-
ACTAUNIVERSITATISUPSALIENSISUPPSALA2006
Digital Comprehensive Summaries of Uppsala Dissertationsfrom the
Faculty of Science and Technology 185
Hydrolytic and OxidativeMechanisms Involved in
CelluloseDegradation
ANU NUTT
ISSN 1651-6214ISBN 91-554-6571-4urn:nbn:se:uu:diva-6888
-
To my family
-
List of Papers
This thesis is based on the following papers, which will be
referred to in the text by their Roman numerals:
I Nutt, A., Sild, V., Pettersson, G. and Johansson, G. (1998)
Pro-gress curves. A mean for functional classification of
cellulases. European Journal of Biochemistry, 258, 200-206.
II Väljamäe, P., Sild, V., Nutt, A., Pettersson, G. and
Johansson, G. (1999) Acid hydrolysis of bacterial cellulose reveals
different modes of synergistic action between cellobiohydrolase I
and en-doglucanase I. European Journal of Biochemistry, 266,
327-334.
III Henriksson, G., Nutt, A., Henriksson, H., Pettersson, B.,
Ståhl-berg, J., Johansson, G. and Pettersson, G. (1999)
Endoglucanase 28 (Cel12A), a new Phanerochaete chrysosporium
cellulase. European Journal of Biochemistry, 259, 88-95.
IV Nutt, A., Salumets, A., Henriksson, G., Sild, V. and
Johansson, G. (1997) Conversion of O2 species by cellobiose
dehydrogenase (cellobiose oxidase) and glucose oxidase - a
comparison. Bio-technology Letters, 19, 379-383.
V Nutt, A., Nilsson, M., Väljamäe, P., Ståhlberg, J., Isaksson,
R. and Johansson, G. o-nitrophenyl cellobioside as an active site
probe for family 7 cellobiohydrolases. Manuscript.
Papers I-III were reproduced with kind permission of Blackwell
Publishing. Paper IV was reproduced with kind permission of
Springer Science and Business Media. The typographical corrections
were approved by the pub-lisher and the Editor-in-Chief of the
Journal.
-
Contents
Introduction.....................................................................................................9Cellulose.....................................................................................................9
Chemical structure
.................................................................................9Crystalline
structure.............................................................................10
Other components of plant cell
wall.........................................................11Cellulases
.................................................................................................12
Biological degradation of cellulose
.....................................................13Classification
of cellulases
..................................................................13Fungal
cellulases..................................................................................16Specific
aspects of cellulase kinetics
...................................................17Hypocrea
jecorina
cellulases...............................................................20Phanerochaete
chrysosporium
cellulases............................................23Phanerochaete
chrysosporium cellobiose dehydrogenase ..................25
Present investigation
.....................................................................................27Aims
of the present
study.........................................................................27Action
of cellulases on end-labelled cellulose (Paper I)
..........................28Influence of acid pre-treatment of
bacterial cellulose on mode of synergy between Cel7A (CBH I) and
Cel7B (EG I) (Paper II)................29Endoglucanase 28 (Cel12A),
a new Phanerochaete chrysosporiumcellulase (Paper III)
..................................................................................31Conversion
of oxygen species by cellobiose dehydrogenase (Paper
IV).33o-nitrophenyl cellobioside as an active site probe for family
7 cellobiohydrolases (Paper V)
...................................................................36Conclusions
..............................................................................................38
Summary in Swedish
....................................................................................39
Acknowledgements.......................................................................................41
References.....................................................................................................43
-
Abbreviations
BC bacterial cellulose BMCC bacterial microcrystalline cellulose
CBD cellulose binding domain CBH Cellobiohydrolase CBM carbohydrate
binding module CD catalytic domain CDH cellobiose dehydrogenase CMC
carboxymethyl cellulose DP degree of polymerisation EG
Endoglucanase GH glycosyl hydrolase GOX glucose oxidase kcat
catalytic constant Ki inhibition constant KD dissociation constant
KM Michaelis-Menten constant MeUmb (Glc)2 methylumbelliferyl
cellobioside oNPC o-nitrophenyl cellobioside pNPC p-nitrophenyl
cellobioside pNPL p-nitrophenyl lactoside
-
9
Introduction
Cellulose Cellulose is the most abundant biopolymer on earth. It
is the main structural component of plant cell walls, constituting
up to 50% of the mass in trees. Apart from vascular plants,
cellulose is also produced by most groups of algae, the slime mold
Dictyostelium, a number of bacterial species and by tunicates.
Despite the simple chemical structure, the physical properties of
cellulose, such as the crystalline state, degree of crystallinity
and molecular weight are highly variable.
Chemical structure Cellulose is a linear polymer composed of
D-glucose residues joined by -1,4-glucosidic bonds. The cellulose
molecule forms a straight, almost fully extended chain, where
glucose residues are rotated 180° relative to each other along the
main axis, which means that the repetitive unit is the glucose
dimer, cellobiose, rather than glucose (Fig. 1). Although glucose
is a highly water-soluble molecule, the solubility of
cellodextrines decreases rapidly with the degree of polymerisation,
cellohexaose already being only slightly soluble. Each chain is
stabilised by intrachain hydrogen bonds formed be-tween the
pyranose ring oxygen in one residue and the hydrogen of the
OH-group on C3 in the next residue (O5...H-O3’) and between the
hydroxyls on C2 and C6 in the next residue (O2-H...O6’) [Gardner
and Blackwell, 1974].
Figure 1. Molecular structure of the cellulose. Reproduced from
[Hildén and Jo-hansson, 2004] with kind permission from the
publisher.
-
10
Crystalline structure Cellulose, as all carbohydrates, has both
hydrophilic (from the HO-groups) and hydrophobic (from the
HC-groups) character [Sundari et al., 1991]. Strong inter- and
intramolecular O H···O bonds retain the chains straight and stacked
in a sheet-like structure. Individual chains co-crystallise
together shortly after biosynthesis into highly crystalline
microfibrils held together by hydrogen bonds, hydrophobic
interactions and van der Waals forces [Brown, Jr. and Saxena,
2000].
The shape of the cellulose microfibril, where the chains are
running in parallel, is determined by the geometry of the cellulose
synthase complex and by the local environment [Doblin et al.,
2002]. In plants, the unit mi-crofibrils are about 3 nm wide and
contain around 36 cellulose chains and are often packed into
larger, 20-100 nm microfibril bundles in the secondary cell wall.
Interestingly, in certain algae, the microfibril width has been
re-ported to be up to 20 nm [Jarvis, 2003]. Bacterial cellulose
(BC) synthesised by the bacterium Acetobacter xylinum is a long
ribbon with a diameter of about 40-60 nm, consisting of
microcrystals with a width of about 3.0 x 6.8 nm [White and Brown,
Jr., 1981].
Figure 2. A schematic diagram representing the differences
between the monoclinic and triclinic forms of cellulose I. Each
rectangle represents a single glucose unit. In the monoclinic form,
the cellobiose units stagger with a shift of a quarter of the c
axis period, whereas the triclinic form exhibits a diagonal shift
of the same amount. Two spacings and angles are given, the first
referring to the (100) face and the sec-ond to the (010) face of
the triclinic crystal. Reproduced with the permission from [Baker
et al., 1997].
-
11
In nature, most cellulose is produced as crystalline and is
defined as cellu-lose I. It is composed of two distinct crystalline
forms (triclinic cellulose Iand monoclinic cellulose I , which
differ from each other in their crystal packing, molecular
conformation and intermolecular hydrogen bonding pat-terns (Fig. 2)
[Atalla and Vanderhart, 1984; Sugiyama et al., 1991; Heiner et al.,
1995]. These differences may influence the physical properties of
the cellulose [Nishiyama et al., 2003]. The ratio of the two phases
depends on the origin of the cellulose. The I form is dominant in
cellulose produced by primitive organisms, such as the bacterium
Acetobacter xylinum and the alga Valonia macrophysa, whereas the I
form dominates in cellulose produced by higher plants [Atalla and
Vanderhart, 1984]. The I form is more stable than the I form, which
has been reported to be more susceptible to enzy-matic hydrolysis
[Hayashi et al., 1997].
The size of an exposed cellobiose unit on the cellulose surface
is ca 1x0.5 nm. Native cellulose also contains less ordered,
amorphous or paracrystalline regions. Amorphous cellulose has not
been studied widely, but it is thought to be held together by
hydrogen bonds between the C2, C3 and C6 hydroxyl groups [Kondo and
Sawatari, 1996].
Other components of plant cell wall In the plant cell wall, the
cellulose microfibrils are embedded into a matrix, cross-linked
mainly by hemicellulose and pectin (Fig. 3). Wood cells contain
also lignin, a non-polysaccharide polymer. In wood, hemicellulose
and lig-nin comprise 20 to 25 and 5 to 30% of the plant dry weight
[Sjöström, 1993].
HemicelluloseHemicellulose is used as a common name for a large
number of different carbohydrate heteropolymers, of which xylans
and glucomannans are the main components. It is a heterogeneous
mixture of different polysaccharides and the composition varies
depending on the plant type. In contrast to cellu-lose, which
itself is crystalline, strong and resistant to hydrolysis,
hemicellu-lose is a highly branched and amorphous structure with
little inherent strength. Apart from glucose, it may contain
mannose, xylose, arabinose, rhamnose and L-fucose.
LigninLignin is a highly branched random polymer of coniferyl,
sinapyl and p-coumaryl alcohols generated by radical
polymerisation. In wood, lignin is bound covalently to the side
groups of different hemicelluloses by ester- or ether bonds and
forms a matrix surrounding the cellulose microfibril.
-
12
PectinPectin is an important part of a fruit cell wall, but it
is present in all plant cell walls. Pectin is composed of “smooth
regions”, consisting of -1,4 linked galacturonic acid residues and
“hairy regions”, consisting mainly of rham-nogalacturonan I, II and
xylogalactouronan.
Figure 3. The structure of wood. Adapted with modifications from
[Harrington, 1998].
Cellulases Cellulases are O-glycosyl hydrolases (GHs) that
hydrolyse -1,4-glucosidic bonds in cellulose. Cellulose degradation
is brought about mainly by bacte-ria, fungi and protozoa, but the
production of cellulases is documented also in plants and in a
number of invertebrate taxa that includes insects, crusta-ceans,
annelids, molluscs, mussels and nematodes [Watanabe and Tokuda,
2001; Davison and Blaxter, 2005].
-
13
Biological degradation of cellulose In wood, crystalline
cellulose microfibrils are tightly packed in a complex network of
hemicellulose constituents and lignin. Most cellulolytic
micro-organisms produce, in addition to cellulases that hydrolyse
the -1,4-glucosidic bonds, a number of other cell-wall-degrading
enzymes, e.g. ligni-nases, xylanases, pectinases, etc. Only a few
micro-organisms produce a complete set of enzymes capable of
degrading native cellulose efficiently. Aerobic and anaerobic
micro-organisms use different strategies to feed on cellulose.
