Functional studies and engineering of family 1
carbohydrate-binding modules
Janne Lehtiö
Royal Institute of Technology
Department of Biotechnology
Stockholm 2001
ISBN 91-7283-152-9
Functional studies and engineering of family 1
carbohydrate-binding modules
Janne Lehtiö
Royal Institute of Technology
Department of Biotechnology
Stockholm 2001
Cover photograph: Recombinant Staphylococcus carnosus, expressing Trichoderma reesei Cel6A
CBM1 on the surface, immobilized on cotton fibers.
 Janne Lehtiö
Department of Biotechnology
Royal Institute of Technology (KTH)
SE-106 91
Sweden
Printed at Universitetsservice US AB, Stockholm 2001
ISBN 91-7283-152-9
CARBOHYDRATE-BINDING MODULES
TABLE OF CONTENTS
INTRODUCTION
4
1. STRUCTURE- FUNCTION RELATIONS OF CARBOHYDRATE- PROTEIN INTERACTIONS
4
1.1 Group I carbohydrate-binding proteins
5
1.2 Group II carbohydrate-binding proteins
7
1.2.1 Carbohydrate-binding modules
2. CELLULOSE- BINDING MODULES
7
8
2.1 Cellulose Structure
11
2.2 Binding-modules with affinity towards crystalline cellulose
12
2.4 Role of the cellulose-binding modules
19
3. FAMILY I CELLULOSE- BINDING DOMAINS
20
3.1 The fold of the family 1 cellulose-binding modules
21
3.2 Characterization and Function of family 1 CBMs
22
4. AIMS OF THIS STUDY
24
RESULTS AND DISCUSSION
24
5. FAMILY 1 CELLULOSE- BINDING MODULE FUNCTION AND INTERACTION
WITH CELLULOSE (I,
II)
24
5.1 The importance of being tryptophan: mutation studies of family 1 CBMs
25
5.2 Binding specificity of family 1 CBMs: electron microscopic experiments
29
5.3 CBM function as a part of an intact enzyme
32
6. CELLULOSE- BINDING MODULES IN AFFINITY APPLICATIONS (III, IV)
33
6.1 Role of the linker peptide in native and engineered enzymes
33
6.2 Bacterial cell immobilization
34
7. PROTEIN ENGINEERING OF FAMILY 1 CARBOHYDRATE- BINDING MODULES (V, VI)
7.1 Phage display of CBM1 for novel binding properties
36
38
7.1.1 Selection of a protein scaffold
38
7.1.2 Construction and characterization of the CBM1 library
39
7.1.3 Selection of binders from the randomized CBM1 library
40
8. CONCLUDING REMARKS
41
9. ABBREVIATIONS
43
10. ACKNOWLEDGEMENTS
44
11. REFERENCES
46
1
JANNE LEHTIÖ
Janne Lehtiö (2001). Fuctional studies and engineering of family 1 carbohydrate-binding
modules
Department of Biotechnology, Royal Institute of Technology (KTH), Stockholm, Sweden
ISBN 91-7283-152-9
Abstract
The family 1 cellulose-binding modules (CBM1) form a group of small, stable carbohydratebinding proteins. These modules are essential for fungal cellulose degradation. This thesis
describes both functional studies of the CBM1s as well as protein engineering of the modules
for several objectives.
The characteristics and specificity of CBM1s from the Trichoderma reesei Cel7A and Cel6A,
along with several other wild type and mutated CBMs, were studied using binding
experiments and transmission electron microscopy (TEM). Data from the binding studies
confirmed that the presence of one tryptophan residue on the CBM1 binding face enhances its
binding to crystalline cellulose. The two T. reesei CBM1s as well as the CBM3 from the
Clostridium thermocellum CipA were investigated by TEM experiments. All three CBMs
were found to bind in linear arrangements along the sides of the fibrils. Further analyses of
the bound CBMs indicated that the CBMs bind to the exposed hydrophobic surfaces, the so
called (200) crystalline face of Valonia cellulose crystals.
The function and specificity of CBM1s as a part of an intact enzyme were studied by
replacing the CBM from the exo-acting Cel7A by the CBM1 from the endoglucanase Cel7B.
Apart from slightly improved affinity of the hybrid enzyme, the module exchange did not
significantly influence the function of the Cel7A. This indicates that the two CBM1s are
analogous in their binding properties and function during cellulose hydrolysis.
The CBM1 was also used for immobilization studies. To improve heterologous expression of
a CBM1-lipase fusion protein, a linker stability study was carried out in Pichia pastoris. A
proline/threonine rich linker peptide was found to be stable for protein production in this host.
For whole bacterial cell immobilization, the T. reesei Cel6A CBM1 was expressed on the
surface of the gram-positive bacteria, Staphylococcus carnosus. The engineered S. carnosus
cells were shown to bind to cellulose fibers.
To exploit the stable CBM1 fold as a starting point for generating novel binders, a phage
display library was constructed. Binding proteins against an amylase as well as against a
metal ion were selected from the library. The amylase-binding proteins were found to bind
and inhibit the target enzyme. The metal binding proteins selected from the library were
cloned on the surface of the S. carnosus and clearly enhanced the metal binding ability of the
engineered bacteria.
Keywords: cellulose-binding, family 1 carbohydrate-binding module, phage display, bacterial surface
display, combinatorial protein library, metal binding, protein engineering, Trichoderma reesei,
Staphyloccus carnosus.
 Janne Lehtiö
2
CARBOHYDRATE-BINDING MODULES
LIST OF PUBLICATIONS
This thesis is based on the following papers, which in text are referred to by their Roman
numerals:
I. Lehtiö J., Sugiyama J., Gustavsson M., Fransson L. and Teeri T.T. The binding
specificity and affinity determinants of the family I cellulose-binding domains
from Trichoderma reesei Cel6A and Cel7A. Manuscript.
II. Srisodsuk M., Lehtiö J., Linder M., Margolles-Clark, E., Reinikainen T. and Teeri
T.T. Trichoderma reesei cellobiohydrolase I with an endoglucanase cellulosebinding domain: action on bacterial microcrystalline cellulose. Journal of
Biotechnology 1997 (57) 49-57.
III. Gustavsson M., Lehtiö J., Denman S., Teeri T.T., Hult K. and Martinelle M.
Stable linker peptides for a cellulose-binding domain–lipase fusion protein
expressed in Pichia pastoris. In press: Protein engineering.
IV. Lehtiö J., Wernérus H., Samuelson P., Teeri T.T. and Ståhl S. Directed
immobilization of recombinant staphylococci on cotton fibers by functional
display of a fungal cellulose-binding domain. FEMS Microbiol Lett. 2001 Feb
20;195(2):197-204.
V. Lehtiö J., Teeri T.T. and Nygren P.- Å. Alpha-amylase inhibitors selected from a
combinatorial library of a cellulose binding domain scaffold. Proteins; Structure,
Function and Genetics 2000 (41) 316-322.
VI. Wernérus H., Lehtiö J., Teeri T., Nygren P- Å. and Ståhl S. Generation of metalbinding staphylococci through surface display of combinatorially engineered
cellulose-binding domains. In press: Appl. Environ. Microbiol.
3
JANNE LEHTIÖ
INTRODUCTION
Carbohydrates play a key part in essentially all living organisms where their roles expand
from structural elements of cell walls to the delicate control of protein folding. The proteincarbohydrate interactions regulate countless activities in living cells. Glycoproteins in
eucaryotic cells maintain and regulate cell- cell interactions, control the cellular immune
system, are involved in protein targeting as well as in glycoprotein turnover in the blood
stream, to name a few examples of the different functions. In all living organisms
carbohydrates are used for energy storage and in many cases as a structural component of the
cells. In plants, the energy captured by photosynthesis is converted to chemical energy and
stored in carbohydrates. Marine and terrestrial plant cell walls, composed of mainly cellulose
and hemicellulose, form the greatest source of biomass on earth.
The ability of proteins to create specific interactions with soluble and solid carbohydrates is a
key to the sugar metabolism. Proteins that specifically recognize various sugar structures
maintain the above mentioned functions and are essentia l in the metabolic pathways for the
generation and degradation of carbohydrates. Lectins, monosaccharide transport proteins and
substrate binding modules of carbohydrate-active enzymes form a large and heterogeneous
group of proteins capable of specific recognition and binding of saccharides and
polysaccharides. This introduction will concentrate on non-catalytic carbohydrate-protein
interactions. The focus will be on carbohydrate-binding modules, which form a defined part
of a carbohydrate-active enzyme, with special emphasis on cellulose-binding modules. The
carbohydrate-active enzymes will be only briefly discussed as far as the substrate binding is
concerned.
1. Structure-function relations of carbohydrate-protein
interactions
Carbohydrate-binding proteins can be divided into two broad groups on the basis of their
binding site topology and location (Drickamer, 1989; Quiocho, 1986; Rini, 1995). The first
group consists of proteins which have a buried binding site for carbohydrates. This group
contains for example periplasmic monosaccharide-binding proteins, membrane bound sugar
transporters and enzymes such as hexokinases and exoglucanases. The second group of
4
CARBOHYDRATE-BINDING MODULES
binding proteins is composed of binders that feature a more open binding site topology on the
surface of the protein. This second group contains proteins such as lectins, endoglucanases,
immunoglobulins and substrate-binding modules of enzymes. Numerous proteins and
domains in this second category bind to insoluble carbohydrates. The binding sites of such
binders are by and large a shallow groove or a flat surface on the binding domain. The
topology of the binding site reflects the binding affinity and function of the protein. High
affinity binders tend to have sugar binding sites in a deep pocket or tunnel. This is typically
seen for example in the processive enzymes, which need to hold on to the substrate during the
hydrolytic process. On the other hand, proteins with a broad specificity, such as lectins, have
a shallow binding site.
Despite the various binding-site topologies of the sugar-binding proteins, shared principles as
to how they achieve molecular recognition and affinity towards saccharides can be found.
Van der Waals interactions arising from stacking aromatic residues of the protein onto the
hydrophobic pyranose rings of the sugar often have a critical role in the binding. The other
common binding interaction is based on hydrogen bonding between the polar groups on
saccharides and amino acid side chains.
1.1 Group I carbohydrate-binding proteins
Proteins with a deep cleft or a tunnel for sugar binding exhibit somewhat different binding
interactions than the proteins with a shallow cleft or a planar binding surface. Hydrogen
bonding is extensive and the hydroxyl groups of the oligosaccharides are often involved in
multiple hydrogen bonds with several amino acid side chains (Quiocho et al., 1997).
Aromatic stacking is also involved, often resulting in high affinity and in some cases also
contributing to specificity by steric exclusion. The dissociation constants (K d) of this type of
carbohydrate-binding interactions are typically in the nanomolar range. This is significantly
higher than that observed with other types of mono- and disaccharide-binders, such as lectins
(Table I). For example, studies of the bacterial maltodextrin/maltose-binding protein have
revealed a high affinity sugar protein interaction (Quiocho et al., 1997; Spurlino et al., 1992).
In this case, two of the studied sugar molecule substrates, maltose and maltotriose, are almost
totally engulfed by the protein, exposing less than 5% of the bound molecule to the solvent. A
relatively large substrate induced conformational change in the protein is postulated to occur
upon binding. Both, extensive hydrogen bond networks and aromatic stacking contribute to
5
JANNE LEHTIÖ
the binding. Very similar binding interactions are reported with other sugar-binding protein
receptors, such as D-allose binding protein (Chaudhuri et al., 1999) and D- glucose Dgalactose-binding protein (Appleyard et al., 2000). The binding event in these proteins seems
to be enthalpy driven and steric fitting of the substrate in the tight binding pocket is an
important asset in creating specificity.
Table I: Examples of carbohydrate-protein interactions and their affinities
Protein
Substrate
Affinity
Referens
Kd (µM)
Maltodextrin-binding protein E.
Maltotriose
0.16
(Quiocho, et al., 1997)
L-Arabinose
0.098
(Quiocho, et al., 1997)
Convalin A D. grandiflora
Mannotriose
3900
(Dam et al., 2000a)
CBM17Cel5A C. cellulovorans
PASA
29
(Boraston et al., 2000)
CBM22Xyn10B C. Thermocellum
Xylotriose
24
(Charnock et al., 2000)
CBM9Xyn10A T. maritima
PASC
18
(Boraston et al., 2001)
CBM2aXyn10A C.fimi
BMCC
3
(McLean et al., 2000)
coli
L-Arabinose- binding protein E.
coli
Carbohydrate-active enzymes synthesizing and degrading sugars bind the substrate in their
active sites following the same principles for the sugar binding. Their active sites often
contain aromatic and polar residues that create the affinity to the substrate. As far as the
binding site topology is concerned, similar structure-function relationship evident for the noncatalytic carbohydrate-binding proteins can also be detected on carbohydrate active enzymes.
This is clearly demonstrated in cellulose-degrading enzymes, for which several structures are
available for detailed structure-function studies of these enzymes (reviewed by Davies &
Henrissat, 1995). In cellulases, the topology of the substrate-binding site typically reflects the
functional characteristics of the enzyme. In the processive exoglucanases, the substratebinding site is buried in a long tunnel within the catalytic module. Such an active site allows
the enzyme to maintain the substrate chain attached to the active site hence enabling the
processive hydrolysis (Divne et al., 1994; Divne et al., 1998). According to the current
hypothesis, a sugar chain is guided into and held in the active site tunnel by several aromatic
residues and an extensive hydrogen bonding network. In related endoglucanases, the binding
6
CARBOHYDRATE-BINDING MODULES
site is located in a cleft. With the more open active site, the enzyme can cleave the substrate
from the middle of the chain and then supposedly dissociate to find the next cleavage site.
The β-glucosidases frequently have a pocket- like binding site topology, which is optimal for
more accessible chains and facilitates more specific recognition of small soluble substrates.
1.2 Group II carbohydrate-binding proteins
The second major group of proteins binding carbohydrates consists of lectin families, viral
proteins (e.g. hemagglutinin, foot-and- mouth disease capsid proteins), toxins, for example the
cholera toxin, and substrate-binding modules, such as the cellulose-binding modules. The
topology of the sugar-binding site in these proteins is a shallow groove or a flat surface.
Lectins form a large group of proteins with diverse biological functions. They are known to
be essential in cell adhesion, cell activation, growth, apoptosis, to initiate immune response
among other functions. The evolutionary diverse lectins are a good example of the common
principles of the carbohydrate-protein interactions achieved by proteins with different origin
and fold. Whatever the origin, the binding site of lectins is typically a shallow groove. Similar
to the monosugar-binding proteins, the affinity and specificity in ligand binding is achieved
through Van der Waals interactions using aromatic side chains and hydrogen bonds. Metal
coordination is also used for carbohydrate interactions in some lectins, such as the
mammalian type C-lectins (Feinberg et al., 2000). The affinity of lectins is relatively low,
often in a millimolar range. For example, the type 1 galectins have affinities around 1 x 104 5 x 104 M-1 (Ramkumar et al., 1995). Even with a relatively low affinity, lectins can
contribute to a tight interaction in cases when several lectins on the cell surface bind
simultaneously to numerous glycans on the surface of another cell (Dam et al., 2000b;
Sacchettini et al., 2001). Substrate-binding modules constitute another major subgroup of the
group II carbohydrate-binding proteins. These modules are linked to the catalytic modules of
carbohydrate-active enzymes. Characteristic to the binding modules so far studied is that they
have a shallow binding groove or a flat binding surface.
1.2.1 Carbohydrate-binding modules
Carbohydrate-active enzymes hydrolyzing insoluble substrates commonly have a modular
structure. The catalytic module or modules are coupled to one or often to several other
modules. In some cases more than ten modules with either a known or an unknown function
are linked together in one enzyme. One of the most frequently occurring modules in these
7
JANNE LEHTIÖ
enzymes are carbohydrate-binding modules. By definition, a carbohydrate-binding module
(CBM) is a contiguous amino acid sequence within a carbohydrate-active enzyme with a
discreet fold and independent carbohydrate-binding activity (Coutinho & Henrissat, 1999).
