X-ray crystallography and structural stability of digestive
lysozyme from cow stomach
Yasuhiro Nonaka
1
, Daisuke Akieda
1
, Tomoyasu Aizawa
1
, Nobuhisa Watanabe
1,2
, Masakatsu
Kamiya
3
, Yasuhiro Kumaki
1
, Mineyuki Mizuguchi
4
, Takashi Kikukawa
1
, Makoto Demura
3
and
Keiichi Kawano
1
1 Division of Biological Sciences, Graduate School of Science, Hokkaido University, Sapporo, Japan
2 Department of Biotechnology and Biomaterial Chemistry, Graduate School of Engineering, Nagoya University, Nagoya, Japan
3 Division of Molecular Life Science, Graduate School of Life Science, Hokkaido University, Sapporo, Japan
4 Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama, Japan
C-type lysozyme (EC 3.2.1.17), represented by hen
egg-white lysozyme (HEWL), is one of the most
well-known enzymes. It has been found in various ver-

crystallography
Correspondence
K. Kawano, Graduate School of Science,
Hokkaido University, North 10, West 8,
Kita-ku, Sapporo, Hokkaido 060 0810,
Japan
Fax: +81 11 706 2770
Tel: +81 11 706 2770
E-mail:
(Received 12 November 2008, revised 22
January 2009, accepted 4 February 2009)
doi:10.1111/j.1742-4658.2009.06948.x
In ruminants, some leaf-eating animals, and some insects, defensive lyso-
zymes have been adapted to become digestive enzymes, in order to digest
bacteria in the stomach. Digestive lysozyme has been reported to be resis-
tant to protease and to have optimal activity at acidic pH. The structural
basis of the adaptation providing persistence of lytic activity under severe
gastric conditions remains unclear. In this investigation, we obtained the
crystallographic structure of recombinant bovine stomach lysozyme 2
(BSL2). Our denaturant and thermal unfolding experiments revealed that
BSL2 has high conformational stability at acidic pH. The high stability in
acidic solution could be related to pepsin resistance, which has been previ-
ously reported for BSL2. The crystal structure of BSL2 suggested that
negatively charged surfaces, a shortened loop and salt bridges could pro-
vide structural stability, and thus resistance to pepsin. It is likely that BSL2
loses lytic activity at neutral pH because of adaptations to resist pepsin.
Abbreviations
BSL2, bovine stomach lysozyme 2; DSC, differential scanning calorimetry; HEWL, hen egg-white lysozyme.
2192 FEBS Journal 276 (2009) 2192–2200 ª 2009 The Authors Journal compilation ª 2009 FEBS
properties, e.g. low optimal pH and resistance to pro-

the data collection, processing and refinement statistics
are summarized in Table 1. BSL2 was crystallized in
the space group P2
1
2
1
2
1
. The structure was refined at
1.5 A
˚
to an R-factor of 17.8% and an R-free of
22.1%. The average B-value for all protein atoms is
10.17 A
˚
2
, and that for all main chain atoms is 9.25 A
˚
2
.
The electron density map was sufficiently clear to build
a molecular model, and most of the side chain confor-
mations were determined unequivocally, although
some residues showed multiple conformers.
This lysozyme is composed of an a-domain and a
b-domain, both of which are common in the previ-
ously reported structures for other c-type lysozymes.
The a-domain is composed of four a-helices (A–D),
and the b-domain is composed of a large loop and a
three-strand antiparallel b-sheet. Figure 1B is a super-

a 31.257
b 56.065
c 64.050
Resolution (A
˚
) 50.00–1.50 (1.55–1.50)
a
No. observations 126 692
I ⁄ r(I) 28.085 (17.272)
No. unique reflections 17833 (1662)
R
merge
0.046 (0.088)
Completeness (%) 95.0 (90.7)
Multiplicity 7.1 (7.1)
Refinement data
Resolution (A
˚
) 17.94–1.50
No. reflections 16 849
R-factor 0.178
R
free
0.221
Rmsd from ideal values
Bond lengths (A
˚
) 0.009
Bond angles (°) 1.261
a