Whereas aerobes generally secrete a set of individual cellulases,
some anaerobes have evolved a multi-enzyme complex- cellulosome
which is associated with the cell surface of the micro-organism,
reviewed recently by Bayer et al. [Bayer et al., 2004].
The cellulolytic enzyme systems in fungi can be divided into two
groups. The white-rot fungi, such as Phanerochaete chrysosporium
and soft-rot fungi, such as Hypocrea jecorina (formerly known as
Trichoderma reesei)and Penicillum pinophilum have complete
cellulolytic enzyme systems ca-pable of the breakdown of
crystalline cellulose to glucose. They consist of several secreted
enzymes acting at the ends (exoglucanases) or in the middle
(endoglucanases) of the cellulose chains. The released cellobiose
is hydro-lysed to glucose by -glucosidases. The second group of
fungi reportedly degrade cellulose by means of oxidative components
together with endoglu-canases, but lack the strict
cellobiohydrolases. A representative of this mechanism is the
cellulolytic system of the brown-rot fungus Postia pla-centa
[Kleman-Leyer et al., 1992].
Classification of cellulases Cellulases can be classified by
different means, according to their substrate specificities,
reaction mechanisms or structural similarities.
Functional classification Cellulases have traditionally been
classified into two distinct classes: cello-biohydrolase (1,4-
-D-glucan cellobiohydrolase, EC 3.2.1.91) and endoglu-canase (1,4-
-D-glucan glucanohydrolase, EC 3.2.1.4), based on their activ-ity
toward a wide range of substrates. This is rather difficult, since
the en-zymes have overlapping specificities toward substrates which
themselves are poorly defined.
By definition, cellobiohydrolases release cellobiose from the
non-reducing ends of the cellulose chain, but the experimental
evidence for this assumption is obscure. Enzyme kinetics on soluble
oligosaccharides and structural data on enzyme-oligosaccharide
complexes show that some cello-biohydrolases may have opposite
chain-end preferences [Barr et al., 1996; Divne et al., 1998;
Koivula et al., 1998]. Cellobiohydrolases are thought to
-
14
work processively, that is, one enzyme molecule can release
several cello-biose units from the cellulose chain without leaving
the substrate. Cellobio-hydrolases show little or no activity on
substituted celluloses, such as CMC, but microcrystalline cellulose
with relatively low DP is relatively rapidly degraded [Beguin and
Aubert, 1994].
Endoglucanases cut cellulose chains at random positions in less
crystal-line regions, creating new chain ends. Extreme
endoglucanases, often called CM-cellulases
(carboxymethyl-cellulases) have little activity towards
crys-talline cellulose, but hydrolyse readily CMC, acid-swollen
cellulose and even barley -glucan in a random fashion, resulting in
a rapid fall in the degree of polymerisation [Kleman-Leyer et al.,
1994].
The classification of cellulases as purely endoglucanases or
exogluca-nases is not absolute and is an over-simplification, since
several studies indi-cate that several cellobiohydrolases can
attack also the internal glucosidic bonds of the cellulose chain
[Ståhlberg et al., 1993; Armand et al., 1997; Boisset et al.,
2000]. Also, several endoglucanases have been shown to hy-drolyse
cellulose processively, which is a common property of
cellobiohy-drolases [Reverbel-Leroy et al., 1997; Gilad et al.,
2003; Cohen et al., 2005; Zverlov et al., 2005]. As a result,
cellulases seem to have a more or less con-tinuous spectrum of
properties ranging from virtually random endogluca-nases to highly
processive strict cellobiohydrolases [Teeri, 1997; Hildén and
Johansson, 2004].
Hydrolytic mechanism In glycosyl hydrolases, enzymatic
hydrolysis of the glycosidic bond usually takes place via general
acid/base catalysis, which requires two critical resi-dues: a
proton donor (HA) and a nucleophile/base (B-). This catalytic
activity is provided by two aspartic- or glutamic acid
residues.
Two different mechanisms can be distinguished- retaining and
inverting mechanisms. In both cases, the acid-base (HA) protonates
the leaving glyco-sidic oxygen with the concomitant formation of a
partial positive charge on the C1 carbon.
In the inverting mechanism, the base (B-) deprotonates a water
molecule, which then attacks the C1 carbon of the glucose ring in
an Sn2 type dis-placement reaction, resulting in inversion of the
configuration at the ano-meric carbon C1.
In the retaining mechanism, a glycosidic bond is hydrolysed via
two sin-gle displacement steps. First, the nucleophile (B-) attacks
directly the C1 carbon, resulting in a covalent intermediate
between the enzyme and the substrate, the first product is
released. In the second step, the acid-base acti-vates a water
molecule by abstracting a proton from it, promoting an attack on
the C1 carbon.
-
15
OR
A
B-
HOOH
A
B-
HO
OR
A
B-
HO
A-
B
H
HROHO
A-
B
OO
HHO
ROH
A-
BH
O
HO
5.5 Å
10 Å
A
B
Figure 4. The two major mechanisms of enzymatic hydrolysis of
the glycosidic bond as first proposed by Koshland [Koshland D.E.,
1953]. (A) The retaining mechanism. (B) The inverting
mechanism.
Recently, a fundamentally different glycosidase mechanism has
been un-veiled for NAD+- and divalent metal ion-dependent GH4
glycosidases whereby hydride abstraction at C3 generates a ketone,
followed by deproto-nation of C2 accompanied by acid-catalysed
elimination of the glycosidic oxygen and formation of a
1,2-unsaturated intermediate. This - -unsaturated species undergoes
a base-catalysed attack by water to generate a 3-keto derivative,
which is then reduced by NADH to complete the reaction cycle [Lodge
et al., 2003; Rajan et al., 2004; Varrot et al., 2005].
Glycoside hydrolase families The glycoside hydrolases can be
classified into structurally related families based on similarities
in the distribution of hydrophobic amino acids in their sequences
[Henrissat, 1991]. Up to date, 106 families have been
distin-guished and the continuously updated information is
available on Carbohy-drate Active Enzymes Database server
(http://www.cazy.org/CAZY) [Coutinho and Henrissat, 1999].
Cellulolytic enzymes are grouped into at least 14 families. The
family classification reflects the structural features of the
enzymes and the evolution of glycoside hydrolases. Some families
con-tain enzymes with different substrate specificities. For
example, families 5, 6, 7, 8, 9 and 48 contain both
cellobiohydrolases and endoglucanases. Fam-ily 7 contains only
fungal hydrolases, whereas family 8 contains only bacte-rial
hydrolases. Some families are evolutionally deeply rooted,
containing cellulases from bacteria, fungi, and plants. Also,
cellulases from different families are found in the same organism.
So far, the hydrolysis mechanism
-
16
seems to be conserved among the members of a given glycosyl
hydrolase family.
Fungal cellulases Cellulases face the difficult problem of
working on a solid substrate. Most of the fungal cellulases share a
common molecular organisation where a large catalytic domain (CD)
is connected by a highly glycosylated linker-peptide to a small
carbohydrate-binding module (CBM). Upon limited proteolysis with
papain the enzymes can, in many cases, be cleaved easily into the
two functional domains [Tomme et al., 1988].
Figure 5. A general sketch of a fungal cellulase. The three
hexagons in the CBM indicate the aromatic residues responsible for
interaction with the hydrophobic face of every second pyranose
ring. The grey area in the figure represents the loops cov-ering
the substrate-binding sites. Reprinted with permission from [Hildén
and Jo-hansson, 2004].
The active site of a cellulase consists of multiple binding
sites for glucose units, which enhances the probability for the
enzyme to remain bound to the substrate after a catalytic cycle and
thereby work processively [Divne et al., 1994]. These binding
subsites are labelled, according to convention, from -nto +n, with
-n at the nonreducing end and +n at the reducing end. The cleav-age
occurs between the -1 and +1 subsites [Davies et al., 1997].
Generally, cellobiohydrolases have a tunnel-shaped active site,
whereas the active site for endoglucanases is more open, forming a
cleft or groove, allowing the enzyme to bind to the middle of the
substrate chain and cleave it. Since some cellobiohydrolases also
can perform these internal cuts, the loops closing the tunnel must
be flexible to allow a cellulose chain to enter the active
site.
Most fungal cellulases contain, in addition to the CD, a
carbohydrate binding module (CBM), more specifically called
cellulose-binding domain (CBD). The CBDs are believed to play an
important role in cellulose hy-drolysis. Although these domains do
not affect the activity of cellulases to-
-
17
ward soluble and amorphous substrates, they significantly
enhance the ca-pacity of the enzymes to hydrolyse crystalline
cellulose. Currently, the CaZy classification lists 45 families of
characterised CBMs based on amino acid sequence similarity (see
http://www.cazy.org/CAZY/), but this number will most probably
increase [Davies et al., 2005]. These families have been re-viewed
recently by Boraston et al., [Boraston et al., 2004]. All fungal
CBDs belong to the family I containing 35 to 40 amino acids and
show strong se-quence similarity with an overall amino acid
identity of 60%, some residues being completely conserved and some
displaying conservative substitutions [Gilkes et al., 1991].
The first structure of a fungal CBD was determined by nuclear
magnetic resonance [Kraulis et al., 1989]. The CBDs of the fungal
cellulases have a wedge-shaped fold containing a basic structure of
a distorted -sheet of three short antiparallel strands. One face of
the wedge is planar and contains three conserved aromatic amino
acids separated by a distance corresponding to the length of the
repeating unit in cellulose, cellobiose [Tomme et al., 1995]. This
interaction is often supplemented by polar residues forming
hydrogen bonds [Tormo et al., 1996]. The other surface is rougher
and less hydrophilic in character.
Specific aspects of cellulase kinetics The enzymatic degradation
of solid cellulose is a complicated process which takes place at a
solid-liquid phase boundary where the enzymes are the mo-bile
components. Several properties of the substrate influence the
kinetics of enzymatic hydrolysis of cellulose: the crystallinity
and probably also the type of the cellulose crystals, the degree of
polymerisation, the distribution of the molecular weight, the
accessible surface for the enzymes and the mi-crostructure of the
cellulose surface [Zhang and Lynd, 2004]. To study the individual
effects of these parameters on the enzymatic hydrolysis is a
com-plicated task, because within any given cellulose sample there
is a great de-gree of variability. Recently, bacterial cellulose
produced by Acetobacter xylinum has become a widely used substrate
for cellulose studies. The ad-vantages of using bacterial cellulose
as a substrate are that it consists of pure cellulose, is
relatively well-defined and is available in never-dried form.