Modules that are part of a structural protein, such as the non-catalytic scaffolding protein in
the cellulosomes, represent an exception to this definition (for a recent review look Shoham et
al., 1999). The first characterized CBMs showed affinity towards cellulose (Gilkes et al.,
1988; Meinke et al., 1991; Tomme et al., 1988). Sequencing of an ever-growing number of
cellulose-binding modules (previously called domains) has led to their classification into
families (Coutinho et al., 1992; Gilkes et al., 1991a and b; Tomme et al., 1995a). Since
numerous new modules binding to a variety of carbohydrates have been discovered in recent
years, the classification has been broadened by Henrissat and Coutinho to include all
carbohydrate binding modules (Coutinho and Henrissat, 1999). Today, the CBMs are
classified in 28 families based on amino acid sequence similarities among over 800 sequences
of known or putative CBMs (Table II) (For the new carbohydrate-binding module
nomenclature see page 9). The known CBMs bind to various carbohydrates, ranging from
crystalline cellulose to soluble sugar residues and monosaccharides (Table II). Some of the
families are further divided into subclasses. For example the subclass CBM2a binds
crystalline cellulose while CBM2b binds oligosaccharides, amorphous cellulose and xylan
(Bolam et al., 2001; Simpson et al., 2000; Xie et al., 2001). The CBMs are usually linked to
the enzyme catalytic module by a flexible linker peptide. The CBMs show independent
binding to the substrate when separated from the rest of the protein (Din et al., 1994b; Linder
et al., 1995b). In some cases, a binding module forms a continued substrate-binding site with
the catalytic module. Such arrangements are seen for example in starch-binding modules of
amylases (MacGregor et al., 2001) and in the Thermofibida fusca endo/exo-acting cellulase
(Sakon et al., 1997).
2. Cellulose-binding modules
Insoluble polycarbohydrates, such as cellulose, represent challenging substrates for enzymatic
degradation. This is in part due to the poor accessibility of the tightly packed substrate to the
catalytic site of the enzyme and partly due to heterogeneity of the physical structure, which
varies from highly crystalline cellulose to isolated sugar chains. These substrate
characteristics set special requirements for the organisms and enzymes hydrolyzing it. One of
the evolutionary solutions for the enzymes degrading insoluble substrates has been to generate
8
CARBOHYDRATE-BINDING MODULES
distinct substrate-binding modules. Examples of such modules can be found commonly in
carbohydrate-active enzymes such as cellulases, hemicellulases and chitinases, but solid
substrate-binding domains are also found in other type of enzyme s, such as collagenases
(Matsushita et al., 2001; Overall, 2001). Several structures of the carbohydrate-binding
modules have been solved (Table III and IV). So far, predominantly the β-sheet secondary
structure is found in cellulose-binding modules. With known structures and substrate
specificities, studies of the structure-function relations of the CBM-cellulose interactions have
become possible.
CBM nomenclature
Old name
T. reesei CBHI CBD (CBDCBHI)
T. reesei CBHII CBD (CBDCBHII)
T. reesei EgI CBD (CBDEgI )
N. patriciarum CBHII
C. thermocellum CipA CBD
New name
T. reesei Cel7A CBM1 (CBM1 Cel7A)
T. reesei Cel6A CBM1 (CBM1 Cel6A)
T. reesei Cel7B CBM1 (CBM1Cel7B)
N. patriciarum Cel6A CBM1
C. thermocellum CipA CBM3
The nomenclature of cellulose-binding modules has been changed recently (previously called
cellulose-binding domains). In the new nomenclature, the family number follows the CBMabbreviation (for example family 1 modules are CBM1). For enzymes with more than one
CBM, the CBMs get a second serial number starting from the N-terminus of the protein (for
example CBM4-1 and CBM4-2). Some of the families are divided into sub- families. This is
indicated by a lower case letter following the family number (such as CBM3a or CBM2b).
9
JANNE LEHTIÖ
Table II: Carbohydrate binding module classification
Family
3D structures
55
17
Approximate size
of the CBM (aa)
30-40
110
90
150
150
5
6
3
20
60
120
yes
yes
7
8
9
- (deleted entry)
1
16
150
170
no
yes
10
11
12
13
10
3
63
94
50
180-200
40-60
150
yes
no
no
yes
14
15
16
91
2
7
70
40
160
no
no
no
17
8
200
no
18
19
20
144
4
60
40
60-70
100
no
no
yes
21
22
23
24
25
26
27
28
21
38
1
3
9
9
5
8
100
160
200
80
110
100
100
200
no
yes
no
no
yes
no
no
no
Summary
841
30-200
Referens:
1
2a
2b
3
4
10
Present number of
entries
71
88
yes
yes
yes
yes
yes
Demonstrated binding
substrates
Cellulose, Chitin
Cellulose, Chitin
Xylan
Cellulose, chitin
Xylan,
Amorphous cellulose
Cellulose
Amorphous cellulose,
Xylan
Cellulose
Cellulose, mono-, di-,
and Ologosaccharides
Cellulose
Cellulose
Chitin, Cellulose
Xylan, Mannose,
GalNac
Chitin, Cellulose
Cellulose
Xylan, Amourphous
cellulose
Amorphous cellulose,
Cellooligosaccharides
Chitin, Cellulose
Chitin
Granular starch.
Cyclodextrins
Starch
Xylan
Soluble mannan
α-1,3-glucan
Starch
Starch, Mannan
Mannan
Amorphous cellulose,
Cellooligosaccharides
(Coutinho and
Henrissat, 1999)
CARBOHYDRATE-BINDING MODULES
2.1 Cellulose Structure
Cellulose is a homopolymer of β-1,4-linked D-glucose. The smallest repeating unit in
cellulose is cellobiose, due to a rotation of every glucose ring by 180° relative to the adjacent
glucose ring (Figure 1). The degree of polymerization of cellulose varies from about 100 to as
many as 15 000 glucose units (Hon, 1994). The long linear glucose polymers crystallize to
form microfibrils. Several electron microscopic studies of the microfibrils demonstrate that
one microfibril can be considered as a single crystalline entity (Helbert et al., 1998; Sugiyama
et al., 1985). Apart from highly crystalline cellulose, more disordered semicrystalline and
amorphous cellulose are also present in various cellulose preparations and in the plant cell
walls. Recent data has confirmed that naturally occurring cellulose crystals have a chain
polarity with all reducing ends pointing in the same direction in the crystal (Koyama et al.,
1997). Relatively pure cellulose is produced by bacteria such as Acetobacter xylium.
OH
OH
OH
O
HO
HO
O
HO
O
OH
( 1-10)
HO
OH
O
HO
O
OH
H
OH
OH
O
O
OH
(110)
(200)
Figure 1. Cellulose structure. The glucose units are linked by β-1,4-linkages, with cellobiose
as the repeating unit. The glucose chains are packed into cellulose crystal. The crystal faces
are indicated on the figure. The gray short lines illustrate the packing of the individual chains
within the crystal.
In nature, cellulose is most often found as a part of the plant cell walls where it forms
complex structures with other wall polysaccharides, lignin and structural proteins. This
heterogeneity of the cellulose and cellulose composites presents a structurally varying target
for the binding modules and the enzymes acting on the cell walls.
11
JANNE LEHTIÖ
2.2 Binding-modules with affinity towards crystalline cellulose
Cellulose-binding modules (CBM) can be roughly divided into two groups according to their
substrate specificity: crystalline cellulose-binding modules and modules binding to
amorphous cellulose and cellooligosaccharides. Some CBMs, such as the CBM2a from P.
fluorescens and C. fimi, bind crystalline cellulose with high affinity but are also able
recognize amorphous cellulose (McLean et al., 2000; Nagy et al., 1998). With several
experimental CBM structures available, interesting features of the binding site topology
related to substrate specificity can be noted.
At present, there are several 3D structures for binding modules, which have demonstrated
affinity towards crystalline cellulose (families 1, 2, 3, 5, 9 and 10) (Table III). The structures
of all of these CBMs exhibit a planar binding face containing three aromatic amino acids,
mainly tryptophans and tyrosines. The importance of these aromatic side chains for the
binding to cellulose has been demonstrated by numerous studies (Bray et al., 1996; Din et al.,
1994; Linder et al., 1995a; Linder et al., 1995b; McLean et al., 2000; Nagy et al., 1998; Ponyi
et al., 2000; Tormo et al., 1996). Apart from the aromatic amino acids, a number of potential
hydrogen bonding polar amino acids, such as glutamines and asparagines, are often conserved
at close proximity to the aromatic side chains. In alanin scanning experiments these polar
amino acids have been shown to play a part in the binding interaction. However, the effects of
changing the polar residues are not as dramatic as the replacement in one of the aromatic
amino acids (Linder et al., 1995a; Linder et al., 1995b; McLean et al., 2000).
An interesting observation concerning the CBMs binding to crystalline cellulose is that some
of them exhibit seemingly irreversible binding or at least a very slow off- rate from crystalline
cellulose. This phenomenon has been shown for certain family 1, family 2 and family 3
CBMs, when the bound CBM has been eluted in non-equilibrium dilution experiments or in
exchange studies using radioactively labeled and non- labeled CBMs (Carrard & Linder, 1999;
Creagh et al., 1996; Ong et al., 1991). Some of the studied cellulose-binding modules, such as
T. reesei CBM1Cel7A and T. maritima CBM9, bind fully reversibly and return to equilibrium
as soon as the bound CBM is diluted with buffer (Boraston et al., 2001; Linder & Teeri,
1996). In one study, where the C. fimi CBM2a was bound to cellulose, the tryptophans were
nevertheless found to be accessible to oxidation with N-bromosuccinimide, although with a
lowered reaction rate than those in an unbound CBM (Creagh et al., 1996). The fact that these
12
CARBOHYDRATE-BINDING MODULES
aromatic side chains are not totally protected indicates that the binding is not fully
irreversible. The authors postulate that, at a given time point, two of the three aromatic side
chains are bound simultaneously, while the third aromatic side chain is not bound and
therefore remains more accessible to oxidation. In another study, the C. fimi CBM2a was
labeled with a fluorescent compound and adsorbed onto the crystalline cellulose surface,
thereafter a part of the surface was photo-bleached. It was found that the bound CBMs
diffused laterally along the surface, moving to the bleached parts, but without clear
dissociation from the cellulose (Jervis et al., 1997). These experiments indicate that the
crystalline cellulose binding modules are maybe capable of “sliding” over a surface
containing a very high density of overlapping binding sites.
Table III: Published crystalline cellulose-binding module structures
Family CBM
T. reesei Cel7A CBM1
Crystal/solution
structure
NMR
T. reesei Cel7B CBM1
NMR
(Mattinen et al., 1998)
2a
C. fimi Xyn10A CBM2a
NMR
(Xu et al., 1995)
3a
C. thermocellum CipC
CBM3
E. crysanthemi Cel5A
CBM5
P. fluorescens Xynl0A
1
5
10
Crystal structure
Referens
(Kraulis et al., 1989)
(Tormo et al., 1996)
NMR
(Brun et al., 1997)
NMR
(Raghothama et al., 2000)
CBM10
There are currently several hypothesis on how the CBM-crystalline cellulose interaction is
achieved. It seems that many of these modules need a continuous crystalline surface for
effective binding. It is generally accepted that the planar ring-ring stacking interaction
between the aromatic side chains of the protein and the apola r face of the sugar pyranose rings
governs the binding interaction. According to one binding model, the crystalline cellulose
binding modules interact with the 200-surface of the crystal, where the pyranose rings are
fully exposed to solvent (Brun et al., 1997; Mattinen et al., 1997; Reinikainen et al., 1995;
Tormo et al., 1996). However, owing to the small estimated surface area of the (200) surface
on a perfect crystal, another model has been proposed recently. According to this model, the
CBMs bind onto the (110) and (1-10) faces of the cellulose crystal (McLean et al., 2000). On
these faces, the pyranose rings are only partially solvent-exposed and form a sort of staircase,
which might represent a binding site for the CBMs.
13
JANNE LEHTIÖ
It has been demonstrated that the orientation of the aromatic side chains is critical for the
binding interaction of the family 2b xylan-binding domain. In this case, a single mutation
(R262G) changing the orientation of a binding site tryptophan cha nges substrate specificity of
the CBM from xylan to cellulose (Simpson et al., 2000). This demonstrates the importance of
structural complementary of the binding site on the protein and on the substrate. Also, when
examining the structures, it seems compelling that the aromatic amino acids aligned on the
planar surface of a CBM have almost exactly the same spacing as every second glucose unit
on the cellulose surface (Reinikainen et al., 1992; Tormo et al., 1996) (Figure 2).
Thermodynamics of the C. fimi CBM2a-crystalline cellulose interaction has been studied by
microcalorimetry experiments (Creagh et al., 1996). It has been postulated that the
dehydration of the CBM binding surface and the cellulose crystal surface are the entropic
driving force of the binding (Creagh et al., 1996). Furthermore, the rigidity of the CBM
binding surface seems to be an important asset, since a disulfide deletion in the T. reesei
CBM1Cel6A weakens the affinity to crystalline cellulose (Carrard and Linder, 1999). This
disulfide bridge stabilizes the loop carrying one of the essential aromatic amino acids. The
deletion may thus alter the orientation of the tryptophan and reduce rigidity of the binding
surface. The entropy loss in a binding event, where a rigid planar CBM surface binds to a flat
crystalline substrate, is probably small. Therefore, the entropy gain in the dehydration of the
surfaces can give an overall beneficial entropy in the binding. Further, crystalline cellulose
presents a contiguous surface of binding sites for the CBM. This can give a motional freedom
for the bound CBM, which may also contribute to the entropy of the binding. The evidence on
the surface diffusion of the CBM2a supports this theory (Jervis et al., 1997).
14
CARBOHYDRATE-BINDING MODULES
A
C
B
D
Figure 2. Structures of crystalline cellulose-binding modules indicating how different folds
nevertheless maintain the same orientations of the essential aromatic side chains. A) T. reesei
Cel7A CBM1 (Kraulis et al., 1989), B) E. crysanthemi Cel5A CBM5 (Brun et al., 1997), C)
P. fluorescens Xynl0A CBM10 (Raghothama et al., 2000), D) a superposition of the three
structures. Side chains are shown only for the aromatic amino acids on the binding surface of
the molecules. In the panels A, B and C β-sheets are coloured green while loops and a short
α-helix in the family 10 CBM (panel C) are shown in yellow. In panel D, each structure is
colo ured using the colour of the aromatic side chains in the panels A, B and C to indicate its
identity.
15
JANNE LEHTIÖ
A
B
Figure 3. Solvent accessible structures of CBMs with different binding site topologies: A)
The structure of the C. fimi Cel9a CBM4-2 (Brun et al., 2000). The sugar-binding site is
located in the groove visable on top of the molecule. B) The structure of T. maritima Xyn10A
CBM9 (Notenboom et al., 2001). This family 9 CBM binds to reducing end of the sugar chain
and the binding site topology is a ”blind canyon”.
16
CARBOHYDRATE-BINDING MODULES
2.3 Binding modules recognizing amorphous cellulose and cellooligosaccharides.
The modules that bind to soluble carbohydrates or amorphous parts of the solid substrate
seem to have evolved binding sites in shallow grooves accommodating the interacting amino
acid side chains (Figure 3A). Such binding grooves are generally able to accommodate a
single sugar chain. The CBMs in the families 2, 4, 9, 13, 17 and 28 have so far shown binding
to amorphous cellulose or cello-oligosaccharides. Currently there are several structures solved
of CBMs with affinity towards isolated chains (Table IV).
