Joint X-ray crystallographic and molecular dynamics study of cellobiohydrolase i from Trichoderma harzianum: Deciphering the structural features of cellobiohydrolase catalytic activity

Article abstract of DOI:10.1111/febs.12049

Aiming to contribute toward the characterization of new, biotechnologically relevant cellulolytic enzymes, we report here the first crystal structure of the catalytic core domain of Cel7A (cellobiohydrolase I) from the filamentous fungus Trichoderma harzianum IOC 3844. Our structural studies and molecular dynamics simulations show that the flexibility of Tyr260, in comparison with Tyr247 from the homologous Trichoderma reesei Cel7A, is enhanced as a result of the short side-chains of adjacent Val216 and Ala384 residues and creates an additional gap at the side face of the catalytic tunnel. T. harzianum cellobiohydrolase I also has a shortened loop at the entrance of the cellulose-binding tunnel, which has been described to interact with the substrate in T. reesei Cel7A. These structural features might explain why T. harzianum Cel7A displays higher k_cat and K_m values, and lower product inhibition on both glucoside and lactoside substrates, compared with T. reesei Cel7A. 2012 The Authors Journal compilation

Full text of DOI:10.1111/febs.12049

Joint X-ray crystallographic and molecular dynamics study
of cellobiohydrolase I from Trichoderma harzianum:
deciphering the structural features of cellobiohydrolase
catalytic activity
Larissa C. Textor1,†, Francieli Colussi1,†, Rodrigo L. Silveira2,†, Viviane Serpa1, Bruno L. de Mello1,
1
3
4
2
Joao Renato C. Muniz , Fabio M. Squina , Nei Pereira Jr , Munir S. Skaf and Igor Polikarpov1
~
~
1 Instituto de Fısica de Sao Carlos, Universidade de Sao Paulo, Sao Carlos, SP, Brazil
~
~
ꢀ
2 Institute of Chemistry, State University of Campinas – UNICAMP, Campinas, SP, Brazil
ꢀ
3 Laboratorio Nacional de Ciencia e Tecnologia do Bioetanol (CTBE), Centro Nacional de Pesquisa em Energia e Materiais (CNPEM),
^
Campinas, SP, Brazil
ꢀ
4 Centro de Tecnologia, Escola de Quımica, Laboratorio de Desenvolvimento de Bioprocessos (LaDeBio), Universidade Federal do Rio de
ꢀ
Janeiro, RJ, Brazil
Keywords
Aiming to contribute toward the characterization of new, biotechnological-
ly relevant cellulolytic enzymes, we report here the first crystal structure of
biomass; cellobiohydrolase; enzymatic
hydrolysis; molecular dynamics; X-ray
the catalytic core domain of Cel7A (cellobiohydrolase I) from the filamen-
tous fungus Trichoderma harzianum IOC 3844. Our structural studies and
molecular dynamics simulations show that the flexibility of Tyr260, in com-
structure
Correspondence
parison with Tyr247 from the homologous Trichoderma reesei Cel7A, is
enhanced as a result of the short side-chains of adjacent Val216 and
Ala384 residues and creates an additional gap at the side face of the cata-
lytic tunnel. T. harzianum cellobiohydrolase I also has a shortened loop at
the entrance of the cellulose-binding tunnel, which has been described to
interact with the substrate in T. reesei Cel7A. These structural features
might explain why T. harzianum Cel7A displays higher kcat and Km values,
and lower product inhibition on both glucoside and lactoside substrates,
compared with T. reesei Cel7A.
~
I. Polikarpov, Instituto de Fısica de Sao
ꢀ
~
Carlos, Universidade de Sao Paulo, Av.
~
Trabalhador Saocarlense 400, Sao Carlos,
SP, 13560-970, Brazil
Fax: +55 16 3373 9881
Tel: +55 16 3373 8088
E-mail: ipolikarpov@ifsc.usp.br
†Equally contributing authors.
(Received 21 August 2012, revised 18
October 2012, accepted 19 October 2012)
doi:10.1111/febs.12049
Introduction
It is now generally accepted that environmental
changes have become a serious problem worldwide
and might drastically affect the Earth’s climate, with
severe consequences for humankind if not addressed
properly. Several studies have associated global warm-
ing with increasing levels of atmospheric CO2, gener-
ated, among other sources, by the burning of fossil
fuels [1–3]. Hence, extensive scientific efforts have been
invested toward utilization and recovery of natural
renewable resources for bioenergy production [4].