The studies of enzymatic attack of cellulose have focused
primarily on the release of reducing sugars from insoluble
cellulose or soluble cellulose de-rivatives. Relatively little is
known about the effect of the individual en-zymes on the
macromolecular structure of insoluble cellulose substrates.
The rate of enzymatic hydrolysis of the cellulosic materials
always de-creases rather quickly. Generally, enzymatic cellulose
degradation is charac-terised by a rapid initial phase followed by
a slow secondary phase that may last until all substrate is
consumed. This has been explained most often by the rapid
hydrolysis of the readily accessible fraction of cellulose,
strong
-
18
product inhibition and slow inactivation of absorbed enzyme
molecules [Converse et al., 1988]. The erosion of the cellulose
surface by the cellulases has been proposed as one of the rate
retardation factors [Väljamäe et al., 1998].
It has been shown that the surface area of cellulose which is
accessible to cellulase enzymes is the most important factor in
determining initial rates of hydrolysis [Thompson et al., 1992;
Helle et al., 1993]. The efficiency of cellulose hydrolysis by an
individual enzyme is dependent on the degree of polymerisation and
crystallinity of the substrate. Generally, cellobiohy-drolases are
relatively more active towards highly crystalline substrates with
relatively low DP, such as BMCC; endoglucanases have only very
limited action on crystalline substrates, but hydrolyse readily
amorphous cellulose [Henrissat et al., 1985].
The interplay of the cellulose domains on crystalline cellulose
degradation, the role of the CBD The adsorption of cellulase to
cellulose is a prerequisite step for hydrolysis. The overall
binding efficiency of the cellulases to the cellulose is enhanced
by the presence of the CBM, and this correlates clearly with higher
activity towards insoluble cellulose [Tomme et al., 1988; Gilkes et
al., 1988; Reini-kainen et al., 1992; Ståhlberg et al., 1993;
Reinikainen et al., 1995]. At the same time, the strong binding via
CBD to the cellulose surface can lead to a population of
nonproductively bound enzymes [Ståhlberg et al., 1991]. In addition
to anchoring the enzyme molecules to the cellulose surfaces, the
disruption of cellulose microfibrils by family II CBDs has been
reported [Din et al., 1991]. Simultaneous addition of separated CBD
and catalytic domain resulted in synergy between these domains in
the hydrolysis of cot-ton cellulose [Din et al., 1994]. Similar
results have not so far been obtained with the CBDs from other
families and even with family II CBDs the effect was seen only
using cotton as substrate and not on microcrystalline cellulose
[Esteghlalian et al., 2001].
It is probable that different CBDs bind to different regions on
the cellu-lose surface. CBDs can promote the enzyme activity
towards different re-gions on the cellulose surface, thereby
determining the substrate specificity [Carrard et al., 2000].
Lately, CBMs from families 1 and 3 were shown to bind
preferentially to the obtuse corners of Valonia cellulose
microcrystals, which expose the hydrophobic phase [Lehtio et al.,
2003].
Many cellulases, like other enzymes acting on polymeric
substrates, have been thought to work processively, i.e. they can
perform several hydrolytic events without dissociating from the
substrate. Generally, the kcat value for oligosaccharide hydrolysis
increases together with the DP of the substrate. In 1983, Lee et
al. found that the cellulase catalysis does not significantly
affect the DP of a solid cellulose substrate. These authors
proposed that cellulose chains are peeled off progressively from
the fibrils by the cellulase enzymes,
-
19
since the DP remains constant during the course of hydrolysis.
[Lee et al., 1983]. Similar findings have been described for H.
jecorina cellobiohy-drolase Cel6A acting on cotton fibers
[Kleman-Leyer et al., 1996]. The thin-ning of cellulose
microfibrils by the action of cellobiohydrolases has been observed
several times [Chanzy and Henrissat, 1985; Boisset et al., 2000;
Lee et al., 2000]. The processivity hypothesis is supported by the
structure of the CD of cellulases, which contain multiple binding
sites for the glycosyl units. The cellulose chains are held
together in microcrystals by van der Waals interactions and
hydrogen bonds. In order to separate a cellulose chain from the
cellulose crystal, a cellulase molecule has to overcome an energy
barrier in breaking the interaction between cellulose chains, which
may slow down the rate of hydrolysis [Skopec et al., 2003]. Since
the active site of a cellobiohydrolase often is a long tunnel, the
cellulose chain is en-closed and held in the active site of the
enzyme by numerous interactions, which makes the enzyme less likely
to dissociate after each hydrolytic step and thus compete more
efficiently with the interactions driving the cellulose chain back
onto the cellulose crystal [von Ossowski et al., 2003]. Deletion in
the loops covering the active site of cellobiohydrolase Cel7A
resulted in loss of activity on crystalline cellulose, whereas the
activity on amorphous cellu-lose and soluble substrates remained
the same or increased [von Ossowski et al., 2003]. In the case of
processive endoglucanases, the active-site covering loop undergoes
a large “loop-flip” conformational change to enclose the active
site upon substrate binding [Davies et al., 1995; Varrot et al.,
2000]. Movements in the loops have been shown also for GH6 family
cellobiohy-drolases, and probably facilitate substrate gliding into
the tunnel to allow occasional endo type of cleavages [Zou et al.,
1999; Varrot et al., 1999]. The efficient hydrolysis of cellulose
needs interplay between those two domains. Experiments with H.
jecorina Cel7A mutants with deletions in the hinge region
connecting the CD and CBD have shown that sufficient distance
be-tween CD and CBD is needed in the cellulases for efficient
hydrolysis of crystalline cellulose [Srisodsuk et al., 1993].
Recently, Mulakala and Reilly presented a model based on the
interaction energies and forces on cello-oligosaccharides
computationally docked to CD and CBD, where CBD wedges itself under
a free chain end on the crystalline cellulose surface and feeds it
to the CD active site tunnel [Mulakala and Reilly, 2005]. The
energy for cellulose structure disruption comes ultimately from the
chemical energy of glycosidic bond breakage [Sinnott, 1998].
Synergism of cellulases Effective degradation of crystalline
cellulose requires cooperation between different types of
cellulases. This cooperation, resulting in higher total activ-ity,
is called synergism. The synergy factor (SFp) is defined as the
ratio of the activity of combined enzymatic action to the sum of
the activities of in-dividual components. Two classes of synergism
between cellulases have
-
20
been described: the cooperation between endoglucanases and
cellobiohy-drolases (endo-exo synergism) and that between two
cellobiohydrolases (exo-exo synergism). Generally, synergism
between endo- and exoenzymes is highest on semicrystalline
cellulose of high DP, lower on amorphous cel-lulose and
non-existent on soluble cellulose derivatives [Henrissat et al.,
1985; Nidetzky et al., 1993; Samejima et al., 1997]. The molecular
basis of synergism is not yet completely understood, largely
because the modes of action of the individual enzymes are not
clear. The synergism has been found to be dependent on the relative
proportions of the enzyme components [Henrissat et al., 1985] and
also on the degree of saturation of the substrate with the enzymes,
decreasing at higher enzyme concentration [Woodward et al., 1988a;
Woodward et al., 1988b].
It is generally assumed that the mechanism of endo-exo synergism
can be discussed in terms of sequential action where by the random
endoglucanase initiates attack and the new chain ends generated are
then hydrolysed by the endwise-acting cellobiohydrolase.
Furthermore, -glucosidases can work in synergy with cellulases by
removing the cellobiose produced.
Hypocrea jecorina cellulases The filamentous soft-rot fungus
Hypocrea jecorina (formerly known as Trichoderma reesei) is one of
the most studied cellulolytic micro-organisms. It degrades plant
litter in its natural environment, in the soil. H. jecorinaproduces
a complete cellulolytic enzyme system and is capable of very
effi-cient degradation of crystalline cellulose.
Table 1. Hypocrea jecorina (Trichoderma reesei) cellulases.
Enzyme Old name Molecular weight (kDa) Isoelectric point
(pI)
Position of the CBM
Cel7A CBH I 57 3.9 C Cel6A CBH II 53 5.9 N Cel7B EG I 55 4.5 C
Cel5A EG II 50 5.5 N Cel12A EG III 25 7.5 - Cel61A EG IV 55 C
Cel45A EG V 36 2.9 C
Hitherto, two cellobiohydrolases (Cel7A and Cel6A), five
endoglucanases (Cel7B, Cel5A, Cel12A, Cel61A and Cel45A) and two
-glucosidases have been isolated from H. jecorina culture medium.
In addition, transcription analysis and genome sequencing have
additionally identified three putative endoglucanases belonging to
the families GH5, GH61 and GH74 and five putative -glucosidases
(one belonging in family GH 1 and four in family GH3) [Foreman et
al., 2003]. Each H. jecorina cellulase is expressed from a single
gene, and a simple on-off co-regulation results in constant
ratios
-
21
among the major enzymes, regardless of the growth conditions
[Ilmen et al., 1997]. All isolated H. jecorina cellulases except
Cel12A have a multidomain structure based on a catalytic domain and
a cellulose-binding domain. Both domains bind to cellulose, but the
affinity of the CD is, in most cases, much lower than that of the
CBD [Ståhlberg et al., 1991].
Three-dimensional structures of the catalytic modules of the
cellobiohy-drolases Cel7A, Cel6A and for the endoglucanases Cel7B
and Cel12A have been solved [Rouvinen et al., 1990; Divne et al.,
1994; Kleywegt et al., 1997; Sandgren et al., 2001]. The overall
shape of Cel7A and Cel6A has been determined by low-resolution
small-angle X-ray scattering analysis (SAXS). Both enzymes were
shown to have similar structures with a large ellipsoidal head and
an elongated cylindrical tail with the average dimen-sions 4.5x18
nm for Cel7A and 4.5x21.5 nm for Cel6A [Abuja et al., 1988; Abuja
et al., 1989].
Cel7ACel7A (CBH I) is the major cellulase produced by H.
jecorina. It comprises about 45-50% of the total cellulolytic
protein of H. jecorina and hydrolyses crystalline cellulose,
liberating cellobiose as the main product [Fägerstam and
Pettersson, 1980]. The crystal structure of the CD revealed a
-sandwichstructure with a 50Å-long substrate-binding tunnel formed
by the inner -sheets and the extensive loops covering the active
site [Divne et al., 1994; Divne et al., 1998]. Ten glycosyl-unit
binding subsites have been identified, 3 of these at the product
side. Four tryptophan residues form a glycosyl-binding platform in
sites –7, –4, –2 and +1 in the tunnel of Cel7A. Cello-biose has its
highest affinity towards the +1,+2 subsites of the enzyme, the
experimentally determined value for Kd, based on competitive
binding or inhibition experiments using various chromogenic or
fluorogenic substrates (p-nitrophenyl glycosides and
methylumbelliferyl glycosides) being about 20 µM [Claeyssens et
al., 1989; van Tilbeurgh et al., 1989; Henriksson et al., 1999b].