The first CBM discovered with a clearly demonstrated binding specificity towards soluble
sugar chains was the C. fimi Cel9B CBM4-1. This binding module binds to soluble cellulose
derivatives such as HEC, CMC and β-glucan as well as cellooligosaccharides with a DP of 3
or greater. No added affinity was observed beyond DPs above 5, indicating that the binding
site can accommodate 5 sugar residues (Johnson et al., 1996b; Kormos et al., 2000; Tomme et
al., 1996a). Alanin scanning of the amino acids in the binding groove demonstrates the
importance of two tyrosines (Y19, Y84) for the carbohydrate-protein interaction: mutation of
these residues decreases the affinity by 50-fold (Kormos et al., 2000). Some of the
asparagines and glutamines have also been shown to contribute for the affinity of the CBM41. Similar results were obtained with the homologo us CBM4-2 from the same enzyme (C.
fimi Cel9B CBM4-2) (Brun et al., 2000). It is noteworthy, that effectively the same amino
acid side chains that contribute in the affinity for the crystalline cellulose are also essential for
the binding of isolated sugar chains. As shown earlier, the specificity is achieved by adjusting
the topology of the binding site and the orientation of the aromatic side chains (Simpson, et
al., 2000). The newly characterized family 17 CBM from the Clostridium cellulovorans
cellulase 5A has similar binding specificity as the C. fimi CBM4s, but exhibits an order of
magnitude greater affinity (Boraston et al., 2000). Interestingly, two newly characterized
binding modules, in family 4, found in the Rhodothermus maritus Xyn10A, bind both to
xylan and to amorphous cellulose (Abou Hachem et al., 2000). In a xylan chain, the adjacent
sugar residues rotate 120° (as compared to the 180° in cellulose) creating a helical backbone
structure and the structural basis of this wide binding specificity has not been investigated in
detail. There is another case, where the two family 2 xylan-binding modules from the C. fimi
Xyn11A bind synergistically, although quite weakly, to amorphous cellulose, even though
neither of the isolated domains have affinity to this substrate (Bolam et al., 2001).
17
JANNE LEHTIÖ
An interesting structure was recently published for the Xyn10A family 9 CBM from
Thermotoga maritima (Figure 3B) (Notenboom et al., 2001). This CBM9 shows specificity
for a wide range of substrates: BMCC, Avicel, amorphous cellulose (PASA), β-glucan,
cellobiose and also weak binding to xylan (Boraston et al., 2001). Binding experiments and
an enzyme complex structure with cellobiose indicate that the binding site can accommodate
two residues (Notenboom et al., 2001). When the cellulose was treated with sodium
borohydride, which reduces the hemiacetal linkage at C-1 of the glucose ring, the binding was
reduced or erased totally. This shows that the CBM9 binds to the reducing end of the sugar
chain, explaining its broad specificity (Boraston et al., 2001). The solved crystal structure of
the module reveals that the sugar-binding site is located in a “blind canyon”, a groove that is
blocked at one end (Figure 3B) (Notenboom et al., 2001). The complex structure with
cellobiose confirms that the reducing end of the saccharide is involved in several hydrogen
bonds thus contributing to the affinity (Notenboom et al., 2001).
Table IV: Published amorphous cellulose / cello-oligosaccharide binding module structures
Family CBM
Crystal/solution
Referens
structure
2b
C. fimi Xyn11A CBM2b*
3
T. fusca Cel9 CBM3
4
C. fimi Cel9B CBM4-1
NMR
(Johnson et al., 1996a)
C. fimi Cel9B CBM4-2
NMR
(Brun, et al., 2000)
Crystal Structure
(Notenboom, et al.,
9
T. maritima Xyn10A CBM9
NMR
Crystal structure
(Simpson et al., 1999)
(Sakon, et al., 1997)
2001)
22
C. Thermocellum Xyn10B
Crystal Structure
(Charnock, et al., 2000)
CBM22*
* Xylan binding module
The thermodynamics of the protein-carbohydrate bind ing seems to gradually shift from the
entropy-driven binding of the crystalline cellulose binders to the enthalpy-driven interaction
of the proteins using pocket- or deep cleft- like binding sites. Calorimetric studies of the C.
fimi Cel9B CBM4-1 and CBM4-2 indicate enthalpy-driven binding interactions with a
favorable standard enthalpy upon binding (Brun et al., 2000; Tomme et al., 1996a). In such
binding interactions the hydrogen bond network has an increased importance for generating
affinity compared with CBMs recognizing crystalline cellulose. However, several different
18
CARBOHYDRATE-BINDING MODULES
mutation studies have shown that the aromatic amino acids also contribute significantly to the
binding interactions when the CBMs bind to soluble substrates. On the other hand, mutations
on the hydrogen bonding polar residues of CBMs do not decrease the binding in a very
dramatic fashion (Kormos et al., 2000; Xie et al., 2001). This could be explained by the fact
that the CBMs binding to cellooligosaccharides have a quite shallow binding groove, which
allows a certain degree of motional freedom for the substrate. This might contribute for the
enthalpy-entropy compensation with loss of a hydrogen bond as postulated in the study of
family 2 xylan-binding module (Xie et al., 2001). Thermodynamic studies of the family 9
CBM, with a lectin like binding site topology, show that the binding event is highly
exothermic with favorable enthalpy changes, as indeed seen in several sugar- lectin
interactions (Boraston et al., 2001; Dam et al., 2000). The irreversible binding, which is
discussed with the crystalline cellulose binding modules, has not been reported with soluble
saccharide binding modules (Boraston et al., 2001; Tomme et al., 1996a).
2.4 Role of the cellulose-binding modules
Arguably the best-characterized binding modules are those with specificity to cellulose.
Cellulose binding modules have been found in at least 15 of the over 28 currently known
CBM families (Table II). Cellulose-binding modules range from small fungal binding
domains of approximately 36 amino acids in family 1, to modules consisting of over 200
residues in families 11 and 17. These CBMs have been shown to have spectrum of
specificities from strictly crystalline cellulose to amorphous cellulose and hemicellulose as
well as monosaccharides, as in lectins (reviewed by Linder & Teeri, 1997). In many cases, the
binding specificity of a CBM is not definite, but it is rather a spectrum of affinities towards
various substrates.
The obvious function of cellulose-binding modules is to bring and maintain the catalytic
module close to the substrate hence increasing the local substrate concentration for the
enzyme active site. This is evident in the CBM deletion experiments where the ability of the
isolated catalytic module to hydrolyze insoluble substrates is significantly decreased, while
the activity towards soluble substrate is not affected (Bolam et al., 1998; Gilkes et al., 1988;
Reinikainen et al., 1992; Reinikainen et al., 1995; Tomme et al., 1988). These experiments
indicate that the CBMs do not take part in the actual hydrolysis event but merely increase the
effective substrate concentration and prolong the enzyme association with the substrate
19
JANNE LEHTIÖ
surface. Similar effects have been observed with other carbohydrate-active enzymes when the
domain function has been studied. For example, in the C. thermocellum Xyn10B, the
substrate-binding module enhances the activity towards insoluble xylan (Charnock et al.,
2000). However, there are also indications that some CBMs can have other functions and
properties during cellulose degradation. The isolated C. fimi CBM2 seems to promote enzyme
activity on solid substrates even when not covalently linked to the catalytic core (Din et al.,
1994a). One of the postulated mechanisms in this synergistic action of the CBM, is that it
disrupts or disperses the cellulose fibers and may thereby assist the hydrolysis by increasing
the effective cellulose surface area. There is also evidence that the family 3 CBM increases
the processivity of the endo/exo-acting T. fusca Cel9A, possibly by helping to feed the
cellulose chain into the active site (Irwin et al., 1998). It has been observed that purified C.
fimi family 2 CBM inhibits the flocculation of the BMCC in the buffer solution, probably by
weakening the intercrystalline interactions (Gilkes et al., 1993). A similar effect has been
observed with the intact T. reesei Cel7B endoglucanase (Srisodsuk et al., 1993).
Most of the characterization of the CBM binding properties have been done on model
substrates such as BMCC, Avicel and Valonia cellulose. Relatively little work has been done
to study the CBM interactions with polycarbohydrate composites and plant cell walls, which
are the natural substrates for the enzymes containing cellulose binding modules. It should be
noted that CBMs with affinity to cellulose are also present in enzymes with the catalytic
domain active on another substrate than cellulose (Black et al., 1996; Bolam et al., 1998;
McKie et al., 2001; McLean et al., 2000). It is likely that the cellulose-binding modules in
these cases will facilitate the degradation of complex plant cell wall substrates instead of
cellulose per se (Black, et al., 1996; Bolam, et al., 2001). It is quite possible that more diverse
roles of the CBMs will be found when the enzyme action on the complex composite
materials, such as cell walls, will be more extensively studied.
3. Family I cellulose-binding domains
Family 1 CBMs form a distinctive group of small peptides with demonstrated activity towards
crystalline cellulose and, at least in one case, to chitin (Linder et al., 1996). The CBM1s are
almost exclusively of fungal origin. Only one algal cell wall protein has been found with four
repeats of putative CBM1s (Liu et al., 1996). Furthermore, all CBMs found in aerobic fungi
are exclusively from family 1. Structures solved for the family 1 CBMs have revealed a fold
20
CARBOHYDRATE-BINDING MODULES
composed of three short β-sheets, which are held together by 2 or 3 disulfide bridges (Kraulis
et al., 1989; Mattinen et al., 1998). The cellulose-binding surface of the family 1 CBMs is a
planar surface with three aromatic amino acids and few conserved polar residues (Figure 2A
and Figure 5) (Kraulis et al., 1989; Linder et al., 1995a; Linder et al., 1995b; Mattinen et al.,
1998; Reinikainen et al., 1992). The sequence alignments show that the three aromatic amino
acids and four to six cysteins are well conserved within the family 1 CBMs (Figure 4). Some
members of the family 1 CBMs, such as Agaricus bisporus CBM1, have a phenylalanine at
the “tip” of the wedge shaped molecule (position 31 in T. reesei CBM1 Cel7A) instead of a
tryptophan or a tyrosine. It remains to be seen, how well these putative CBMs with a
phenylalanine bind to carbohydrates and what is their substrate specificity.
Tr
Hi
Fo
Np
Pc
Pp
Cel7A
Cel7B
Cel6B
Cel5A
Cel45A
Cel45A
Cel5A
Cel7A
Cel7A
Cel45A
Cel10
Cel6A
Cel6A
Xyn10
Cel7A
Bgl3
Dom1
Dom2
Dom3
Dom4
PTQSHYGQCGGI...GYSGPTVCASGTTCQVLNPYYSQCL
CTQTHWGQCGGI...GYSGCKTCTSGTTCQYSNDYYSQCL
ACSSVWGQCGGQ...NWSGPTCCASGSTCVYSNDYYSQCL
QQTVWGQCGGI...GWSGPTNCAPGSACSTLNPYYAQCL
QQTLYGQCGGA...GWTGPTTCQAPGTCKVQNQWYSQCL
CTAERWAQCGGN...GWSGCTTCVAGSTCTKINDWYHQCL
AQGGAWQQCGGV...GFSGSTSCVSGYTCVYLNDWYSQCQ
PKAGRWQQCGGI...GFTGPTQCEEPYICTKLNDWYSQCL
GSVDQWGQCGGQ...NYSGPTTCKSPFTCKKINDFYSQCQ
SVVPAYYQCGGSKSAYPNGNLACATGSKCVKQNEYYSQCV
AQAPIWGQCGGN...GWTGATTCASGLKCEKINDWYYQCV
CSNGVWAQCGGQ...NWSGTPCCTSGNKCVKLNDFYSQCQ
CGGAAWAQCGGE...NFHGDKCCVSGHTCVSINQWYSQCQ
NCAAKWGQCGGN...GFNGPTCCQNGSRCQFVNEWYSQCL
VTVPQWGQCGGI...GYTGSTTCASPYTCHVLNPYYSQCY
AQSGLYQQCGGI...GWTGATTCVSGATCTVLNPYYSQCL
ACGVLYEQCGGI...GFDGVTCCSEGLMCMKMGPYYSQCR
QVKPPYGQCGGM...NYSGKTMCSPGFKCVELNEFFSQCD
VCGKEYAACGGE...MFMGAKCCKFGLVCYETSGKWQSQC
GEVGRYAQCGGM...GYMGSTMCVGGYKCMAISEGSMYKQ
Figure 4. Sequence alignment of CBMs and putative CBMs in family 1. The abbreviations
are Tr: Trichoderma reesei, Hi: Humicola insolens, Fo Fusarium oxysporum, Np
Neocallimastix patriciarum, Pc Phanerochaete chrysosporium, Pp Porphyra purpurea. The
conserved aromatic amino acids on the binding surface are indicated in bold.
3.1 The fold of the family 1 cellulose-binding modules
Family 1 cellulose-binding modules occur as structurally independent, well-defined domains
usually attached to a catalytic module via a flexible linker peptide at either the N- or Cterminal end of the catalytic module. Their molecular structure very much resembles that
21
JANNE LEHTIÖ
found in the inhibitory cystine knot (ICK) family (Norton & Pallaghy, 1998; Pallaghy et al.,
1994). Distinctive characteristics in the ICK-motifs are short β-sheets, which are held together
by up to four disulfide bridges. There is no significant hydrophobic core in these proteins, the
disulfide bridges alone seem to stabilize the structure. The proteins found in this structural
family are predominantly toxins interacting with ion channels or as enzyme inhibitors (Norton
& Pallaghy, 1998). Interestingly, a few lectins have also been found to adopt the ICK motif
structure. The venom of the chinese bird spider Selenocosmia huwena is a lectin with three
disulfide bridges and a three-stranded antiparallel β-sheet (Lu et al., 1999). There are also two
chitin binding proteins, which have similar structural features: the hevein, which is a chitinbinding protein from the rubber tree Hevea brasiliensis (Asensio et al., 1995), and the chitinbinding protein from Amaranthus caudatus (Martins et al., 1996). However, these proteins do
not require crystalline chitin for binding, but bind to oligochitin with five or more residues
with micromolar affinity (Asensio et al., 2000a; Asensio et al., 2000b). The similarity of these
proteins with the family 1 CBMs is limited to their structure being stabilized by disulfide
bridges and forming small compact and very stable domains. The spacing of the cysteins and
the pattern of the disulfide bridges seems to be conserved in the ICK-family and similar
conservation can be seen in the family 1 CBM sequence alignments (Flinn et al., 1999;
Norton and Pallaghy, 1998). However, many of the CBMs contain only two disulfide bridges,
instead of the four found in ICK-family.
3.2 Characterization and Function of family 1 CBMs
The best-characterized members of the family 1 CBMs are the Trichoderma reesei Cel7A
CBM1 and Cel6A CBM1 (for the new and the old nomenclature see page 9). The importance
of the aromatic side chains on the planar surface is well demonstrated in several studies
(Table V, in the results part) (Linder et al., 1995a; Linder et al., 1995b; Reinikainen et al.,
1992). Studies of these CBMs show that the binding characteristics rely on a few amino acids.
Single point mutations on the aromatic amino acids of the CBM1Cel7A can elevate the affinity
close to that of the significantly better binding of the CBM1Cel7B, or to erase it almost
completely (Linder et al., 1995a; Linder et al., 1995b). Mutation of the aromatic amino acid
(Y31A) to an alanine in T. reesei CBMCel7A is the least disturbing for the overall structure of
the molecule, but does nevertheless abolish its binding affinity to cellulose (Linder et al.,
1995b; Mattinen et al., 1997). Histidines occur on the binding face of the family 3 bacterial
CBMs but have not been observed in the family 1 CBMs (Sakon et al., 1997; Tormo et al.,
22
CARBOHYDRATE-BINDING MODULES
1996). However, a histidine residue has been succesfully introduced on the binding surface of
T. reesei CBM1Cel7A by mutagenesis, resulting in a pH-dependent binding module (Linder et
al., 1999; Reinikainen et al., 1992). Furthermore, when the binding surface of the Cel7A
CBM1 was randomized and displayed on a bacteriophage, one of the binders obtained, after
biopanning aga inst Avicel, contained a histidine residue as a part of the binding surface
(Smith et al., 1998).
