The chitinolytic system of Lactococcus lactis ssp. lactis
comprises a nonprocessive chitinase and a chitin-binding
protein that promotes the degradation of a- and b-chitin
Gustav Vaaje-Kolstad, Anne C. Bunæs, Geir Mathiesen and Vincent G. H. Eijsink
Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, A
˚
s, Norway
Chitin is a widespread biopolymer composed of b(1,4)-
linked N-acetylglucosamine that provides structural
and chemical resistance in the exoskeleton of crusta-
ceans and arthropods, as well as in the cell wall of
fungi. Chitin exists almost exclusively in an insoluble
crystalline form that complexes with proteins and ⁄ or
minerals to form a robust composite material. Three
naturally occurring crystalline polymorphs have been
described in the literature: the dominant polymorph
a-chitin (antiparallel packing of the chitin chains);
b-chitin (parallel packing of the chitin chains); and the
minor polymorph c-chitin (mixture of parallel and
antiparallel chain packing) [1,2]. In nature, chitin is
only exceeded in abundance by the structural biopoly-
mers of plants (cellulose and hemicellulose) and is an
important source of energy for a variety of organisms.
The primary degraders of chitin are microorganisms
that secrete one or several chitin-degrading enzymes
(chitinases). On the basis of sequence and structure,
chitinases are classified into two distinct families (18
Keywords
chitin; chitin binding; chitinase; lactic acid
bacterium; Lactococcus lactis
Correspondence

b-chitin, LlCBP33A was found to bind to both a- and b-chitin. LlCBP33A
increased the hydrolytic efficiency of LlChi18A to both a- and b-chitin.
These results show the general importance of chitin-binding proteins in
chitin turnover, and provide the first example of a family 33 chitin-binding
protein that increases chitinase efficiency towards a-chitin.
Abbreviations
CBM, carbohydrate-binding module; CBP, chitin-binding protein; FnIII, Fibronectin-III; LAB, lactic acid bacterium; 4MU-(GlcNAc)
3,
4-methylumbelliferyl-b-D-N,N¢,N¢¢-diacetylchitobioside; TEV, tobacco etch virus.
2402 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
and 19) of glycoside hydrolases [3,4]. Recently, a com-
plete survey of Trichoderma chitinases suggested a fur-
ther classification of family 18 chitinases into
subgroups A (bacterial ⁄ fungal), B (plant ⁄ fungal) and
C (killer toxin-like chitinases) [5]. Family 18 chitinases
are represented in most living organisms, whereas fam-
ily 19 enzymes are mostly found in plants, where they
contribute to defence against chitinous pathogens.
As a result of the recalcitrance of chitinous matrices,
microorganisms have devised a variety of complemen-
tary strategies to gain access to and degrade individual
polymer chains. First, the chains are degraded by both
endochitinases, that attack the chitin chain randomly,
and exochitinases, that attack the chitin chains from
either the reducing or nonreducing end [6,7]. As endo-
acting enzymes increase substrate availability for exo-
acting enzymes, synergistic effects are observed [8–10].
Second, some chitinases act processively, that is, they
remain associated with one and the same polymer
chain whilst cleaving off consecutive dimers (also

E7 and E8 from Thermobifdia fusca [26], showing a
large diversity of binding preferences. The function of
family 33 CBPs was first demonstrated for CBP21
from S. marcescens [10], and a second example has
been described recently in a study on carbohydrate-
binding proteins and domains from T. fusca [26].
Genes encoding family 33 CBPs occur even in bacte-
ria containing otherwise seemingly simple chitinolytic
machineries, such as in the lactic acid bacterium
(LAB) Lactococcus lactis ssp lactis IL1403. LABs are
Gram-positive, facultatively, anaerobic, fermentative
bacteria that are of major importance in the food
industry for the generation of fermented products. In
general, there is not much known about the ability of
LABs to degrade chitin, but one study has shown that
L. lactis is able to grow on a minimal medium contain-
ing N-acetylglucosamine oligomers as the sole carbon
source [27]. According to the CAZy database [3], only
a few of the sequenced LAB genomes contain genes
that together encode a complete chitinolytic machin-
ery. The genome sequence of L. lactis [28] shows three
genes potentially involved in chitin turnover, coding
for the following: a secreted family 18 chitinase (gene
name chiA; protein referred to as LlChi18A); a
secreted family 33 CBP (yucG; protein referred to as
LlCBP33A); and a family 20 N-acetylhexosaminidase
(LnbA). The chiA and yucG genes are separated by
19 bp in an operon starting with a putative transcrip-
tional regulator positioned 166 bp upstream from the
chitinase start codon. In this study, we have followed