The action of Cel7A on cellulose is much less sensitive to
inhibition by cellobiose, with an apparent Ki around 1.5 mM [Gruno
et al., 2004].
Cel7A, like the other GHs belonging to family 7, hydrolyse the
-1,4 glu-cosidic bond of cellulose with retention of the anomeric
carbon configura-tion [Knowles et al., 1988; Claeyssens et al.,
1990]. Site-directed mutagene-sis confirmed that Glu217 acts as the
proton donor and Glu212 as the nu-cleophile in a
double-displacement mechanism, whereas Asp214 is likely to be
involved in maintenance of the appropriate pKa values of the other
cata-lytic residues [Divne et al., 1994; Ståhlberg et al., 1996;
Kleywegt et al., 1997]. It has been shown that Cel7A hydrolyses
soluble oligosaccharides from the reducing end [Biely et al., 1993;
Barr et al., 1996]. The kcat values for Cel7A have been shown to
increase with the DP of the substrate, having values of 4.0 s-1 for
cellotetraose and 9.5 s-1 for cellohexaose, whereas KMvalues
decrease from 7 µM to 3 µM [Nidetzky and Claeyssens, 1994].
Struc-
-
22
tural and kinetic data, taken together, indicate strongly that
Cel7A is a highly processive enzyme. Recently, the processivity
values for Cel7A acting on cellulosic substrates labelled at the
reducing end with anthranilic acid were determined and found to be
88±10 for bacterial cellulose, 42±10 for bacterial microcrystalline
cellulose and 34±2.0 for endoglucanase-pretreated bacterial
cellulose, respectively [Kipper et al., 2005].
Cel6ACel6A (CBH II) is another cellobiohydrolase produced by H.
jecorina, con-stituting approximately 20% of the secreted protein.
The crystal structure of the Cel6A catalytic domain was the first
cellulase structure solved [Rouvinen et al., 1990]. Cel6A CD is an
/ barrel protein, similar to triose phosphate isomerase (TIM) with
the exception that it contains seven instead of eight -strands. The
active site forms a 20 Å long tunnel with four glyco-syl unit
binding subsites. Two additional subsites close to the tunnel
entrance have been identified [Koivula et al., 1998]. A tryptophan
residue (Trp 272) at the +4 subsite is critical in the crystalline
cellulose degradation by Cel6A. Its mutation leads to an overall
decrease in activity by at least an order of magnitude. Compared to
H. jecorina Cel7A, the active-site tunnel of Cel6A is shorter and
more open. The tunnel-covering loops can undergo move-ments,
resulting in the closing or opening of the tunnel [Zou et al.,
1999; Varrot et al., 1999]. This is apparently the reason for the
observed endoactiv-ity and lower processivity of Cel6A.
Cel6A hydrolyses the glucosidic bond with inversion of the
configuration at the anomeric carbon by a single displacement
mechanism [Knowles et al., 1988; Claeyssens et al., 1990]. The bond
cleavage takes place from the non-reducing end of the substrate.
Two catalytically important aspartate residues have been
identified, of which Asp 221 acts as proton donor and Asp 175
stabilises the positively-charged transition state [Koivula et al.,
2002].
EndoglucanasesCel7B (EG I) is the main endoglucanase of H.
jecorina, accounting for 5-10% of the total cellulase [Bhikhabhai
et al., 1984] and shows 45% sequence homology to Cel7A [Penttilä et
al., 1986]. Cel7B also shows a very similar fold. However, four
loops covering the tunnel in Cel7A are partially deleted in Cel7B,
resulting in an open-groove-shaped active site [Kleywegt et al.,
1997]. Cel7B cleaves the polymeric substrates in random fashion and
pos-sesses transglycosylation activity [Claeyssens et al., 1990;
Biely et al., 1991].
Another endoglucanase, Cel5A (EG II) is produced by H. jecorina
in comparable amounts [Saloheimo et al., 1988]. Cel5A hydrolyses
the gluco-sidic bond via the double-displacement mechanism and has
slightly lower activity than Cel7B on substituted celluloses and
-glucan [Penttilä et al., 1987]. Its three-dimensional structure is
not yet established, but the kinetic
-
23
studies of oligosaccharide hydrolysis by Cel5A suggest that the
active site of Cel5A consists of five glucosyl binding subsites
[Macarron et al., 1993; Biely et al., 1993].
Cel12A (EG III) is a small, 25 kDa endoglucanase that does not
have a CBD. It can hydrolyse, in addition to cellulose, also
-1,3-1,4 glucan, xy-loglucan and xylan [Hayn et al., 1993; Karlsson
et al., 2002]. The crystal structure revealed that Cel12A consists
of two largely anti-parallel -sheets which pack on top of each
other. The substrate-binding cleft is approxi-mately 35 Å long, 8 Å
wide and 15 Å deep with at least six sugar-binding subsites, from
–4 to +2 [Sandgren et al., 2001; Sandgren et al., 2005]. Cel12A
hydrolyses the glucosidic bonds in cellulose using the
double-displacement mechanism with Glu 116 as nucleophile and Glu
200 as gen-eral acid/base [Okada et al., 2000].
The expression of Cel61A (EG IV) is induced together with the
other cel-lulases in H. jecorina [Saloheimo et al., 1997]. However,
the specific en-doglucanase activity of Cel61A (EG IV) is several
orders of magnitude lower than that of Cel7B toward both
cello-oligosaccharides and amorphous and substituted celluloses,
and its role remains obscure [Karlsson et al., 2001].
Cel45A (EG V) seems to have quite unique hydrolytic properties.
It does not hydrolyse cellotriose, cellotetraose and cellopentaose
and has lower ac-tivity toward cellulosic substrates than do other
H. jecorina endoglucanases [Karlsson et al., 2002]. The main
product of Cel45A (EG V) cellulose hy-drolysis is cellotetraose,
with significant amounts of cellotriose and cel-lopentaose. It
hydrolyses readily glucomannan, being able to cleave a glyco-sidic
bond between glucose and a mannose unit, which indicates that
Cel45A is a glucomannanase rather than a strict endoglucanase
[Karlsson et al., 2002].
Phanerochaete chrysosporium cellulases The white-rot
basidomycete, Phanerochaete chrysosporium, employs an array of
extracellular enzymes capable of completely degrading the major
polymers of wood: cellulose, hemicellulose and lignin. P.
chrysosporiumexhibits a system of synergistically-acting cellulases
homologous to H. je-corina [Uzcategui et al., 1991a; Uzcategui et
al., 1991c]. Sequencing of the whole P. chrysosporium genome
revealed at least 40 putative endogluca-nase-encoding genes (in
families GH5, GH9, GH12, GH61 and GH74), six genes encoding GH7
cellobiohydrolases and one gene for a cellobiohy-drolase belonging
to the GH6 family [Martinez et al., 2004]. The data con-cerning
cellobiohydrolases are in accordance with previous findings [Sims
et al., 1988; Covert et al., 1992a; Tempelaars et al., 1994].
Hitherto, three cellobiohydrolases: Cel7D (CBH 58), Cel7C (CBH
62) and Cel6A (CBH 50), and three endoglucanases: Cel5A (EG 38),
Cel5B (EG
-
24
44) and Cel12A (EG 28) have been isolated and characterised from
P. chry-sosporium [Uzcategui et al., 1991a; Uzcategui et al.,
1991c; Henriksson et al., 1999a]. The fourth previously
characterised endoglucanase, EG 36, is probably a truncated isoform
of EG 38 lacking a cellulose-binding module [Uzcategui et al.,
1991a]. Recently, from cellulose-grown medium of P. chrysosporium,
two peptides with previously identified cDNAs matching to cel7e and
cel7f were identified for the first time [Wymelenberg et al.,
2005]. In addition, several peptides matching putative
endoglucanases from GH families 12, 45, 61 and 74 were identified
[Wymelenberg et al., 2005].
Table 2. Phanerochaete chrysosporium cellulases
Enzyme Old name Molecular weight (kDa) Isoelectric point
(pI)
Position of the CBM
Cel7D CBH 58 58 3.8 C Cel7C CBH 62 62 4.9 C Cel6A CBH 50 50 4.9
N Cel5B EG 44 44 4.3 N Cel5A EG 38 38 5.6-5.7 N Cel12A EG 28 28 5.2
-
In contrast to H. jecorina, where the expression of cellulases
is co-regulated, the cellulase genes of P. chrysosporium are
differentially transcribed, de-pending on the substrate and the
stage of degradation. During growth on cellulose powder, the
highest expression is observed for Cel7D and Cel6A, with lower
levels of Cel7C and GH5 endoglucanases [Uzcategui et al., 1991b;
Vanden Wymelenberg et al., 1993]. Using aspen wood chips as the
growth medium, transcripts of cel6A, cel7C and cel7E dominate
[Vallim et al., 1998]. During growth on minimal medium containing
glucose, only cel7A and cel7B transcripts were seen [Covert et al.,
1992b].
Cel7DCel7D is the major secreted cellulase in the cultures grown
on cellulose powder as a carbon source [Szabo et al., 1996]. It has
55% amino acid se-quence identity to H. jecorina Cel7A. The main
architecture of the Cel7D catalytic module, including most aspects
of substrate-binding- and catalytic machinery, resembles the
structure of Cel7A, the main differences being deletions and other
changes in the loops covering the substrate-binding tun-nel, which
makes the tunnel more open without any direct contacts between one
side and the other [Munoz et al., 2001]. In total, 11 substrate
binding subsites have been identified, three of them at the product
side. Cel7D has higher activity than H. jecorina Cel7A toward both
soluble and insoluble substrates [von Ossowski et al., 2003].
-
25
Phanerochaete chrysosporium cellobiose dehydrogenase Cellobiose
dehydrogenase (CDH) is an extracellular enzyme produced by many
lignocellulose-degrading soft-rot, white rot and brown-rot fungi,
in-cluding P. chrysosporium [Westermark and Eriksson, 1974;
Schmidhalter and Canevascini, 1993; Roy et al., 1996; Fang et al.,
1998; Schou et al., 1998; Temp and Eggert, 1999]. The enzyme is
produced at relatively high levels (0.5% of the secreted protein)
when cellulose is the main carbon source [Szabo et al., 1996]. CDH
oxidises cellobiose, lactose and longer cello-oligosaccharides to
the corresponding lactones using a wide spectrum of electron
acceptors, including quinones, phenoxyradicals, Fe3+, Cu2+ and
molecular oxygen. CDH exhibits strong discrimination against
glucose, as indicated by the 87,000-fold larger specificity
constant (kcat/KM) for cello-biose compared to glucose [Henriksson
et al., 1998]. This feature has been used to construct amperometric
biosensors for the measurement of cello-biose, lactose and
cellooligosaccharides [Nordling et al., 1993].
