Báo cáo khoa học: Role of disulfide bonds in goose-type lysozyme potx

Role of disulfide bonds in goose-type lysozyme
Shunsuke Kawamura
1
, Mari Ohkuma
1
, Yuki Chijiiwa
1
, Daiki Kohno
1
, Hiroyuki Nakagawa
1
, Hideki
Hirakawa
2
, Satoru Kuhara
2,3
and Takao Torikata
1
1 Department of Bioscience, School of Agriculture, Tokai University, Aso, Kumamoto, Japan
2 Graduate School of Systems Life Sciences, Kyushu University, Hakozaki, Higashi-ku, Fukuoka, Japan
3 Graduate School of Genetic Resource Technology, Kyushu University, Hakozaki, Higashi-ku, Fukuoka, Japan
Lysozyme, one of the best characterized carbohydro-
lases, cleaves the glycosidic linkage between N-acetyl-
glucosamine (GlcNAc) and N-acetylmuramic acid in
bacterial cell walls. This enzyme is classified into
six types, chicken type (C-type) [1–3], phage type
(T4-type) [4,5], goose type (G-type) [6–8], invertebrate
type [9–11], bacteria type [12], and plant type [13], on
the basis of the similarity in amino acid sequences.
These different classes of lysozymes have overall simi-
larities in tertiary structure [7,14–17], although their

amino acid sequences are almost entirely different.
Much information on the structural properties and
enzymatic mechanisms of C-type and T4-type lyso-
zymes has been accumulated thus far. In particular,
hen egg-white lysozyme and human lysozyme, which
belong to a class of C-type lysozymes with four disul-
fide bonds, and T4 phage lysozyme with no disulfide
bonds, have been intensively studied as model proteins
for elucidating enzymatic function and protein stabil-
ity. In contrast to C-type and T4-type lysozymes,
information on G-type lysozyme is limited. In verte-
brates, the primary structure has been reported for five
Keywords
disulfide bonds; goose-type lysozyme;
ostrich; site-directed mutagenesis; structural
stability
Correspondence
S. Kawamura, Department of Bioscience,
School of Agriculture, Tokai University, Aso,
Kumamoto 869-1404, Japan
Fax: +81 967 67 3960
Tel: +81 967 67 3918
E-mail:
(Received 15 November 2007, revised 3
March 2008, accepted 25 March 2008)
doi:10.1111/j.1742-4658.2008.06422.x
The role of the two disulfide bonds (Cys4–Cys60 and Cys18–Cys29) in the
activity and stability of goose-type (G-type) lysozyme was investigated
using ostrich egg-white lysozyme as a model. Each of the two disulfide
bonds was deleted separately or simultaneously by substituting both Cys

residues with either Ser or Ala. No remarkable differences in secondary
structure or catalytic activity were observed between the wild-type and
mutant proteins. However, thermal and guanidine hydrochloride unfolding
experiments revealed that the stabilities of mutants lacking one or both of
the disulfide bonds were significantly decreased relative to those of the
wild-type. The destabilization energies of mutant proteins agreed well with
those predicted from entropic effects in the denatured state. The effects of
deleting each disulfide bond on protein stability were found to be approxi-
mately additive, indicating that the individual disulfide bonds contribute to
the stability of G-type lysozyme in an independent manner. Under reducing
conditions, the thermal stability of the wild-type was decreased to a level
nearly equivalent to that of a Cys-free mutant (C4S ⁄ C18S ⁄ C29S ⁄ C60S) in
which all Cys residues were replaced by Ser. Moreover, the optimum tem-
perature of the catalytic activity for the Cys-free mutant was downshifted
by about 20 °C as compared with that of the wild-type. These results indi-
cate that the formation of the two disulfide bonds is not essential for the
correct folding into the catalytically active conformation, but is crucial for
the structural stability of G-type lysozyme.
Abbreviations
(GlcNAc)
n
, b-1,4-linked oligosaccharide of GlcNAc with a polymerization degree of n; C-type, chicken type; GEL, goose egg-white lysozyme;
GlcNAc, N-acetylglucosamine; G-type, goose type; MD, molecular dynamics; OEL, ostrich egg-white lysozyme; T4-type, phage type; b-ME,
b-mercaptoethanol.
2818 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
G-type lysozymes, i.e. those from ostrich [18], black
swan [19], embden goose [6], cassowary [20], and rhea
[21], and five from chicken [22], flounder [23], carp
[24], salmon [25], and orange-spotted grouper [26].
Additionally, Irwin & Gong [27] reported that mam-

mals and zebrafish carry two G-type lysozyme genes.
Recently, invertebrate G-type lysozyme genes and ⁄ or
enzyme activity from scallop [28,29] and tunicate [30]
were also reported. G-type lysozyme differs from the
C-type in that it is much more specific for peptide-
substituted substrate [31]. C-type lysozyme hydrolyzes
a homopolymer (chitin) effectively, whereas G-type
lysozyme is a poor catalyst of the hydrolysis of this
substrate. The differences in substrate specificity
between these lysozymes and the mechanistic details of
the catalytic reaction of G-type lysozyme remain
unclear. Previously, Honda & Fukamizo reported the
mode of binding of GlcNAc oligomer to goose egg-
white lysozyme (GEL), and postulated that GEL has
six substrate-binding subsites (sites B–G) [32]. This
subsite structure was partly visualized in terms of the
crystal structure of the GEL–(GlcNAc)
3
complex [14];
however, part of the subsite structure (sites E–G)
remains unknown.
On the basis of sequence comparison of G-type lyso-
zymes, we have shown that the amino acid sequences
of three a-helices (a5, a7, and a8) are highly conserved
in this enzyme group [20,21]. These three a-helices are
located at the center of the protein molecule, and form
a hydrophobic core in the overall structure of G-type
lysozyme (Fig. 1). Recently, using ostrich egg-white
lysozyme (OEL) as a model, we demonstrated the
involvement of Glu73 on a5 (Ala64–Glu73) as a criti-

cal catalytic residue and also indicated the crucial role
of Glu73 in the structural stability of G-type lysozyme,
probably through the interhelical hydrogen bond with
Tyr169 on a8 (Tyr169–Gln182) [33]. These observa-
tions suggest that the core elements (a5, a7, and a8)
play an important role in the maintenance of the
three-dimensional structure of G-type lysozyme.
In addition to the three a-helices, the Cys4–Cys60
and Cys18–Cys29 disulfide bonds (numbering from
bird sequences), which are located in the N-terminal
region, are completely conserved in avian and mamma-
lian G-type lysozymes, although another three Cys res-
idues are conserved from mouse to human [20,21,27].
In the crystal structure of GEL (Fig. 1), which shares
83% amino acid identity with OEL, these two disulfide
bonds are located on the molecular surface, although
Cys60 is partially buried in the interior of the protein.
It has been shown that the integrity of the native
three-dimensional structure of many proteins is pro-
moted by the presence of disulfide bonds, because
removal of one or more of these linkages results in a
reduction in the stability of the native state relative to
the denatured state [34–44]. On the other hand, none
of the four disulfide bonds was reported to be impor-
tant in stabilizing the native structure of the Pseudoal-
teromonas haloplanktis a-amylase [45]. It was also
shown that the Cys191–Cys220 disulfide bond, which
is highly conserved in the trypsin family of serine pro-
teases, is not essential for the catalytic function, struc-
ture and stability of trypsin [46]. Therefore, analysis of

the role of the disulfide bonds in activity and stability
will be useful for our understanding of the structure–
function relationship of G-type lysozyme. The present
article describes a site-directed mutational analysis of
OEL to address the role of the disulfide bonds in
G-type lysozyme.
Results and Discussion
Choice of residues for mutagenesis
A striking difference within G-type lysozymes is the
variation in Cys content. There are four conserved Cys
residues in avian and mammalian G-type lysozymes,
which form two intramolecular disulfide bonds in the
mature proteins [20,21,27]. The crystal structure of
GEL shows that the Cys18–Cys29 disulfide bond con-
nects the N-terminus of a-helix 1 with a loop between
a-helices 1 and 2, and the Cys4–Cys60 disulfide bond
connects the N-terminal long loop with a loop between
a-helices 4 and 5. The G-type lysozymes found in fish
have either no Cys residue, as in flounder and grouper
[23,26], one, as in carp and salmon [24,25], or two (no
potential to form an intramolecular disulfide bond),
as in zebrafish [27]. The absence of intramolecular
Fig. 1. The three-dimensional structure of GEL. The structure was
created using the coordinate file, Protein Data Bank entry 153L
[14]. The two disulfide bonds (Cys4–Cys60 and Cys18–Cys29) and
three a-helices (a5, a7, and a8) are shown in blue and green,
respectively. The side chain of Glu73 is also shown in red. The
figure was generated using
MOLSCRIPT (v. 2.1.2).
S. Kawamura et al. Disulfide bonds in goose-type lysozyme

FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2819
disulfide bonds seems to be a common characteristic
among the fish G-type lysozymes. The G-type lyso-
zymes in invertebrates have six to 13 Cys residues
other than the four Cys residues conserved in avian
and mammalian G-type lysozymes, which renders the
disulfide patterns of invertebrate G-type lysozymes
quite different from those of the bird and mammalian
lysozymes [28–30]. As the locations of disulfide bonds
in invertebrate G-type lysozymes have not yet been
identified, we decided to focus our attention on the
Cys18–Cys29 and Cys4–Cys60 disulfide bonds, which
are absolutely conserved in avian and mammalian
G-type lysozymes.
Expression and characterization of mutant
proteins
To investigate the contribution of the two disulfide
bonds to the activity and stability of G-type lysozyme,
three mutant proteins (C4S ⁄ C60S, C18S ⁄ C29S, and
C4S ⁄ C18S ⁄ C29S ⁄ C60S), in which each of the disulfide
bonds was singly or together disrupted by Cys fi Ser
mutations, were initially constructed. The Ser residue
was chosen because it is structurally similar to Cys
except that it contains a hydroxyl group instead of a
thiol group. The mutant proteins were expressed and
purified in the same manner as used for the wild-type
[33]. The yields of the mutant proteins were com-
parable to that of the wild-type, approximately
60–70 mgÆL
)1

. The purified proteins were found to be
homogeneous on analysis by SDS ⁄ PAGE, and gave a
single peak on RP-HPLC (data not shown). The N-ter-
minal sequence of each mutant protein was determined
to be Ser-Arg-Thr-Gly, which coincided with that of
the wild-type, indicating that each mutant was correctly
processed at the C-terminus of the a-factor signal. The
integrity of the mutant proteins was confirmed by mea-
surements of far-ultraviolet CD. The CD spectra of the
three mutants were almost indistinguishable from that
of the wild-type (Fig. 2), indicating that the backbone
conformation of the mutant proteins is practically the
same as that of the wild-type. Thus, it appears that
none of the disulfide bonds are critically important to
the folding process of G-type lysozyme.
Effects of mutations on catalytic activity
We previously reported that the recombinant OEL
preferentially hydrolyzes the third glycosidic linkage
from the nonreducing end of (GlcNAc)
6
, and the
cleavage pattern seen for (GlcNAc)
5
is similar to that
seen for (GlcNAc)
6
[47]. To examine the effects of the
mutations on the catalytic activity of G-type lysozyme,
we initially analyzed the activities of the wild-type and
its mutants by monitoring the enzyme-catalyzed lysis

of Micrococcus luteus cells, which is a high molecular
mass polymeric substrate with a highly negative
charge. Mutant C4S ⁄ C60S had lytic activity to the
same extent as the wild-type, exhibiting 99.0% activity.
The lytic activities of mutants C18S ⁄ C29S and
C4S ⁄ C18S ⁄ C29S ⁄ C60S were 76.5% and 70.6% of that
of the wild-type, respectively (data not shown). As the
substrate used for lytic activity is chemically heteroge-
neous, the activities of the wild-type and three mutant
proteins were more precisely evaluated by measuring
the enzyme-catalyzed hydrolysis of (GlcNAc)
5
(Fig. 3).
Consistent with the results obtained in the M. luteus
assay, the wild-type and mutant C4S ⁄ C60S hydrolyzed
the initial substrate (GlcNAc)
5
almost completely after
240 min of reaction, and (GlcNAc)
5
was hydrolyzed
mainly to (GlcNAc)
2
+ (GlcNAc)
3
with much less
cleavage to (GlcNAc)
1
+ (GlcNAc)
4

. Mutants
C18S ⁄ C29S and C4S ⁄ C18S ⁄ C29S ⁄ C60S hydrolyzed
(GlcNAc)
5
to produce (GlcNAc)
2
and (GlcNAc)
3
,as
in the case of the wild-type, although the overall rates
of hydrolysis were slightly affected by each of the
mutations: the two mutants took 420 min to hydrolyze
most of the (GlcNAc)
5
. These results are consistent
with the fact that each of the two disulfide bonds is
–30
–20
–10
0
10
20
30
40
50
60
190 210 230 250
Wavelen
g
th (nm)

[θ] × 10
3
(deg cm
2
· dmol
–1
)
Wild type
C18S/C29S
C4S/C60S
C4S/C18S/C29S/C60S
Fig. 2. CD spectra of the wild-type and its three mutant proteins in
the far-UV region.
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2820 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
far apart from the catalytic glutamate, Glu73, in the
crystal structure of GEL (even the nearest two resi-
dues, Cys18 and Glu73, are located 16.8 A
˚
apart from
each other). Although the reason for the slight reduc-
tion in the activity of the two mutants is presently
obscure, none of the disulfide bonds is considered to
be critically important to the catalytic function of
G-type lysozyme. This observation, together with the
result obtained by CD analysis, implies that G-type
lysozyme can fold and function in the absence of both
disulfide bonds. This is consistent with the findings
that the flounder and salmon G-type lysozymes, which
have no disulfide bonds in their native forms, and the

scallop G-type lysozyme, which has six Cys residues
other than the well-conserved four Cys residues in
birds, possessed lytic activity against Micrococcus
lysodeikticus [23,25,29].
Effects of mutations on stability
Disulfide bonds have been suggested to play an impor-
tant role in maintaining structural integrity and protein
stabilization. This conclusion has been supported by
characterization of mutants of various proteins in
which disulfide bonds have been either deleted or mod-
ified. For example, Pace et al. [34] reported that dis-
ruption of one and two disulfide bonds in
ribonuclease T1 caused decreases in conformational
stability by 3.4 and 7.2–9.3 kcalÆmol
)1
, respectively.
Many other examples can be found in the reviews by
Wetzel [48] and Bets [49]. It is widely accepted that the
stabilizing effect of a disulfide bond can be attributed
principally to the destabilization of the unfolded form
by the loss of conformational entropy imposed by the
crosslink [34,36,39,50–54]. However, the mechanism of
protein stabilization by disulfide bond formation is
difficult to resolve, because the disulfide bond may
influence the enthalpy and entropy of both the native
and unfolded states of the protein [49].
The thermal unfolding of the wild-type protein has
been investigated using CD and fluorescence spectros-
copy, and led to the observation that the unfolding
transition of the wild-type is well represented by a two-

state mechanism at pH 5.0 in the presence of 0.5 m
guanidine hydrochloride [33]. Like the wild-type pro-
tein, the three mutants (C18S ⁄ C29S, C4S ⁄ C60S, and
C4S ⁄ C18S ⁄ C29S ⁄ C60S) reversibly unfolded in a single
cooperative fashion under these conditions. The ther-
mal unfolding curves of the wild-type and three mutant
proteins obtained with fluorescence measurements are
shown in Fig. 4A. Replacing Cys18 and Cys29 or Cys4
and Cys60 with a pair of Ser residues had significant
effects on the thermal unfolding of the mutant proteins.
The T
m
values for mutants C18S ⁄ C29S and C4S ⁄ C60S
were decreased by 6.3 °C and 9.5 °C, respectively, as
compared to 60.6 °C for the wild-type (Table 1). The
Cys mutations reduced the thermostability of the pro-
teins by 3.11 and 4.29 kcalÆmol
)1
for mutants
C18S ⁄ C29S and C4S ⁄ C60S, respectively, at 60.6 °C.
The combination of these destabilizing mutations
(C4S ⁄ C18S ⁄ C29S ⁄ C60S) caused a further decrease in
thermostability relative to the wild-type by 15.3 °C
(DDG: )6.14 kcalÆmol
)1
) (Table 1).
The contribution of the disulfide bonds to the struc-
tural stability of OEL was further assessed by means
of unfolding experiments with guanidine hydrochloride
as a denaturant. Figure 4B shows the guanidine hydro-