of carbon. In order to further analyse whether L. lactis
would look for chitin as an alternative source of car-
bon, the bacterium was grown in a medium containing
only 0.1% (w ⁄ v) glucose. During the growth period
and starvation period, culture samples were taken and
assayed for chitinolytic activity. Chitinolytic activity
was detected, peaking 7 h after inoculation (Fig. 1).
After 7 h, chitinolytic activity declined, but still remained
significant. We could not detect chitinolytic activity in
uninoculated culture medium or in cultures grown with
normal glucose concentrations.
Cloning and purification of LlChi18A and
LlCBP33A
The gene fragments coding for the mature proteins of
LlChi18A and LlCBP33A were successfully cloned into
the pETM11 and pET30 Xa ⁄ LIC expression vectors,
respectively.
When expressed in Escherichia coli BL21 DE3, both
proteins were produced in large amounts, although
partly (LlChi18A) or almost exclusively (LlCBP33A)
in an insoluble form (inclusion bodies). The culture
conditions (temperature, isopropyl thio-b-d-galactoside
concentrations and duration of culture) were varied in
an attempt to obtain soluble protein. For LlChi18A,
this resulted in the production of sufficient amounts
of soluble protein. Soluble LlCBP33A was obtained
through refolding of protein obtained from the inclu-
sion bodies. After testing several denaturation and
refolding protocols, we adopted a protocol based on
denaturation in 8 m urea, pH 8.0 for 3 h and refolding

of the TIM-barrel fold (Fig. 2A), which is responsible
for deepening the substrate-binding groove in many
family 18 chitinases [29]. A deep substrate-binding
groove is considered to be characteristic of enzymes
that act in an exo-fashion and ⁄ or that tend to stick
tightly to the substrate whilst degrading it in a proces-
sive manner [30,31]. Enzymes lacking the a + b-fold
insertion have a shallow catalytic cleft, as illustrated
by the crystal structure of the plant family 18 sub-
group B chitinase hevamine [32]. Such shallow cata-
lytic clefts are typically seen amongst endo-acting,
nonprocessive carbohydrate-degrading enzymes.
Detailed studies using chitosan as substrate have
shown that ChiC1 from S. marcescens is indeed a non-
processive endo-acting enzyme [30,33]. A model of
LlChi18A automatically generated by 3d-jigsaw [34]
using the structure of hevamine (Protein Data
Bank code: 2HVM) as template suggested that the two
Fig. 1. Chitinolytic activity produced by cultured L. lactis. Bar chart
of chitinolytic activity measured in the culture supernatant of a
starved L. lactis culture at specific time points. The bar labelled as
‘LM17’ indicates the chitinolytic activity present in fresh culture
medium. Activity was recorded by measuring the hydrolysis of the
fluorogenic substrate 4MU-(GlcNAc)
3
. All experiments were run in
triplicate.
L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2404 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
A

residues: serine in the SXGG motif and a tyrosine resi-
due. Although LlChi18A lacks serine, it does contain
this tyrosine residue (Tyr48, corresponding to Tyr10 in
S. marcescens ChiB). A multiple sequence alignment of
the 50 family 18 catalytic modules that are most simi-
lar to the LlChi18A catalytic module (not shown)
showed that about one-half of the proteins had a sub-
stitution at either the conserved serine or tyrosine,
whereas none had substitutions at both positions.
Thus, it appears that family 18 glycosyl hydrolases are
tolerant to substitutions of either of the discussed
amino acids, as long as both are not substituted.
Another conspicuous sequence characteristic of
LlChi18A is the presence of an asparagine residue at
position 230. The presence of an asparagine at this
position is characteristic for family 18 chitinases with
acidic pH optima for activity, whereas enzymes with
more neutral pH optima have an aspartic acid at this
position. For the latter type of enzyme, it has been
shown that mutation of aspartic acid to asparagine
leads to a drastic acidic shift of the pH optimum [35].
Indeed, LlChi18A was found to have an acidic pH
optimum for activity (see below).
LlCBP33A is a family 33 CBP. The only available
three-dimensional structure of a family 33 CBP is that
of CBP21 from S. marcescens, which binds exclusively
to b-chitin [17,20]. The combination of sequence and
structural information with the results of site-directed
mutagenesis studies showed that the surface of fam-
ily 33 CBPs contains a patch of highly conserved,