CDH is a monomeric enzyme with molecular weight of 90 kDa
consisting of two distinct domains: a flavin domain, which contains
FAD and a cyto-chrome b type heme-carrying domain. These two
domains are connected via a 25-residue peptide linker which is
susceptible to cleavage by papain [Henriksson et al., 1991]. The
overall shape of CDH is reported to be “cigar shaped” with a length
of 180 Å and a maximal width of about 50 Å [Lehner et al., 1996].
The crystal structures of both domains have been solved [Hallberg
et al., 2000; Hallberg et al., 2002]. The flavin domain is
peanut-shaped and consists of two structurally distinct subdomains;
one that binds the FAD cofactor and one that binds the substrate,
cellobiose. The interface between these two subdomains forms a 12
Å-long funnel-shaped tunnel that leads down to the active site. Two
glycosyl-binding subsites were identified; the innermost
glycosyl-binding subsite (C) adjacent to the flavin ring and the
binding subsite (B) close to the tunnel entrance. The architecture
of sugar-binding subsites supports prevhious findings [Henriksson
et al., 1998] that the specificity of the CDH is determined mostly
by the configuration of the C2 carbon in the C subsite, whereas in
the B subsite, configurations at C2, C3 and possibly also C6
carbons appear to be important. The tight binding of the substrate
at the B subsite partly explains the observed strong glucose
discrimination.
To date, CDH is the only known extracellular flavocytochrome. It
is still unclear how the electrons are transferred between these
two domains, but according the crystal structures, both domains
where the active site is located display a high degree of surface
complementarity allowing the cofactors to communicate over a
distance relevant for inter-domain electron transfer [Hallberg et
al., 2002].
CDH can transfer electrons both to one- and two-electron
acceptors. The reduction rate of two-electron acceptors is
virtually unaffected by the loss of
-
26
the heme domain [Samejima and Eriksson, 1992; Henriksson et al.,
1993]. For one-electron acceptors, both the sink model, where the
heme domain acts as electron sink, enhancing the rate of reduction
of one-electron accep-tors, and the electron chain model, where the
heme directly reduces the elec-tron acceptor after obtaining the
electrons from FADH2, have been dis-cussed. Recent results indicate
that both electrons from two-electron-reduced CDH are transferred
to one-electron acceptor cytochrome c via the heme [Igarashi et
al., 2005]. However, this does not have to be the case for all
one-electron acceptors.
CDH can produce both hydrogen peroxide and superoxide as primary
re-duction products of molecular oxygen [Morpeth, 1985; Kremer and
Wood, 1992a] and can also degrade hydrogen peroxide under the same
conditions [Henriksson et al., 1993].
CDH binds to cellulose with an estimated binding constant in the
submi-cromolar range. The cellulose binding is probably of
hydrophobic nature and the putative binding site is located on the
flavin domain [Henriksson et al., 1997]. However, in the crystal
structure there is no obvious substructure or surface patch that
can be assigned as the cellulose-binding site [Hallberg et al.,
2002].
The physiological role of CDH in wood degradation has not been
estab-lished unambiguously. Several roles for CDH in wood
degradation have been proposed and these hypotheses have been
reviewed by Henriksson et al. [Henriksson et al., 2000]. Some
examples include preventing the repoly-merisation of lignin by
reducing phenoxyl radicals produced by lignolytic enzymes;
participating in lignin degradation by supporting manganese
per-oxidase, relieving product inhibition of cellulases by
oxidating cellobiose- the main hydrolysis product cellulases, etc.
One of the suggested functions of the CDH is involved in generating
reactive oxygen species, such as super-oxide and hydroxyl radicals.
Kremer and Wood proposed that the Fe3+ -reducing activity of CDH
may be important for the production of hydroxyl radicals [Kremer
and Wood, 1992b]. Reactive oxygen species may acceler-ate the
depolymerisation of cellulose by attacking its crystalline
structure, thereby making it more accessible for hydrolytic
enzymes.
-
27
Present investigation
Aims of the present study This thesis has focused on the
characterisation of the dynamics in the enzy-matic degradation of
cellulose microfibrils, including both the action of the individual
enzymes and their synergistic interplay. More specifically, the
following tasks were addressed:
To determine the end-preference of cellobiohydrolases on their
natural substrate - crystalline cellulose.
To investigate synergistic mechanisms between endo- and
exo-acting cellulases with specific focus on its dependence on the
nature of the substrate.
To evaluate the conversion of oxygen species by cellobiose
dehydro-genase.
In addition, a new endoglucanase from Phanerochaete
chrysosporium was isolated and characterised. The interactions
between GH7 family cellobiohy-drolases and o-nitrophenyl
cellobioside were investigated and employed for indirect binding
studies.
-
28
Action of cellulases on end-labelled cellulose (Paper I) The aim
of paper I was to determine the end-preference of
cellobiohy-drolases acting on crystalline cellulose.
It was assumed for a long time that cellobiohydrolases hydrolyse
cellu-lose starting from the non-reducing end of the cellulose
chain. An increasing amount of both structural and kinetic data
using soluble substrates show indeed strong evidence that
cellobiohydrolases have different end-preferences. However, the
behaviour of the enzymes on crystalline cellulose is not only
active-site mediated, but may be influenced by the binding to the
bulk cellulose, especially if a binding domain is involved.
To study the end-preference of cellulases we used
reducing-end-labelled bacterial microcrystalline cellulose (BMCC)
as substrate. We followed the release of the labelled end group in
relation to the total reaction course (total fraction of cellulose
solubilised). Conclusions were drawn with the help of
computer-simulated progress curves.
Figure 6. Some examples of the simulated progress curves. The
first example shows a progress curve for a strict exoenzyme that
hydrolyses the substrate from the la-beled end. The second example
shows the progress curve for a highly processive strict exoenzyme
hydrolyzing the substrate from the non-labelled end.
The computer simulations were carried out to mimic the actual
experimental situation. The following parameters, which may
influence the shape of the progress curve, were taken into account:
end-preference of the enzyme,
-
29
probability for endo/exoactivity, processivity towards labelled
ends, proces-sivity towards non-labelled end and enzyme
concentration (non-productively bound enzymes can disturb
productively bound ones). Some examples of the simulated progress
curves are shown in Figure 6. In short, one can conclude that the
initial slope of the progress curve over the 1:1 line indicates
that the enzyme prefers the labelled end. The slope is also
dependent on the enzyme processivity. An exo-acting enzyme with
end-preference for the labelled end and low processivity performs
only a few hydrolytic events before dissociat-ing from the
substrate, and the resulting progress curve for such an enzyme is
strongly convex with a high initial slope. If the processivity of
an enzyme is higher than the average DP of the substrate, a
productively bound enzyme will hydrolyse whole cellulose chains
without dissociating from the sub-strate, releasing one labelled
end group per hydrolysed cellulose chain. In this case, the
resulting progress curve is a straight line.
Experimental progress curves showed clearly distinguishable
patterns for GH6 and GH7 family cellobiohydrolases. The progress
curves for Cel7A and Cel7D were virtually identical, showing a
strong preference for the la-belled end, whereas the progress
curves for Cel6 enzymes were close to the 1:1 line, showing no
obvious end-preference. These data can be explained by combined
endo/exo activity. In order to clarify the end-preference of Cel6
enzymes, a substrate labelled at both ends is needed.
Influence of acid pre-treatment of bacterial cellulose on mode
of synergy between Cel7A (CBH I) and Cel7B (EG I) (Paper II) In
this study, the separate, sequential and simultaneous actions of
Cel7A (CBH I) and Cel7B (EG I) were investigated using well-defined
substrates derived from bacterial cellulose.
The high synergy between cellobiohydrolases and endoglucanases
is common among cellulolytic enzymes. According to the conventional
model, endoglucanases create new chain ends at more amorphous
substrate areas upon which cellobiohydrolases can start to
hydrolyse the cellulose proces-sively. The aim of our study was to
evaluate the synergistic mechanism be-tween Cel7A and Cel7B on
substrates with different physical properties, crystallinity and
DP.
The bacterial cellulose was hydrolysed in boiling 1 M
hydrochloric acid, the samples were withdrawn at certain time
points, neutralised and washed. From these samples, the
crystallinity index and DP were determined and these samples were
used as substrate.
Some characteristic properties of different cellulose samples
are shown in Table 3.
-
30
Table 3. Parameters of cellulose samples. BC, appearance common
to bacterial cellulose (large fibrous bundles); BMCC, appearance
common to bacterial micro-crystalline cellulose (suspension with a
milky consistency)
Property Sample
BC A25 A40 BMCC Time in 1M HCl, min 0 25 40 760 Weight loss by
HCl, % 0 0.26 1.0 5.2 Crystallinity index, % 87.9 90.3 91.3 92.4
DP, glucose unit 2620 212 151 114 Appearance BC BC BC BMCC Relative
activity Cel7A, % 37 69 100 66 Cel7B, % 95 53 57 100 Cel7A/Cel7B, %
100 56 50 33 Maximum synergy factor 7.8 4.1 2.3 1.7
Maximum effect of pretreatment with Cel7B
2.12 - 1.09 -
The most dramatic changes during cellulose hydrolysis by acid
occurred within the first 25 minutes of the acid treatment, when
the DP decreased more than ten-fold and the synergy by a factor of
two.
Similarly to the results obtained by Samejima et al. [Samejima
et al., 1997], the synergism between endo- and exoglucanases was
highest on un-treated bacterial cellulose and decreased during the
course of the acid hy-drolysis. The high synergy on BC is in
accordance with the classical endo/exo synergy model, where the
endoglucanases create new starting points for
cellobiohydrolases.
Interestingly, the relative activities of Cel7A on different
cellulose sam-ples did not reflect the changes in cellulose DP, as
would be expected if the number of chain ends were the limiting
factor for activity. Cel7A did not have maximal activity on the
cellulose sample of lowest DP (the sample with the highest
concentration of end-groups). Also, if the number of chain ends
were the main limiting factor, one should expect much lower
activity for Cel7A on BC than on BMCC, but the activities of Cel7A
differed only by a factor of two on these substrates.
The activity of endoglucanase Cel7B was found to be almost equal
on BC and BMCC. However, since we compared the release of
solubilised sugar and not the concentration of the end-groups on
the cellulose, the amount of released soluble sugar does not
necessarily reflect the total hydrolytic activ-ity of an
endoenzyme.
Pre-treatment of cellulose with Cel7B increased the activity of
Cel7A on BC about two-fold, whereas the effect of EG pre-treatment
on BMCC was neglible. Therefore, the synergy observed between EG
and CBH on BMCC cannot be explained by the sequential attack of EG
and CBH. We propose a
-
31
new, more interactive mechanism of synergy between EG and CBH on
BMCC, whereby EG “polishes” the cellulose surface from small
cellulose chains, supporting in this way the processive action of
CBH.