Fig. 3, BSL2 is more resistant than HEWL to pepsin.
Pepsin readily digested HEWL in acidic conditions
with physiological ionic strength (150 mm NaCl),
whereas BSL2 remained intact after 4 h. This result
corresponded to that for natural BSL2 from bovine
stomach, based on residual activity [9].
In one report, protease resistance was correlated with
protein thermostability [12]. To evaluate the structural
stability of BSL2, denaturant-induced unfolding and
thermal unfolding were monitored. Figure 4 shows the
guanidinium hydrochloride-unfolding curves of BSL2
and HEWL, as determined by CD ellipticity at 222 nm,
indicating the disruption of the native structure. The
parameters derived from these unfolding curves are
shown in Table 2. At pH 6.0, BSL2 and HEWL were
similar in their midpoints (C
m
), Gibbs free energies
without denaturant (DG
w
), and m values indicative of
cooperativity. At pH 2.0, in contrast, BSL2 unfolded at
a higher concentration of guanidinium hydrochloride
than HEWL. The Gibbs free energy of BSL2 at low
pH was much greater than that of HEWL, indicating
the high conformational stability of BSL2. The transi-
tion temperatures (T
m
) and unfolding enthalpy values
(DH

pH 7.0 as 1.0.
A
B
C
Fig. 3. SDS ⁄ PAGE of pepsin-treated BSL2 and HEWL with (A)
0m
M NaCl (B) 150 mM NaCl, and (C) 500 mM NaCl. Aliquots of the
solution were sampled at intervals of 1 h. M is the marker lane.
Structure and stability of bovine stomach lysozyme Y. Nonaka et al.
2194 FEBS Journal 276 (2009) 2192–2200 ª 2009 The Authors Journal compilation ª 2009 FEBS
Discussion
Although BSL2 has an acidic optimal pH, the relative
activity level is lower than or comparable to that of
HEWL, even at acidic pH (Fig. 2). BSL2, like many
acidophilic proteins [13–15], possesses a greater num-
ber of acidic residues than nondigestive lysozymes
(Table 3). An increase in acidic residues would result
in low lytic activity, because the electrostatic attraction
between the lysozyme and the negatively charged bac-
terial membrane becomes weaker, especially at neutral
pH. BSL isozymes are considered to function below
pH 6 in nature [9]. It is likely that BSL2 has lost lytic
activity at neutral pH and retains it below pH 6.
In the case of house fly digestive lysozyme, the crys-
tallographic analysis and catalytic activity experiments
indicated that the catalytic residues have lower pK
a
values than those of HEWL, and thus the optimal pH
is shifted to the acidic range [11]. Using the crystallo-
graphic structures, we calculated the pK

DH (kJÆmol
)1
)
a
406.4 386.4
a
The unfolding enthalpies at transition temperature T
m
.
W64
A
B
W111
Normalized intensityNormalized intensity
11.0 10.8 10.6 10.4 10.2
p.p.m.
10.0 9.8
11.0 10.8 10.6 10.4 10.2
p.p.m.
10.0 9.8
W63
W34
W108
W108
W62
W63
W111
W123
Fig. 5. 1D
1

a
, whereas the polarity of Thr110 reduces the pK
a
for house fly lysozyme. In the case of Asp52, the pK
a
is modulated by the hydrogen bond network. There
are hydrogen bonds formed by Asp52, Asn46 and
Asp48 in HEWL. House fly lysozyme has an aspara-
gine at position 48, and the absence of the negative
charge should reduce the pK
a
of Asp52 as compared
to HEWL [11]. Asn46 in BSL2 is distant from Asp52,
and the absence of this hydrogen bond network would
reduce the pK
a
. However, Asp52 in BSL2 is more
exposed to solvent than that in HEWL, and this raises
the pK
a
. As a result, the calculated pK
a
values for
BSL2 were comparable to those for HEWL. The result
suggests that the catalytic activity of BSL2 is not
adapted to acidic conditions, unlike the case with
house fly lysozyme.
BSL2 and other vertebrate digestive lysozymes have
been reported to be resistant to pepsin digestion, as is
also shown in Fig. 3. The efficiency of peptide bond

