ C, with trehalose. The suggested three-step
 P. Sá-Pereira et al. / Enzyme and Microbial Technology 34 (2004) 278–282 281

1 Table 1
with 0.5M trehalose Xyl I deactivation parameters, at 60 and 70 ◦ C, obtained from application
of the kinetic deactivation schemes proposed (a first-order kinetic and
three-step mechanism considering Scheme 1)
Deactivation 60 ◦ C 60 ◦ C + 70 ◦ C 70 ◦ C +
Activity a

parameters trehalose trehalose

0.1 k1 (h−1 ) 2.89 2.33 3.20 6.26
k2 (h−1 ) 7.48 8.36 n.a. 10.07
k3 (h−1 ) 0.04 0.00 n.a. 0.00
without additive
β1 1.33 1.22 n.a. 1.36
β2 0.23 0.47 n.a. 0.32

60ºC t1/2 (h) 0.57 1.41 0.22 0.39

0.01
n.a.: not applicable.
0 0.5 1 1.5 2 2.5 3 3.5

1
similar to that at 60 ◦ C, without additive. At 60 and 70 ◦ C,
with 0.5M trehalose

with this disaccharide, the Xyl I half-lives increased 2.5- and

0.1 2-fold, respectively.
The experimental deactivation data for Xyl II as function
Activity a