chloride-induced unfolding curves of the wild-type and
Fig. 3. Time course plots of (GlcNAc)
5
degradation by the wild-type and its three
mutant proteins. The enzymatic reaction
was performed in 10 m
M sodium acetate
buffer (pH 4.0) at 40 °C. Numerals in the
figures are the polymerization degrees of
the reaction product species. Relative error
indicates the recovery of the observed value
at each reaction time calculated as
described in Experimental procedures. The
solid lines were drawn by roughly following
the experimental data points.
S. Kawamura et al. Disulfide bonds in goose-type lysozyme
FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2821
mutant proteins obtained with fluorescence measure-
ments. The transitions of the three mutant proteins
were highly cooperative. The unfolding transitions of
the mutant proteins occurred at lower concentrations
of guanidine hydrochloride than that of the wild-type:
the C
m
values were reduced, as compared with that of
the wild-type, by 0.47, 0.70 and 1.07 m for mutants
C18S ⁄ C29S, C4S ⁄ C60S, and C4S ⁄ C18S ⁄ C29S ⁄ C60S,
respectively (Table 2). The DG
H2O
values of unfolding

indicated that the three mutants, C18S ⁄ C29S,
C4S ⁄ C60S, and C4S⁄ C18S ⁄ C29S ⁄ C60S, were destabi-
lized by 2.33, 3.33 and 4.86 kcalÆmol
)1
, respectively, at
0 m guanidine hydrochloride in comparison to the
wild-type.
The decrease in the stability of the least stable
C4S ⁄ C18S ⁄ C29S ⁄ C60S mutant was further confirmed
by CD measurements: its thermal and guanidine
hydrochloride transition curves derived from the CD
and fluorescence data completely coincided, and the
T
m
and C
m
values obtained with CD were in good
agreement with the data determined with fluorescence
(Fig. 5, and Tables 1 and 2). The coinciding transitions
derived from these two different methods also indicate
that the Cys-free mutant undergoes thermal and guani-
dine hydrochloride-induced denaturation, which is con-
sidered to be a reversible two-state process as observed
in the wild-type. It is thus likely that the fold-
ing ⁄ unfolding pathway of G-type lysozyme does not
significantly change in the absence of one or both of
the disulfide bonds, which supports the notion that the
presence of a complete set of disulfide bonds is not
required for the folding process of G-type lysozyme.
To corroborate the importance of the disulfide

bonds in the structural stability of OEL, two mutant
proteins (C18A ⁄ C29A and C4A ⁄ C60A), in which
Cys18 and Cys29 or Cys4 and Cys60 were replaced
by Ala, respectively, were constructed and analyzed
with respect to their guanidine hydrochloride dena-
turation (Fig. 4B and Table 2). It was found that
mutants C18A⁄ C29A and C4A ⁄ C60A exhibited the
same stabilities as mutants C18S ⁄ C29S and
C4S ⁄ C60S, respectively, which strongly suggests that
the observed destabilization arising from the Cys to
Ser or Ala mutations is not due to negative side
–0.2
0
0.2
0.4
0.6
0.8
1.0
1.2
30 35 40 45 50 55 60 65 70 75 80
Temperature (ºC)
Fraction unfolded
Wild-type
C18S/C29S
C4S/C60S
C4S/C18S/C29S/C60S
Wild-type (reduced)
–0.2
0
0.2

0.4
0.6
0.8
1.0
1.2
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
GdnHCl (M)
Fraction unfolded
Wild-type
C18S/C29S
C4S/C60S
C4S/C18S/C29S/C60S
C18A/C29A
C4A/C60A
A
B
Fig. 4. Thermal and guanidine hydrochloride (GdnHCl)-induced
unfolding curves of the wild-type and its mutant proteins obtained
by fluorescence measurements. (A) and (B) show the thermal and
guanidine hydrochloride unfolding curves of the wild-type and its
mutant proteins, respectively. Experimental details are described in
Experimental procedures. The thermal unfolding curve of the wild-
type treated with 0.1
M b-ME for reduction is indicated as ‘wild-
type (reduced)’.
Table 1. Parameters characterizing the thermal denaturation of the wild-type and the three mutants. Thermodynamic parameters were cal-
culated from the thermal unfolding curves presented in Figs 4A and 5A. All values are the averages of at least two determinations. Data for
the wild-type protein were reported by Kawamura et al. [33]. Errors are within ± 0.3 °C for T
m
, ± 5.5 kcalÆmol

)1
for DH
m
, and ± 0.016 kca-
lÆmol
)1
K for DS
m
for the wild-type protein, determined from four independent experiments. Flu, fluorescence.
Protein Method DH
m
(kcalÆmol
)1
) DS
m
(kcalÆmol
)1
ÆK) T
m
(°C) DT
m
(°C) DDG (kcalÆmol
)1
)
Wild-type Flu 170.1 0.510 60.6 – –
C18S ⁄ C29S Flu 161.8 0.494 54.3 )6.3 )3.11
C4S ⁄ C60S Flu 145.5 0.452 51.1 )9.5 )4.29
C4S ⁄ C18S ⁄ C29S ⁄ C60S Flu 127.6 0.401 45.3 )15.3 )6.14
Wild-type (reduced) Flu 107.6 0.337 46.2 )14.4 )4.85
Wild-type CD 166.9 0.501 60.5 – –

C4S ⁄ C18S ⁄ C29S ⁄ C60S CD 125.0 0.393 45.3 )15.2 )5.97
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2822 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
effects of the introduction of the present mutations,
but is due mainly to deletion of the disulfide
bond(s).
We also examined the thermal stability of the wild-
type treated with 0.1 m b-mercaptoethanol (b-ME) for
reduction and compared it with those of the wild-type
and mutant C4S ⁄ C18S ⁄ C29S ⁄ C60S (Fig. 4A and
Table 1). The reduced wild-type exhibited a marked
decrease in thermostability relative to the nonreduced
wild-type of 14.4 °C(DDG: )4.85 kcalÆmol
)1
). The T
m
value for the reduced wild-type was 46.2 °C, a value
almost identical to that of the Cys-free mutant
(T
m
45.3 °C). No change was observed for the Cys-free
mutant when it was melted in the presence of 0.1 m
b-ME (data not shown).
In addition, the temperature dependence of the cata-
lytic activity against (GlcNAc)
5
for the Cys-free
mutant was examined, and the result was compared
with that for the wild-type (Fig. 6). The wild-type
protein exhibited the highest activity at 60 °C, and the

activity was drastically reduced at 65 °C or above. In
contrast, the optimum temperature for the Cys-free
mutant was decreased to 40 °C, which was about
20 °C lower than that of the wild-type, and a remark-
able drop of the activity was observed above this
temperature. All of these results indicate that the two
disulfide bonds are directly involved in the structural
stability of G-type lysozyme. Interestingly, the
optimum temperature of the lytic activity for the floun-
der G-type lysozyme was shown to be 25 °C by lyso-
plate assay [23]. Our results suggest that the low
optimum temperature observed for the flounder lyso-
zyme could be a consequence of the absence of the
two intrachain disulfide bonds. In the case of inverte-
brate G-type lysozymes, the high content of Cys
residues suggests their importance in protein stability.
We noted that the decrease in the T
m
value for mutant
C4S ⁄ C18S ⁄ C29S ⁄ C60S ()15.3 °C) agreed well with the
sum of the decreases in the T
m
values for mutants
C4S ⁄ C60S and C18S ⁄ C29S ()15.8 °C). The DC
m
value
for the Cys-free mutant ()1.07 m) was also found to be
Table 2. Parameters characterizing the guanidine hydrochloride denaturation at pH 5.0 and 30 °C. Parameters were calculated from the gua-
nidine hydrochloride unfolding curves presented in Figs 4B and 5B. All values are the averages of at least two determinations. Data for the
wild-type protein were reported by Kawamura et al. [33]. Errors are within ± 0.03