Abbreviations
BMCC, bacterial microcrystalline cellulose; CBM, cellulose binding module; CCD, catalytic core domain; CMC, carboxymethyl cellulose;
cNPL, 2-chloro-4-nitrophenyl-b-lactoside; DP, degree of polymerization; GH, glycosyl hydrolase; Ma_Cel7B, Melanocarpus albomyces Cel7B;
MD, molecular dynamics; Pc_Cel7D, GH family 7 cellobiohydrolase from Phanerochaete chrysosporium; PDB, Protein Data Bank; pNPC,
p-nitrophenyl-D-cellobioside; Te_Cel7A, GH family 7 cellobiohydrolase from Talaromyces emersonii; Th_Cel7A, Trichoderma harzianum
cellobiohydrolase I; Tr_Cel7A, catalytic core domain of Cel7A (CBHI) from Trichoderma reesei.
56
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
Environment-friendly alternatives, such as wind, solar
power and biomass, represent important sources of
renewable energy. As lignocellulosic material comprises
most of the biomass on Earth, biodegradation of cellu-
lose and hemicellulose has become one focus of
research on renewable liquid biofuels. Hydrolysis of
cellulose represents a considerable challenge because
the substrate is insoluble and recalcitrant against depo-
lymerization and enzymatic action.
accessible reducing and nonreducing ends of cellulose
chains are attacked by types I and II cellobiohydrolas-
es, respectively. The cellobiose units released are then
hydrolyzed into glucose units by b-glucosidases [8].
Cellulases can be classified into glycosyl hydrolase
(GH) families based on their amino-acid sequence and
folding similarities [9]. Together with some endoglu-
canases and chitosanases, type I cellobiohydrolases
belong to the GH family 7 [10]. The exoglucanase,
Cel7A, is the most important single enzyme compo-
nent for cellulose depolymerization and conversion
into cellobiose in fungi [11]. It is considered to be the
key enzyme in the hydrolysis process because it is able
to hydrolyze crystalline cellulose extensively, albeit at
a slow rate [12]. Moreover, Cel7A is the predominant
enzyme secreted by the fungus Trichoderma sp. when
grown under conditions suitable for the production of
cellulases, reaching up to 60% of the total secreted
protein [13].
The biotechnological use of filamentous fungi is a
growing field with promising practical applications in
bioenergy and biotechnology because these microor-
ganisms are known to express several hemicellulose-
and cellulose-degrading enzymes at high yields [4]. A
wide range of industrial technologies use cellulases,
mainly in starch processing, grain alcohol fermenta-
tion, malting and brewing, extraction of fruit and veg-
etable juices, pulp, paper and textile [5,6]. Future
promising applications for cellulases include the pro-
duction of fuel ethanol from the abundant lignocellu-
losic biomass through the enzymatic breakdown of
cellulose [7].
In general, the cellulases produced by T. reesei are
glycoproteins that frequently contain both O- and
N-linked glycosylation sites [14]. Most cellulases dis-
play a similar structural organization, consisting of
two distinct functional and structural domains
(Fig. 1A): a cellulose-binding domain and an N-gly-
cosylated catalytic core domain (CCD) [15]. The
domains are joined together by a long and highly
O-glycosylated linker region. Structural studies have
shown that N-glycosylation of Cel7A from T. reesei
strains QM9414 and ALKO2877 (derived from
QM9414) is present as single N-acetylglucosamine
modification on three (Asn45, Asn270 and Asn384)
out of four potential sites located in the protein cata-
lytic domain [16]. The occurrence of such modifica-
tions depends on several factors, including strain and
The well-studied filamentous fungus Trichoder-
ma reesei expresses a number of cellulases that can be
broadly divided into three types: endoglucanases
(endo-1,4-b-D-glucanases); cellobiohydrolases or exo-
glucanases (exo-1,4-b-D-glucanases); and b-glucosidases
(1,4-b-D-glucosidases) [5,8]. The classical view of cellu-
lose hydrolysis is described as a simultaneous and syn-
ergistic action of these three types of enzymes. The
hydrolysis process starts with endoglucanases that ran-
domly hydrolyze internal b-1,4-glycosidic linkages in
the amorphous region of cellulose microfibrils, reduc-
ing significantly the degree of substrate polymerization
and creating new terminal ends. Simultaneously, the
Fig. 1. Structure of cellobiohydrolase I
from Trichoderma harzianum. (A)
Structural organization. Predicted N-
glycosylation sites are identified (N) and
the O-glycosylated linker is shown. (B)
Ribbon diagram of the crystallographic
structure of the CCD from Th_Cel7A
highlighting the loops of the tunnel.