4-methylumbelliferyl-b-d-N, N¢,N¢-diacetylchitobioside
[4MU-(GlcNAc)
3
] showed that LlChi18A has a narrow
pH activity profile with an optimum at pH 3.8
(Fig. 4A). Studies on pH stability showed that the
AB
Fig. 3. Structural comparison of CBP21 and LlCBP33A. Illustrations
of the CBP21 structure (A) and a structural model of LlCBP33A (B)
shown in a surface representation. The surface thought to be
involved in chitin binding is coloured blue. The side-chains of resi-
dues marked with an asterisk in the sequence alignment of Fig. 2B
are shown as blue sticks. Residues important for chitin binding and
the function of CBP21 [10,17], but not conserved in LlCBP33A, are
shown as blue sticks and labelled. For illustration purposes only,
the figure also shows the small inserts in LlCBP33A (orange) as
they were rendered by the structure prediction program. Note that,
as no template structure residues are available for modelling the
inserts, the structural prediction of these inserts is highly inaccu-
rate. Phe55 is coloured magenta and its side-chain is shown. Trp
(Trp51) in the LlCBP33A insert is hidden from view. The model of
LlCBP33A was generated by SwissModel (http://swissmodel.
expasy.org//SWISS-MODEL.html; [50]), using CBP21 (Protein Data
Bank code: 2BEM) as structural template. The model of LlCBP33A
is deposited in the PMDB database (PMDB code: PM0075054).
L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2406 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
enzyme was unstable at pH 3.8 and below, whereas
enzyme activity remained stable for more than a week
at bench temperature when dissolved in buffers with a

chitinases with shallow substrate-binding clefts [37–39].
Processive chitinases with their characteristic deep sub-
strate-binding grooves usually have about 10-fold
higher k
cat
and 10-fold lower K
m
values for oligomeric
substrates [39].
Initial rate measurements with (GlcNAc)
3
and (Glc-
NAc)
4
as substrates yielded specific activities of 0.64
and 11.6 s
)1
, respectively (Fig. 4C), within the range of
other results reported in the literature (e.g. ChiC1 from
S. marcescens [30]). The products observed for (Glc-
NAc)
3
degradation were GlcNAc and (GlcNAc)
2
.
(GlcNAc)
4
degradation resulted in the exclusive forma-
tion of (GlcNAc)
2

represents a nonhydrolysable background oligosaccharide that is also seen (with equal peak area) in control samples without enzyme. Glc-
NAc was not observed before all (GlcNAc)
6
was degraded. Although the experiments in (D) were not conducted to preserve anomeric ratios
generated by the enzyme, one important trend is still visible: the combination of a relative predominance of b-anomers for the (GlcNAc)
2
product and the approximately equilibrium anomeric ratio for the tetrameric product suggests that the conversion of (GlcNAc)
6
to (GlcNAc)
2
and (GlcNAc)
4
primarily results from binding of the nonreducing end of the substrate in subsite )2.
G. Vaaje-Kolstad et al. L. lactis chitinase and chitin-binding protein
FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS 2407
subsites )2 to +2. Analysis of the initial degradation
products formed from (GlcNAc)
6
showed a 1 : 1 ratio
of (GlcNAc)
2
to (GlcNAc)
4
, which indicates a nonpro-
cessive mode of action (Fig. 4D). Processive chitinases
tend to convert (GlcNAc)
6
processively into three (Glc-
NAc)
2