M for C
m
, ± 0.09 kcalÆmol
)1
ÆM for m, and ± 0.08 kcalÆmol
)1
for DG
H2O
for the wild-type protein, determined from four independent experiments. Flu, fluorescence.
Protein Method m (kcalÆmol
)1
ÆM) C
m
(M) DC
m
(M) DG
H2O
(kcalÆmol
)1
) DDG
H2O
(kcalÆmol
)1
)
Wild-type Flu 5.48 2.21 – 12.11 –
C18S ⁄ C29S Flu 5.67 1.74 –0.47 9.88 )2.33
C4S ⁄ C60S Flu 5.84 1.51 )0.70 8.78 )3.33
C4S ⁄ C18S ⁄ C29S ⁄ C60S Flu 6.38 1.14 )1.07 7.25 )4.86
C18A ⁄ C29A Flu 5.68 1.75 )0.46 9.93 )2.18
C4A ⁄ C60A Flu 5.90 1.51 )0.70 8.93 )3.18

Wild-type CD 5.32 2.20 – 11.73 –
C4S ⁄ C18S ⁄ C29S ⁄ C60S CD 6.37 1.14 )1.06 7.25 ) 4.48
–0.2
0
0.2
0.4
0.6
0.8
1.0
1.2
Temperature (ºC)
Fraction unfolded
Wild-type (Flu)
Wild-type (CD)
C4S/C18S/C29S/C60S (Flu)
C4S/C18S/C29S/C60S (CD)
–0.2
0
0.2
0.4
0.6
0.8
1.0
1.2
GdnHCl(M)
Fraction unfolded
Wild-type (Flu)
Wild-type (CD)
C4S/C18S/C29S/C60S (Flu)
C4S/C18S/C29S/C60S (CD)

30 35 40 45 50 55 60 65 70 75 80
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
A
B
Fig. 5. Thermal and guanidine hydrochloride (GdnHCl)-induced
unfolding curves of the wild-type and mutant C4S ⁄ C18S ⁄
C29S ⁄ C60S obtained by CD and fluorescence measurements. (A)
and (B) show the thermal and guanidine hydrochloride unfolding
curves, respectively. Experimental details are described in Experi-
mental procedures. The unfolding curves of the wild-type and
mutant C4S ⁄ C18S ⁄ C29S ⁄ C60S obtained with CD are indicated as
‘wild-type (CD)’ and ‘C4S ⁄ C18S ⁄ C29S ⁄ C60S (CD)’, respectively,
and those of the wild-type and mutant C4S ⁄ C18S ⁄ C29S ⁄ C60S
obtained with fluorescence are indicated as ‘wild-type (Flu)’ and
‘C4S ⁄ C18S ⁄ C29S ⁄ C60S (Flu)’, respectively.
S. Kawamura et al. Disulfide bonds in goose-type lysozyme
FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2823
nearly equal to the sum of the DC
m
values for the
corresponding double mutants ()1.17 m). These findings
indicate that the effects on the protein stability of delet-
ing the disulfide bonds are approximately additive. As
the locations of the two disulfide bonds are far from
each other in the crystal structure of GEL (Fig. 1), the
individual disulfide bonds probably contribute to the
structural stability of G-type lysozyme in an indepen-
dent manner rather than in a cooperative manner.
We can assume two mechanisms by which the disul-
fide bonds affect stability. One is the entropic effect on

the unfolded forms, and the other is the effect of the
amino acid substitutions on the native forms. Theoreti-
cal approaches have suggested that the entropic effect
(DS) will be related to the size of the loop enclosed by
the crosslink (n). A commonly used approximation is
that derived by Pace et al. [34]: DS = )2.1–1.5 ·
R · ln n (calÆmol
)1
ÆK), where R is the gas constant.
According to the equation, the increases in entropy in
the unfolded proteins caused by deletion of the Cys18–
Cys29 and Cys4–Cys60 disulfide bonds are 9.51 and
14.15 calÆmol
)1
ÆK, respectively. In terms of free energy,
the expected entropic destabilization ()TDS)of
mutants C18S ⁄ C29S and C4S ⁄ C60S is )3.17 and
)4.72 kcalÆmol
)1
, respectively, at the T
m
value of the
wild-type at pH 5.0 (60.6 °C). These theoretical values
are in good agreement with the observed DDG values
for mutants C18S ⁄ C29S () 3.11 kcalÆmol
)1
) and
C4S ⁄ C60S ()4.29 kcalÆmol
)1
) (Table 1). The sum of

the theoretical values for the two double mutants
()7.40 kcalÆmol
)1
) is also comparable to the observed
DDG value for the Cys-free mutant ()6.14 kcalÆmol
)1
).
These results suggest that the increase in entropy of
the unfolded state is a dominant factor determining
the DDG. In the case of the guanidine hydrochloride
denaturation, the expected entropic destabilization at
30 °Cis)2.88 and )4.29 kcalÆmol
)1
for mutants
C18S ⁄ C29S and C4S ⁄ C60S, respectively. These theo-
retical values are close to but lower than the observed
DDG
H2O
values for mutants C18S ⁄ C29S ( )2.33 kcalÆ
mol
)1
) and C4S ⁄ C60S ()3.33 kcalÆmol
)1
) (Table 2).
However, as a small error in m results in a large devia-
tion in DG
H2O
, due to a long extrapolation to 0 m
denaturant, and as the concentration of guanidine
hydrochloride at the midpoint of the denaturation

(C
m
) is regarded as the most reliable parameter for
estimation of protein stability [55], the destabilization
energies (DDG
D
) of the three mutants (C18S ⁄ C29S,
C4S ⁄ C60S, and C4S ⁄ C18S ⁄ C29S ⁄ C60S) were recalcu-
lated using the equation [56] DDG
D
= m¢(C¢
m
) C
m
),
where m¢ and C¢
m
are the values for the mutant, and
C
m
is the value for for the wild-type. The destabiliza-
tion energies thus obtained are )2.66 and )4.09 kcalÆ
mol
)1
for mutants C18S ⁄ C29S and C4S ⁄ C60S, respec-
tively, which are in good agreement with the respective
theoretical values. The DDG
D
value for the Cys-free
mutant ()6.83 kcalÆ mol

)1
) is also nearly equal to the
sum of the theoretical values for the corresponding
double mutants ()7.17 kcalÆmol
)1
). These findings sug-
gest that deletion of one or both of the disulfide bonds
increases the conformational entropy of the unfolded
state, thereby stabilizing the unfolded state and, as a
result, destabilizing the protein thermodynamically.
This is in agreement with the proposal that stabiliza-
tion of proteins by disulfide bonds can be essentially
ascribed to a decrease of the conformational entropy
of the unfolded state. Cooper et al. [36] showed that
the reduction in T
m
resulting from selective disruption
and modification of the Cys6–Cys127 disulfide bond of
hen egg-white lysozyme is totally attributable to an
increase in the entropy difference between the native
and denatured states. In contrast, Doig & Williams
[57] reported, from a thermodynamic analysis on six
small proteins, that the dominant effect of disulfide
bonds on stability is enthalpic. The thermodynamic
characterization of a mutant human lysozyme lacking
the Cys77–Cys95 disulfide bond showed that the
decrease in DG
H2O
for the mutant protein was caused
Tem

p
erature (ºC)
Concentration of hydrolyzed (GlcNAc)
5
at each
temperature (mM)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
10 20 30 40 50 60 70 80 90
Wild-type
C4S/C18S/C29S/C60S
Fig. 6. Temperature dependence of the catalytic activity for the
hydrolysis of (GlcNAc)
5
by the wild-type and mutant
C4S ⁄ C18S ⁄ C29S ⁄ C60S. The concentrations of hydrolyzed (Glc-
NAc)
5
at each temperature were calculated by subtracting the
concentration of the remaining (GlcNAc)
5