Endoglucanase 28 (Cel12A), a new Phanerochaetechrysosporium
cellulase (Paper III) Paper III describes the isolation and
characterisation of a new small en-doglucanase from Phanerochaete
chrysosporium.
Protein purification Phanerochaete chrysosporium strain K3 was
cultivated and the culture fil-trate was precipitated with ammonium
sulphate as described by Szabo et al. [Szabo et al., 1996]. After
dissolving and desalting the sample, the initial separation of the
proteins on DEAE-Sepharose was performed as described by Uzcatequi
et al. [Uzcategui et al., 1991b]. Material corresponding to pool B
was collected and after adjusting the pH to 5 and adding ammonium
sul-phate to 2 M concentration, the protein was applied to a Phenyl
Sepharose®CL-4B column equilibrated with 0.5 M ammonium sulphate in
0.1 M am-monium acetate buffer, pH 5. Proteins were eluted by two
linear gradients, first from 0.5 M ammonium sulphate in 0.1 M
ammonium acetate, pH 5, to 0.1 M ammonium acetate buffer, pH 5,
followed by a second gradient from 0.1 M ammonium acetate buffer,
pH 5 to the same buffer containing 50% ethylene glycol. EG activity
was eluted in the first gradient, corresponding to Cel5A (EG 38).
In the second gradient, two protein peaks were isolated
cor-responding to Cel7C (CBH 62) and Cel6A (CBH 50). The latter of
these coincided with the second peak of EG activity. This last peak
was collected, concentrated and applied to a Biogel P100 column
equilibrated with 0.05 M ammonium acetate buffer, pH 5. Two major
peaks were obtained, the first corresponding to Cel6A and the
second to an enzyme with high activity to-wards CMC.
Protein characterisation The purified enzyme showed a single
band on SDS/PAGE with an estimated molecular weight of 28 kDa and
an isoelectric point at pH 5.2. Deglycosyla-tion analysis together
with amino acid analysis showed that this protein con-tains 1-2
N-glycosylation sites. Peptide mapping after cleavage with the V8
protease showed a pattern clearly distinguishable from the
previously char-acterised P. chrysosporium endoglucanases Cel5A (EG
38) and Cel5B (EG 44) (Fig.7). A sequence obtained from a peptide
showed strong homology with endoglucanases belonging to the GH12
family.
-
32
Figure 7. Peptide mapping of P. chrysosporium endoglucanases on
SDS/PAGE using V8 protease. The peptide pattern of EG28 is totally
different from the other endoglucanases, indicating that this is
enzyme is not a fragment of either of the others.
The Cel12A protein sequence translated from P. chrysosporium
cel12a gene (protein data bank access code: AAU12276) [Vanden
Wymelenberg et al., 1993] shows about 40% homology with H. jecorina
Cel12A. The peptide isolated and sequenced in protein mapping
belongs apparently to the -helixof the protein located near to the
C-terminus. The catalytically important amino acids seem to be
conserved also in P. chrysosporium Cel12A.
Kinetic properties Cel12A did not bind to crystalline cellulose
to any detectable extent. In con-trast to many other cellulases,
pNPL was not a substrate for this enzyme. The comparison of the
catalytic constants on pNPC as substrate are shown in Table 4.
Table 4. Kinetic constants on p-nitrophenyl cellobioside.
Enzyme kcat, min-1 KM, mM kcat/KM, min-1/mM-1
Cel12A 1.61 12.6 0.13 Cel7B 87.4 3.46 2.5 * 103
Cel7A 0.146 97.4 * 10-3 1.6
Cel12A showed high activity towards CMC and amorphous cellulose,
with lower activity on xylan and glucan. Interestingly, Cel12A did
not produce soluble sugars using Avicel (a microcrystalline
cellulose derived from wood sample) as substrate. Cel12A showed
synergy both with Cel7A and Cel6A from H. jecorina using filter
paper and Avicel as substrate, but not with en-doglucanases.
-
33
Cel12A swells efficiently filter paper and disperses the filter
paper struc-ture, releasing short fibres. This phenomenon led us to
propose a physiologi-cal function for Cel12A enzymes in cellulose
degradation. We speculate that Cel12A has an important function in
an early stage of cellulose degradation by degrading amorphous
material located in the narrow regions between or on the surface of
the microfibrils that are sterically inaccessible to the larger
two-domain cellulases, leading to the swelling and separation of
microfibrils.
Figure 8. Suggested cellulolytic strategy in P. chrysosporium.
1) Crystalline mi-crofibrils are held together by more amorphous
material. 2) Cel12A diffuses into pores and nicks the amorphous
chains, thereby releasing the microfibrils. 3) “Classi-cal”
endoglucanases now have access to the substrate and can create
nicks. 4) Cello-biohydrolases degrade the crystalline cellulose
processively.
Conversion of oxygen species by cellobiose dehydrogenase (Paper
IV) Cellobiose dehydrogenase produces hydrogen peroxide in the
presence of an electron donor and molecular oxygen. Typically,
during the reaction course, cellobiose and molecular oxygen are
consumed in equimolar amounts, whereas the level of hydrogen
peroxide produced is significally lower and seems to reach
pseudo-steady state conditions. It has been shown that CDH can
degrade hydrogen peroxide under the same conditions [Henriksson et
al., 1993].
The aim of this paper was to find out what actually happens to
the hydro-gen peroxide produced by CDH and into which species it is
converted. We compared CDH to a well-characterised enzyme – glucose
oxidase (GOX).
-
34
This enzyme catalyses the oxidation of glucose into
gluconolactone by re-ducing molecular oxygen to hydrogen
peroxide:
GOX produced H2O2 in equimolar amounts to the cellobiose and O2
con-sumed (Fig. 9B), whereas the level of hydrogen peroxide
produced by CDH was much lower than would be expected from the
basic stochiometry (Fig. 9A).
Figure 9. Balance of saccharide oxidation by CDH (A) or glucose
oxidase (B). On A, empty circles show the molar amount of
cellobiose consumed at the given time point, filled triangles show
the consumed O2 and filled rectangles show hydrogen peroxide
formation. On B, empty circles show consumed glucose, filled
triangles consumed O2 and filled squares formed hydrogen
peroxide.
Catalase catalyses the decomposition of hydrogen peroxide into
water and O2 according the following scheme:
If all of the hydrogen peroxide produced by CDH is decomposed
either by catalase or by CDH into water and dioxygen, the total
balance between the cellobiose and oxygen consumption rates should
be 2:1.
Adding catalase to the reaction mixture indeed caused the ratio
between the consumption rates of cellobiose and O2 to become 2:1,
which supports the idea that hydrogen peroxide is the primary
product of the oxygen reduc-tion by CDH.
-
35
CDH is known to reduce ferric ions:
These divalent ferrous ions are known to decompose hydrogen
peroxide into a hydroxyl radical and a hydroxyl ion in Fenton’s
reaction according the following scheme:
The hydroxyl radicals produced are known to react with
cellulose, causing its depolymerisation. Kremer and Wood [Kremer
and Wood, 1992b] have proposed the following reaction scheme, where
HROH is a part of a saccha-ride:
Since the solutions inevitably contain traces of ferric ions, we
repeated the experiments using 1 mM desferrioxamine mesylate to
inactivate any traces of ferric ions. Under these conditions,
hydrogen peroxide was formed stoichiometrically from cellobiose and
O2.
The reduction rate of Fe3+ ions by CDH is much higher than for
O2 and if the Fe2+ ions produced by CDH are oxidised back by
hydrogen peroxide, this makes the reduction of Fe3+ ions by CDH
even more kinetically favourable.
Taking all results together, we can conclude that hydrogen
peroxide is a primary product of cellobiose oxidation by CDH under
aerobic conditions and that hydrogen peroxide is not decomposed
further when traces of ferrous ions are eliminated from the system.
The inevitable traces of metal ions pre-sent in standard test
conditions and in nature seem to be sufficient for an enhanced
Fenton reaction, where CDH produces both components needed- reduced
metal ions and hydrogen peroxide.
-
36
o-nitrophenyl cellobioside as an active site probe for family 7
cellobiohydrolases (Paper V) Nitrophenyl glycosides are widely used
chromogenic substrates for glycosyl hydrolases, since they contain
good leaving groups with favourable spectral properties, which make
the reaction easy to monitor. In many cases, how-ever, the kcat
observed has been orders of magnitude lower than that observed for
the cleavage of oligosaccharides. For cellulase studies, mostly
p-nitro-phenyl cellobioside and lactoside have been used as model
substrates. Inter-estingly, many of the GH12 family endoglucanases
can readily hydrolyse oNPC, some of them being more than 40 times
active on oPNC as compared to pNPC [Sandgren et al., 2005]. In this
study, interactions between o-nitro-phenyl cellobioside (oNPC) and
GH7 family cellobiohydrolases were inves-tigated.
Table 5. Comparison of kinetic constants of Cel7A and Cel7D on
oNPC and pNPC.
Enzyme / substrate kcat, s-1 KM, µM kcat /KM, s-1M-1
Cel7A / oNPC (66±15)*10-6 7.0±4.5 9.5 Cel7A / pNPC 0.0026±0.0001
26±3 100 Cel7D / oNPC 0.015±0.002 3200±100 4.6 Cel7D / pNPC
0.046±0.0021 1300±160 35
For both Cel7A and Cel7D, the observed rate constants for oNPC
are lower than those for pNPC, as shown in Table 5. Moreover, oNPC
was found to be an inhibitor of both enzymes. Since the active site
of the cellu-lases consists of multiple binding sites for sugar
units, it is likely that these model substrates can bind to the
active site of the enzyme in more than one mode, some of these
being non-productive. In certain cases where some binding modes may
coexist, cooperativity can occur for productive or non-productive
binding. This may result in devations from the standard
Micha-elis-Menten kinetics, which has been observed for certain
cellobiohydrolase-model substrate combinations (data not
published). The modelling data show that apart from the competition
between productive and non-productive binding modes, there may also
be some sterical constraints imposed by the o-nitro group that give
a real decrease in the kcat.We found that oNPC quenches the natural
fluorescence of Cel7A and Cel7D. The fluorescence quenching
increased upon addition of the oNPC and followed a Langmuir
isotherm. The fluorescence of the Cel7A could be recovered by
adding cellobiose. The dissociation constant for cellobiose
determined from competitive binding experiments was in fair
agreement with that obtained earlier [Claeyssens et al., 1989;
Henriksson et al., 1999b], suggesting that oNPC binds to the active
site of the enzyme with a strong preference for the +1, +2 subsites
for which cellobiose is known to have
-
37
highest affinity [Divne et al., 1998] The fluorescence of Cel7D
did not re-cover upon adding up to 1 mM cellobiose, suggesting that
oNPC may bind possibly also closer to the entrance of the cellulose
binding tunnel.