C-terminal substrate-binding domains are likely to be
located on this side, which again suggests that this side
of the enzyme is optimized for interacting with the
longer (polymeric) part of the substrate. In conclusion,
these experimental data and the inferences made from
the sequence and structural comparisons above indi-
cate that LlChi18A is a nonprocessive endo-acting
chitinase, with overall properties that are quite simi-
lar to those of, for example, the nonprocessive endo-
chitinase ChiC1 from S. marcescens.
Binding preferences for LlCBP33A
Some family 33 CBPs bind to a broad selection of
insoluble carbohydrates (e.g. ChbB, which binds both
a- and b-chitin [21], and Chb3 from St. coelicolor,
which binds a-chitin, b-chitin, colloidal chitin and
chitosan [22]), whereas others bind only to a specific
substrate variant (e.g. CBP21 from S. marcescens
which strictly binds to b-chitin [20] and CHB1 from
St. olivaceoviridis [23] and CHB2 from St. reticuli [24]
which strictly bind to a-chitin). A common property is
that binding is influenced by pH (e.g. CBP21 from
S. marcescens does not bind at pH < 4.5 [20]).
The binding preferences of LlCBP33A were investi-
gated by incubating the protein with various types of
chitin and other insoluble polymeric substrates. As
noncrystalline ⁄ amorphous chitin variants, chitin beads
(re-acetylated chitosan beads) and colloidal chitin (chi-
tin processed with strong acid to disrupt the ordered
crystalline properties of native chitin to render it amor-
phous) were used. Preliminary experiments showed

SDS-PAGE sample buffer after the substrates had been washed to
remove nonspecifically bound protein. Note that the samples in (B)
are approximately sixfold concentrated compared with the corre-
sponding samples in (A) (A shows 20 lL of a 300 lL supernatant;
B shows 20 lL samples of bound protein resolubilized in 50 lLof
SDS-PAGE sample buffer). (C) Bar chart quantifying the binding of
LlCBP33A to a variety of insoluble substrates. Bound protein was
determined indirectly by measuring the concentration of free
protein in the supernatants after 24 h of incubation.
L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2408 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
binding the ordered, crystalline chitin forms rather
than individual chitin chains. Interestingly, LlCBP33A
also showed some binding to Avicel (microcrystalline
cellulose,  20% bound), as has also been observed
for other family 33 CBMs [21,41].
In terms of binding to the various chitin forms, the
characteristics of LlCBP33A are similar to those of
ChbB from B. amyloliquefaciens, in that both proteins
bind well to both a- and b-chitin. As noted above,
ChbB is the closest homologue of Ll CBP33A and the
two proteins share sequence characteristics that sepa-
rate them from the ‘one-substrate binders’ such as
CBP21 [17,20] and CHB1 [23]. It is conceivable that
the above-mentioned two mutations in the binding sur-
face and the two insertions that are putatively close to
this surface (Fig. 3) endorse LlCBP33A and ChbB
with the ability to bind a wider variety of substrates
than do CBP21 and CHB1.
Degradation of a- and b-chitin

LlCBP33A, the end-point of the reaction (i.e. solubili-
zation of all chitin) was reached after approximately
2 weeks. When LlCBP33A was present in the reaction,
the degradation rate was substantially higher, the
end-point being reached after approximately 48 h
(Fig. 6B). Thus, LlCBP33A clearly acts synergistically
with LlChi18A in the degradation of b-chitin. The
increase in LlChi18A efficiency on addition of
LlCBP33A is comparable with the increase observed
when adding CBP21 during the degradation of b-chitin
with ChiC1 from S. marcescens [10].
Although the occurrence of family 33 CBPs has been
known for some time [23], the present results provide
only the third demonstration of the accessory function
of these proteins. The effect of LlCBP33A on b-chitin
degradation is of the same order of magnitude as the
effect of CBP21. The effect on a-chitin degradation is
unique for LlCBP33A, but is rather modest (Fig. 6A).
It should be noted that, in nature, chitin is often found
as a composite where layers ⁄ sheets of chitin are inter-
woven with proteins and ⁄ or minerals in a recalcitrant
A
B
180
160
140
120
100
80
GlcNAc