Observed glycosylation sites at N126,
N283 and N397 are displayed.
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
57

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
culture conditions, and generates several enzyme iso-
forms with similar catalytic and adsorption properties
[17,18].
compared with those produced by T. reesei. Moreover,
T. harzianum secretes an effective and well-balanced
enzymatic system that is able to completely hydrolyze
cellulosic substrates into monomeric glucose [27].
T. harzianum has a higher efficiency of hydrolysis com-
pared with T. reesei, owing to the higher b-glucosidase
and xylanase activities; consequently this represents a
potential advantage for T. harzianum in terms of more
efficient biomass deconstruction.
X-ray structures of four GH family 7 cellobiohydro-
lases have been published to date: two correspond to
the CCDs of Trichoderma reesi Cel7A [Protein Data
Bank (PDB) entry 1CEL) [19] and of Phanerochae-
te chrysosporium Cel7D (PDB entry 1GPI) [20],
whereas two other structures represent the single-
domain thermostable filamentous fungi cellobiohydro-
lases Cel7A from Talaromyces emersonii (PDB id code
1Q9H) [21] and Cel7B from Melanocarpus albomyces
(PDB id code 2RFW) [22]. In addition, the crystallo-
graphic structure of cellobiohydrolase Cel7A from
Heterobasidion annosum is available at the PDB under
the code 2XSP, but has not yet been published. Fur-
thermore, solution structures of several cellulose bind-
ing modules (CBMs) have been determined, including
that of T. reesei Cel7A (PDB entry code 1CBH [23]).
Crystallographic structures of the full-length pro-
teins have not yet been determined, presumably
because of the presence of the highly mobile linker
region, which hampers crystallization. At the same
time, CCD structures of T. reesei cellobiohydrolase I
(Tr_Cel7A) have been determined in complex with a
number of oligosaccharides and ligands [24]. The CCD
fold consists of two large antiparallel b-sheets, which
form a b-sandwich tertiary structure. An elongated cel-
lulose-binding tunnel surrounded by disulfide-bridged
loops is formed along the concave side of the b-sand-
wich. The tunnel encloses the active site at its far end
where the catalytic residues aspartate and both gluta-
mates are located [19]. The presence of the elongated
tunnel mediates the enzyme specificity that is achieved
by a substantial number of interactions between the
cellulose chain ends and the residues of the tunnel [25].
The polysaccharide chain slides through the substrate-
binding tunnel and every second glycosidic bond is
correctly presented to the catalytic apparatus, which
explains why these enzymes progressively liberate
disaccharide units (cellobiose) from the reducing end
of a cellulose chain [19]. After enzymatic cleavage, the
cellobiose is found bound to sites +1 and +2 near the
exit of the tunnel [19]. The catalytic triad of carboxyl-
ate groups is strictly conserved among the enzymes of
GH family 7 [26].
In order to investigate, in greater detail, the enzymes
present in the cellulolytic complex secreted by T. har-
zianum, we studied the enzymatic kinetics and inhibi-
tion by cellobiose of Cel7A from T. harzianum and
compared these data with enzymatic parameters of the
homolog T. reesei Cel7A (81% amino acid sequence
identity with T. harzianum Cel7A). We also deter-
mined the crystallographic structure of T. harzianum
cellobiohydrolase I (Th_Cel7A), and performed molec-
ular dynamics (MD) computer simulations on these
structures to obtain more detailed information on the
structural features of the enzymes and their dynamics.
Our analyses provide insights into the molecular basis
of the observed differences in the enzymatic activity
and kinetics between Tr_Cel7A and Th_Cel7A.
Results
Kinetic constants and enzyme specificity for
different soluble and insoluble substrates
Assuming Michaelis–Menten kinetics of the full-length
T. harzianum Cel7A, the experimentally determined
Km and kcat values were, respectively, 3.4 mM and
23.2 minÀ1 for p-nitrophenyl-D-cellobioside (pNPC)
and 3.7 mM and 44.6 minÀ1 for 2-chloro-4-nitrophe-
nyl-b-lactoside (cNPL) (Table 1). In comparison with
the reported kinetic parameters of Tr_Cel7A, which
ranged from 0.4 to 1.23 mM [4,26,28–30], our results
showed that Th_Cel7A exhibits significantly lower
affinity for these substrates, whereas the turnover
value, kcat, was approximately 3.5 times larger than
that of Tr_Cel7A on the lactoside substrate
( ꢀ 12 minÀ1) [26,28]. Moreover, our inhibition assays
showed that Th_Cel7A was competitively inhibited by
cellobiose, although with a much higher inhibition
constant of 7.2 Æ 1.2 mM compared with the Ki of
0.02 mM for Tr_Cel7A [31]. Low inhibition by the
end-product and by the competitive inhibitor, cellobi-
ose, might be potentially important for biotechnologi-
cal applications of the enzyme.