from that of the initial
(GlcNAc)
5
substrate.
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2824 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
by both entropic and enthalpic factors [37]. Therefore,
detailed calorimetric measurements of the mutant pro-
teins will be required to determine the thermodynamic
parameters precisely.
Heat inactivation
Next, we examined the stability against irreversible
heat inactivation for the wild-type and mutant
C4S ⁄ C18S ⁄ C29S ⁄ C60S. Unexpectedly, no difference
was observed in the irreversible heat inactivation of
these enzymes (data not shown). Both enzymes
regained 60% and 20% of their activities after 2 h at
80 °C and 90 °C, respectively. This is consistent with
the recent finding that the salmon G-type lysozyme
with no disulfide bonds, which has an optimum tem-
perature of 30 °C, regained 30% of its activity immedi-
ately after incubation at 90 °C for 3 h [25]. As
suggested by Kyomuhendo et al. [25], it appears that
OEL, as in the case of the salmon protein, possesses
an extraordinary capacity to correctly refold its struc-
ture. When considered together with the result
obtained for the salmon protein, this indicates that,
probably, no disulfide bonds are required for the fold-
ing process of G-type lysozyme.
Homology modeling and MD (molecular

dynamics) simulation
The tertiary structures of the wild-type and mutant
C4S ⁄ C18S ⁄ C29S ⁄ C60S were constructed by homology
modeling on the basis of the X-ray structure of GEL
[14]. The Ramachandran plot provided by procheck
ensured very good confidence for the wild-type and the
mutant protein. The reliabilities of the constructed
structures were with 94.3% residues in the most
favored regions and 5.7% in additional allowed
regions. There were no residues in the generously
allowed regions and disallowed regions in the two
proteins.
Figure 7A shows the distribution of the B-factors
of the main chain atoms (N, C
a
, C, and O) in the
MD-generated average structural models of the wild-
type and mutant C4S ⁄ C18S ⁄ C29S ⁄ C60S. The mole-
cules are colored according to the B-factor, from
dark blue for low B-factor to red for high B-factor.
Although the overall structure of the mutant was
quite similar to that of the wild-type, an increase in
B-factors throughout the molecule was observed in
the Cys-free mutant. This increase indicates that the
conformation of the mutant protein becomes more
flexible. Figure 7B shows the plots of the B-factors
of the main chain atoms versus each residue for the
wild-type and the Cys-free mutant. All B-factors of
the Cys-free mutant were substantially higher than
those of the wild-type. The average B-factors of the

main chain atoms for the wild-type and the Cys-free
mutant in the 200 samples obtained by the MD sim-
ulations were 36.4 and 63.1 A
˚
2
, respectively. These
findings suggest that the two disulfide bonds in
G-type lysozyme act to keep the protein molecule
folded tightly.
A
B
Fig. 7. (A) Residue flexibilities calculated
for the wild-type and mutant C4S ⁄ C18S ⁄
C29S ⁄ C60S. The ribbon diagram was
colored on the basis of the amplitudes of
fluctuations (B-factor) of individual residues.
A blue-to-red color spectrum is used to
represent different levels of flexibilities,
where the smallest motions are in blue and
the highest ones are in red. (B) Amplitudes
of B-factors of each amino acid residue. The
B-factors for the wild-type and mutant
C4S ⁄ C18S ⁄ C29S ⁄ C60S are shown as blue
and pink lines, respectively. The locations of
a-helices and b-sheets in the wild-type are
shown as red and sky-blue lines, respec-
tively. The locations of the four Cys residues
are shown as green bars.
S. Kawamura et al. Disulfide bonds in goose-type lysozyme
FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2825

Conclusions
Our results demonstrated experimentally that the indi-
vidual disulfide bonds contribute significantly to the
structural stability of G-type lysozyme in an indepen-
dent manner. The results also suggested that the
reduced stabilities caused by deletion of one or both
of the disulfide bonds are due mainly to entropic
effects. The MD data showed that deletion of the
two disulfide bonds makes the protein conformation
more flexible. It is thus speculated that the formation
of the disulfide bonds is involved in the structural
stability of G-type lysozyme by reducing the confor-
mational entropy of the unfolded state and by
increasing the rigidity of the protein molecule. We
also found that deletion of both disulfide bonds does
not prevent the proper folding into the catalytically
active conformation of G-type lysozyme. This sug-
gests that the formation of the two disulfide bonds of
G-type lysozyme occurs late in the folding process,
and that these disulfide bonds can be formed inde-
pendently rather than sequentially. It is therefore sug-
gested that the structure around the mutation sites,
the N-terminal region in this case (Arg1–Cys60), may
not be responsible for the folding initiation site of
G-type lysozyme. When the overall results are taken
into account, the two disulfide bonds of G-type lyso-
zyme may confer stability after the protein reaches its
final folded form in the absence of both disulfide
bonds. On the other hand, we have shown that
G-type lysozyme has a structurally invariant core

composed of three a-helices (a5, a7, and a8) [20,21].
Previous investigations of protein folding suggested
that a-helical structures are formed at an early stage
in protein folding [58–63]. As G-type lysozyme was
considered to require no disulfide bonds for folding
and function, we suppose that the three a-helices may
pack together in the early stage of the folding process
and act as nucleation sites around which the structure
can be formed.
Experimental procedures
Materials
All enzymes used for DNA manipulation were purchased
from TaKaRa (Otsu, Japan) and Toyobo (Osaka, Japan).
The oligonucleotides used were from Hokkaido System Sci-
ence (Sapporo, Japan). Escherichia coli strain JM109 was
used for the transformation and propagation of recombi-
nant plasmids. Multi-Copy Pichia Expression kits, including
expression plasmid pPIC9K and host strain GS115, were
obtained from Invitrogen (Carlsbad, CA, USA). N-Acetyl-
glucosamine oligosaccharides [(GlcNAc)
n
] were prepared by
acid hydrolysis of chitin followed by charcoal celite column
chromatography [64]. M. luteus cells were from Sigma (St
Louis, MO, USA). Other reagents were of analytical or
biochemical grade.
Preparation of mutant proteins
Recombinant OEL with an extra Ser at the N-terminus
was prepared as described previously [33] and used
throughout this study as the wild-type. It should be

noted that the additional Ser residue at the N-terminus
had little effect on the secondary structure, substrate-
binding ability, lytic activity and structural stability of
OEL [33]. A plasmid harboring the wild-type sequence
(pGSer–OEL) [33] was used as a template DNA for site-
directed mutagenesis. The oligonucleotide primers used
were 5¢-GCCCTCGAGAAAAGATCTAGAACTGGATCT
TACGGAG-3¢ for C4S, 5¢-GCCCTCGAGAAAAGATC
TAGAACTGGAGCTTACGGAG-3¢ for C4A, 5¢-CAAA
AGCTTTCTGTCGATCCAGC-3¢ for C60S, 5¢-CAAAAG
CTTGCTGTCGATCCAGC-3¢ for C60A, 5¢-TCTTCTAA
GTCTGCTAAGCCAGAAAAGCTGAACTACTCT GGA
GTTG-3¢ for C18S ⁄ C29S, and 5¢-TCTGCTAAGTCTGC
TAAGCCAGAAAAGCTGAACTACGCTGGAGTT G-3¢
for C18A ⁄ C29A. Mutant genes were constructed by
PCR-based site-directed mutagenesis (megaprimer
method [65]) and verified by DNA sequencing. Mutants
C4S ⁄ C60S and C4A ⁄ C60A were produced by two con-
secutive rounds of mutagenesis. To create the gene
encoding OEL with a quadruple mutation (C4S ⁄ C18S ⁄
C29S ⁄ C60S), the C18S ⁄ C29S mutation was introduced
into the gene encoding C4S ⁄ C60S. Expression and puri-
fication o f mutant proteins were carried out as
described for the wild-type [33]. The purity was con-
firmedbySDS⁄ PAGE and RP-HPLC on a YMC-Pack
C4 column (4.6 · 250 mm; YMC, Tokyo, Japan). The
N-terminal amino acid sequence was determined with a
Shimadzu model PPSQ21 sequencer (Shimadzu C o.,
Kyoto, Japan). The protein concentration was measured
by amino acid analysis with a Hitachi Model L-8500A