Irreversible binding behavior reported on crystalline cellulose binding of some bacterial
CBMs, has also been observed with the T. reesei CBM1Cel6A (Carrard & Linder, 1999). The
mechanism of such irreversible binding is not yet clear, but this property can be altered to
fully reversible binding by single point mutations (Carrard & Linder, 1999; Linder and Teeri,
1996). Additionally, it has been shown that the two T. reesei CBMs (CBM1Cel7A and
CBM1Cel6A) demonstrate synergistic binding when the CBMs are covalently coupled to each
other with a linker peptide (Linder et al., 1996). The catalytic module and CBM1 have also
been shown to have synergistic binding onto cellulose, especially at concentrations resulting
in a lower surface coverage (Palonen et al., 1999; Srisodsuk et al., 1997).
A CBM1 is clearly critical for crystalline cellulose degradation by the fungal enzymes:
deletion of the CBM1 does not have an effect on the enzyme activity on soluble substrate but
reduces the enzyme activity on crystalline substrate considerably (Reinikainen et al., 1992;
Reinikainen et al., 1995; Tomme et al., 1988). It is clear that these CBMs enhance the local
concentration of the catalytic module in the proximity of the substrate, similar to some
bacterial CBMs (Gilkes et al., 1988). It has also been proposed, that the CBM1s contribute to
the disruption of the cell wall texture or even on cellulose crystals. So far, the only evidence
supporting such theory is a recent report showing that the T. reesei endoglucanase CBM1Cel7B
weakens the hydrogen bonding of the cellulose crystals, which leads to detectable changes on
an NMR-spectra (Xiao et al., 2001). Earlier reports have shown that the same intact enzyme
prevents the flocculation of BMCC in the solution, which may also be mainly a result of the
CBM binding (Srisodsuk et al., 1998). There is no clear evidence that CBM1s would promote
enhanced hydrolysis on cellulose when not covalently linked to the catalytic module. A
relatively long linker peptide coupling the binding- and catalytic module seems to be needed
for effective enzyme action on BMCC (Srisodsuk et al., 1993). An interesting observation
concerning the T. reesei Cel7A and Cel6A enzymes, is that the amount of bound enzyme
increases as a function of the hydrolysis on BMCC (Palonen et al., 1999).
23
JANNE LEHTIÖ
4. Aims of this study
The general aim of this work was to study family 1 carbohydrate-binding modules in order to
understand their binding mechanism, function as a part of a hydrolytic enzyme and their
engineering for different applications. The specific aims were:
1) To investigate the adsorption of the CBM1Cel7A and CBM1Cel6A on cellulose crystals
(publication I).
2) To study the function of family 1 CBMs as a part of an intact cellulase (publication
II).
3) To investigate the use of a CBM1 as an immobilization tag for another enzyme
(publication III) and for immobilization of microbial cells (publication IV).
4) To use phage display technology to engineer variants of the CBM1 with novel binding
specificities (publications V and VI).
RESULTS AND DISCUSSION
5. Family 1 cellulose-binding module function and
interaction with cellulose (I, II)
The solution structure of T. reesei CBM1Cel7A reveals a compact wedge shaped molecule,
which is stabilized by two disulfide bridges (Figure 5). Mutational studies have demonstrated
that the aromatic amino acids 5, 31 and 32 (numbering according to Kraulis et al., 1989) on
the planar surface are crucial for binding to crystalline cellulose (Linder et al., 1995b;
Reinikainen et al., 1992). In addition, it has been shown that the molecular rigidity can play
an essential role in the CBM1 binding, since removal of one of the disulfide bridges in the
Cel6A CBM1 reduces its binding (Carrard & Linder, 1999). In publication I, complementary
mutation studies have been carried out for the Cel7A CBM1. Mutation work performed in this
and other studies of the T. reesei CBM1Cel7A and CBM1Cel6A are summarized in Table V.
In order to study the CBM1 binding to crystalline cellulose, several CBM1s were fused to the
double Z-domain of the staphylococcal protein A (SpA) (Nilsson et al., 1987). The ZZ-tag
24
CARBOHYDRATE-BINDING MODULES
was used for several reasons. Firstly, the antibody binding ability of the ZZ-tag was utilized in
the affinity purification of the ZZ-CBM fusion proteins and in detection of the proteins during
the cultivations using the western blot technique. Furthermore, the tag allowed direct
immunogold labeling of the CBM fusions in the TEM studies (publication I). The wild type T.
reesei CBM1Cel7A, CBM1Cel6A and CBM1Cel7B, Neocallimastix patriciarum CBM1Cel6A and
two mutated CBM1Cel7A variants (CBM1Cel7AY5W and CBM1Cel7AY5W;Y31W) were cloned
and produced as fusion to the ZZ-domain (Figure 1, publication I). Also the family 3 CBM
from the C. thermocellum scaffolding protein CipA (CBM3CipA) was produced and tested in
some of the experiments (publication I). Binding of the CBM1- ZZ fusion proteins was tested
on BMCC and Valonia cellulose (equilibrium binding isoterms, Figure 2, publication I) and
the reversibility of the binding was investigated in non-equilibrium dilution experiments
(Table VI). To determine the binding site of the CBM1 on the cellulose crystal surfaces the
fusion proteins were immunogold labeled and studied by transmission electron microscopy
(TEM).
5.1 The importance of being tryptophan: mutation studies of family 1
CBMs
In the family 1 CBMs the three aromatic amino acids, that interact directly with cellulose, are
on a planar surface of the wedge shaped molecule (Figure 5) (Linder et al., 1995b;
Reinikainen et al., 1992). The Cel7A CBM1 interaction with cellohexose, shown in NMR
studies, seems to rely on the glucose residues 1, 3 and 5 interacting with the three aromatic
side chains on the CBM1 (Mattinen et al., 1997). The combined surface area of both the
aromatic side chains and the pyranose rings are postulated to contribute directly to the binding
energy in the hydrophobic interaction (Doig & Williams, 1992; Williams et al., 1993; Xie et
al., 2001). Apart from the aromatic amino acids, a glutamine (Q34) in CBMCel7A contributes
to the affinity (Linder et al., 1995b).
A single point mutation (Y5W) in the CBM1Cel7A increases its affinity (publication I) (Linder
et al., 1995b). Mutating two of the tyrosines to tryptophans (CBM1Cel7A Y5W:Y31W) does
not further enhance the affinity. All the CBM1s with at least one tryptophan (T. reesei Cel6A,
Cel7B and N. patriciarum Cel6A as well as the mutated Cel7A CBM1s) on their binding face
feature similar binding (publication I). Better binding of the CBM1 containing one tryptophan
might be due to increased surface area of the hydrophobic face stacking onto the pyranose
25
JANNE LEHTIÖ
rings on the cellulose surface. The fact that the second tryptophan on the double mutant does
not further enhance the binding is somewhat puzzling. Perhaps the exposed hydrophobic
surfaces are unable to form complementary interactions because the glucose residue spacing
on the cellulose chain does not perfectly match that of the aromatic residues on the CBM. The
spacing seems to be an important asset, since similar spacing can be seen in all the crystalline
cellulose-binding modules that have their 3D structure available (Figure 2) (Tormo et al.,
1996). Structural changes of the double mutant binding surface can not be ruled out in the
absence of the experimental structure. However, the binding of the N. patriciarum Cel6A,
containing two tryptophans, was shown to be similar to the other CBM1 with one or more
tryptophans.
While the T. reesei Cel7A CBM1 is fully reversible, the CBM1Cel6A has shown irreversible
binding on BMCC (Carrard & Linder, 1999). All of the CBM1s tested, fused to a double Z –
domain, show fully reversible binding and returned to equilibrium in less than 30 minutes
(Table VI). In the case of Cel6A CBM1, the irreversible binding earlier observed could be
canceled by a single point mutation (W7Y), by deleting one of the disulfide bridges or by
increasing the temperature (Carrard & Linder, 1999). Deletion of the disulfide bridge
probably makes the binding surface less rigid and would thus have a negative effect on the
entropy of the binding interaction. Changing a tryptophan to a tyrosin diminishes the surface
area needed to create the Van der Waals interaction and has been shown to reduce the affinity
of Cel6A CBM1 (Carrard & Linder, 1999). It is interesting that, while the isolated Cel6A
CBM binds irreversibly, the intact Cel6A is partially reversible (Carrard & Linder, 1999;
Palonen et al., 1999). This difference in the binding characteristics is puzzling since the two
modules of the intact enzyme bind synergisticly in low concentrations (Palonen et al., 1999).
It is possible that the added mass of the catalytic domain nevertheless weakens the binding
from seemingly irreversible to partially reversible. In the case of the CBM-ZZ fusion proteins,
the double Z- domain does not bind to cellulose at all. Again, it might be that the added mass
of the ZZ-fusion partner leads to the fully reversible binding of the protein (Table VI). The
fact that the CBM1 reversibility is temperature dependent and seems to be sensitive to the
mass changes raise questions of its biological relevance. In nature T. reesei grows in a tropical
climate, at temperatures where the CBMs show reversible binding. Furthermore, the mass of a
catalytic module changes the characteristics of the binding towards reversibility.
26
CARBOHYDRATE-BINDING MODULES
Y5
Q34
Y32
N29
Y31
Table V: CBM1 mutations and their relative effects on affinity
CBM1Cel7A
Y5A
---
Y5W
Q34A -
Y31A -
--
++
Y31W
++
Y5H
--
Y31H
-
Y5H
H4V
--
Y31D* -
--
P16R
-
T17K
V18T
A20T
-
-
Y32A -
V27Y
L28S
P30D
--
N29A
-
-
CBM1Cel6A
W7Y
--
∆SS3
--
27
JANNE LEHTIÖ
Figure 6. The structure of the Cel7A CBM1 shown in two different orientations (Kraulis et
al., 1989). In the combinatorial CBM1 library eleven positions were randomized. The amino
acids at the randomized positions are coloured red on the wild type structure, illustrating the
binding surface topology.
28
CARBOHYDRATE-BINDING MODULES
Consequently, the irreversible binding phenomenon is not likely to be relevant with intact
enzymes in nature. This could explain why the processively acting T. reesei Cel6A has a
seemingly irreversible CBM.
Table V: Non-equilibrium dilution experiment of the ZZ-CBM fusion proteins
Fusion protein
T. reesei
T. reesei
T. reesei
N. patriciarum
T. reesei
T. reesei
Recovery to isoterm in 30 minutes
Cel7A ZZ-CBM1
Cel6A CBM1-ZZ
Cel7B ZZ-CBM1
Cel6A CBM1-ZZ
Cel7A ZZ-CBM1Y5W
Cel7A ZZ-CBM1Y5W:Y31W
99,0±1,1
99,4±0,2
99,8±0,1
99,5±0,3
99,8±0,9
99,0±0,5
5.2 Binding specificity of family 1 CBMs: electron microscopic experiments
There is an ongoing discussion concerning possible specific binding sites of different CBMs
on the crystalline cellulose. It is difficult to detect any obvious differences on the binding
faces of the family 1 CBMs when comparing the sequences (alignment, Figure 4) or
experimental or modeled structures of the family 1 CBMs (Hoffren et al., 1995; Kraulis et
al., 1989; Mattinen et al., 1998). However, there is one study suggesting that CBMs may have
at least partly separate binding sites on crystalline cellulose (Carrard et al., 2000). In that
study, several CBMs were linked one by one to the CelD endoglucanase catalytic module
from C. thermocellum. It was shown that a complex with a new CBM promotes enhanced
hydrolysis after 50 h, whereas the addition of the original CBM-complex did not have this
boost effect (Carrard et al., 2000). However, it was found that the added uncoupled catalytic
module, used as a control, also promoted the boost effect in some cases, but not all.
Furthermore, with the error of 5-12% the differences were quite small and it remains to be
seen what is the actual specificity difference of these CBMs.
To address the question of the specificity of CBMs, binding of different CBM-ZZ fusion
proteins on Valonia cellulose was studied by transmission electron microscopic experiments.
The CBMs studied were family 1 CBMs from T. reesei Cel7A and Cel6A and the family 3
CBM from C. thermocellum. In the TEM visualisation, distribution of the gold particles on
isolated crystals indicated binding of the CBMs in one or two rows along the crystal edges
(publication I). For a more detailed picture of the CBM adsorption to Valonia cellulose
crystals, a set of TEM experiments was carried out with isolated, centred crystals carrying the
29
JANNE LEHTIÖ
bound immuno-gold labeled CBMs. Two-dimensional coordinates of the gold particles were
recorded with three different rotations of individual crystals (0°, -45° and 45°) and this data
was used to calculate the three dimensional location of the labeled proteins around the
cellulose crystal (Figure 4, publication I). The residual error for the coordinate transformation,
as defined in Eq. 1 (publication I) was always less than 0.02, indicating that the
experimentally measured locations of the labels fit acceptably in the mathematical equation,
and that the tilting and centering of the fiber were accurate. The calculated coordinates of the
labels indicated that the CBMs were bound in two distinct groups. The fiber orientation of the
same crystals were determined by diffraction experiments as described in (Imai et al., 1998;
Koyama et al., 1997). The diffraction pattern is distinctive to the direction of the glucose
chains in the crystal, hence it is possible to determine the orientation of the isolated crystal.
The mapping of the labels combined with the diffraction pattern analysis, indicate that the
CBMs adsorbed onto the surface with fully exposed pyranose rings, i.e. the (200) crystal face
of the cellulose surface.
The intriguing question is the amount of (200) face exposed on the crystal surface. In a
perfectly shaped crystal this face is only one glucose chain wide, which would lead to a very
limited surface area for the binding of the CBMs. The electron microscopic experiments of
Halocynthia papillosa and Valonia cellulose reveal that cellulose crystals are not perfectly
shaped. Instead, the (200) face is often blunt exposing a larger area of this surface to the
solvent (Helbert et al., 1998; Sugiyama et al., 1985). This observation could explain why
earlier surface coverage calculations of the CBMs on crystalline cellulose disagree with the
experimental binding capacity of the CBMs (Mattinen et al., 1997; Reinikainen et al., 1995;
Tormo et al., 1996). In case of imperfect crystals, the exposed (200) surface area increases
dramatically. The distribution of the labels into two distict groups this experiment is obvious.
Knowing that the distance of the gold particle to the actual binding site of the CBM can be
anything from 0.5 nm up to 5 nm judging from the sizes of the CBM-ZZ fusion protein and
the gold-conjugated antibody, the grouping of labeling was significant. The estimated size of
the isolated Valonia crystal is 20 x 20 nm and the clear grouping of the labels implies that the
specific binding site is to a limited surface on the crystal. We did not detect any significant
binding on the (110) or (1-10) surfaces as proposed by (McLean et al., 2000).
In the experiments where T. reesei enzymes Cel6A and Cel7A binding was monitored as a
function of the hydrolysis, the relative amount of bound enzyme increased despite the fact
30
CARBOHYDRATE-BINDING MODULES
that the total cellulose surface area diminishes (Palonen et al., 1999). This agrees with our
TEM visualization data since the cellobiohydrolases peeling off chains from the (200) face
would simultaneously reveal more of the (200) surface for the CBM binding. Other
experiments have suggested that bacterial CBM2s have primary and secondary binding sites
on the BMCC (Creagh et al., 1996) (personal communication with Larry Walker). It may be
that some CBMs first bind to (110) face of the crystal as suggested in (McLean et al., 2000).
Binding on the (110) surface should conceivably have lower affinity than the binding on the
(200) face. In the above mentioned recent publication by (Carrard et al., 2000), the CBM3CipA
fused to a catalytic module by cohesin-dockerin fusion, is postulated to cover a wide range of
binding sites on the cellulose surface. In the case of CBM3, an addition of the same or another
type of CBM, fused to the same catalytic module, does not promote an additional boost effect.