Trichoderma harzianum is a filamentous fungus that
exhibits both cellulolytic and mycoparasitic activities,
and, unlike T. reesei, its cellulolytic activity and bio-
technological potential are still largely unexplored.
Enzymatic studies of several strains of T. harzianum
have shown that these fungi are able to produce cellu-
lolytic complexes with higher b-glucosidase activity
Specific activity and substrate-hydrolysis assays of
the full-length Th_Cel7A and its CCD are shown in
Fig. 2. Our experiments revealed that Th_Cel7A is
58
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
Table 1. Kinetic parameters of Th_Cel7A. Activity measurements were performed using 700 nM enzyme in 50 mM sodium acetate buffer,
pH 5.0, at 50 °C
pNPC
cNPL
Cellobiose
Km (mM)
3.4 Æ 0.5
Vmax (mMÁminÀ1
)
kcat (minÀ1
)
Km (mM)
3.7 Æ 1.2
Vmax (mMÁminÀ1
)
Kcat (minÀ1
)
Ki (mM)
Th_Cel7A
0.016 Æ 0.001
23.2 Æ 1.1
0.031 Æ 0.004
44.6 Æ 4.7
7.2
active and very efficient in degrading bacterial micro-
crystalline cellulose (BMCC), achieving almost 90%
conversion of the latter substrate in 24 h under the
conditions of our experiment. As expected, the CCD is
significantly less efficient in the hydrolysis of crystal-
line cellulose (Avicel, Sigmacell 20 and BMCC) com-
pared with the full-length enzyme containing both
CCD and CBM domains (Fig. 2A,B). Furthermore,
Th_Cel7A is capable of the enzymatic cleavage of
amorphous carboxymethyl cellulose (CMC); however,
in this case, the enzymatic efficiency was low and there
was no significant difference in specific activity
between the CCD and the full-length enzyme (Fig. 2A,
B). Furthermore, neither full-length enzyme nor the
CCD was able to hydrolyze more than 10% of this
substrate after 24 h of reaction. Conversion of CMC
rapidly reached a plateau and the reaction did not pro-
ceed any further (Fig. 2C,D). Both specific activities
Fig. 2. Activities of purified cellobiohydrolase I from Trichoderma harzianum (Th_Cel7A FL) and its CCD (Th_Cel7A CCD) on different
cellulosic substrates. Assays were performed in 50 mM sodium citrate buffer, pH 5.0, at 50 °C. (A) Specific activities of Th_Cel7A and (B)
percentage substrate conversion in 24 h. (C) Time course of substrate conversion by Th_Cel7A and (D) by its CCD. [Note the change in
scale between (C) and (D)].
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
59

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
and hydrolytic yields of Th_Cel7A on Avicel and Sig-
macell 20 were at intermediate levels (between those
observed for BMCC and CMC) and there was a con-
siderable difference in activity and hydrolysis yields
between complete enzyme and its CCD (Fig. 2). The
measured specific activities of Th_Cel7A were compa-
rable with that of Tr_Cel7A (Fig. 2A). According to
the reported data, the specific activities of Tr_Cel7A
against Avicel or Sigmacell 20 were 0.014–0.3 UÁmgÀ1
and 0.04–0.07 UÁmgÀ1, respectively, for CMC under
similar experimental conditions [15,32]. These results
show that the exoglucanase, Cel7A, from T. harzianum
is efficient in hydrolyzing crystalline cellulosic sub-
strates, such as BMCC, Avicel and Sigmacell 20, but is
not effective in the hydrolysis of substituted substrates,
such as CMC.
L6) that enclose the long cellulose-binding tunnel
(Fig. 1B).
The CCD contains a total of 20 cysteine residues,
all of which are involved in the disulfide bridges
known to stabilize the loops around the cellulose-bind-
ing cleft. As commonly found in enzymes secreted by
fungi [19–22,24,26], the N-terminal glutamine is cyc-
lized into a pyroglutamate residue. Three of the four
glycosylation sites predicted according to the sequence
motif N-X-S/T (where X is any amino acid residue,
except proline) could be identified in the electron den-
sity maps, and single N-acetyl D-glucosamine residues
were modeled covalently bound to Asn126, Asn283
and Asn397. All N-acetyl D-glucosamine residues make
contact with symmetry-related protein molecules.