amino acid analyzer (Hitachi High-Technologies Co.,
Tokyo, Japan).
CD spectra
CD spectra were obtained at 25 °C with a Jasco J-600 spec-
tropolarimeter (Japan Spectroscopic Co., Tokyo, Japan).
Proteins were dissolved to a final concentration of
0.15 mgÆmL
)1
in 10 mm sodium acetate buffer (pH 5.0).
The data were expressed in terms of mean residue elliptic-
ity. The path-length of the cells was 0.1 cm for far-UV CD
spectra (190–260 nm). Each spectrum was corrected by
subtracting the spectrum of the buffer.
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2826 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
Assay of enzymatic activity
Bacteriolytic activity (lytic activity) of lysozyme was
assayed using lyophilized cells of M. luteus as a substrate.
One hundred microliters of lysozyme (final concentration
0.015 lm) was added to 3 mL of a suspension of M. luteus
adjusted to A 0.9 at 540 nm with 0.1 m sodium phosphate
buffer (pH 7.0) at 25 °C. The activity was determined from
the first 5 min of linear decrease in absorbance at 540 nm.
Lysozyme-catalyzed hydrolysis of (GlcNAc)
5
was ana-
lyzed at 40 °C, when all proteins are folded. The reaction
mixture, containing 0.1 mm lysozyme and 1 mm (GlcNAc)
5
,

was incubated in 10 m m sodium acetate buffer (pH 4.0).
After a given reaction time, 200 lL of the reaction mixture
was withdrawn and rapidly chilled in a KOOL KUP
(Towa, Tokyo, Japan). The reaction mixture was centri-
fuged at 4000 g for 1h with Ultrafree C3LGC (Millipore,
Billerica, MA, USA), and the filtrate was lyophilized. The
dried sample was dissolved in 50 lL of ice-cold water,
and then 10 lL of the solution was applied to a TSKgel
G-Oligo-PW column (7.8 · 600 mm; Tosoh, Tokyo, Japan)
in a JASCO 800 series HPLC column. Elution was per-
formed with distilled water at room temperature and a flow
rate of 0.3 mLÆ min
)1
. Each chito-oligosaccharide concentra-
tion was calculated from the peak area monitored as the
UV absorption at 220 nm, using the standard curve
obtained for authentic saccharide solutions. The relative
error was defined as (y ) x) ⁄ x · 100, where x is the concen-
tration of the initial substrate in (GlcNAc)
1
units, and y is
the recovered concentration of all chito-oligosaccarides in
(GlcNAc)
1
units.
The temperature dependence of the catalytic activity for
the hydrolysis of (GlcNAc)
5
was assayed at temperatures
ranging from 20 °Cto80°C. The substrate was dissolved

in 10 mm sodium acetate buffer (pH 4.0) and incubated at
various temperatures for 5 min. Then, the enzyme dissolved
in the same buffer was added, and the activity was mea-
sured at the designated temperature for 30 min (wild-type)
or 40 min (C4S ⁄ C18S ⁄ C29S ⁄ C60S). The reaction time was
chosen so that about 50% of the initial substrate
(GlcNAc)
5
was hydrolyzed at 40 °C.
Thermal unfolding
Reversible thermal unfolding was monitored by CD and
fluorescence measurements as described previously [33]. CD
measurements at 222 nm were performed with a Jas-
co J-600 spectropolarimeter using a 0.1 cm cuvette. The flu-
orescence intensities at 360 nm, excited at 280 nm, were
measured with a Hitachi F-4500 Fluorescence Spectropho-
tometer using a 1 cm cuvette. Samples (CD, 0.15 mgÆmL
)1
;
fluorescence, 0.015 mgÆmL
)1
) were dissolved in 0.1 m
sodium acetate buffer (pH 5.0) containing 0.5 m guanidine
hydrochloride. These conditions were chosen for complete
reversibility of the thermal denaturation [33]. In the
solution of the reduced wild-type, 0.1 m b-ME was also
added. The water-jacketed cell containing each sample was
heated for 5 min at a given temperature by a thermostati-
cally regulated circulating-water bath. All samples were
fully equilibrated at each temperature before measurement.

The temperature of sample solutions was directly measured
using a TX1001 thermometer (Yokokawa M&C Co.,
Tokyo, Japan). To facilitate comparison between the two
sets of unfolding curves, the experimental data were
normalized as follows. The fraction of unfolded protein
was calculated from either the CD values or fluorescence
intensities by linearly extrapolating the pretransition and
post-transition base lines into the transition zone, and then
plotted against temperature. Assuming that the unfolding
equilibrium involves a two-state mechanism, the unfolding
curves were subjected to a least squares analysis to deter-
mine the midpoint temperatures (T
m
) and thermodynamic
parameters. The enthalpy and entropy changes at T
m
(DH
m
and DS
m
) were calculated using van’t Hoff analysis. Under
the assumption that the DC
p
values for the mutant proteins
are negligible compared to DS
m
, the difference in the free
energy change of unfolding (at T
m
of the wild-type protein)

between the mutant and wild-type proteins (DDG) was esti-
mated by the relationship DDG = DT
m
· DS
m
(mutant
protein), given by Becktel & Schellman [66], where DT
m
is
the difference in T
m
between the mutant and wild-type
proteins.
Guanidine hydrochloride unfolding
Guanidine hydrochloride-induced unfolding curves were
also determined by monitoring two different parameters at
30 °C [33]. One is the CD value at 222 nm, and the other
is the intrinsic fluorescence (excitation at 280 nm and
emission at 360 nm). Samples (CD, 0.15 mgÆmL
)1
; fluores-
cence, 0.015 mgÆmL
)1
) were incubated in 0.1 m sodium
acetate buffer (pH 5.0) with varying concentrations of gua-
nidine hydrochloride at 30 °C for 1 h. All samples were
fully equilibrated at each denaturant concentration before
measurement. Denaturation was completely reversible
under these conditions, and the unfolding data were
analyzed on the basis of a two-state model. From the

guanidine hydrochloride unfolding profiles, the difference
in free energy change between the folded and unfolded
states (DG) was calculated according to Pace [55]. The free
energy change in water (DG
H2O
) and the dependence of
DG on the guanidine hydrochloride concentration (m) were
determined by least squares fitting of the data for the
transition region using the equation DG =DG
H2O
) m[gua-
nidine hydrochloride]. The guanidine hydrochloride con-
centration at the midpoint of the transition (DG = 0) was
defined as C
m
. The differences in C
m
(DC
m
) between the
wild-type and mutant proteins were calculated by subtract-
ing the value of the wild-type from those of mutant
proteins.
S. Kawamura et al. Disulfide bonds in goose-type lysozyme
FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2827
Homology modeling and MD simulation
The 50 models of the wild-type OEL were initially con-
structed on the basis of the crystal structure of GEL (Pro-
tein Data Bank code: 153L) using the homology modeling
program modeller 8v2 [67]. The structure of mutant

C4S ⁄ C18S ⁄ C29S ⁄ C60S was obtained by replacing the four
Cys residues of the model with the highest reliability for the
wild-type using the autorotamer feature in the biopolymer
module of insight II (Accelrys Co., San Diego, CA, USA).
Stereochemical qualities of the models for the wild-type and
the mutant protein were examined by the program
procheck [68].
MD simulations of the wild-type and mutant C4S ⁄ C18S ⁄
C29S ⁄ C60S were performed using the Amber ff99 force
field in amber version 7 [69]. The system was solvated in a
periodic cubic box and filled with the TIP3 water mole-
cules. Long-range nonbonded interactions were truncated
by using a cutoff of 12.0 A
˚
. The initial structure for the
MD simulation was constructed by the following four steps:
(a) optimization of the positions of the water molecules using
energy minimization and MD simulation; (b) optimization
of the positions of the hydrogen atoms using energy minimi-
zation; (c) optimization of the conformation of the side chain
using energy minimization; and (d) optimization of the posi-
tions of all atoms in the system using energy minimization.
In these steps, energy minimization was performed by the
steepest descent and conjugate gradient methods until the
gradient vector was < 0.01 kcalÆmol
)1
ÆA
˚
)1
) or until 3000