b-domain and a positively charged a-domain. The
electrostatic repulsion on the surface will be weaker,
and this could contribute to the higher stability. There
are fewer charged residues on the surface of the house
fly lysozyme, and the electrostatic repulsion will be
smaller. The house fly lysozyme may have achieved
structural stability by decreasing the positively charged
residues.
The increase in acidic residues is also expected to
result in an increase in the number of salt bridges. The
numbers of the salt bridges in BSL2 and HEWL, how-
ever, are comparable (Table 3). It is noteworthy that
BSL2 contains a complex salt bridge (Glu83–Lys91–
Glu86) that is absent in the three other lysozymes. A
triangular salt bridge formed by two acidic residues
and one basic residue can be more strong than the
sum of simple salt bridges [25–27]. The loop located
between Glu83 and Lys91 connects the b-domain and
the a-domain. In the case of calcium-binding lysozyme,
calcium binding at this loop stabilizes the native struc-
ture [28,29]. By analogy, the electrostatic interaction at
this loop is considered to contribute to the overall
structural stability.
The overall structures of these lysozymes are very
similar (Fig. 1B), and the numbers of hydrogen bonds
are comparable (Table 3). A marked difference is
observed in the region from the C-terminus of the
C-helix to the following loop, residues 100–103 in
HEWL (Fig. 1B). The C-helices of human lysozyme
and HEWL are terminated at residue 101 followed by

the region that includes residues 100–102 could be
associated with resistance to pepsin. The replacement
of residue 21 could also be an adaptation to stabilize
this region.
Experimental procedures
Expression and purification of BSL2
In an Escherichia coli expression system, removal of an
extra methionine residue at the N-terminus does not take
place in the case of lysozyme [33]. We obtained recom-
binant BSL2 with a perfect sequence using the methylo-
trophic yeast Pichia pastoris, basically as described by
Digan et al. [34].
The cDNA was ligated to the expression vector pPIC3
(Invitrogen, Carlsbad, CA, USA). To secrete BSL2 into the
culture, we incorporated the native signal sequence of
BSL2. The plasmid was linearized by SalI, and transformed
into P. pastoris GS115 by electroporation. Genotypic selec-
tion and phenotypic screening were performed on a mini-
mal dextrose plate (1.34% yeast nitrogen base, 4 · 10
)5
%
biotin, 1% dextrose, and 1.5% agar) and on a minimal
methanol lysoplate (1.34% yeast nitrogen base, 4 · 10
)5
%
biotin, 0.061% Micrococcus lysodeikticus, and 1.5% agar,
in 10 mm potassium phosphate buffer, pH 5.0), as previ-
ously reported, except for pH and buffer concentration
[35]. Colonies on a minimal dextrose plate were inoculated
onto a minimal methanol lysoplate, and 200 lL of metha-

and freeze-dried.
Assay of lytic activity
The lytic activities of BSL2 and HEWL against M. lys-
odeikticus were estimated using the turbidimetric method
[39]. Lyophilized M. lysodeikticus was purchased from
Sigma-Aldrich (St Louis, MO, USA). Suspensions of
M. lysodeikticus were prepared in sodium acetate (for pH 4
and 5) and sodium phosphate (for pH 6 and 7) buffer. The
ionic strength of each buffer was adjusted to 0.1 [40]. Lyso-
zyme solution and M. lysodeikticus suspension were mixed,
and the decrease in absorbance was monitored at 540 nm
with a thermostatically controlled cell holder at 25 °C. The
relative activity was calculated from the speed of the absor-
bance decrement.
Pepsin digestion
Pepsin was obtained from Sigma-Aldrich. HEWL was
obtained from Seikagaku Corp. (Tokyo, Japan). Lysozymes
were dissolved in 10 mm HCl (pH 2), and the final protein
concentration was 0.5 mgÆmL
)1
. The digestion experiment
was carried out in the presence of pepsin at 37 °C. The
aliquots were sampled at intervals of 1 h and then frozen
until electrophoresis.
X-ray crystallography
A crystal of BSL2 was obtained by the vapor diffusion (sit-
ting drop) method, using 0.1 m sodium Hepes buffer at
pH 7.5, containing 0.2 m NaCl and 30% 2-methyl-2,4-penta-
nediol. The space group of the crystal was P2
1