Overall comparison of the T. harzianum Cel7A
CCD with other Cel7A CCD structures
Overall structure of T. harzianum Cel7A CCD
The refined Cel7A CCD model (Table 2), comprises
residues 18–445 and assumes the characteristic fold of
Cel7A enzymes [19–22,24,26], with the b-strands of the
b-sandwich structure connected by the loops (L1 to
Consistent with the high sequence homology between
the Cel7A from T. harzianum and the Cel7A from
T. reesei, structural superposition showed that both
molecules have a very similar global fold, with an
ꢁ
rmsd of 0.5 A for 426 Ca aligned residue pairs. As
expected, the largest structural deviations were associ-
ated with the loop regions that display high tempera-
ture factors and therefore are mobile or structurally
disordered. The first important structural difference
between Th_Cel7A and Tr_Cel7A was associated with
loop 3 at the entrance of the tunnel (residues Ser99 to
Gln101 in the Tr_Cel7A sequence). Loop 3 is critical
in defining the overall opening profile of the ligand-
binding site. Some residues present in this loop in
Tr_Cel7A are absent in Th_Cel7A and consequently
the enzyme has a more open entrance to the binding
tunnel (Figs 3 and 4). Gln101 from Tr_Cel7A loop 3
is a key residue in the substrate recognition at subsites
À6 and À5 [24] and such interactions are thus not
available in Th_Cel7A.
Table 2. X-ray refinement statistics
Diffraction data
Space group
ꢁ
Unit cell parameters (A)
P212121
a = 56.7, b = 69.2, c = 90.6
45.3–1.67 (1.71–1.67)
42070 (4077)
6.1 (5.5)
a
ꢁ
Resolution limits (A)
Number of unique reflections
Redundancy
Completeness (%)
Rmerge (%)b
I/sigI (I)b
99.8 (98.8)
9.1 (32.9)
24.4 (4.7)
Refinement
factor/Rfree (%)
R
13.1/16.2
Number of atoms
Protein
3281
42
N-acetyl-D-glucosamine
Poly(ethylene glycol)
Water
A similar shortening of loop 3 was also described
for GH family 7 cellobiohydrolases from T. emersonii
(Te_Cel7A) and P. chrysosporium (Pc_Cel7D). The
CCDs of Te_Cel7A and Pc_Cel7D display a sequence
similarity of 70% and 60%, respectively, with the
24
694
2
ꢁ
B factors (A )
Protein
10.5
28.3
33.0
27.2
N-acetyl-D-glucosamine
PEG
ꢁ
CCD of Th_Cel7A, and a structural rmsd of 0.8 A for
Water
rmsd
ꢁ
417 Ca Te_Cel7A aligned residues and 0.9 A for 415
ꢁ
Bond length (A)
Ca Pc_Cel7D aligned residues, respectively. The equiv-
alent loop 3 in both structures was similar to that of
Th_Cel7A and consequently they also exhibit a more
open tunnel entrance when compared with Tr_Cel7A
[20,24].
0.011
Bond angle (º)
1. 494
Ramachandran (preferred/allowed)% 98.1/1.9
a
Values in parentheses are for the highest resolution shell.
b
Rmerge = ∑|Ij À <I>|/∑ <Ij>, where Ij is the observed intensity of
A second important feature of the Th_Cel7A CCD
structure is related to the loops that form the sidewalls
an individual reflection and <I> is the average intensity of that
reflection.
60
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
access to the binding tunnel at the substrate site
(Fig. 5A). Whereas the equivalent loop 5 in Pc_Cel7D
was significantly shorter, resulting in a totally exposed
tunnel (Fig. 5B), the side face of the cellulose channel
from Tr_Cel7A was almost completely covered
(Fig. 5C). The entrance to the catalytic tunnel of
Th_Cel7A was also more exposed compared with that
of Th_Cel7A (although not as much as in Pc_Cel7D)
because Th_Cel7A residues Ile203 and Tyr371 are
substituted by shorter side-chains of Val216 and
Ala384 in Th_Cel7A (Fig. 5A). As discussed later,
MD simulations demonstrate that the presence of
Ala384 in loop 6 of Th_Cel7A in place of Tyr371 in
Tr_Cel7A is responsible for notable effects on the
dynamics of the loops enclosing the substrate-binding
site. Our crystal structure and MD simulations
revealed that Tyr260 can adopt different conforma-
tions in the Th_Cel7A CCD. Conformations in which
the Tyr260 side chain is more exposed to the solvent
can partially obstruct the enhanced access to the cellu-
lose-binding channel in Th_Cel7A. Regarding the cata-
lytic residues at the cellulose-binding cleft, the only
observed conformational difference was associated
with Glu217, which is H-bonded to the glucosyl unit
at the product +1 site in Tr_Cel7A, whereas the equiv-
alent Glu230 of Th_Cel7A assumed an alternative con-
formation, forming an H-bond to Asp154.