The competitive binding approach was used to study the binding
of the cellobiose to the catalytically inactive Cel7A mutants. The
catalytic group Cel7A mutants D214N, E212Q and E217Q all bind oNPC
with an affinity very similar to that observed for the wild type
enzyme (Table 6), strongly suggesting that it binds at the same
position. The dissociation constants thus determined for cellobiose
were in the same range as that for the wild type but generally
somewhat lower. A possible explanation for the difference observed
could be that the lower charge density found in the active site of
the mutants is favourable for the binding of a neutral ligand, such
as cello-biose, especially since the carboxylate-amide substitution
allows hydrogen bond interactions to be retained.
Table 6. Binding constants for o-nitrophenyl cellobioside and
cellobiose for Cel7D, Cel7A and its catalytically inactive
mutants.
Enzyme Kd for oNPC, µM Kd for cellobiose, µM
Cel7A wt 7.4±0.4 23±4 D214N 7.1±0.7 8.9±1.1E212Q 4.7±0.4
8.1±0.3E217Q 3.9±0.4 13.5±3Cel7D 110±10 -
-
38
Conclusions
GH7 family cellobiohydrolases Cel7A and Cel7D degrade
crystalline cellu-lose processively with a strong preference for
the reducing end.
There is a common pattern of hydrolysis for GH7 family
cellobiohydrolases and another clearly distinguishable pattern for
GH6 family cellobiohy-drolases.
Synergistic action between cellobiohydrolases and endoglucanases
cannot be explained using only the classical endo-exo model. A new
concept whereby endoglucanase “polishes” the cellulose surface is
proposed.
A new endoglucanase, Cel12A from P. chrysosporium, was isolated
and characterised. This enzyme may have a role in disintegrating
larger struc-tures.
Hydrogen peroxide produced by CDH is decomposed via an enhanced
Fen-ton’s reaction.
o-nitrophenyl cellobioside has a potential as an active site
probe for GH7 family cellobiohydrolases.
-
39
Summary in Swedish
Cellulosa är den dominerande kolhydraten som produceras av
växter, och kan i träd utgöra över 50% av vikten. Kemiskt utgörs
cellulosa, liksom stär-kelse, av långa kedjor av hopkopplade
glukos-(druvsocker)molekyler. Skill-naden mellan cellulosa och
stärkelse bygger väsentligen på att glukosenhe-terna är hopkopplade
på olika sätt. Den nedbrytning av cellulosabaserad biomassa som
åstadkommes av mikroorganismer är följaktligen en mycket viktig
faktor i det totala kolkretsloppet i naturen.
I anslutning till de hela tiden stigande oljepriserna har ett
ökande antal länder uppmärksammat problemet med fossila bränslen,
och intresset för att hitta alternativa förnyelsebara energikällor
som kan ersätta olja är större än någonsin. Sveriges regering har
beslutat att skapa förutsättningar för att bryta Sveriges beroende
av fossila bränslen till år 2020. Biologiskt styrda proces-ser
förefaller lovande för energiomsättning, i synnerhet då för
omvandling av lignocellulosamaterial till bekvämare bränslen.
Därmed är en fördjupad för-ståelse av mekanismerna för enzymatisk
cellulosanedbrytning av högsta betydelse och kommer att bidra till
förbättrade processer för storskalig och miljövänlig biologisk
nedbrytning av cellulosabaserad biomassa, vilket ännu är alltför
kostsamt för att kunna konkurrera med fossila bränslen under
rim-liga villkor.
I den här avhandlingen ligger fokus på cellulosanedbrytande
enzymer från två mögelsvampar, nämligen Hypocrea jecorina, en
rötsvamp som lever i tropisk förna och där bryter ner
cellulosarester samt Phanerochaete chrysos-porium, en
tränedbrytande s.k.vitrötesvamp som har isolerats från flishögar
och är kapabel att bryta ned alla komponenter i ved. Studien har
omfattat såväl undersökningar av separata enzymer som deras förmåga
att samverka synergistiskt, dvs att effekten när enzymerna arbetar
tillsammans blir större än summan av deras individuella
effekter.
I det första delarbetet har metoder utvecklats för att avgöra
från vilken ände av cellulosakedjorna som enzymerna fördrar att
arbeta och sedan an-vänts för att funktionellt karakterisera ett
antal viktiga enzymer. Här kunde vi avgöra att de hos svampar ofta
förekommande enzymerna från struktur-familj 7 startade
nedbrytningen av cellulosakedjorna från den s.k. reduceran-de änden
och sedan kunde frisätta ett stort antal lösliga sockermolekyler
utan att lämna kedjan, vilket kallas processiv funktion.
I nästa arbete studerades speciellt samverkan mellan enzymer med
delvis olika sätt att attackera cellulosakedjorna. Ett klassiskt
sätt för samverkan är
-
40
här den så kallade endo-exosynergismen, där en typ av enzym har
förmågan att klippa en lång kedja internt till ett fåtal kortare
kedjor, som sedan stegvis bryts ned av enzymer som endast verkar
från ändarna och därmed gynnas av när flera fria kedjeändar skapas.
Våra resultat tydde dock på att samverkan även följde andra
principer, till exempel att ett enzym "städar bort" korta rester av
kedjor, vilket därmed gör andra kedjor mera tillgängliga.
En annan aspekt på samverkan av enzymer belyses av ett nytt
endogluka-nas med ovanligt små dimensioner, som isolerades från P.
chrysosporiumoch studerades funktionellt. Dess släktskap med andra
kända enzymer fast-slogs också. Detta enzym var jämförelsevis
ineffektivt i sin förmåga att fri-sätta lösligt socker från
cellulosa, men visade stor förmåga att luckra upp strukturer i t.ex
filterpapper. Vi tolkade detta som att enzymet har som sin främsta
uppgift att försvaga kontakterna mellan mikrofibrerna i naturliga
cellulosastrukturer så att andra enzymer lättare kommer åt att
genomföra en fullständig nedbrytning. En sådan uppluckrande roll
skulle gynnas av att enzymet är litet, och lättare kan ta sig in i
en från början kompakt struktur, något som är svårt/omöjligt för de
flesta andra cellulosanedbrytande enzy-mer.
Hos P. chrysosporium och många andra vitrötesvampar hittar man
förut-om de enzymer som bryter ner cellulosakedjorna, också ett
enzym som kal-las cellobiosdehydrogenas vilket i stället styr
elektronöverföringar, främst från sockret cellobios till ett brett
spektrum av mottagarmolekyler. Intressant nog har man inte lyckats
fastställa den grundläggande biologiska rollen för detta enzym. En
föreslagen roll var att enzymet skulle producera väteperox-id,
vilket motsades av andra resultat som antydde att sådan inte kunde
pro-ducerades. Vi kunde, bland annat med hjälp av syrgaselektrod,
påvisa att enzymet verkligen kunde producera väteperoxid, men att
det också, under realistiska förhållanden med spårmängder av
järnjoner närvarande, också hade förmågan att förbruka väteperoxid,
vilket kunde förklara de tidigare motstridiga resultaten.
Ett avslutande arbete är inriktat på en del speciella
mätresultat som kan fås när enzymer med uppgift att bryta ned långa
kedjemolekyler studeras med hjälp av små, lösliga modellsubstanser.
Resultaten kompliceras av att dessa substanser ofta kan bindas till
enzymet på flera sätt än det som ligger till grund för en reaktion.
Många gånger kan olika bindningslägen leda till ömsesidig
utestängning, dvs bara ett av bindningssätten är möjligt, medan
andra bindningssätt kan vara samtidiga. Vi har med olika metoder
studerat sådana effekter, men också visat att en substans,
orto-nitrofenylcellobiosid, som blott mycket långsamt bryts ned av
enzymet, vid bindning förändrar enzymets förmåga att fluorescera,
så att den också kan användas som en "reportermolekyl" för att
studera bindning av andra molekyler som inte ger några lätt
observerbara effekter, men som konkurrerar med reportermoleky-len
om att bindas på samma ställe.
-
41
Acknowledgements
This study was carried out at the Department of Biochemistry and
Organic Chemistry, Uppsala University. This work was supported by
the Swedish Research Council for Engineering Sciences, the Swedish
Natural Science Research Council, The Royal Academy of Sciences,
the Swedish Institute, the Swedish Pulp and Paper Research
Institute and Wood Ultrastructure Research Center (WURC).
I would like to express my sincere gratitude to all the people
who have sup-ported me throughout this long journey.
My very special gratitude belongs to Gunnar Johansson, my
supervisor, for Your constant support and encouragement (not
mentioning Your natural talent to generate creative ideas) during
all these years, especially for allow-ing me to compromise between
science and family.
Thank You, all my co-supervisors: Göran Pettersson, The grand
father of the “cellulasgruppen”, for making
our lab a warm and friendly place, and of course, for sharing
Your enormous knowledge about cellulases and nature in general;
Late Veljo Sild, for teaching me to look into the fascinating
world of cel-lulases in a unique way, being a friend, teacher and
supervisor in one person;
Jerry Ståhlberg, for being always so positive and helpful.
Special thanks for Your contribution to the pending manuscript;
I would like to thank also: Professor Bengt Mannervik, for
always finding time for discussions. Priit Väljamäe, for Your
critical mind, discussions and comments. My closest co-authors,
Andres Salumets, Gunnar Henriksson, Hongbin
Henriksson, Bert Pettersson, Roland Isaksson, Mikael Nilsson for
fruitful collaboration.
David Eaker for linguistic revisions of my manuscripts and this
thesis. All of the people at our department for creating nice and
friendly envi-
ronment. Special thanks to Per-Axel for finding solutions to ALL
technical problems in Your smart way. I am deeply impressed of Your
skills. The people at the amino-acid lab for Your excellent work
(as always, indeed). Lilian, Your assistance at the course lab is
indispensable. Gun and
-
42
Brigitta T., it has been always a pleasure to assist in teaching
at Your courses. All other PhD students at the department for all
fun and support. Present and former members of our group,
especially Jing, my room-mate, for Your friendship and all stories
about Chinese culture; Lisa and Caroline, for pleasant time
together after moving into new lab; Lars, Istvan and Gabriel, for
truly great company.
My fellow-Estonians in Uppsala, especially 2xAnneli, Mart,
Sirje, 2xTanel, Taavo, Teet and Triin. My friends here and there.
Special thanks goes to Linda for keeping things together during my
stays in Estonia.
My mother and father, for the foundation. My sister and brother,
for al-ways being there for me.
Finally, my dearest: Toomas, for all Your support and love and
my children, Tuuli and Siim, who added a whole new dimension to my
life.