We also tested the same C. thermocellum CBM3CipA fused to the ZZ-domains in the TEM
experiments (publication I). The labeling pattern obtained with CBM3CipA on the Valonia
cellulose crystals was very similar to the T. reesei CBM1 Cel7A and CBM1Cel6A. Also the
CBM3CipA mapping with rotation experiment gave similar results, indicating binding on the
(200) face of the crys tal. This suggests, that the crystalline cellulose-binding modules use a
universal strategy for creating affinity to crystalline cellulose.
Linking of the T. reesei Cel7A and Cel6A CBM1s together with a linker peptide results in
significant synergistic binding, indicating that the binding sites of these two CBMs on the
BMCC have to be able to bind very close to one another on the crystal surface (Linder et al.,
1996). We have also constructed a different double CBM with the two CBM1s separated by a
double Z-domain (unpublished data). This double CBM was found to exhibit synergistic
binding to BMCC similar to the first double CBM. A domain exchange experiment, where the
CBM1 of a processive cellobiohydrolase (T. reesei Cel7A) was replaced with the higher
affinity CBM1 from endo-acting enzyme (T. reesei CEL7B), indicated no great change or
even slightly enhanced hydrolytic capacity of the hybrid enzyme (Figure 3a, publication II).
This indicates that the different endoglucanase CBM1 did not direct the cellobiohydrolase
catalytic module to unfavorable locations on cellulose (publication II). Furthermore, the
presence of the Cel7B CBM did not hinder the processivity of the exo-acting enzyme with its
higher affinity. Also these experiments suggest that the family 1 CBMs bind to similar sites
on a cellulose crystal.
31
JANNE LEHTIÖ
5.3 CBM function as a part of an intact enzyme
The function of the CBMs in the overall process of cellulose hydrolysis is an interesting
question. A CBM obviously enhances the effective local concentration of the catalytic
module, hence allowing it to react with the insoluble substrate. In the case of crystalline
cellulose-degrading enzymes, the task of breaking the crystal is likely to be the rate- limiting
step instead of the bond cleavage event. The CBM might act by maintaining the catalytic
module in close proximity of the crystal allowing the substrate to enter the active site (Teeri et
al., 1998). The domain switch experiment in publication II raises the question why T. reesei
Cel7A has CBM that binds weaker when the chimeric enzyme with the higher affinity
CBM1Cel7B can hydrolyse BMCC somewhat more effectively (Figure 3a, publication II). Why
has this domain switch not occurred in nature? The natural substrate of these enzymes is the
plant cell wall and the fungus produces a series of enzymes to degrade the wall
polycarbohydrates. The probable answer lies in the interaction of the synergistically acting
enzyme mixture on the complex substrates like plant cell walls.
The action of the Cel7A and Cel6A has been shown to be processive, with Cel6A acting on
the non-reducing and with Cel7A reducing ends of that crystal (Boisset et al., 2000; Brooks et
al., 1992; Imai et al., 1998). However, we could not detect any distinct preference of the
studied cellobiohydrolase CBM1s to the cellulose chain ends in the microscopic studies
(publication I). Instead, as the TEM experiments indicated that both CBM1s bind to the
exposed (200) face along the crystal, which is likely to be exposed on the disrupted parts of
plant cell wall. In that way, the proposed binding specificity could in fact direct the processive
exoglucanases towards accessible chain ends for productive binding. Interestingly, the
recently discovered family 9 T. maritima CBM9Xyn10A has been shown to bind selectively to
the reducing end of the cellulose (Boraston et al., 2001; Notenboom et al., 2001). This
binding specificity could indeed bring a catalytic module in close proximity of the chain ends.
However, in the case of the processive cellobiohydrolases, with the active site in a tunnel
(Divne et al., 1993; Divne et al., 1994; Divne et al., 1998), it is probable that this kind of
binding would block the productive bind ing or stop the processive action of the enzyme.
32
CARBOHYDRATE-BINDING MODULES
6. Cellulose-binding modules in affinity applications (III,
IV)
Today, there are countless studies in which a specific binding ability of a protein has been
used for various biotech applications. The traditional uses of affinity proteins include ligands
for affinity purification, different detection methods employing antibodies and other binding
proteins, enzymes and cell immobilization. The cellulose binding modules present an
attractive immobilization tag for biotechnical applications due to their natural ability to
spontaneously bind to cellulose across a wide range of conditions. Cellulose provides an
inexpensive, biodegradable and medically safe immobilization matrix, and proteins fused to
CBMs can be immobilized and purified in a one step procedure. There is a vast number of
studies where a CBM has been used as an affinity tag and an immobilization anchor (Bayer et
al., 1994; Ong, et al., 1991) (Linder et al., 1998; Rotticci-Mulder et al., 2001; Tomme et al.,
1998; Tomme et al., 1996b). In publication III, we have carried out a study to optimize a
linker peptide for heterologous protein expression of the Candida antarctica lipase B (CALB)
in yeast. We have also studied the direct immobilization of recombinant Staphylococcus
carnosus cells onto cellulose fibers by surface expression of the CBM1Cel6A (publication IV).
6.1 Role of the linker peptide in native and engineered enzymes
Carbohydrate active enzymes are typically modular proteins with a catalytic module attached
to various other modules (Henrissat, 1999; Henrissat & Davies, 1997). Especially in enzymes
active on insoluble substrates, these modules are for carbohydrate binding. These CBMs are
frequently linked to the rest of the enzyme through a long and probably flexible linker peptide
(Beguin & Aubert, 1994; Gilkes et al., 1991b; Teeri et al., 1987). The role of the linker
peptide in natural enzymes has been studied in several linker deletion studies (Srisodsuk, et
al., 1993) (Black et al., 1997; Ferreira et al., 1990). In the T. reesei Cel7A, both the
absorption and the activity were reduced by deleting major parts of the linker (Srisodsuk et
al., 1993). It seems in this case, that for effective hydrolysis of the insoluble substrate, the
catalytic module has to be anchored to the cellulose by a CBM with a relatively long linker to
allow productive binding and hydrolysis. Similar long linker sequences are found in cellulosebinding modules with catalytic modules active on hemicellulose. The length of the linker is
shown to be important when degrading the complex cell wall or composite structures, but the
importance of the linker le ngth is not always apparent when the enzyme action is studied
33
JANNE LEHTIÖ
using model substrates (Black et al., 1997; Ferreira et al., 1990). There are also some reports
that truncated forms of the enzymes occur in the culture supernatant under certain culture
conditions. Under these conditions, the catalytic core of the enzyme is probably released by
linker-specific proteases (Knowles et al., 1988; Miller et al., 1988). It might be that in some
cases it is beneficial to release the catalytic module from the binding module to allow
additional mobility. It has also been speculated that the linker sequences are beneficial for a
rapid evolution of these modular enzymes (Ferreira et al., 1990).
Protein engineering of enzymes for biotechnical applications often requires linker peptides as
spacers to facilitate protein-protein fusion. The most important feature of such linkers is
proteolytic stability. Another requirement of these linkers is solubility, in order to sustain the
extended conformation in solution and flexibility, in order to allow independent function of
the fused protein domains. Linker peptides found in nature frequently consist of small polar
residues such as glycine, presumably to give flexibility and enhance the solubility (Argos,
1990). In fungal cellulases O-glucosylated linkers are common, probably to improve the
solubility and proteolytic stability (Srisodsuk et al., 1993). In an earlier study, a lipase linked
to the CBM1Cel6A from Neocallimastix patriciarum showed good binding on cellulose and the
CBM-lipase fusion protein was shown to maintain full lipase activity (Rotticci-Mulder et al.,
2001). However, the linker peptide used for the fusion suffered significant proteolytic
degradation during cultivations of the Pichia pastoris host. As described in publication III, we
tested a series of engineered linkers to find a stable, well-expressed linker peptide for protein
production in P. pastoris (publication III, Table I). The native N. patriciarum linker (linker
construct 2) was susceptible to proteolysis like the earlier linker derived from the T. reesei
Cel6A linker (construct 1). We found that the proline/threonine rich linker, which mimicked
the C. fimi cellulase linker, was stable in P. pastoris protein production. This linker was
somewhat shorter (13 amino acids) than the linkers 1, 2 and 3, which were degraded during
the cultivation. The 13 amino acid long linker is, in all probability, enough to give sufficient
spatial separation from the cellulose surface and allows optimal orientation of the catalytic
site for interaction with the substrate.
6.2 Bacterial cell immobilization
Immobilization of microorganisms and recombinant surface expression of selected affinity
proteins have considerable potential in biotechnology, applied microbiology and
34
CARBOHYDRATE-BINDING MODULES
immunology. In addition to immobilization of various enzymes and antibodies, CBMs have
been employed also for whole cell immobilization (Francisco et al., 1993; Wierzba et al.,
1995). In recent years there have also been several reports of bacterial surface display aiming
for various applications (for reviews see Georgiou et al., 1997; Ståhl et al., 2000; Benhar,
2001; Hansson et al., 2001). A widely studied area of bacterial surface display is live vaccinedelivery systems, but biotech applications such as whole-cell bioabsorbents for environmental
use (Samuelson et al., 2000; Sousa et al., 1996), diagnostic tools development (Fuchs et al.,
1991; Gunneriusson et al., 1999; Gunneriusson et al., 1996) and display of combinatorial
libraries on bacteria are also growing areas (Daugherty et al., 1999; Lu et al., 1995). Here we
have used the T. reesei CBM1Cel6A for immobilizing the gram-positive Staphylococcus
carnosus on cellulose fibers (publication IV). This food-grade bacteria is traditionally used
for food biotech applications like starter cultures in meat and fish fe rmentations as well as in
numerous vaccine studies. Consequently, it is regarded safe in food products and
environmental applications (Hammes et al., 1995). Immobilization of recombinant S.
carnosus cells on to inert and medically safe cellulose matrixes could be used for example in
the development biofilters, fermentation immobilization or possibly in development of fast
and easy diagnostic tools using a “dip-stick” format.
The CBM1Cel6A from T. reesei has been well characterized. It has a typical family 1 CBMstructure with three disulfide bridges stabilizing the domain and a flat surface featuring three
exposed aromatic amino acids (Carrard & Linder, 1999; Hoffren et al., 1995). The Cel6a
CBM1 provides a robust and stable immobilization tag with specific affinity towards
cellulose. The CBM1Cel6A has higher affinity to cellulose than CBM1Cel7A and the isolated
CBM1Cel6A does not dilute from the cellulose in non-equilibrium experiments at low
temperatures (Carrard & Linder, 1999). These features are beneficial for proteins utilized as
immobilization tags.
The Cel6A CBM1 was cloned into the surface expression vector (Samuelson et al., 1995) and
transformed to S. carnosus cells. Surface accessibility of the CBM, and the albumin binding
protein (ABP) fused to CBM on the surface receptor chimera, were detected in whole-cell
ELISA assays, as described in the materials and methods (publication IV). The chimeric
surface receptor was also purified from the cell wall of the S. carnosus and shown to exhibit
the expected molecular weight on western blot analyses. Functionality of the displayed
CBM1Cel6A was detected by cellulose binding experiments on cotton fibers as well as by light
35
JANNE LEHTIÖ
microscopic experiments (publication IV). The recombinant S. carnosus cells bound readily
to the cotton fibers, and in higher bacterial densities, the fibers were covered with a
monolayer of bacteria. The bacteria lacking surface expressed CBM were not associated with
fibers. The good immobilization of the bacteria is probably due to a avidity effect of the
CBM1Cel6A, since these chimerical receptors are shown to have high surface density on S.
carnosus (Robert et al., 1996). Synergistic, high affinity binding of a double CBM has also
been demonstrated in a study where two family 1 CBMs were coupled with a linker peptide,
(Linder et al., 1996). In an earlier study recombinant S. carnosus cells were immobilized
successfully on fibronectin by surface expression of different fibronectin-binding domains
(Liljeqvist et al., 1999). These two studies propose that this heterologous surface display
method provides a good general way to create whole-cell immobilization for various
applications.
7. Protein engineering of family 1 carbohydrate-binding
modules (V, VI)
Traditionally, the immune system has been employed for the generation of high-affinity
antibodies to diverse target molecules for vast application areas including therapeutic,
diagnostic and biotechnological applications. More than 108 antibodies with different binding
specificities circulate in the human body (Skerra, 2000). This diversity has the potential to
recognize virtually any foreign structure that the body encounters. However, the stability,
heterologous production, size and complexity of the antibodies can present some difficulties
and limitations for their use in biotechnology applications. This has led to efforts to recruit
alternative protein scaffolds for creating diversity for the selection of novel binders (for
reviews see Benhar, 2001; Nygren & Uhlen, 1997). In such work, proteins with the desired
binding properties are selected from pools (libraries) of related but different candidates using
a suitable selection system. Libraries are typically constructed using combinatorial
mutagenesis, at the DNA level, of the molecular surface of the chosen parental protein
scaffold.
36
CARBOHYDRATE-BINDING MODULES
A
P
S X
gene III
1) Transformation of E. coli
with phagemid DNA
2) Infection with helper phage
protein X
(phagemid encoded)
pIII
(helper phage encoded)
406 aa/ 3-5 copies
Phagemid DNA
®
pVIII
50 aa/ ca 2700 copies
single stranded DNA
6.4 kb
B
Production of
phage particles
Target
Selection
Wash
Cultivation of E. coli cells
containing phagemid DNA
Elute binding
phages
Identify binder from
insert DNA sequence
Reinfect E. coli
Phage display
A) Principle of the phagemid system “3
E. coli cells are transformed with the phagemid
containing the foreign gene, in this case a CBM1 containing the randomized residues (X),
fused to pIII- gene. Thereafter, the growing cells are coinfected with a helper phage to produce
new phage particles, which are now decorated with the library variants, and contain the
respective gene in the phagemid. The helper phages are required for phage assembly.
Schematic drawing of phage particle is shown on the right. B) A schemmatic presentation of
the selection procedure (biopanning). A standard selection for novel binders involves multible
rounds of selection (typically 4-5 biopanning rounds) combined with amplification. The
variants with affinity towards the desired target molecule are selected and enriched. After
several rounds, the selected variants are indentified by DNA sequencing. Thereafter, the
variants are produced for further characterization and application.
37
JANNE LEHTIÖ
7.1 Phage display of CBM1 for novel binding properties
Combinatorial protein libraries employing the phage display technology are probably the
most widely used system to generate binders with novel binding specifity (The M13 phage
display principle is shown on the page 38). The display of heterologous proteins fused to
filamentous phage coat protein was first described in the 1980s (Smith, 1985). This technique
has since been combined with combinatorial protein chemistry to create a powerful means of
selecting peptides and proteins with novel properties (Geysen et al., 1984). Phage display
enables the coupling of the genotype and phenotype allowing the selection of the binding
protein together with the respective gene encoding the protein. Numerous new techniques
have been developed in recent years for peptide and protein display (for recent reviews see
Benhar, 2001; Forrer et al., 1999; Nygren & Uhlen, 1997; Skerra, 2000).
7.1.1 Selection of a protein scaffold
The selection of a suitable scaffold for combinatorial mutagenesis is not trivial and there are
several criteria that set limitations. The fold should be stable enough to tolerate an extensive
randomization of the surface, typically involving 10-20 positions. To overcome the problems
associated with the antibody binders, the scaffold for creating novel binders should be based
on one polypeptide chain and be easily produced in a microbial hosts. The shape and size of
the randomized surface is also important. In this sense, no single protein fold is ideal for all
applications. Use of a folded protein as a starting point for the combinatorial approaches can
provide certain advantages over the use of linear peptides. The smaller loss of motional
freedom for a structured protein compared to that of a flexible peptide during target binding
can give a smaller entropy loss (Ladner, 1995). Also, the display of hydrophobic side chains
required for the hydrophobic interactions with the target might be easier to achieve in a prefixed protein surface than on a peptide where the side chains may become buried upon
display.