steps had been completed. MD simulation was performed as
follows. First, solvent in the systems was equilibrated for
20 ps while the temperature was increased from 0 K to
300 K. Following equilibration, MD simulation was per-
formed by running a 200 ps simulation with a 2 fs time step
at 300 K in an NVT (constant volume dynamics) ensemble,
and the temperature was maintained through a ‘weak-cou-
pling’ scheme. After that, MD simulation was performed by
running a 500 ps simulation with a 2 fs time step at 300 K in
an NPT (constant pressure dynamics) ensemble. A pressure
relaxation time of 1.0 ps (taup) was applied. Electrostatics
were calculated using the particle mesh Ewald method with a
cutoff of 12.0 A
˚
. During the MD simulation, the last 200
conformations were sampled at 1 ps intervals, and used
for structural analysis. The atomic positional fluctuations
for all atoms were calculated as B-factors using amber
version 7 [69].
Acknowledgements
This work was supported by a Grant-in-Aid for Young
Scientists (B) from the Ministry of Education, Culture,
Sports, Science, and Technology of Japan (18780080),
in part by Research and Study Program of Tokai
University Educational System General Research
Organization, and in part by the Science Research
Promotion Fund from the Promotion and Mutual Aid
Corporation for Private Schools of Japan. The authors
are grateful to Yuta Mizukami, Department of Bio-
science, Kyushu Tokai University, for his technical

assistance.
References
1 Canfield RE (1963) The amino acid sequence of egg
white lysozyme. J Biol Chem 238, 2698–2707.
2 Jolle
`
s J, Jauregui-Adell J, Bernier I & Jolle
`
s P (1963) La
structure chimique du lysozyme de blanc d’oeuf de
poule: etude detailee. Biochim Biophys Acta 78, 668–
689.
3 Blake CCF, Koenig DF, Mair GA, North ACT, Phil-
lips DC & Serma VR (1965) Structure of hen egg white
lysozyme. A three-dimensional Fourier synthesis at 2 A
˚
resolution. Nature 206, 757–761.
4 Inoue M, Imada M & Tsugita A (1970) The amino acid
sequence of T4 phage lysozyme. IV. Diluted acid hydro-
lysis and order of tryptic peptides. J Biol Chem 245 ,
3479–3484.
5 Matthews BW & Remington SJ (1974) The
three-dimensional structure of the lysozyme from
bacteriophage T4. Proc Natl Acad Sci USA 71, 4178–
4182.
6 Simpson RJ & Morgan FJ (1983) Complete amino acid
sequence of embden goose (Anwer anwer) egg-white
lysozyme. Biochim Biophys Acta 744 , 349–351.
7 Gru
¨

tter MG, Weaver LH & Matthews BW (1983)
Goose lysozyme structure: an evolutionary link
between hen and bacteriophage lysozyme? Nature 303,
828–831.
8 Isaacs NW, Machin KJ & Masakuni M (1985) Three-
dimensional structure of goose-type lysozyme from the
egg white of the Austalian black swan, Cygnus atratus.
Aust J Biol Sci 38, 13–22.
9 Jolle
`
s J & Jolle
`
s P (1975) The lysozyme from Asterias
rubens. Eur J Biochem 54, 19–23.
10 Ito Y, Yoshikawa A, Hotani T, Fukuda S, Sugimura K
& Imoto T (1999) Amino acid sequences of lysozymes
newly purified from invertebrates imply wide distribu-
tion of a novel class in the lysozyme family. Eur J
Biochem 259, 456–461.
11 Nilsen IW, Øverbø K, Sandsdalen E, Sandaker E, Slet-
ten K & Myrnes B (1999) Protein purification and gene
isolation of chlamysin, a cold-active lysozyne-like
enzyme with antibacterial activity. FEBS Lett 464,
153–158.
12 Ho
¨
ltje J-V (1996) Bacterial lysozymes. In Lysozymes:
Model Enzymes in Biochemistry and Biology (Jolle
`
sP,

ed), pp. 65–74. Birkha
¨
user-Verlag, Basel, Switzerland.
13 Beintema JJ & Terwisscha van Scheltinga AC (1996)
Plant lysozymes. In Lysozymes: Model Enzymes in
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2828 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS
Biochemistry and Biology (Jolle
`
s P ed), pp. 75–86.
Birkha
¨
user-Verlag, Basel, Switzerland.
14 Weaver LH, Gru
¨
tter MG & Matthews BW (1995) The
refined structures of goose lysozyme and its complex
with a bound trisaccharide show that the ‘goose-type’
lysozymes lack a catalytic aspartate residue. J Mol Biol
245, 54–68.
15 Goto T, Abe Y, Kakuta Y, Takeshita K, Imoto T &
Ueda T (2007) Crystal structure of Tapes japonica
lysozyme with substrate analogue. J Biol Chem 282,
27459–27467.
16 Evrard C, Fastrez J & Declercq JP (1998) Crystal struc-
ture of the lysozyme from bacteriophage lambda and its
relationship with V and C-type lysozymes. J Mol Biol
276, 151–164.
17 Marcotte EM, Monzingo AF, Ernst SR, Brezezinski R
& Robertus JD (1996) X-ray structure of an anti-fungal

chitosanase from Dtreptomyces N174. Nat Struct Biol 3,
155–162.
18 Schoentgen F, Jolle
`
s J & Jolle
`
s P (1982) Complete
amino acid sequence of ostrich (Struthio camelus) egg-
white lysozyme, a goose type lysozyme. Eur J Biochem
123, 489–497.
19 Richard JS, Geoffrey SB, Donna SD & Francis JM
(1980) Complete amino acid sequence of the goose-type
lysozyme from the egg white of the black swan.
Biochemistry 19, 1814–1819.
20 Thammasirirak S, Torikata T, Takami K, Murata K &
Araki T (2002) The primary structure of Cassowary
(Casuarius casusrius ) goose type lysozyme. Biosci Bio-
technol Biochem 66, 147–156.
21 Pooart J, Torikata T & Araki T (2004) The primary
structure of a novel goose-type lysozyme from Rhea egg
white. Biosci Biotechnol Biochem 68, 159–169.
22 Nakano T & Graf T (1991) Goose type lysozyme
gene of the chicken: sequence genomic organization
and expression reveals major differences to chicken-
type lysozyme gene. Biochim Biophys Acta 1090, 273–
276.
23 Hikima J, Minagawa S, Hirono I & Aoli T (2001)
Molecular cloning, expression and evolution of the Jap-
anese flounder goose-type lysozyme gene, and the lytic
activity of its recombinant protein. Biochim Biophys

Acta 1520, 35–44.
24 Savan R, Aman A & Sakai M (2003) Molecular cloning
of G type lysozyme cDNA in common carp (Cyprinus
carpio L.). Fish Shellfish Immunol 15, 263–268.
25 Kyomuhendo P, Myrnes B & Nilsen IW (2007) A
cold-active salmon goose-type lysozyme with high heat
tolerance. Cell Mol Life Sci 64, 2841–2847.
26 Yin S, Weng S, Deng W & He J (2002) Molecular clon-
ing, expression on the goose-type lysozyme cDNA from
orange-spotted grouper (Epinephelus coioides), and the
lytic activity of its recombinant protein. Aquac Asia 3,
21–22.
27 Irwin DM & Gong Z (2003) Molecular evolution of
vertebrate goose-type lysozyme genes. J Mol Evol 56,
234–242.
28 Zou H, Song L, Xu W & Yang G (2005) Molecular
cloning and characterization analysis of cDNA encod-
ing g-type lysozyme from scallop (Argopecten irradians).
High Tech Lett 15, 101–106.
29 Zhao J, Song L, Li C, Zou H, Ni D, Wang W & Xu W
(2007) Molecular cloning an invertebrate goose-type
lysozyme gene from Chlamys farreri, and lytic activity
of the recombinant protein. Mol Immunol 44, 1198–
1208.
30 Nilsen IW, Myrnes B, Edvardsen RB & Chourrout D
(2003) Urochordates carry multiple genes for goose-type
lysozyme and no genes for chicken- or invertebrate-type
lysozymes. Cell Mol Life Sci 60, 2210–2218.
31 Arnheim N, Inouye M, Law L & Laudin A (1973)
Chemical studies on the enzymatic specificity of goose