Time (h)
200
180
160
140
120
100
80
60
40
20
0
200
Fig. 6. Chitin degradation by LlChi18A in the absence and presence
of LlCBP33A at pH 6.0, 37 °C. (A) Full lines show the degradation
of 0.5 mgÆmL
)1
a-chitin by LlChi18A ( ) and LlChi18A in the pres-
ence of LlCBP33A (d) with nonstatic incubation. (B) Full lines show
the degradation of 0.1 mgÆmL
)1
b-chitin by Ll Chi18A ( ) and
LlChi18A in the presence of LlCBP33A (d) with static incubation.
For comparison, the production of the minor end-product GlcNAc is
also shown (dotted lines through squares for LlChi18A; dotted lines
through circles for LlChi18A in the presence of LlCBP33A). The
production of GlcNAc in the reaction with a-chitin could not be
quantified accurately, but was of the same order of magnitude.
G. Vaaje-Kolstad et al. L. lactis chitinase and chitin-binding protein
FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS 2409

tion is that a disulfide bridge on the surface, close to
the important Tyr54 in CBP21 (Cys41–Cys49 in
CBP21), is missing in LlCBP33A, which has the 50–57
insert in this area (Fig. 2B). This could affect the bind-
ing properties of the protein, as it may introduce flexi-
bility and ⁄ or structural changes in this crucial region.
Conclusions
The present data show that the putative chitinase
and CBP genes in L. lactis code for a functional
chitinolytic machinery capable of converting chitin to
GlcNAc and (GlcNAc)
2
. The primary product of this
machinery is (GlcNAc)
2
, which can be converted to
mono-sugars by the putative N-acetylglucosaminidase
encoded by LnbA. We were able to show that L. lactis
indeed produces chitinolytic activity under certain
conditions. However, further work is needed to analyse
the role and regulation of the chitinolytic system of
this bacterium. LlChi18A was shown to be active and
relatively stable at low pH, which agrees with the
ability of L. lactis to grow and thrive in mildly acidic
environments.
The finding of nonhydrolytic accessory proteins for
chitinases has reinforced interest in the question as to
whether such proteins may also exist for cellulose. The
existence of substrate-disrupting accessory proteins and
domains that act synergistically with cellulases has

),
magnesium sulfate (0.25 gÆL
)1
) (Sigma), disodium glycerol-
ophosphate (19 gÆL
)1
) (Sigma) and manganese sulfate
(0.05 gÆL
)1
) (Sigma). As a carbon source, b-chitin isolated
from squid pen (France Chitin, Marseille, France), a-chitin
isolated from shrimp (Hov-Bio, Tromsø, Norway), colloidal
chitin and glucose were used (all chitin variants at a final
concentration of 1% w ⁄ v and glucose at final concentra-
tions of 0.1% or 0.4% w ⁄ v). The cultures were incubated
at 30 °C and samples were taken at various time points (4,
7, 8 and 10.5 h) in order to assay for chitinolytic activity in
the culture supernatants (see below for assay details).
Cloning of L. lactis chitinases and CBP
Genomic DNA from L. lactis was isolated from an over-
night culture using a midi-prep genomic DNA isolation kit
(Qiagen, Venlo, The Netherlands) and stored at ) 20 °C. A
3392 bp long region of the genome containing a putative
transcription regulator (GenBank ID: AAK06047.1), chitin-
ase gene (GenBank ID: AAK06048.1) and gene encoding a
family 33 CBP (GenBank ID: AAK06049.1) was amplified
L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2410 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
by PCR using primers flanking 100 bp upstream of the first
ORF and 100 bp downstream of the third ORF (forward

excised from the TOPO vector using XhoI and BsaI, and
ligated to NdeI–XhoI-digested pETM11 vector (Gu
¨
nter
Stier, EMBL Heidelberg, Germany). The pETM11 vector
contains a T7 promoter sequence for expression and an
N-terminal His
6
tag for immobilized metal affinity chroma-
tography purification.
The putative CBP was cloned using the pET30 Xa ⁄ LIC
kit (Merck Chemicals Ltd, Nottingham, UK), which
provides a ligation-independent method for cloning a gene
of interest. The expression vector (pET30 Xa ⁄ LIC) provides
an N-terminal His
6
tag that can be removed from the
N-terminus of the purified protein using activated factor X
leaving no non-native amino acids. Cloning primers were
designed according to the suppliers’ instructions, containing
ends compatible with the expression vector (forward
primer, 5¢-GGTATTGAGGGTCGCCATGGTTATGTTC
AATCACCA-3¢; reverse primer, 5¢-AGAGGAGAGTTAG
AGCCTTACAAGAAGGGTCCAAAGA-3¢). The PCR
product was purified, treated with T4 exonuclease to
create vector-compatible overhangs and annealed to a
prepared expression vector (pET30 Xa ⁄ LIC) provided by
the supplier.
The final constructs (pETM11-LlChi18A and pET30Xa-
LIC-LlCBP33A) were transformed into E. coli BL21Star