Fig. 3. Structure-based alignment of cellobiohydrolase
I from
Trichoderma harzianum (Uniprot Q9P8P3), Talaromyces emersonii
(Uniprot Q8TFL9, PDB 1Q9H), Phanerochaete chrysosporium
(Uniprot Q7LIJ0, PDB 1GPI) and Trichoderma reesei (Uniprot
P07981.1, PDB 1CEL). The alignment was carried out using Strap
(http.//3d-alignment.eu/). Numbering of protein sequences is
according to the deposited PDB structures. The position of Ala384
and gaps in loops 3 and 5 are depicted in rectangular boxes.
MD simulations of T. harzianum and T. reesei
Cel7A CCDs
of the ligand-binding tunnel (loops 4–6), in which the
conformation and sequence composition of the loops
are crucial for determining the shape of the channel
side-faces. Loops 5 and 6 of Th_Cel7A enclose part of
the side tunnel surface and create a gap that gives
To gain further insight into the mobilities of
Th_Cel7A and Tr_Cel7A CCDs, we performed MD
simulations of these enzymes starting from their apo
structures. According to the results of our simulations,
Fig. 4. Structural superposition of Cel7A
from Trichoderma harzianum (gray) and
Cel7A from Trichoderma reesei (blue, PDB
code 8CEL). The larger boxed area
represents a closer view of the loop
absent in T. harzianum, highlighting
Gln101 and the glucosyl residues from
Tr_Cel7A (blue) at the entrance of the
cellulose-binding tunnel. H-bonds are
shown as dashed lines.
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
61

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
Fig. 5. Molecular surface representation of Cel7A from (A) Trichoderma harzianum (Th; gray), (B) Phanerochaete chrysosporium (Pc; cyan)
and (C) Trichoderma reesei (Tr; slate). The side face of the tunnel is depicted in a rectangular box. Residues and loops composing the
tunnel side-face are indicated. Regions in red and dark blue represent negatively and positively charged residues, respectively.
the presence of Ala384 in loop 6 of Th_Cel7A in place
of Tyr371 in Tr_Cel7A was responsible for consider-
able differences in the dynamics of the loops enclosing
the substrate-binding site. In particular, Tyr260 (loop
5) experienced significant conformational changes in
Th_Cel7A, consistent with the higher values of B-fac-
tors obtained from the crystal structures. Even more
interestingly, the MD simulations showed that the
side-chain conformation of Tyr260 plays an important
role in modulating the degree of exposure of the sub-
strate-binding site in the unliganded Th_Cel7A CCD.
Figure 6A displays the time evolution of the rmsd val-
ues of Tyr247 (in loop 5 of Tr_Cel7A) and Tyr260 (in
loop 5 of Th_Cel7A), where we observed two states.
State 1 denotes protein conformations whose rmsd val-
becomes completely exposed and Tyr260 adopts a
conformation which very closely resembles that in our
ꢁ
2.9-A resolution structure (Structure 1). By contrast,
the side chains of Tyr247 (loop 5) and Tyr371 (loop 6)
in Tr_Cel7A remain constantly in contact and the
loops enclosing the binding site are never found open.
Even though Tyr247 in Tr_Cel7A is also capable of
undergoing conformational changes, as shown by the
peak in the Tyr247 rmsd near 60 ns, the fluctuations
are not sufficient to break the contact between loops 5
and 6, thus preventing the binding site from being
completely exposed (opened). Thus, the MD simula-
tions suggest that side-chain conformations of Tyr260
in loop 5 may break contacts with Ala384 in loop 6
and expose the binding site in Th_Cel7A, whereas its
counterpart residues Tyr247 (loop 5) and Tyr371 (loop
6) in Tr_Cel7A are never observed to be opened except
in the course of our simulations, thus preventing expo-
sure of the binding site.
ꢁ
ues relative to the average structure are below 3.0 A,
whereas State 2 refers to those protein conformations
ꢁ
whose rmsd values are above 3.0 A. Figure 6B shows
representative snapshots from States 1 and 2 in the
region of loops 5 and 6. The results showed that the
side chain of Th_Cel7A Tyr260 is substantially more
mobile than the corresponding Tyr247 in Tr_Cel7A.