egg white lysozyme. J Biol Chem 248 , 233–236.
32 Honda Y & Fukamizo T (1998) Substrate binding sub-
sites of chitinase from barley seeds and lysozyme from
goose egg white. Biochim Biophys Acta 1388, 53–65.
33 Kawamura S, Ohno K, Ohkuma M, Chijiiwa Y &
Torikata T (2006) Experimental verification of the cru-
cial role of Glu
73
in the catalytic activity and structural
stability of goose type lysozyme. J Biochem 140, 75–85.
34 Pace CN, Grimsley GR, Thomson JA & Barnett BJ
(1988) Conformational stability and activity of ribonu-
clease T
1
with zero, one, and two intact disulfide bonds.
J Biol Chem 263, 11820–11825.
35 Wetzel R, Perry LJ, Baase WA & Becktel WJ (1988)
Disulfide bonds and thermal stability in T4 lysozyme.
Proc Natl Acad Sci USA 85, 401–405.
36 Cooper A, Eyles SJ, Radford SE & Dobson CM (1992)
Thermodynamic consequences of the removal of a disul-
phide bridge from hen lysozyme. J Mol Biol 225, 939–
943.
37 Kuroki R, Inaka K, Taniyama Y, Kidokoro S, Matsu-
shima M, Kikuchi M & Yutani K (1992) Enthalpic
destabilization of a mutant human lysozyme lacking a
disulfide bridge between cysteine-77 and cysteine-95.
Biochemistry 31, 8323–8328.
38 Taniyama Y, Ogasahara K, Yutani K & Kikuchi M
(1992) Folding mechanism of mutant lysozyme

C77 ⁄ 95A with increased secretion efficiency in yeast.
J Biol Chem 267, 4619–4624.
39 Vogl T, Brengelmann R, Hinz H-J, Scharf M, Lo
¨
tzbe-
yer M & Engels JW (1995) Mechanism of protein stabil-
ization by disulfide bridges: calorimetric unfolding
studies on disulfide-deficient mutants of the a-amylase
inhibitor Tendamistat. J Mol Biol 254, 481–496.
40 Zhu H, Dupureur CM, Zhang X & Tsai M-D (1995)
Phospholipase A
2
engineering. The roles of disulfide
bonds in structure, conformational stability, and cata-
lytic function. Biochemistry 34, 15307–15314.
S. Kawamura et al. Disulfide bonds in goose-type lysozyme
FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS 2829
41 Guzzi R, Sportelli L, Rosa CL, Milardi D, Grasso D,
Verbeet MP & Canters GW (1999) A spectroscopic and
calorimetric investigation on the thermal stability of the
Cys3Ala ⁄ Cys26Ala azurin mutant. Biophys J 77, 1052–
1063.
42 Klink TA, Woycechowsky KJ, Taylor KM & Raines
RT (2000) Contribution of disulfide bonds to the
conformational stability and catalytic activity of ribo-
nuclease A. Eur J Biochem 267, 566–572.
43 Parisi M, Mazzini A, Sorbi RT, Ramoni R, Grolli S &
Favilla R (2005) Role of the disulphide bridge in fold-
ing, stability and function of porcine odorant binding
protein: spectroscopic equilibrium studies on

C63A ⁄ C155A double mutant. Biochim Biophys Acta
1750, 30–39.
44 Ramazzotti M, Parrini C, Stefani M, Manao G &
Degl’Innocenti D (2006) The intrachain disulfide bridge
is responsible for the unusual stability properties of
novel acylphosphatase from Escherichia coli. FEBS Lett
580, 6763–6768.
45 Siddiqui KS, Poljak A, Guilhaus M, Feller G, D’Amico
S, Gerday C & Cavicchioli R (2005) Role of disulfide
bridges in the activity and stability of a cold-active
a-amylase. J Bacteriol 187, 6206–6212.
46 Wang ECW, Hung S-H, Cahoon M & Hedstrom L
(1997) The role of the Cys191–Cys220 disulfide bond in
trypsin: new targets for engineering substrate specificity.
Protein Eng 10, 405–411.
47 Kawamura S, Fukamizo T, Araki T & Torikata T
(2003) Expression of a synthetic gene coding for ostrich
egg-white lysozyme in Pichia pastoris and its enzymatic
activity. J Biochem 133, 123–131.
48 Wetzel R (1987) Harnessing disulfide bonds using
protein engineering. Trends Biochem Sci 12, 478–482.
49 Betz SF (1993) Disulfide bonds and the stability of
globular proteins. Protein Sci 2, 1551–1558.
50 Flory PJ (1956) Theory of elastic mechanism in fibrous
proteins. J Am Chem Soc 78, 5222–5235.
51 Poland DG & Scheraga HA (1965) Statistical mechanics
of noncovalent bonds in polyamino acids. VIII. Cova-
lent loops in proteins. Biopolymers 3, 379–399.
52 Goldenberg DP (1985) Dissecting the roles of individual
interactions in protein stability: lessons from a circular-

ized protein. J Cell Biochem 29, 321–335.
53 Creighton TE (1988) Disulfide bonds and protein stabil-
ity. Bioessays 8, 57–63.
54 Harrison PM & Sternberg MJ (1994) Analysis and
classification of disulphide connectivity in proteins. The
entropic effect of cross-linkage. J Mol Biol 244,
448–463.
55 Pace CN (1990) Measuring and increasing protein
stability. Trends Biotechnol 8 , 93–98.
56 Clarke J, Henrick K & Fersht AR (1995) Disulfide
mutants of barnase I: changes in stability and structure
assessed by biophysical methods and X-ray crystallogra-
phy. J Mol Biol 253, 493–504.
57 Doig AJ & Williams DH (1991) Is the hydrophobic
effect stabilizing or destabilizing in proteins? The contri-
butions of disulphide bonds to protein stability. J Mol
Biol 217, 389–398.
58 Udgaonkar JB & Baldwin RL (1988) NMR evidence
for an early framework intermediate on the folding
pathway of ribonuclease A. Nature 335, 694–699.
59 Roder H, Elove GA & Englander SW (1988) Structural
characterization of folding intermediates in cyto-
chrome c by H-exchange labeling and proton NMR.
Nature 335, 701–704.
60 Bycroft M, Matoushcek A, Kllis J, Serrano L & Fersht
AR (1990) Detection and characterization of a folding
intermediate in barnase by NMR. Nature 346, 488–490.
61 Miranker A, Radford SE, Karplus M & Dobson CM
(1991) Demonstration by NMR folding domains in
lysozyme. Nature 349, 633–636.

62 Chaffotte AF, Guillou Y & Goldberg ME (1992)
Kinetic resolution of peptide bond and side chain
far-UV circular dichroism during the folding of hen
egg white lysozyme. Biochemistry 31, 9694–9702.
63 Ueda T, Nakashima A, Hashimoto Y, Miki T, Yamada
H & Imoto T (1994) Formation of a-helix 88–98 is
essential in the establishment of higher-order structure
from reduced lysozyme. J Mol Biol 235, 1312–1317.
64 Rupley JA (1964) The hydrolysis of chitin by concen-
trated hydrochloric acid, and the preparation of
low-molecular weight substrates for lysozyme. Biochim
Biophys Acta 83, 245–255.
65 Sarkar G & Sommer SS (1990) The ‘megaprimer’
method of site-directed mutagenesis. Biotechniques 8,
404–407.
66 Becktel WJ & Schellman JA (1987) Protein stability
curves. Biopolymers 26, 1859–1877.
67 Fiser A & Sali A (2003) Modeller: generation and
refinement of homology-based protein structure models.
Methods Enzymol 374, 461–491.
68 Laskowski RA, MacArthur MW, Moss DS & Thornton
JM (1993) PROCHECK: a program to check the stereo
chemical quality of protein structures. J Appl Crystal-
logr 26, 283–291.
69 Pearlman DA, Case DA, Caldwell JW, Ross WS,
Cheatham TE III, DeBolt S, Ferguson D, Seibel G
& Kollman P (1995) AMBER, a package of computer
programs for applying molecular mechanics, normal
mode analysis, molecular dynamics and free energy
calculations to simulate the structural and energetic

properties of molecules. Comp Phys Commun 91, 1–41.
Disulfide bonds in goose-type lysozyme S. Kawamura et al.
2830 FEBS Journal 275 (2008) 2818–2830 ª 2008 The Authors Journal compilation ª 2008 FEBS