of time, at 60 and 70 ◦ C, in the absence and in the presence

0.01
of 0.5 M of trehalose, were adjusted using several deactiva-
without additive
tion steps (Fig. 2). Xyl II thermal deactivation, at 60 and
0.001
70 ◦ C, in the absence and presence of trehalose, showed se-
quential enzyme deactivation in two steps (Schemes 2 and
70ºC 3). The experimental data fits well with a two-step mech-
0.0001 anism. Without trehalose, the enzyme deactivation process
0 0.5 1 1.5 2 2.5 3 3.5 lead to a final state Ed without activity. In its presence a final
Incubation time (h) state, E2 , with a residual activity equal to β2 was obtained
(Fig. 2). The disaccharide stabilized the enzyme inducing
Fig. 2. Deactivation kinetics of Xyl II at 60 and 70 ◦ C, in the absence and
in the presence of trehalose (experimental and predicted data—markers
the appearance of a final state with activity.
and lines, respectively). Studies were performed by pre-incubating the The deactivation parameters obtained, at 60 and 70 ◦ C,
Xyl II (0.7 ␮g ml−1 ) in 0.05 M phosphate buffer pH 6.0, without and with with and without trehalose, using the models described by
0.5 M of trehalose, at 60 and 70 ◦ C. Deactivation profiles were obtained Schemes 2 and 3 are presented in Table 2. At 60 ◦ C, the
through the application of the proposed models of Sadana [15]. deactivation coefficient, k1 , was not changed by the addi-
tion of trehalose. However, the k2 deactivation coefficient
decreased from 1.2 to 0.2, indicating that trehalose reduced
the deactivation rate of the second step. Thus, the half-lives
model, represented by Scheme 1 (Section 2), correlated well
obtained increased about eightfold, with trehalose.
with experimental data. At 70 ◦ C in the absence of trehalose,
At 70 ◦ C, the deactivation coefficients k1 and k2 , decreased
the data fits well with a single-step, first-order deactivation
significantly by adding trehalose. This showed that the stabi-
mechanism (Eq. (4)).
lization effect of trehalose was due to a decrease in the deac-
The deactivation parameters for Xyl I obtained at 60 and
tivation rates of both steps of the proposed scheme (Scheme
70 ◦ C, with and without trehalose, using a three-step model
2). The half-life increased slightly (about twofold) when tre-
are presented in Table 1. At 60 ◦ C, Xyl I gave a β1 value
halose was added. However, a final state, E2 , with a residual
greater than 1, indicating that the intermediate state E1 had
a higher specific activity than the initial state, E. The value
obtained for β2 was less than 1, and thus the intermediate Table 2
state E1 was inactivated to an intermediate state E2 , with a Xyl II deactivation parameters, at 60 and 70 ◦ C, obtained from application
specific activity even lower that the initial enzyme activity. of the kinetic deactivation schemes proposed (Schemes 2 and 3)
The deactivation process lead to final state Ed . Addition of Deactivation 60 ◦ C 60 ◦ C + 70 ◦ C 70 ◦ C +
trehalose, at 60 ◦ C, did not affect the deactivation coefficient parameters trehalose trehalose
rates k1 and k2 , although it induced a decrease in k3 . This k1 (h−1 ) 7.31 7.11 19.46 4.32
slowed the third deactivation rate step by about 13-fold. At k2 (h−1 ) 1.18 0.20 2.29 0.00
70 ◦ C, in the presence of trehalose, the Xyl I deactivation β1 0.61 0.83 0.57 0.30
β2 n.a. 0.01 n.a. 0.01
kinetics changed from a one-step to a three-step mechanism
(Scheme 1). The addition of trehalose gave an increase of t1/2 (h) 0.35 2.78 0.14 0.29
10 ◦ C in the half-life. That is, the half-life at 70 ◦ C was n.a.: not applicable.
282 P. Sá-Pereira et al. / Enzyme and Microbial Technology 34 (2004) 278–282
activity of about 35% was obtained. Trehalose made it pos-
sible to obtain a final state, E2 , with residual activity at both
temperatures. Without trehalose, following 3 h of incubation
at 70 ◦ C, the final state, Ed , had no activity.
4. Conclusions
In this work, for both Xyl I and Xyl II, trehalose increased
the initial activities and induced the appearance of interme-
diate stages with residual activities, leading to the increase
of enzyme half-lives. Trehalose stabilized both xylanases by
reducing the deactivation rates of the steps proposed in the
deactivation schemes, thus leading to final enzyme states
with activity.
The deactivation data obtained for thermal inactivation of
the two xylanases, with and without trehalose was treated
using a model involving several deactivation steps. The pro-
posed sequential deactivation models involving more than
one-step, correlated well with the experimental data, sug-
gesting the existence of multiple enzyme forms. The models
provide valuable insights into xylanase deactivation mecha-
nisms.
Such a strategy to explain and modulate the experimental
results by proposing suitable schemes for xylanase thermal
deactivation had not been previously reported. The stabiliz-
ing effect of trehalose could allow applications of these xy-
lanases in industrial processes that are thermo-dependent.
References
[1] Törrönen A, Rouvinen J. Structural and functional properties of low
molecular weight endo-1,4-␤-xylanases. J Biotechnol 1997;57:137–
49.
[2] Sá-Pereira P, Paveia H, Costa-Ferreira M, Aires-Barros M. A new
look at xylanases: an overview of purification strategies. Mol
Biotechnol 2003;24(3):257–81.
[3] Illanes A. Stability of biocatalysts. Electron J Biotechnol
1999;2(1):7–15.
[4] Nikolova PV, Creagh AL, Duff SJ, Haynes CA. Thermostability
and irreversible activity loss of exoglucanase/xylanase Cex from
Cellulomonas fimi. Biochemistry 1997;36(6):1381–8.
[5] Davoodi J, Wakarshuk WW, Surewicz WK, Carey PR. Scan-rate
dependence in protein calorimetry: the reversible transitions of
Bacillus circulans xylanase and a disulfide-bridge mutant. Protein
Sci 1998;7:1538–44.
[6] George SP, Ahamad A, Rao MB. A novel thermostable xylanase
from Thermomonospora sp.: influence of additives on thermostability.
Bioresour Technol 2001;7(3):221–4.
[7] Sá-Pereira P, Duarte J, Costa-Ferreira M. Electroelution as a simple
and fast protein purification method: isolation of an extracellular
xylanase from Bacillus sp. CCMI 966. Enzyme Microb Technol
2000;27:95–9.
[8] Sá-Pereira P, Costa-Ferreira M, Aires-Barros MR. Enzymatic
properties of a neutral endo-1,3(4)-␤-xylanase Xyl II from Bacillus
subtilis. J Biotechnol 2002;94:265–75.
[9] Sá-Pereira P, Mesquita A, Duarte JC, Aires-Barros MR, Costa-
Ferreira M. Rapid production of thermostable cellulase-free xylanase
by a strain of Bacillus subtilis and its properties. Enzyme Microb
Technol 2002;30(7):924–33.
[10] Colacino F, Crichton RR. 8-Enzyme thermostabilization: the state of
the art. Biotechnol Genet Eng Rev 1997;14:211–77.
[11] Sá-Pereira P. Characterization of a ␤-1,4-endoglucanase codifying
gene from Cellvibrio mixtus, expressed in Escherichia coli (abstract
in English). M.Sc. Thesis, Technical University of Lisbon; 1994.
[12] Miller L. Use of dinitrosalicylic acid reagent for determination of
reducing sugar. Anal Chem 1959;31:426–8.
[13] Bailey M, Biely P, Poutanen K. Interlaboratory testing of methods
for assay of xylanase activity. J Biotechnol 1992;23:257–70.
[14] Park MT, Lee MS, Choi JY, Kim SC, Lee GM. Orthophosphate anion
enhances the stability and activity of endoxylanase from Bacillus sp.
Biotechnol Bioeng 2001;72(4):434–40.
[15] Sadana A. Biocatalysis. Fundamental of enzyme deactivation kinetics.
In: Gretchen KC, editor. Prentice-Hall; 1991. p. 33–95.
[16] Melo EP, Faria TQ, Martins LO, Gonçalves AM, Cabral
JMS. Cutinase unfolding and stabilization by trehalose and
mannosylglycerate. Proteins 2001;42:542–52.
[17] Sola-Penna M, Meyer-Fernandes JR. Stabilization against thermal
inactivation promoted by sugars on enzyme structure and function:
why is trehalose more effective than other sugars. Arch Biochem
Biophys 1998;360(1):10–4.