In addition, Tyr260 interactions with the residues of
loop 6, mainly Ala384, are frequently broken. When-
ever such interactions are broken, the binding site
Analysis of root-mean-square-fluctuations of the
a-carbons depicted in Figure 7 also provided valuable
information on the dynamics of these enzymes. The
data showed that the flexibilities of most loops of
Th_Cel7A are greater than those of Tr_Cel7A. In par-
ticular, the data showed that loop 4 is substantially
62
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
of Te_Cel7A in terms of tunnel entrance and shape
(Figs 3 and 5).
The identified tunnel side cleft, which gives direct
access to the enzyme’s active site and exposes the cata-
lytic glutamates of Th_Cel7A to the solvent, is directly
dependent on the conformation of the Tyr260 side
chain. The cleft is partially obstructed when this resi-
due is solvent-exposed. The observed flexibility of
Th_Cel7A Tyr260 in the absence of ligands is not
unexpected.
A similar effect has been previously
reported for the equivalent Tyr247 in unliganded
Tr_Cel7A [19,26]. Nevertheless, our MD simulations
in the absence of substrate indicate that the presence
of the short side-chain from Ala384 allows almost free
motion of Tyr260 toward the inner and solvent-
exposed conformations in Th_Cel7A. The movements
of the Tyr247 side chain in Tr_Cel7A are restrained by
the interactions with Tyr371 (Ala384 in Th_Cel7A)
and are accomplished simultaneously with conforma-
tional changes of Tyr371, which points towards the
inner part of the tunnel when Tyr247 is in the solvent-
exposed conformation, and vice versa. Despite the rel-
ative conformational freedom of Tyr247, our MD
results indicate that this residue makes persistent con-
tacts with Tyr371 at loop 6 of Tr_Cel7A. Thus, the
side opening that exists in Th_Cel7A structure, does
not exist in Tr_Cel7A since it is mostly covered either
by Tyr247 or Tyr371 (Figs 3 and 6).
Fig. 6. (A) Time evolution of the rmsd for Tyr260 (Th_Cel7A) and
ꢁ
Tyr247 (Tr_Cel7A), showing State 1 (rmsd below 3.0 A) and State
ꢁ
2 (rmsd above 3.0 A). (B) Snapshots of the region between loops 5
and 6, representative of States 1 and 2. Molecular surfaces are
depicted light red. Tyr260 is more mobile than Tyr247, and the
conformational changes of Tyr260 in Th_Cel7A create a completely
exposed binding site. Although Tyr247 (Tr_Cel7A) can also undergo
significant conformational changes, these are not sufficient to
create an opened binding site owing to the presence of Tyr371 at
loop 6, as opposed to Ala384 at loop 6 in Th_Cel7A.
Sequence analyses of cellobiohydrolases I from dif-
ferent organisms show that the position occupied by
Ala384, although moderately conserved, displays pre-
dominantly certain types of residues, amongst which
alanine or tyrosine are the most frequent. Interestingly,
most of the Cel7A enzymes that have a gap at loop 3,
and thus exhibit a wider tunnel entrance, also contain
an alanine residue, which in T. harzianum is observed
to guarantee additional access to the cellulose-binding
tunnel. The dynamically enhanced side opening to the
substrate-binding tunnel, alluded to above, is structur-
ally novel. While the crystallographic structure of
Te_Cel7A lacks part of the side tunnel face owing to
the presence of a flexible loop [21], the Pc_Cel7D
structure shows a completely opened tunnel [20]. This
is a result of a shorter loop 5 that completely exposes
the tunnel and may be associated with higher rates of
hydrolysis, reduced processivity and considerable
endoglucanase activity of Pc_Cel7D [20,21].
more mobile in Th_Cel7A. This is interesting because
loop 4 is the main loop that covers the tunnel and is
located at the entrance of the binding site. A larger
flexibility in this region might facilitate the entry of the
substrate into the tight binding tunnel of the enzyme.