We have investigated the use of the non-toxic, ICK-like fold of the CBM1Cel7A from
Trichoderma reesei as a protein scaffold in a combinatorial library to create novel binding
proteins (publication V). The ICK fold is characterized by the virtual absence of a
hydrophobic core and is instead stabilized by disulfide bridges (Figure 7). The family 1
CBMs are very stable and tolerate organic solvents as well as boiling. Sequence alignments of
the family 1 CBMs show that the amino acids forming the planar cellulose-binding surface
38
CARBOHYDRATE-BINDING MODULES
and the cysteins forming the disulfide bridges are well conserved, while the rest of the
sequence is generally highly variable (Figure 4). The ICK-fold seems to be robust and not
very sensitive for sequence variation and gap size between the half cysteins (Norton &
Pallaghy, 1998; Pallaghy et al., 1994). There is also one study where the cellulose-binding
surface of this CBM has been randomized successfully, indicating that the fold is tolerant to
changes if only the positions of the disulphide bridges are maintained (Smith, et al., 1998). In
nature this fold is found in proteins with diverse biological functions. Many of the ICKstructural motif proteins are inhibitors and receptor binders interacting with the target proteins
active site or binding pocket (Norton & Pallaghy, 1998).
When selecting a scaffold for the library construction we considered the properties of the
CBM1 fold and especially the bind ing surface topology. The rationale was to randomize a
surface area around the tip of the wedge-shaped molecule in order to obtain a topology of the
novel binders, which could interact with the binding cavities or active sites of the target
proteins.
XCXXXXYXQCGGX---XYXGXTCCXXXXXCXXXNXYYSQCX
Figure 7. A schematic view of the general pattern of the postulated disulfide bridge formation
of the family 1 CBMs.
7.1.2 Construction and characterization of the CBM1 library
A phage display library of CBM1Cel7A was constructed by randomization of eleven positions
located around the “tip” (Figure 6) (publication V). The estimated library size was 42-46
million variants (publication V). We exploited the phagemid system allowing monovalent
display of the library variant thus minimizing avidity effects (Nord et al., 1995). 50 different
variants of yet unselected clones were analyzed on the DNA level by sequencing (results
indicating over 92% correctly constructed genes) of which five randomly picked variants
were produced and purified from E. coli periplasmic fractions. None of the produced variants
showed interdomain disulfide bridge formation. Subsequently, 32 different variants after the
different pannings were produced and none of the proteins formed dimers, indicating that the
formation of the correct intramolecular disulfide bridges.
39
JANNE LEHTIÖ
7.1.3 Selection of binders from the randomized CBM1 library
To select novel binders, we first used porcine pancreatic α-amylase (PPA) as a target protein,
since the structure was known and the active site is located in a groove- like binding pocket.
The novel binding molecules obtained demonstrated specific binding towards the PPA. No
cross reactivity was detected upon binding to another amylase (barley α-amylase) with a
similar structure (publication V). Two of the PPA-binders were characterized further and
shown to inhibit the enzyme to a certain degree. Furthermore, acarbose, an active site binding
inhibitor, did interfere with the binding of the variants. It thus seems likely that the novel
binders bound in close proximity of the enzyme active site.
We also characterized the two PPA binding variants by creating structural molecular models
of the modules using the wild type as a template. The models indicated no significant main
chain conformational perturbations between the mutants and the wild type. In the variant
CBDPPA1:5 one non-randomised residue A20 has been deleted, but even this deletion did not
seem to have a severe impact on the backbone conformation. The subsequent residues, S21
(now S20) and G22 (now G21), which are loop residues, are positionally shifted to
compensate for the deletion, thereby shortening the loop that they constitute. Also in this
variant the C19-C35 SS-bond is still maintained in the model. However, the C19 C-alpha
atom is shifted towards the deletion by 0.6Å nevertheless not significantly affecting the SSbond lengths
In a different set of experiments, biopanning against metal ions (Ni2+) was performed, using
the library, in order to test the scaffold for metal capture purposes (publication VI). Several
Ni2+-binders were obtained after five panning rounds. Since the wild type CBMCel6A had been
successfully displayed on the surface of S. carnosus (publication IV), we choose to
investigate metal capture in a whole bacterial cell format by genetically engineering the
selected metal binders on the surface of the gram-positive bacteria. All of the variants were
readily displayed on the bacterial surface and two of the variants showed significantly
enhanced Ni2+-binding capacity (publication VI). The selected metal binders featured
numerous histidine residues on the surface indicating that the imidazole side chains
coordinated the Ni2+-ions. The same principle of creating affinity is used in immobilized
metal affinity chromatography. In the eight selected variants, histidine residues appear in all
of the randomized positions suggesting that all of the side chains point towards the solution.
40
CARBOHYDRATE-BINDING MODULES
The side chain orientation is important for effective display of various chemical properties
and for the formation of a stable folded protein. This type of recombinant bacteria, with
enhanced metal binding capacity, has potential use in biofilters or biosensors for
environmental applications.
In the different selections we obtained novel binding proteins exploiting the CBM1 fold.
Variants with novel affinities could be isolated from the library and that all of the produced
variants formed single soluble proteins. The randomized positions form a relatively large
surface area on the 36 amino acid residue CBM1Cel7A and it is remarkable that the fold
tolerates such an extensive randomization. The selection of novel binders towards two widely
different types of epitopes shows that the created library is an exellent source of novel binding
molecules.
8. Concluding remarks
Following years of detailed structure- function studies of differnts CBMs, a detailed picture on
the structral basis of their affinity to polycarbohydrates and their role in enzymatic
carbohydrate degradation is beginnig to emerge. Nevertheless, several open questions remain
to be studied. In this work, the function and applications of the small, stable family 1
cellulose-binding modules were studied. To address the questions on CBM1 specificity on the
crystalline cellulose as well as their function in module interactions, several mutagenesis,
electron microscopy and domain exchange studies were performed. For the fist time, direct
experimental evidence was obtained for the postulated CBM-cellulose interaction on the fully
exposed sugar rings on crystalline cellulose. Further experiments will be needed to verify
whether this is the only interaction or one of the possible modes of binding. Details of the
mechanisms of this and the other types of protein-carbohydrate interaction also require
continued efforts. Especially interesting will be comparisons between the constantly growing
numbers of different CBMs as well as studies addressing the mechanism of the seemingly
irreversible binding of some CBMs.
In resent years, several interesting applications exploiting CBM have been reported. CBMs
have not only been used successfully in the traditional affinity applications but also for
example in transgenetic trees to enhance cellulose production. In this work, the use of a CBM
as an immobilization anchor was studied for two different purposes: protein immobilization
41
JANNE LEHTIÖ
and live bacterial cell immobilization. The CBM1 was also used successfully as a protein
scaffold for protein engineering. The CBM1 fold was proven to be very stable and novel
binders were selected against two very different targets. One of these novel binders was
cloned on the surface of bacteria for metal capture purposes. There is potential to further
develop the recombinant CBM1-bacteria created
for
many
biotechnological
and
environmental applications as well as to use the CBM-library to discover further binding
molecules with a variety of affinities.
42
CARBOHYDRATE-BINDING MODULES
9. Abbreviations
CBM
carbohydrate binding module
CBD
cellulose binding domain (old no menclature)
PASC
phosphoric acid swollen cellulose
BMCC
bacterial microcrystalline cellulose
HEC
hydroxyl methyl cellulose
CMC
carboxyl methyl cellulose
TEM
transmission electron microscopy
CALB
Candida antarctica lipase B
ABP
albumin binding protein
PPA
porcine pancreatic α-amylase
CipA
Clostridium thermocellum scaffolding protein
Cel
cellulase enzyme
43
JANNE LEHTIÖ
10. Acknowledgements
This thesis is the result of many late lonely hours in the lab, as well as joyful teamwork and
collaborations. I would like to sincerely thank all of you who I have worked with during these
years.
Most of all I would like to thank Professor Tuula T. Teeri for enthusiastic supervising of this
thesis and for your never-ending optimism, always believing that a great scientific discovery
is just around the corner. I had no idea what the future would bring when I, as an undergraduate student, called to the VTT telephone exchange.
I would like to thank Professor Per- Åke Nygren for your supervision of the phage work and
your ever so philosophic approach to biochemistry. Professor Stefan Ståhl, thanks for your
great, uplifting optimism and positive attitude. Without you I never would have expressed
CBD on staff “for the first time ever”. I also thank Professor Mathias Uhlén for running the
great DNAconer with great style.
I will also like to thank Professor Junji Sugiyama for letting me peek into the electron
microscope. It turned out to be: YATAA!
I have fantastic friends! Thank you Linda, Jussi, Christin, Henrik, Malin and Mark for you
valuable help during these last crazy weeks. Without you I never would have managed to
print this work in time.
Thanks to all of my colleagues at the Department of Biotechnology for your welcoming
attitude. You really made me feel like home from the first day when I arrived from Finland.
I am especially grateful to:
Henrik W for the successful staff- stuff, but most of all for being a good friend. One of these
days you might even beat me on a tennis court. J Malin G, you are always as enthusiastic
and helpful, no matter if it is work or sushi- lunch. Patrik for introducing me to the world of
gram-positive bacteria as well as really making me feel welcome to DNAcorner. Linda F –not
only for your wonderful, sacrifising help with the thesis, but also for the concept-days. Malin
E for great fun both in the lab and outside, maybe we’ll get to have a real business lunch one
of these days. Jenny O for ”skrivarstugan”, lets pretend we still have to do it in future.
I would also like to thank Aspe, Åsa, Ulla, Soraya, Tarja N., Markus L., Elin, Marianne,
Jenny R., Eva, Peter, Savo, Jakob, Karin, Maria W., C. och E, Nina, Anna G. och B, Martin
L, Martin B, Alex, Stuart, Sofia, Fiona, Fredrik, Joke, Kristoffer, Ulla, Urban, Tornbjörn,
Kristina, Joakim W and L and all the rest of you whom I‘ve worked with during these five
years at VTT, DNAconer and at ex-XVe.
44
CARBOHYDRATE-BINDING MODULES
I want to say my thanks to all of the good friend who helped me to stay in touch with the
outside world. Jussi for being my best friend for over 25 years. There just isn’t a thing in this
world that I do which you would not understand. Christin, five years ago I would have never
guessed how much laughter and fun there was in store for me when I met you. You are my
favorite swede! Katarina for without you the Troikka would not be complete. Thanks also
Uwe, Jouko, Timo, Pelle, Pekka, Mark, Al, Jaakko and all the others for the many superb
moments.
I would also like to thank my grand parents for keeping my nutrition well balanced during my
study years with great dinners!
Jyri and Mona and especially Eemeli and Oskari for the lively moments. I hope I can spend
more time with you now from now on.
Finally I would like to thank my parents for your full support. You have provided me with a
shelter and freedom at same time and most of all a steady platform from where I could jump
as high as I possibly can.
45
JANNE LEHTIÖ
11. References
Abou Hachem, M., Nordberg Karlsson, E., Bartonek-Roxa, E., Raghothama, S., Simpson, P. J., Gilbert, H. J.,
Williamson, M. P. & Holst, O. (2000). Carbohydrate-binding modules from a thermostable Rhodothermus
marinus xylanase: cloning, expression and binding studies. Biochem J 345 Pt 1, 53-60.
Appleyard, A. N., Herbert, R. B., Henderson, P. J., Watts, A. & Spooner, P. J. (2000). Selective NMR
observation of inhibitor and sugar binding to the galactose-H(+) symport protein GalP, of Escherichia coli.
Biochim Biophys Acta 1509, 55-64.
Argos, P. (1990). An investigation of oligopeptides linking domains in protein tertiary structures and possible
candidates for general gene fusion. J Mol Biol 211, 943-58.
Asensio, J. L., Canada, F. J., Bruix, M., Rodriguez-Romero, A. & Jimenez-Barbero, J. (1995). The interaction of
hevein with N-acetylglucosamine-containing oligosaccharides. Solution structure of hevein complexed to
chitobiose. Eur J Biochem 230, 621-33.
Asensio, J. L., Canada, F. J., Siebert, H. C., Laynez, J., Poveda, A., Nieto, P. M., Soedjanaamadja, U. M.,
Gabius, H. J. & Jimenez-Barbero, J. (2000a). Structural basis for chitin recognition by defense proteins: GlcNAc
residues are bound in a multivalent fashion by extended binding sites in hevein domains. Chem Biol 7, 529-43.
Asensio, J. L., Siebert, H. C., von Der Lieth, C. W., Laynez, J., Bruix, M., Soedjanaamadja, U. M., Beintema, J.
J., Canada, F. J., Gabius, H. J. & Jimenez-Barbero, J. (2000b). NMR investigations of protein-carbohydrate
interactions: studies on the relevance of Trp/Tyr variations in lectin binding sites as deduced from titration
microcalorimetry and NMR studies on hevein domains. Determination of the NMR structure of the complex
between pseudohevein and N,N',N"-triacetylchitotriose. Proteins 40, 218-36.
Bayer, E. A., Morag, E. & Lamed, R. (1994). The cellulosome--a treasure-trove for biotechnology. Trends
Biotechnol 12, 379-86.
Beguin, P. & Aubert, J. P. (1994). The biological degradation of cellulose. FEMS Microbiol Rev 13, 25-58.
Benhar, I. (2001). Biotechnological applications of phage and cell display. Biotechnol. Adv. 19, 1-33.
Black, G. W., Rixon, J. E., Clarke, J. H., Hazlewood, G. P., Ferreira, L. M., Bolam, D. N. & Gilbert, H. J.
(1997). Cellulose binding domains and linker sequences potentiate the activity of hemicellulases against
complex substrates. J Biotechnol 57, 59-69.
Black, G. W., Rixon, J. E., Clarke, J. H., Hazlewood, G. P., Theodorou, M. K., Morris, P. & Gilbert, H. J.
(1996). Evidence that linker sequences and cellulose-binding domains enhance the activity of hemicellulases
against complex substrates. Biochem J 319, 515-20.
Boisset, C., Fraschini, C., Schulein, M., Henrissat, B. & Chanzy, H. (2000). Imaging the enzymatic digestion of
bacterial cellulose ribbons reveals the endo character of the cellobiohydrolase Cel6A from Humicola insolens
and its mode of synergy with cellobiohydrolase Cel7A. Appl Environ Microbiol 66, 1444-52.
Bolam, D. N., Ciruela, A., McQueen-Mason, S., Simpson, P., Williamson, M. P., Rixon, J. E., Boraston, A.,
Hazlewood, G. P. & Gilbert, H. J. (1998). Pseudomonas cellulose-binding domains mediate their effects by
increasing enzyme substrate proximity. Biochem J 331, 775-81.
Bolam, D. N., Xie, H., White, P., Simpson, P. J., Hancock, S. M., Williamson, M. P. & Gilbert, H. J. (2001).
Evidence for synergy between family 2b carbohydrate binding modules in Cellulomonas fimi xylanase 11A.
Biochemistry 40, 2468-77.
Boraston, A. B., Chiu, P., Warren, R. A. & Kilburn, D. G. (2000). Specificity and affinity of substrate binding by
a family 17 carbohydrate-binding module from Clostridium cellulovorans cellulase 5A. Biochemistry 39, 1112936.
46
CARBOHYDRATE-BINDING MODULES
Boraston, A. B., Creagh, A. L., Alam, M. M., Kormos, J. M., Tomme, P., Haynes, C. A., Warren, R. A. &
Kilburn, D. G. (2001). Binding specificity and thermodynamics of a family 9 carbohydrate-binding module from
Thermotoga maritima xylanase 10A. Biochemistry 40, 6240-7.
Bray, M. R., Johnson, P. E., Gilkes, N. R., McIntosh, L. P., Kilburn, D. G. & Warren, R. A. (1996). Probing the
role of tryptophan residues in a cellulose-binding domain by chemical modification. Protein Sci 5, 2311-8.