refmac5 [45] in the ccp4 suite, and was visually inspected
using coot [46]. Water molecules were found by the func-
tions in refmac5 and coot, and were checked visually using
coot. A sodium ion was added to the model as judged by the
electron density, coordination number, and interatomic dis-
tance. The structure was deposited in the Protein Data Bank
under the code 2Z2F.
Analysis of structural features
A salt bridge in Table 3 was defined as a negative residue
and a positive residue with an interatomic distance of
< 4.0 A
˚
. The hydrogen bonds were detected using the
what if web interface with the following criteria: maximal
distances of 3.5 A
˚
for donor–acceptor and 2.5 A
˚
for
hydrogen–acceptor, and minimal angles of 60° for donor–
hydrogen–acceptor and 90° for hydrogen–acceptor–X.
Water-mediated hydrogen bonds were not included.
CD
CD at 222 nm was measured with a Jasco J-725 spectro-
polarimeter (Japan Spectroscopic, Tokyo, Japan), using
optical cells with path length of 1 mm. The guanidinium
hydrochloride-induced unfolding experiment was carried
out at 298 K using 50 mm KCl ⁄ HCl buffer at pH 2.0, and
50 mm sodium phosphate buffer at pH 6.0. The concentra-
tion of guanidinium hydrochloride was determined by the

Hydrogen–deuterium exchange was measured by 1D
1
H-
NMR performed on a Bruker 500 MHz instrument (Bruker
BioSpin, Rheinstetten, Germany), with a cryogenic probe
and a JEOL ECA-600 instrument (JEOL, Tokyo, Japan).
The exchange was initiated by dissolving protein that had
been lyophilized with pH-adjusted buffer (pH 1.9) in D
2
O
to give a final protein concentration of 0.3 mm in 50 mm
sodium phosphate. The sample was incubated at 298 K. A
total of 32 scans of each sample were collected at 30 or
60 min intervals. To acquire the spectra before hydrogen
exchange, lysozyme solution was subjected to
1
H-NMR in
the same buffer with 95% H
2
O ⁄ 5% D
2
O. The peaks of
unexchangeable hydrogens were used to normalize inten-
sity. The peaks of indole hydrogens were assigned on the
basis of the BMRB database (bmr1093 and bmr4562 for
HEWL and bmr76 for human lysozyme were used), and
using proshift [48], a chemical-shift prediction tool.
Estimation of protein concentration
The protein concentrations were estimated spectrophoto-
metrically by following the extinction coefficients at 280 nm

larval midgut. Insect Biochem 17, 809–817.
6 Kornegay JR, Schilling JW & Wilson AC (1994) Molec-
ular adaptation of a leaf-eating bird – stomach lyso-
zyme of the hoatzin. Mol Biol Evol 11, 921–928.
7 Stewart CB, Schilling JW & Wilson AC (1987) Adap-
tive evolution in the stomach lysozymes of foregut
fermenters. Nature 330, 401–404.
8 Prager EM (1996) Adaptive evolution of lysozyme:
changes in amino acid sequence, regulation of expres-
sion and gene number. In Lysozyme: Model Enzymes in
Biochemistry and Biology (Jolles P, ed.), pp. 323–345.
EXS, Basel.
9 Dobson DE, Prager EM & Wilson AC (1984) Stomach
lysozymes of ruminants. I. Distribution and catalytic
properties. J Biol Chem 259, 11607–11616.
10 Stewart CB & Wilson AC (1987) Sequence convergence
and functional adaptation of stomach lysozymes from
foregut fermenters. Cold Spring Harb Symp Quant Biol
52, 891–899.
11 Cancado FC, Valerio AA, Marana SR & Barbosa JA
(2007) The crystal structure of a lysozyme c from
housefly Musca domestica, the first structure of a diges-
tive lysozyme. J Struct Biol 160, 83–92.
12 Daniel RM, Cowan DA, Morgan HW & Curran MP
(1982) A correlation between protein thermostability
and resistance to proteolysis. Biochem J 207, 641–644.
13 Bonisch H, Schmidt CL, Schafer G & Ladenstein R
(2002) The structure of the soluble domain of an archa-
eal rieske iron–sulfur protein at 1.1 A resolution. J Mol
Biol 319, 791–805.