-
43
References
Abuja PM, Pilz I, Claeyssens M, Tomme P (1988) Domain structure
of cellobiohy-drolase II as studied by small angle X-ray
scattering: Close resemblance to cel-lobiohydrolase I. Bioch
Biophys Res Comm 156: 180-185
Abuja PM, Pilz I, Tomme P, Claeyssens M (1989) Structural
changes in cellobiohy-drolase I upon binding of a macromolecular
ligand as evident by SAXS investi-gations. Bioch Biophys Res Comm
165: 615-623
Armand S, Drouillard S, Schulein M, Henrissat B, Driguez H
(1997) A bifunctional-ized fluorogenic tetrasaccharide as a
substrate to study cellulases. J Biol Chem272: 2709-2713
Atalla RH, Vanderhart DL (1984) Native cellulose: A composite of
2 distinct crys-talline forms. Science 223: 283-285
Baker AA, Helbert W, Sugiyama J, Miles MJ (1997) High-resolution
atomic force microscopy of native Valonia cellulose I
microcrystals. J Struct Biol 119: 129-138
Barr BK, Hsieh YL, Ganem B, Wilson DB (1996) Identification of
two functionally different classes of exocellulases. Biochemistry
35: 586-592
Bayer EA, Belaich JP, Shoham Y, Lamed R (2004) The cellulosomes:
multienzyme machines for degradation of plant cell wall
polysaccharides. Annu Rev Micro-biol 58: 521-554
Beguin P, Aubert JP (1994) The biological degradation of
cellulose. FEMS Micro-biol Rev 13: 25-58
Bhikhabhai R, Johansson G, Pettersson G (1984) Isolation of
cellulolytic enzymes from Trichoderma reesei QM 9414. J Appl
Biochem 6: 336-345
Biely P, Vrsanska M, Claeyssens M (1993) Mode of action of
Trichoderma reesei -1,4-glucanases on cellooligosaccharides. In
Proceedings of the second TRICEL symposium on Trichoderma reesei
cellulases and other hydrolases, Suominen P, Reinikainen T (eds) pp
99-108. Foundation for Biotechnical and Industrial Fer-mentation
Research: Helsinki
Biely P, Vrsanska M, Claeyssens M (1991) The
endo-1,4-beta-glucanase I from Trichoderma reesei: Action on
beta-1,4-oligomers and polymers derived from D-glucose and
D-xylose. Eur J Biochem 200: 157-163
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: 1444-1452
Boraston AB, Bolam DN, Gilbert HJ, Davies GJ (2004)
Carbohydrate-binding mod-ules: fine-tuning polysaccharide
recognition. Biochem J 382: 769-781
Brown RM, Jr., Saxena IM (2000) Cellulose biosynthesis: A model
for understand-ing the assembly of biopolymers. Plant Physiology
and Biochemistry 38: 57-67
Carrard G, Koivula A, Soderlund H, Beguin P (2000)
Cellulose-binding domains promote hydrolysis of different sites on
crystalline cellulose. Proc Natl Acad Sci U S A 97: 10342-10347
-
44
Chanzy H, Henrissat B (1985) Unidirectional degradation of
Valonia cellulose mi-crocrystals subjected to cellulase acition.
Febs Lett 184: 285-288
Claeyssens M, Tomme P, Brewer CF, Hehre EJ (1990) Stereochemical
course of hydrolysis and hydration reactions catalysed by
cellobiohydrolases I and II from Trichoderma Reesei. Febs Lett 263:
89-92
Claeyssens M, van Tilbeurgh H, Tomme P, Wood TM, McRae SI (1989)
Fungal cellulase systems: Comparison of the specificities of the
cellobiohydrolases iso-lated from Penicillium pinophilum and
Trichoderma reesei. Biochem J 261:819-826
Cohen R, Suzuki MR, Hammel KE (2005) Processive endoglucanase
active in crys-talline cellulose hydrolysis by the brown rot
basidiomycete Gloeophyllum tra-beum. Appl Environ Microbiol 71:
2412-2417
Converse AO, Matsuno R, Tanaka M, Taniguchi M (1988) A model of
enzyme adsorption and hydrolysis of microcrystalline cellulose with
slow deactivation of the adsorbed enzyme. Biotechnology and
Bioengineering 32: 38-45
Coutinho PM, Henrissat B (1999) Carbohydrate-active enzymes: an
integrated data-base approach. In Recent Advances in Carbohydrate
Bioengineering, Davies G, Henrissat B, Svensson B (eds) pp 3-12.
The Royal Society of Chemistry: Cam-bridge
Covert SF, Bolduc J, Cullen D (1992a) Genomic organization of a
cellulase gene family in Phanerochaete chrysosporium [published
erratum appears in Curr Genet 1993;23(4):374]. Curr Genet 22:
407-413
Covert SF, Vanden Wymelenberg A, Cullen D (1992b) Structure,
organization, and transcription of a cellobiohydrolase gene cluster
from Phanerochaete chryso-sporium. Appl Environ Microbiol 58:
2168-2175
Davies GJ, Gloster TM, Henrissat B (2005) Recent structural
insights into the ex-panding world of carbohydrate-active enzymes.
Curr Opin Struct Biol 15: 637-645
Davies GJ, Tolley SP, Henrissat B, Hjort C, Schulein M (1995)
Structures of oligo-saccharide-bound forms of the endoglucanase V
from Humicola insolens at 1.9 A resolution. Biochemistry 34:
16210-16220
Davies GJ, Wilson KS, Henrissat B (1997) Nomenclature for
sugar-binding subsites in glycosyl hydrolases. Biochem J 321:
557-559
Davison A, Blaxter M (2005) Ancient origin of glycosyl hydrolase
family 9 cellu-lase genes. Mol Biol Evol 22: 1273-1284
Din N, Damude HG, Gilkes NR, Miller RC, Jr., Warren RAJ, Kilburn
DG (1994) D-1-C-x revisted: Intramolecular synergism in a
cellulase. Proc Natl Acad Sci U S A 91: 11383-11387
Din N, Gilkes NR, Tekant B, Miller RC, Jr., Warren AJ, Kilburn
DG (1991) Non-hydrolytic Disruption of Cellulose Fibres by the
binding Domain of a Bacterial Cellulose. Bio/Technology 9:
1096-1099
Divne C, Ståhlberg J, Reinikainen T, Ruohonen L, Pettersson G,
Knowles JKC, Teeri TT, Jones TA (1994) The three-dimensional
crystal structure of the cata-lytic core of cellobiohydrolase I
from Trichoderma reesei. Science Washington265: 524-528
Divne C, Ståhlberg J, Teeri TT, Jones TA (1998) High-resolution
crystal structures reveal how a cellulose chain is bound in the 50
angstrom long tunnel of cello-biohydrolase I from Trichoderma
reesei. J Mol Biol 275: 309-325
Doblin MS, Kurek I, Jacob-Wilk D, Delmer DP (2002) Cellulose
biosynthesis in plants: from genes to rosettes. Plant Cell Physiol
43: 1407-1420
-
45
Esteghlalian AR, Srivastava V, Gilkes NR, Kilburn DG, Warren RA,
Saddle JN (2001) Do cellulose binding domains increase substrate
accessibility? Appl Bio-chem Biotechnol 91-93: 575-592
Fägerstam L, Pettersson LG (1980) The 1-4- beta-glucan
cellobiohydrolases of Trichoderma reesei QM 9414. A new type of
cellulolytic synergism. Febs Lett119: 97-100
Fang J, Liu W, Gao PJ (1998) Cellobiose dehydrogenase from
Schizophyllum com-mune: purification and study of some catalytic,
inactivation, and cellulose-binding properties. Arch Biochem
Biophys 353: 37-46
Foreman PK, Brown D, Dankmeyer L, Dean R, Diener S, Dunn-Coleman
NS, Goe-degebuur F, Houfek TD, England GJ, Kelley AS, Meerman HJ,
Mitchell T, Mitchinson C, Olivares HA, Teunissen PJ, Yao J, Ward M
(2003) Transcrip-tional regulation of biomass-degrading enzymes in
the filamentous fungus Trichoderma reesei. J Biol Chem 278:
31988-31997
Gardner KH, Blackwell J (1974) The structure of native
cellulose. Biopolymers 13:1975-2001
Gilad R, Rabinovich L, Yaron S, Bayer EA, Lamed R, Gilbert HJ,
Shoham Y (2003) CelI, a noncellulosomal family 9 enzyme from
Clostridium thermocellum, is a processive endoglucanase that
degrades crystalline cellulose. J Bacteriol 185:391-398
Gilkes NR, Henrissat B, Kilburn DG, Miller RC, Warren AJ (1991)
Domains in Microbial beta-1,4-Glycanases: Sequence Conservation,
function, and Enzyme Families. Microbiol Rev 55: 303-315
Gilkes NR, Warren RAJ, Miller RC, Jr., Kilburn DG (1988) Precise
excision of the cellulose binding domains from two Cellulomonas
fimi cellulases by a homolo-gous protease and the effect of
catalysis. J Biol Chem 263: 10401-10407
Gruno M, Väljamäe P, Pettersson G, Johansson G (2004) Inhibition
of the Tricho-derma reesei cellulases by cellobiose is strongly
dependent on the nature of the substrate. Biotechnol Bioeng 86:
503-511
Hallberg BM, Bergfors T, Backbro K, Pettersson G, Henriksson G,
Divne C (2000) A new scaffold for binding haem in the cytochrome
domain of the extracellular flavocytochrome cellobiose
dehydrogenase. Structure 8: 79-88
Hallberg BM, Henriksson G, Pettersson G, Divne C (2002) Crystal
structure of the flavoprotein domain of the extracellular
flavocytochrome cellobiose dehydro-genase. J Mol Biol 315:
421-434
Harrington, J. J. MicroFibril Angle in Wood. Butterfield, B. G.
1998. New Zealand, IAWA/IUFRO.
Hayashi N, Sugiyama J, Okano T, Ishihara M (1997) Selective
degradation of the cellulose I-alpha component in Cladophora
cellulose with Trichoderma viridecellulase. Carbohydr Res 305:
109-116
Hayn, M., Klinger, R, and Esterbauer, H. Isolation and partial
characterization of a low molecular weight endoglucanase from
Trichoderma reesei. Suominen, P. and Reinikainen, T. Proceedings of
the second TRICEL symposium on Tricho-derma reesei cellulases and
other hydrolases. 147-151. 1993. Helsinki, Founda-tion for
Biotechnical and Industrial Fermentation Research.
Heiner AP, Sugiyama J, Teleman O (1995) Crystalline cellulose
I-alpha and I-beta studied by molecular dynamics simulation.
Carbohydr Res 273: 207-223
Helle SS, Duff SJB, C