cleared lysate was applied to a 3 cm · 5 cm Ni-NTA col-
umn (Qiagen, Venlo, The Netherlands) equilibrated with
running buffer (100 mm Tris ⁄ HCl, pH 8.0). LlChi18A was
eluted by running four column volumes of elution buffer
(100 mm Tris ⁄ HCl, pH 8.0 and 100 mm imidazole) through
the column. The peak containing chitinase was collected
and concentrated using a Centricon P-20 unit (Millipore,
Billerica, MA, USA) and dialysed overnight in 20 mm
Tris ⁄ HCl, pH 8.0.
For LlCBP33A, the pellet resulting from centrifugation
of the sonicated cells was resuspended in denaturing buffer
containing 8 m urea, 0.1 m NaH
2
PO
4
,10mm Tris ⁄ HCl,
pH 8.0 and 25 mm dithiothreitol, and incubated at room
temperature for 3 h with gentle shaking. Subsequently, the
unfolded protein was purified on an Ni-NTA column under
denaturing conditions, using 8 m urea, 0.1 m NaH
2
PO
4
and
10 mm Tris ⁄ HCl, pH 8.0 as running buffer, and 8 m urea,
0.1 m NaH
2
PO
4
,10mm Tris ⁄ HCl, pH 8.0 and 100 mm

containing 50 lL of Xarrest Agarose (Merck Chemicals
Ltd). Both proteins were dialysed overnight in 20 mm
Tris ⁄ HCl, concentrated using Centricon P-20 units
(Millipore) and stored at 4 °C.
Protein purity was analysed by SDS-PAGE. Protein con-
centrations were determined using the Bradford micro-assay
(Bio-Rad, Hercules, CA, USA) according to the instruc-
tions provided by the supplier, employing purified bovine
serum albumin (New England Biolabs, Beverly, MA, USA)
as standard.
Chitin-binding assays
Binding studies were conducted using powdered a-chitin
from shrimp shells (Hov-Bio), powdered b -chitin from
squid pen (France Chitin), chitin beads (New England Biol-
abs), colloidal chitin and Avicel (microcrystalline cellulose;
Sigma). All chitin variants were suspended in ddH
2
Oto
yield a 20 mgÆmL
)1
stock suspension. Binding was assayed
in 1 mL reactions in Eppendorf tubes containing
5mgÆmL
)1
chitin and 400 lgÆmL
)1
LlCBP33A in 50 mm
citrate–phosphate buffer, pH 6.0. Reactions were mixed by
vertical rotation (60 r.p.m.) at room temperature for 24 h.
Subsequently, the chitin (with the bound protein fraction)

mum substrate concentration used had to be limited to
about twice K
m
because of the occurrence of substrate
inhibition (which is usual in this type of assay; for
example, see [8]). Standard reaction mixtures contained
2.0 nm of LlChi18A, 0.1 mgÆmL
)1
bovine serum albumin
and 0–200 lm of the substrate in 50 mm citrate–phosphate
buffer, pH 3.8. The reaction mixtures were incubated at
37 °C and product formation was monitored by taking out
50 lL samples at different time points (0–20 min), in which
the reaction was stopped by the addition of 1.95 mL of
0.2 m Na
2
CO
3
. The amount of 4MU released was deter-
mined by measuring the fluorescence emitted at 460 nm on
excitation at 380 nm, using a DyNA 200 fluorimeter
(Hoefer Pharmacia Biotech, San Francisco, CA, USA). The
release of 4MU proved to be linear with time for all sub-
strate concentrations, allowing the straightforward calcula-
tion of enzyme velocities by linear regression (all curves
had correlation coefficients above 0.99). Kinetic para-
meters were calculated by directly fitting the data to the
Michaelis–Menten equation by nonlinear regression using
graphpad prism (GraphPad Software Inc., San Diego, CA,
USA).