thermostability. Protein Eng
1, 477–480.
22 Zavodszky P, Kardos J, Svingor A & Petsko GA (1998)
Adjustment of conformational flexibility is a key event
in the thermal adaptation of proteins. Proc Natl Acad
Sci USA 95, 7406–7411.
23 Varley PG & Pain RH (1991) Relation between stabil-
ity, dynamics and enzyme-activity in 3-phosphoglycerate
kinases from yeast and Thermus thermophilus. J Mol
Biol 220, 531–538.
24 Wrba A, Schweiger A, Schultes V, Jaenicke R &
Zavodszky P (1990) Extremely thermostable d-glycer-
aldehyde-3-phosphate dehydrogenase from the
eubacterium Thermotoga maritima. Biochemistry 29,
7584–7592.
25 Mayne L, Englander SW, Qiu R, Yang JX, Gong YX,
Spek EJ & Kallenbach NR (1998) Stabilizing effect of a
multiple salt bridge in a prenucleated peptide. JAm
Chem Soc 120, 10643–10645.
26 Gvritishvili AG, Gribenko AV & Makhatadze GI
(2008) Cooperativity of complex salt bridges. Protein
Sci 17, 1285–1290.
27 Horovitz A, Serrano L, Avron B, Bycroft M & Fersht
AR (1990) Strength and cooperativity of contributions
of surface salt bridges to protein stability. J Mol Biol
216, 1031–1044.
28 Nitta K, Tsuge H & Iwamoto H (1993) Comparative
study of the stability of the folding intermediates of the
calcium-binding lysozymes. Int J Pept Protein Res 41,
118–123.

tion of signal sequence on level of expression in the
Pichia pastoris expression system. Protein Eng 14,
705–710.
36 Brierley RA, Bussineau C, Kosson R, Melton A & Siegel
RS (1990) Fermentation development of recombinant
Pichia pastoris expressing the heterologous gene – bovine
lysozyme. Ann NY Acad Sci 589, 350–362.
37 Katakura Y, Zhang WH, Zhuang GQ, Omasa T,
Kishimoto M, Goto W & Suga KI (1998) Effect of
methanol concentration on the production of human
beta(2)-glycoprotein I domain V by a recombinant
Pichia pastoris: a simple system for the control of meth-
anol concentration using a semiconductor gas sensor.
J Ferment Bioeng 86, 482–487.
38 Koganesawa N, Aizawa T, Shimojo H, Miura K, Ohni-
shi A, Demura M, Hayakawa Y, Nitta K & Kawano K
(2002) Expression and purification of a small cytokine
growth-blocking peptide from armyworm Pseudaletia
separata by an optimized fermentation method using
the methylotrophic yeast Pichia pastoris. Protein Expr
Purif 25, 416–425.
39 Yoshimura K, Toibana A, Kikuchi K, Kobayashi M,
Hayakawa T, Nakahama K, Kikuchi M & Ikehara M
(1987) Differences between Saccharomyces cerevisiae
and Bacillus subtilis in secretion of human lysozyme.
Biochem Biophys Res Commun 145, 712–718.
40 Miller GL & Golder RH (1950) Buffers of pH 2 to 12
for use in electrophoresis. Arch Biochem 29, 420–423.
41 Otwinowski Z & Minor W (1997) Processing of x-ray
diffraction data collected in oscillation mode. In

identification and analysis tools on the ExPASy server.
In The Proteomics Protocols Handbook (Walker JM,
ed.), pp. 571–607. Humana Press, Totowa, NJ.
50 Koradi R, Billeter M & Wuthrich K (1996) Molmol: a
program for display and analysis of macromolecular
structures. J Mol Graph 14, 51–55, 29–32.
Structure and stability of bovine stomach lysozyme Y. Nonaka et al.
2200 FEBS Journal 276 (2009) 2192–2200 ª 2009 The Authors Journal compilation ª 2009 FEBS