Brooks, M. M., Tuohy, M. G., Savage, A. V., Claeyssens, M. & Coughlan, M. P. (1992). The stereochemical
course of reactions catalysed by the cellobiohydrolases produced by Talaromyces emersonii. Biochem J 283, 314.
Brun, E., Johnson, P. E., Creagh, A. L., Tomme, P., Webster, P., Haynes, C. A. & McIntosh, L. P. (2000).
Structure and binding specificity of the second N-terminal cellulose-binding domain from Cellulomonas fimi
endoglucanase C. Biochemistry 39, 2445-58.
Brun, E., Moriaud, F., Gans, P., Blackledge, M. J., Barras, F. & Marion, D. (1997). Solution structure of the
cellulose-binding domain of the endoglucanase Z secreted by Erwinia chrysanthemi. Biochemistry 36, 16074-86.
Carrard, G., Koivula, A., Söderlund, H. & Beguin, P. (2000). Cellulose-binding domains promote hydrolysis of
different sites on crystalline cellulose. Proc Natl Acad Sci U S A 97, 10342-7.
Carrard, G. & Linder, M. (1999). Widely different off rates of two closely related cellulose-binding domains
from Trichoderma reesei. Eur J Biochem 262, 637-43.
Charnock, S. J., Bolam, D. N., Turkenburg, J. P., Gilbert, H. J., Ferreira, L. M., Davies, G. J. & Fontes, C. M.
(2000). The X6 thermostabilizing domains of xylanases are carbohydrate-binding modules: structure and
biochemistry of the Clostridium thermocellum X6b domain. Biochemistry 39, 5013-21.
Chaudhuri, B. N., Ko, J., Park, C., Jones, T. A. & Mowbray, S. L. (1999). Structure of D-allose binding protein
from Escherichia coli bound to D-allose at 1.8 Å resolution. J Mol Biol 286, 1519-31.
Coutinho, J. B., Gilkes, N. R., Warren, R. A., Kilburn, D. G. & Miller, R. C., Jr. (1992). The binding of
Cellulomonas fimi endoglucanase C (CenC) to cellulose and Sephadex is mediated by the N-terminal repeats.
Mol Microbiol 6, 1243-52.
Coutinho, P. M. & Henrissat, B. (1999). Carbohydrate-Active Enzymes server at URL: http://afmb.cnrsmrs.fr/~pedro/CAZY/db.html. .
Creagh, A. L., Ong, E., Jervis, E., Kilburn, D. G. & Haynes, C. A. (1996). Binding of the cellulose-binding
domain of exoglucanase Cex from Cellulomonas fimi to insoluble microcrystalline cellulose is entropically
driven. Proc Natl Acad Sci U S A 93, 12229-34.
Dam, T. K., Cavada, B. S., Grangeiro, T. B., Santos, C. F., Ceccatto, V. M., de Sousa, F. A., Oscarson, S. &
Brewer, C. F. (2000a). Thermodynamic binding studies of lectins from the diocleinae subtribe to deoxy analogs
of the core trimannoside of asparagine-linked oligosaccharides. J Biol Chem 275, 16119-26.
Dam, T. K., Roy, R., Das, S. K., Oscarson, S. & Brewer, C. F. (2000b). Binding of multivalent carbohydrates to
concanavalin A and Dioclea grandiflora lectin. Thermodynamic analysis of the multivalency effect. J Biol Chem
275, 14223-30.
Daugherty, P. S., Olsen, M. J., Iverson, B. L. & Georgiou, G. (1999). Development of an optimized expression
system for the screening of antibody libraries displayed on the Escherichia coli surface. Protein Eng 12, 613-21.
Davies, G. & Henrissat, B. (1995). Structures and mechanisms of glycosyl hydrolases. Structure 3, 853-9.
Din, N., Damude, H. G., Gilkes, N. R., Miller, R. C., Jr., Warren, R. A. & Kilburn, D. G. (1994a). C1-Cx
revisited: intramolecular synergism in a cellulase. Proc Natl Acad Sci U S A 91, 11383-7.
47
JANNE LEHTIÖ
Din, N., Forsythe, I. J., Burtnick, L. D., Gilkes, N. R., Miller, R. C., Jr., Warren, R. A. & Kilburn, D. G. (1994b).
The cellulose-binding domain of endoglucanase A (CenA) from Cellulomonas fimi: evidence for the
involvement of tryptophan residues in binding. Mol Microbiol 11, 747-55.
Divne, C., Sinning, I., Ståhlberg, J., Pettersson, G., Bailey, M., Siika -aho, M., Margolles-Clark, E., Teeri, T.T. &
Jones, T. A. (1993). Crystallization and preliminary X-ray studies on the core proteins of cellobiohydrolase I and
endoglucanase I from Trichoderma reesei. J Mol Biol 234, 905-7.
Divne, C., Ståhlberg, J., Reinikainen, T., Ruohonen, L., Pettersson, G., Knowles, J. K., Teeri, T. T. & Jones, T.
A. (1994). The three-dimensional crystal structure of the catalytic core of cellobiohydrolase I from Trichoderma
reesei. Science 265, 524-8.
Divne, C., Ståhlberg, J., Teeri, T. T. & Jones, T. A. (1998). High-resolution crystal structures reveal how a
cellulose chain is bound in the 50 A long tunnel of cellobiohydrolase I from Trichoderma reesei. J Mol Biol 275,
309-25.
Doig, A. J. & Williams, D. H. (1992). Why water-soluble, compact, globular proteins have similar specific
enthalpies of unfolding at 110 degrees C. Biochemistry 31, 9371-5.
Drickamer, K. (1989). Demonstration of carbohydrate-recognition activity in diverse proteins which share a
common primary structure motif. Biochem Soc Trans 17, 13-5.
Feinberg, H., Park-Snyder, S., Kolatkar, A. R., Heise, C. T., Taylor, M. E. & Weis, W. I. (2000). Structure of a
C-type carbohydrate recognition domain from the macrophage mannose receptor. J Biol Chem 275, 21539-48.
Ferreira, L. M., Durrant, A. J., Hall, J., Hazlewood, G. P. & Gilbert, H. J. (1990). Spatial separation of protein
domains is not necessary for catalytic activity or substrate binding in a xylanase. Biochem J 269, 261-4.
Flinn, J. P., Pallaghy, P. K., Lew, M. J., Murphy, R., Angus, J. A. & Norton, R. S. (1999). Role of disulfide
bridges in the folding, structure and biological activity of omega-conotoxin GVIA. Biochim Biophys Acta 1434,
177-90.
Forrer, P., Jung, S. & Pluckthun, A. (1999). Beyond binding: using phage display to select for structure, folding
and enzymatic activity in proteins. Curr Opin Struct Biol 9, 514-20.
Francisco, J. A., Stathopoulos, C., Warren, R. A., Kilburn, D. G. & Georgiou, G. (1993). Specific adhesion and
hydrolysis of cellulose by intact Escherichia coli expressing surface anchored cellulase or cellulose-binding
domains. Biotechnology (N Y) 11, 491-5.
Fuchs, P., Breitling, F., Dubel, S., Seehaus, T. & Little, M. (1991). Targeting recombinant antibodies to the
surface of Escherichia coli: fusion to a peptidoglycan associated lipoprotein. Biotechnology (N Y) 9, 1369-72.
Georgiou, G., Stathopoulos, C., Daugherty, P. S., Nayak, A. R., Iverson, B. L. & Curtiss, R., 3rd (1997). Display
of heterologous proteins on the surface of microorganisms: from the screening of combinatorial libraries to live
recombinant vaccines. Nat Biotechnol 15, 29-34.
Geysen, H. M., Meloen, R. H. & Barteling, S. J. (1984). Use of peptide synthesis to probe viral antigens for
epitopes to a resolution of a single amino acid. Proc Natl Acad Sci U S A 81, 3998-4002.
Gilkes, N. R., Claeyssens, M., Aebersold, R., Henrissat, B., Meinke, A., Morrison, H. D., Kilburn, D. G.,
Warren, R. A. & Miller, R. C., Jr. (1991a). Structural and functional relationships in two families of beta-1,4glycanases. Eur J Biochem 202, 367-77.
Gilkes, N. R., Henrissat, B., Kilburn, D. G., Miller, R. C., Jr. & Warren, R. A. (1991b). Domains in microbial
beta-1, 4-glycanases: sequence conservation, function, and enzyme families. Microbiol Rev 55, 303-15.
Gilkes, N. R., Kilburn, D. G., Miller, R. C., Jr., Warren, R. A., Sugiyama, J., Chanzy, H. & Henrissat, B. (1993).
Visualization of the adsorption of a bacterial endo-beta-1,4-glucanase and its isolated cellulose-binding domain
to crystalline cellulose. Int J Biol Macromol 15, 347-51.
48
CARBOHYDRATE-BINDING MODULES
Gilkes, N. R., Warren, R. A., Miller, R. C., Jr. & Kilburn, D. G. (1988). Precise excision of the cellulose binding
domains from two Cellulomonas fimi cellulases by a homologous protease and the effect on catalysis. J Biol
Chem 263, 10401-7.
Gunneriusson, E., Samuelson, P., Ringdahl, J., Grönlund, H., Nygren, P. Å. & Ståhl, S. (1999). Staphylococcal
surface display of immunoglobulin A (IgA)- and IgE-specific in vitro-selected binding proteins (affibodies)
based on Staphylococcus aureus protein A. Appl Environ Microbiol 65, 4134-40.
Gunneriusson, E., Samuelson, P., Uhlén, M., Nygren, P. Å. & Ståhl, S. (1996). Surface display of a functional
single-chain Fv antibody on staphylococci. J Bacteriol 178, 1341-6.
Hammes, W., Bosch, L. & Wolf, G. (1995). Contribution of Staphylococcus carnosus and Staphylococcus
piscifermentans to the fermentation of protein foods. J. Appl. Bacteriol. Symp. Suppl. , 765-835.
Hansson, M., Samuelson, P., Gunneriusson, E. & Ståhl, S. (2001). Surface display on gram-positive bacteria.
Comb Chem High Throughput Screen 4, 171-84.
Helbert, W., Nishiyama, Y., Okano, T. & Sugiyama, J. (1998). Molecular imaging of halocynthia papillosa
cellulose. J Struct Biol 124, 42-50.
Henrissat, B. (1999). Classification of chitinases modules. Exs 87, 137-56.
Henrissat, B. & Davies, G. (1997). Structural and sequence-based classification of glycoside hydrolases. Curr
Opin Struct Biol 7, 637-44.
Hoffren, A. M., Teeri, T. T. & Teleman, O. (1995). Molecular dynamics simulation of fungal cellulose-binding
domains: differences in molecular rigidity but a preserved cellulose-binding surface. Protein Eng 8, 443-50.
Hon, D. (1994). Cellulose 1, 1-25.
Imai, T., Boisset, C., Samejima, M., Igarashi, K. & Sugiyama, J. (1998). Unidirectional processive action of
cellobiohydrolase Cel7A on Valonia cellulose microcrystals. FEBS Lett 432, 113-6.
Irwin, D., Shin, D. H., Zhang, S., Barr, B. K., Sakon, J., Karplus, P. A. & Wilson, D. B. (1998). Roles of the
catalytic domain and two cellulose binding domains of Thermomonospora fusca E4 in cellulose hydrolysis. J
Bacteriol 180, 1709-14.
Jervis, E. J., Haynes, C. A. & Kilburn, D. G. (1997). Surface diffusion of cellulases and their isolated binding
domains on cellulose. J Biol Chem 272, 24016-23.
Johnson, P. E., Joshi, M. D., Tomme, P., Kilburn, D. G. & McIntosh, L. P. (1996a). Structure of the N-terminal
cellulose-binding domain of Cellulomonas fimi CenC determined by nuclear magnetic resonance spectroscopy.
Biochemistry 35, 14381-94.
Johnson, P. E., Tomme, P., Joshi, M. D. & McIntosh, L. P. (1996b). Interaction of soluble cellooligosaccharides
with the N-terminal cellulose-binding domain of Cellulomonas fimi CenC 2. NMR and ultraviolet absorption
spectroscopy. Biochemistry 35, 13895-906.
Knowles, J., Teeri, T. T., Lehtovaara, P., Penttilä, M. & Saloheimo, M. (1988). In Biochemistry and Genetics of
Cellulose Degradation, pp. 153-169. Edited by J. P. Aubert, P. Beguin & J. Milet. London: Academic Press.
Kormo s, J., Johnson, P. E., Brun, E., Tomme, P., McIntosh, L. P., Haynes, C. A. & Kilburn, D. G. (2000).
Binding site analysis of cellulose-binding domain CBD(N1) from endoglucanse C of Cellulomonas fimi by sitedirected mutagenesis. Biochemistry 39, 8844-52.
Koyama, M., Helbert, W., Imai, T., Sugiyama, J. & Henrissat, B. (1997). Parallel-up structure evidences the
molecular directionality during biosynthesis of bacterial cellulose. Proc Natl Acad Sci U S A 94, 9091-5.
49
JANNE LEHTIÖ
Kraulis, J., Clore, G. M., Nilges, M., Jones, T. A., Pettersson, G., Knowles, J. & Gronenborn, A. M. (1989).
Determination of the three-dimensional solution structure of the C-terminal domain of cellobiohydrolase I from
Trichoderma reesei. A study using nuclear magnetic resonance and hybrid distance geometry-dynamical
simulated annealing. Biochemistry 28, 7241-57.
Ladner, R. C. (1995). Constrained peptides as binding entities. Trends Biotechnol 13, 426-30.
Liljeqvist, S., Cano, F., Nguyen, T. N., Uhlén, M., Robert, A. & Ståhl, S. (1999). Surface display of functional
fibronectin-binding domains on Staphylococcus carnosus. FEBS Lett 446, 299-304.
Linder, M., Lindeberg, G., Reinikainen, T., Teeri, T. T. & Pettersson, G. (1995a). The difference in affinity
between two fungal cellulose-binding domains is dominated by a single amino acid substitution. FEBS Lett 372,
96-8.
Linder, M., Mattinen, M. L., Kontteli, M., Lindeberg, G., Ståhlberg, J., Drakenberg, T., Reinikainen, T.,
Pettersson, G. & Annila, A. (1995b). Identification of functionally important amino acids in the cellulosebinding domain of Trichoderma reesei cellobiohydrolase I. Protein Sci 4, 1056-64.
Linder, M., Nevanen, T., Söderholm, L., Bengs, O. & Teeri, T. T. (1998). Improved immobilization of fusion
proteins via cellulose-binding domains. Biotechnol Bioeng 60, 642-7.
Linder, M., Nevanen, T. & Teeri, T. T. (1999). Design of a pH-dependent cellulose-binding domain. FEBS Lett
447, 13-6.
Linder, M., Salovuori, I., Ruohonen, L. & Teeri, T. T. (1996). Characterization of a double cellulose-binding
domain. Synergistic high affinity binding to crystalline cellulose. J Biol Chem 271, 21268-72.
Linder, M. & Teeri, T. T. (1996). The cellulose-binding domain of the major cellobiohydrolase of Trichoderma
reesei exhibits true reversibility and a high exchange rate on crystalline cellulose. Proc Natl Acad Sci U S A 93,
12251-5.
Linder, M. & Teeri, T. T. (1997). The roles and function of cellulose-binding domains. J. Biotechnol. 57, 15-28.
Liu, Q. Y., Ross, N., Lanthier, P. & Reith, M. (1996). A gametophyte cell wall protein of the red alga Porphyra
purpurea (Rhodophyta) contains four apparent polysaccharide-binding domains. J Phycol 32, 995-1003.
Lu, S., Liang, S. & Gu, X. (1999). Three-dimensional structure of Selenocosmia huwena lectin-I (SHL-I) from
the venom of the spider Selenocosmia huwena by 2D-NMR. J Protein Chem 18, 609-17.
Lu, Z., Murray, K. S., Van Cleave, V., LaVallie, E. R., Ståhl, M. L. & McCoy, J. M. (1995). Expression of
thioredoxin random peptide libraries on the Escherichia coli cell surface as functional fusions to flagellin: a
system designed for exploring protein-protein interactions. Biotechnology (N Y) 13, 366-72.