were injected using a Gilson 123 autoinjector. Eluted oligo-
saccharides were monitored by recording the absorption at
210 nm. Chromatograms were collected and analysed using
Gilson unipoint software (Gilson). A standard solution
L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2412 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
containing 100 lm of (GlcNAc)
1–4
was analysed at the
start, in the middle and at the end of each series of sam-
ples. The resulting average values of the standards (display-
ing standard deviations of < 5%) were used for
calibration. All measurements were performed in triplicate.
Background was corrected for by subtracting the value of
samples taken at t = 0 min.
The determination of the initial products from (GlcNAc)
6
degradation was performed by incubating 1 nm of
LlChi18A with 200 lm of (GlcNAc)
6
in 50 mm citrate–
phosphate buffer, pH 6.0. Products formed after 2 min of
incubation at 37 °C were analysed using isocratic HPLC as
described above.
The presence of chitinolytic activity in culture super-
natants of L. lactis was assayed using 4MU-(GlcNAc)
2
as
substrate; 50 lL of supernatant was mixed with 50 lLof
50 mm citrate–phosphate buffer, pH 3.8, containing 50 lm

of b-chitin powder
or 0.5 mgÆmL
)1
of a-chitin powder, in 50 mm citrate–phos-
phate buffer, pH 6.0. The reaction was buffered at a higher
pH than in the kinetic experiments as the long-term stabil-
ity (incubations exceeding 1 h) of LlChi18A (in the pres-
ence of bovine serum albumin) was better at a near-neutral
pH (at pH 6.0, there was no detectable loss of activity
under the conditions described below; A. C. Bunæs and
G. Vaaje-Kolstad, unpublished observations). Reaction
mixtures were incubated at 37 °C for up to 2 weeks with
vigorous shaking (a-chitin; 1300 r.p.m. in an Eppendorf
Thermomixer comfort; Eppendorf) or without shaking
(b-chitin). Initial experiments showed that the degradation
rate of a-chitin was slow when using static incubation
(results not shown); thus, to increase the amount of pro-
duct formed, and thereby the reliability of the assay, vig-
orous shaking was applied to promote the chitin–LlChi18A
and ⁄ or chitin–LlCBP33A contact. At time points ranging
from 2 to 340 h, 60 lL of the reaction was taken and
mixed with an equivalent amount of 70% acetonitrile in
an Eppendorf tube (the presence of acetonitrile arrests all
enzyme activity). All reactions were run in triplicate and
all samples were stored at )20 °C until product analysis by
HPLC as described above.
Acknowledgements
We thank Svein J. Horn for helpful discussions. This
work was funded by the Norwegian Research Council,
grants 171991 (GVK), 164653 (GVK), 159058 (GM)

9 Suzuki K, Sugawara N, Suzuki M, Uchiyama T, Kato-
uno F, Nikaidou N & Watanabe T (2002) Chitinases A,
B, and C1 of Serratia marcescens 2170 produced by
recombinant Escherichia coli: enzymatic properties and
synergism on chitin degradation. Biosci Biotechnol
Biochem 66, 1075–1083.
10 Vaaje-Kolstad G, Horn SJ, van Aalten DM, Synstad B
& Eijsink VG (2005) The non-catalytic chitin-binding
G. Vaaje-Kolstad et al. L. lactis chitinase and chitin-binding protein
FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS 2413
protein CBP21 from Serratia marcescens is essential for
chitin degradation. J Biol Chem 280, 28492–28497.
11 Robyt JF & French D (1967) Multiple attach
hypothesis of alpha-amylase action: action of porcine
pancreatic, human salivary, and Aspergillus oryzae
alpha-amylases. Arch Biochem Biophys 122, 8–16.
12 Teeri TT (1997) Crystalline cellulose degradation: new
insight into the function of cellobiohydrolases. Trends
Biotechnol 15, 160–167.
13 von Ossowski I, Stahlberg J, Koivula A, Piens K, Bec-
ker D, Boer H, Harle R, Harris M, Divne C, Mahdi S
et al. (2003) Engineering the exo-loop of Trichoderma
reesei cellobiohydrolase, Ce17A. A comparison with
Phanerochaete chrysosporium Cel7D. J Mol Biol 333,
817–829.
14 Tjoelker LW, Gosting L, Frey S, Hunter CL, LeTrong
H, Steiner B, Brammer H & Gray PM (2000) Structural
and functional definition of the human chitinase chitin-
binding domain. J Biol Chem 275, 514–520.
15 Hashimoto M, Ikegami T, Seino S, Ohuchi N, Fukada