MacGregor, E. A., Janecek, S. & Svensson, B. (2001). Relationship of sequence and structure to specificity in
the alpha-amylase family of enzymes. Biochim Biophys Acta 1546, 1-20.
Martins, J. C., Maes, D., Loris, R., Pepermans, H. A., Wyns, L., Willem, R. & Verheyden, P. (1996). H NMR
study of the solution structure of Ac-AMP2, a sugar-binding antimicrobial protein isolated from Amaranthus
caudatus. J Mol Biol 258, 322-33.
Matsushita, O., Koide, T., Kobayashi, R., Nagata, K. & Okabe, A. (2001). Substrate recognition by the collagenbinding domain of Clostridium histolyticum class I collagenase. J Biol Chem 276, 8761-70.
Mattinen, M. L., Kontteli, M., Kerovuo, J., Linder, M., Annila, A., Lindeberg, G., Reinikainen, T. &
Drakenberg, T. (1997). Three-dimensional structures of three engineered cellulose-binding domains of
cellobiohydrolase I from Trichoderma reesei. Protein Sci 6, 294-303.
Mattinen, M. L., Linder, M., Drakenberg, T. & Annila, A. (1998). Solution structure of the cellulose-binding
domain of endoglucanase I from Trichoderma reesei and its interaction with cello-oligosaccharides. Eur J
Biochem 256, 279-86.
50
CARBOHYDRATE-BINDING MODULES
Mattinen, M. L., Linder, M., Teleman, A. & Annila, A. (1997). Interaction between cellohexaose and cellulose
binding domains from Trichoderma reesei cellulases. FEBS Lett 407, 291-6.
McKie, V. A., Vincken, J. P., Voragen, A. G., van den Broek, L. A., Stimson, E. & Gilbert, H. J. (2001). A new
family of rhamnogalacturonan lyases contains an enzyme that binds to cellulose. Biochem J 355, 167-77.
McLean, B. W., Bray, M. R., Boraston, A. B., Gilkes, N. R., Haynes, C. A. & Kilburn, D. G. (2000). Analysis of
binding of the family 2a carbohydrate-binding module from Cellulomonas fimi xylanase 10A to cellulose:
specificity and identification of functionally important amino acid residues. Protein Eng 13, 801-9.
Meinke, A., Gilkes, N. R., Kilburn, D. G., Mille r, R. C., Jr. & Warren, R. A. (1991). Multiple domains in
endoglucanase B (CenB) from Cellulomonas fimi: functions and relatedness to domains in other polypeptides. J
Bacteriol 173, 7126-35.
Miller, R., Gilkes, N., Greeberg, N., Kilburn, D., Langsford, M. & Warren, R. A. (1988). In Biochemistry and
Genetics of Cellulose Degradation, pp. 235-248. Edited by J. P. Aubert, P. Beguin & J. Milet. London: Acdemic
Press.
Nagy, T., Simpson, P., Williamson, M. P., Hazlewood, G. P., Gilbert, H. J. & Orosz, L. (1998). All three surface
tryptophans in Type IIa cellulose binding domains play a pivotal role in binding both soluble and insoluble
ligands. FEBS Lett 429, 312-6.
Nilsson, B., Moks, T., Jansson, B., Abrahmsen, L., Elmblad, A., Holmgren, E., Henrichson, C., Jones, T. A. &
Uhlén, M. (1987). A synthetic IgG-binding domain based on staphylococcal protein A. Protein Eng 1, 107-13.
Nord, K., Nilsson, J., Nilsson, B., Uhlén, M. & Nygren, P. Å. (1995). A combinatorial library of an alpha-helical
bacterial receptor domain. Protein Eng 8, 601-8.
Norton, R. S. & Pallaghy, P. K. (1998). The cystine knot structure of ion channel toxins and related
polypeptides. Toxicon 36, 1573-83.
Notenboom, V., Boraston, A. B., Kilburn, D. G. & Rose, D. R. (2001). Crystal structures of the family 9
carbohydrate-binding module from Thermotoga maritima xylanase 10A in native and ligand-bound forms.
Biochemistry 40, 6248-56.
Nygren, P. Å. & Uhlén, M. (1997). Scaffolds for engineering novel binding sites in proteins. Curr Opin Struct
Biol 7, 463-9.
Ong, E., Gilkes, N. R., Miller, R. C., Jr., Warren, A. J. & Kilburn, D. G. (1991). Enzyme immobilization using a
cellulose-binding domain: properties of a beta-glucosidase fusion protein. Enzyme Microb Technol 13, 59-65.
Overall, C. M. (2001). Matrix metalloproteinase substrate-binding domains, modules and exosites. Overview and
experimental strategies. Methods Mol Biol 151, 79-120.
Pallaghy, P. K., Nielsen, K. J., Craik, D. J. & Norton, R. S. (1994). A common structural motif incorporating a
cystine knot and a triple -stranded beta-sheet in toxic and inhibitory polypeptides. Protein Sci 3, 1833-9.
Palonen, H., Tenkanen, M. & Linder, M. (1999). Dynamic interaction of Trichoderma reesei cellobiohydrolases
Cel6A and Cel7A and cellulose at equilibrium and during hydrolysis. Appl Environ Microbiol 65, 5229-33.
Ponyi, T., Szabo, L., Nagy, T., Orosz, L., Simpson, P. J., Williamson, M. P. & Gilbert, H. J. (2000). Trp22,
Trp24, and Tyr8 play a pivotal role in the binding of the family 10 cellulose-binding module from Pseudomonas
xylanase A to insoluble ligands. Biochemistry 39, 985-91.
Quiocho, F. A. (1986). Carbohydrate-binding proteins: tertiary structures and protein-sugar interactions. Annu
Rev Biochem 55, 287-315.
51
JANNE LEHTIÖ
Quiocho, F. A., Spurlino, J. C. & Rodseth, L. E. (1997). Extensive features of tight oligosaccharide binding
revealed in high-resolution structures of the maltodextrin transport/chemosensory receptor. Structure 5, 9971015.
Raghothama, S., Simpson, P. J., Szabo, L., Nagy, T., Gilbert, H. J. & Williamson, M. P. (2000). Solution
structure of the CBM10 cellulose-binding module from Pseudomonas xylanase A. Biochemistry 39, 978-84.
Ramkumar, R., Surolia, A. & Podder, S. K. (1995). Energetics of carbohydrate binding by a 14 kDa S-type
mammalian lectin. Biochem J 308, 237-41.
Reinikainen, T., Ruohonen, L., Nevanen, T., Laaksonen, L., Kraulis, P., Jones, T. A., Knowles, J. K. & Teeri, T.
T. (1992). Investigation of the function of mutated cellulose-binding domains of Trichoderma reesei
cellobiohydrolase I. Proteins 14, 475-82.
Reinikainen, T., Teleman, O. & Teeri, T. T. (1995). Effects of pH and high ionic strength on the adsorption and
activity of native and mutated cellobiohydrolase I from Trichoderma reesei. Proteins 22, 392-403.
Rini, J. M. (1995). Lectin structure. Annu Rev Biophys Biomol Struct 24, 551-77.
Robert, A., Samuelson, P., Andreoni, C., Bachi, T., Uhlén, M., Binz, H., Nguyen, T. N. & Ståhl, S. (1996).
Surface display on staphylococci: a comparative study. FEBS Lett 390, 327-33.
Rotticci-Mulder, J. C., Gustavsson, M., Holmquist, M., Hult, K. & Martinelle, M. (2001). Expression in Pichia
pastoris of Candida antarctica lipase B and lipase B fused to a cellulose-binding domain. Protein Expr Purif 21,
386-92.
Sacchettini, J. C., Baum, L. G. & Brewer, C. F. (2001). Multivalent protein-carbohydrate interactions. A new
paradigm for supermolecular assembly and signal transduction. Biochemistry 40, 3009-15.
Sakon, J., Irwin, D., Wilson, D. B. & Karplus, P. A. (1997). Structure and mechanism of endo/exocellulase E4
from Thermomonospora fusca. Nat Struct Biol 4, 810-8.
Samuelson, P., Hansson, M., Ahlborg, N., Andreoni, C., Gotz, F., Bachi, T., Nguyen, T. N., Binz, H., Uhlén, M.
& Ståhl, S. (1995). Cell surface display of recombinant proteins on Staphylococcus carnosus. J Bacteriol 177,
1470-6.
Samuelson, P., Wernerus, H., Svedberg, M. & Ståhl, S. (2000). Staphylococcal Surface Display of MetalBinding Polyhistidyl Peptides. Appl Environ Microbiol 66, 1243-1248.
Shoham, Y., Lamed, R. & Bayer, E. A. (1999). The cellulosome concept as an efficient microbial strategy for the
degradation of insoluble polysaccharides. Trends Microbiol 7, 275-81.
Simpson, P. J., Bolam, D. N., Cooper, A., Ciruela, A., Hazlewood, G. P., Gilbert, H. J. & Williamson, M. P.
(1999). A family IIb xylan-binding domain has a similar secondary structure to a homologous family IIa
cellulose-binding domain but different ligand specificity. Structure Fold Des 7, 853-64.
Simpson, P. J., Xie, H., Bolam, D. N., Gilbert, H. J. & Williamson, M. P. (2000). The structural basis for the
ligand specificity of family 2 carbohydrate-binding modules. J Biol Chem 275, 41137-42.
Skerra, A. (2000). Engineered protein scaffolds for molecular recognition. J Mol Recognit 13, 167-87.
Skerra, A. (2000). Lipocalins as a scaffold. Biochim Biophys Acta 1482, 337-50.
Smith, G. P. (1985). Filamentous fusion phage: novel expression vectors that display cloned antigens on the
virion surface. Science 228, 1315-7.
Smith, G. P., Patel, S. U., Windass, J. D., Thornton, J. M., Winter, G. & Griffiths, A. D. (1998). Small binding
proteins selected from a combinatorial repertoire of knottins displayed on phage. J Mol Biol 277, 317-32.
52
CARBOHYDRATE-BINDING MODULES
Sousa, C., Cebolla, A. & Lorenzo, V. (1996). Enhanced metalloadsorption of bacterial cells displaying poly-His
peptides. Nat. biotechnol. 14, 1117-1120.
Spurlino, J. C., Rodseth, L. E. & Quiocho, F. A. (1992). Atomic interactions in protein-carbohydrate complexes.
Tryptophan residues in the periplasmic maltodextrin receptor for active transport and chemotaxis. J Mol Biol
226, 15-22.
Srisodsuk, M., Kleman-Leyer, K., Keranen, S., Kirk, T. K. & Teeri, T. T. (1998). Modes of action on cotton and
bacterial cellulose of a homologous endoglucanase-exoglucanase pair from Trichoderma reesei. Eur J Biochem
251, 885-92.
Srisodsuk, M., Lehtiö, J., Linder, M., Margolles-Clark, E., Reinikainen, T. & Teeri, T. T. (1997). Trichoderma
reesei cellobiohydrolase I with an endoglucanase cellulose-binding domain: action on bacterial microcrystalline
cellulose. J Biotechnol 57, 49-57.
Srisodsuk, M., Reinikainen, T., Penttilä, M. & Teeri, T. T. (1993). Role of the interdomain linker peptide of
Trichoderma reesei cellobiohydrolase I in its interaction with crystalline cellulose. J Biol Chem 268, 20756-61.
Ståhl, S., Robert, A., Gunneriusson, E., Wernerus, H., Cano, F., Liljeqvist, S., Hansson, M., Nguyen, T. N. &
Samuelson, P. (2000). Staphylococcal surface display and its applications. Int J Med Microbiol 290, 571-7.
Sugiyama, J., Harada, H., Fujiyoshi, Y. & Uyeda, N. (1985). Observation of cellulose microfibrils in Valonia
macrophysa by high resolution electoron microscopy. Mokuzai Gakkaishi 31, 61-67.
Teeri, T. T., Koivula, A., Linder, M., Wohlfahrt, G., Divne, C. & Jones, T. A. (1998). Trichoderma reesei
cellobiohydrolases: why so efficient on crystalline cellulose? Biochem Soc Trans 26, 173-8.
Teeri, T. T., Lehtovaara, P., Kauppinen, S., Salovuori, I. & Knowles, J. (1987). Homologous domains in
Trichoderma reesei cellulolytic enzymes: gene sequence and expression of cellobiohydrolase II. Gene 51, 43-52.
Tomme, P., Boraston, A., McLean, B., Kormos, J., Creagh, A. L., Sturch, K., Gilkes, N. R., Haynes, C. A.,
Warren, R. A. & Kilburn, D. G. (1998). Characterization and affinity applications of cellulose-binding domains.
J Chromatogr B Biomed Sci Appl 715, 283-96.
Tomme, P., Creagh, A. L., Kilburn, D. G. & Haynes, C. A. (1996a). Interaction of polysaccharides with the Nterminal cellulose-binding domain of Cellulomonas fimi CenC. 1. Binding specificity and calorimetric analysis.
Biochemistry 35, 13885-94.
Tomme, P., Gilkes, N. R., Guarna, M. M., Haynes, C. A., Hasenwinkle, D., Jervis, E., Johnson, P., McIntosh, L.,
Warren, R. A. & Kilburn, D. G. (1996b). Cellulose-binding domains. Versatile affinity tags for inexpensive
large-scale purification, concentration, and immobilization of fusion proteins. Ann N Y Acad Sci 799, 418-24.
Tomme, P., Van Tilbeurgh, H., Pettersson, G., Van Damme, J., Vandekerckhove, J., Knowles, J., Teeri, T. &
Claeyssens, M. (1988). Studies of the cellulolytic system of Trichoderma reesei QM 9414. Analysis of domain
function in two cellobiohydrolases by limited proteolysis. Eur J Biochem 170, 575-81.
Tomme, P., Warren, R. A. J., Miller, R. C., Kilburn, D. G. & Gilkes, N. R. (1995a). Cellulose-binding domains:
Classification and properties. In ACS Symp. Ser., pp. 143-163.
Tormo, J., Lamed, R., Chirino, A. J., Morag, E., Bayer, E. A., Shoham, Y. & Steitz, T. A. (1996). Crystal
structure of a bacterial family-III cellulose-binding domain: a general mechanism for attachment to cellulose.
Embo J 15, 5739-51.
Wierzba, A., Reichl, U., Turner, R. F. P., Warren, R. A. J. & Kilburn, D. G. (1995). Production and properties of
a bifunctional fusion protein that mediates attachment of Vero cells to cellulosic matherials. Biotechnol. Bioeng.
47, 147-154.
Williams, D. H., Searle, M. S., Mackay, J. P., Gerhard, U. & Maplestone, R. A. (1993). Toward an estimation of
binding constants in aqueous solution: studies of associations of vancomycin group antibiotics. Proc Natl Acad
Sci U S A 90, 1172-8.
53
JANNE LEHTIÖ
Xiao, Z. Z., Gao, P. J., Qu, Y. B. & Wang, T. H. (2001). Cellulose-binding domain of endoglucanase III from
Trichoderma reesei disrupting the structure of cellulose. BIOTECHNOLOGY LETTERS 23, 711-715.
Xie, H., Bolam, D. N., Nagy, T., Szabo, L., Cooper, A., Simpson, P. J., Lakey, J. H., Williamson, M. P. &
Gilbert, H. J. (2001). Role of hydrogen bonding in the interaction between a xylan binding module and xylan.
Biochemistry 40, 5700-7.
Xu, G. Y., Ong, E., Gilkes, N. R., Kilburn, D. G., Muhandiram, D. R., Harris -Brandts, M., Carver, J. P., Kay, L.
E. & Harvey, T. S. (1995). Solution structure of a cellulose-binding domain from Cellulomonas fimi by nuclear
magnetic resonance spectroscopy. Biochemistry 34, 6993-7009.
54