(2001) Characteristics of a Streptomyces coelicolor A3(2)
extracellular protein targeting chitin and chitosan. Appl
Environ Microbiol 67, 1268–1273.
23 Schnellmann J, Zeltins A, Blaak H & Schrempf H
(1994) The novel lectin-like protein CHB1 is encoded
by a chitin-inducible Streptomyces olivaceoviridis gene
and binds specifically to crystalline alpha-chitin of fungi
and other organisms. Mol Microbiol 13, 807–819.
24 Kolbe S, Fischer S, Becirevic A, Hinz P & Schrempf H
(1998) The Streptomyces reticuli alpha-chitin-binding
protein CHB2 and its gene. Microbiology 144 (Pt 5),
1291–1297.
25 Folders J, Tommassen J, van Loon LC & Bitter W
(2000) Identification of a chitin-binding protein secreted
by Pseudomonas aeruginosa. J Bacteriol 182, 1257–1263.
26 Moser F, Irwin D, Chen S & Wilson DB (2008) Regula-
tion and characterization of Thermobifida fusca carbo-
hydrate-binding module proteins E7 and E8. Biotechnol
Bioeng 100, 1066–1077.
27 Chen HC, Chang CC, Mau WJ & Yen LS (2002)
Evaluation of N-acetylchitooligosaccharides as the main
carbon sources for the growth of intestinal bacteria.
FEMS Microbiol Lett 209, 53–56.
28 Bolotin A, Wincker P, Mauger S, Jaillon O, Malarme
K, Weissenbach J, Ehrlich SD & Sorokin A (2001) The
complete genome sequence of the lactic acid bacterium
Lactococcus lactis ssp. lactis IL1403. Genome Res 11,
731–753.
29 Suzuki K, Taiyoji M, Sugawara N, Nikaidou N, Hen-
rissat B & Watanabe T (1999) The third chitinase gene

L. lactis chitinase and chitin-binding protein G. Vaaje-Kolstad et al.
2414 FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS
36 Zeltins A & Schrempf H (1997) Specific interaction of
the Streptomyces chitin-binding protein CHB1 with
alpha-chitin – the role of individual tryptophan resi-
dues. Eur J Biochem 246, 557–564.
37 Bokma E, Barends T, Terwissch van Scheltingab AC,
Dijkstr BW & Beintema JJ (2000) Enzyme kinetics of
hevamine, a chitinase from the rubber tree Hevea brasil-
iensis. FEBS Lett 478, 119–122.
38 Hoell IA, Klemsdal SS, Vaaje-Kolstad G, Horn SJ &
Eijsink VG (2005) Overexpression and characterization
of a novel chitinase from Trichoderma atroviride strain
P1. Biochim Biophys Acta 1748, 180–190.
39 Horn SJ, Sorlie M, Vaaje-Kolstad G, Norberg AL,
Synstad B, Varum KM & Eijsink VGH (2006) Compar-
ative studies of chitinases A, B and C from Serratia
marcescens. Biocatal Biotransform 24, 39–53.
40 Horn SJ, Sikorski P, Cederkvist JB, Vaaje-Kolstad G,
Sorlie M, Synstad B, Vriend G, Varum KM & Eijsink
VG (2006) Costs and benefits of processivity in enzy-
matic degradation of recalcitrant polysaccharides. Proc
Natl Acad Sci USA 103, 18089–18094.
41 Tsujibo H, Orikoshi H, Baba N, Miyahara M, Miyam-
oto K, Yasuda M & Inamori Y (2002) Identification
and characterization of the gene cluster involved in
chitin degradation in a marine bacterium, Alteromonas
sp. strain O-7. Appl Environ Microbiol 68, 263–270.
42 Reese ET, Siu RGH & Levinson HS (1950) The biologi-
cal degradation of soluble cellulose derivatives and its

marcescens at 1.9-A
˚
resolution. Proc Natl Acad Sci
USA 97, 5842–5847.
50 Schwede T, Kopp J, Guex N & Peitsch MC (2003)
SWISS-MODEL: an automated protein
homology-modeling server. Nucleic Acids Res 31,
3381–3385.
G. Vaaje-Kolstad et al. L. lactis chitinase and chitin-binding protein
FEBS Journal 276 (2009) 2402–2415 ª 2009 The Authors Journal compilation ª 2009 FEBS 2415