Discussion
Our structural characterization of the T. harzianum
Cel7A CCD reveals that this secreted enzyme has the
canonical fold of the GH family 7, with the loops of
the b-sandwich domain forming a typical cellobiohy-
drolase cellulose-binding tunnel instead of the far more
open groove found in endoglucanases [19,24]. Despite
the high sequence homology of this enzyme with the
Cel7A from T. reesei, Th_Cel7A, from a structural
point of view, is intermediate between Tr_Cel7A and
Pc_Cel7D, with significant similarities to the structure
Considering the overall tunnel shape, length and resi-
due conservation, it is expected that the sugar contacts
described for Tr_Cel7A [24] are maintained in
Th_Cel7A. Except for Gln101, all the residues interact-
ing with the glucosyl units in T. reesei are conserved in
T. harzianum, including the four tryptophan residues
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
63

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
Fig. 7. Root-mean-square-fluctuation
(rmsf) of the a-carbons of Th_Cel7A and
Tr_Cel7A. The upper part of the figure
shows the rmsf as a function of the
residue numbers and the lower part of the
figure shows superposed snapshots of the
MD simulations colored according to the
rmsf of each a-carbon (see the color
scale). Loop 4, which covers the tunnel at
the entrance of the binding site and is
substantially more mobile in Th_Cel7A, is
indicated.
that account for the formation of sites À7, À4, À2 and
+1 in Tr_Cel7A, which, together with tyrosine residues,
interact with the hydrophobic glucosyl faces of the
sugar units and presumably act as lubricating agents
facilitating the processive action of the enzyme [24]. It
is important to mention that the corresponding
Tr_Cel7A residue, Q101, is not present in Th_Cel7A,
and the absence of specific H-bonds at sites À6 and À5
lowers the binding affinity of Th_Cel7A to the sub-
strate, which is consistent with the higher Kmax experi-
mentally observed for Th_Cel7A (compared with
Tr_Cel7A). However, the same amino-acid substitution
might be favorable for the sliding of the cellulose chain
into the active site. Moreover, as a result of the less
restricted conformational mobility of Th_Cel7A
Tyr260, which is equivalent to the more conformation-
ally constrained Tr_Cel7A Tyr247, the specific H-bond
interactions with the substrate at site À2 might be
weaker in Th_Cel7A, contributing to the sliding of the
sugar units within the Th_Cel7A catalytic tunnel. This
is consistent with the higher kcat value, as observed in
our enzymatic studies of the enzyme.
degrees of polymerization, a different ratio between Ia
and Ib types of crystalline cellulose between these sub-
strates, higher amorphous content and a presence of
the minor hemicellulose fraction in their composition
[33,34]. BMCC was used in our experiments in a
never-dried form, whereas both Avicel and Sigmacell
20 were purchased in a dried form and rehydrated
before use as substrates for enzymatic hydrolysis. It is
known that dehydration/rehydration cycles of cellu-
losic materials have a significant effect on their poros-
ity and accessibility, both of which tend to show
rather a strong decrease after dehydration/rehydration
[34,35]. This renders substrates that are less susceptible
to enzymatic cleavage [36,37]. Furthermore, aggrega-
tion of cellulose microcrystals, entangled structures,
noncellulosic components attached to the cellulose, as
well as amorphous regions of the cellulose were
recently shown to represent obstacles for processive
cellobiohydrolase action and to slow down Tr_Cel7A
cellulose hydrolysis significantly [38,39]. It is known
that native crystalline cellulose is composed of two dif-
ferent crystalline components: cellulose Ia and cellu-
lose Ib [40]. Bacterial-type celluloses are mostly
composed of the Ia (triclinic) form, whereas the Ib
(monoclinic) form of cellulose is dominant in the cot-
ton/wood type of biomass [40–42]. The less thermody-
namically stable Ia form of cellulose was shown to be
more susceptible to enzymatic degradation by Tricho-
derma cellulases than was Ib cellulose, and in the
mixed substrates the Ia form of cellulose was degraded
first by cellulases [42]. Finally, xylose and xylooligosac-
charides were found to be strong inhibitors of cellulose
hydrolysis by cellulases and they might also have some
detrimental effect on the Th_Cel7A conversion of Avi-
cel and Sigmacell 20 [43]. The presence of hemicellu-
lose was demonstrated to be detrimental for biomass
Our activity studies showed that the cellobiohydro-
lase from T. harzianum is active against cellulosic
substrates with different degrees of crystallinity and
polymerization. Th_Cel7A was efficient in depolymer-
ization of highly crystalline BMCC, less efficient for
hydrolysis of crystalline and insoluble Avicel and Sig-
macell 20, and poorly efficient in CMC hydrolysis
(Fig. 2). The specific hydrolysis profile obtained for
Th_Cel7A in general correlates well with the published
data for Tr_Cel7A [32]. The smaller conversion yields
of Avicel and Sigmacell 20 might be related to several
factors, such as lower accessibility of these substrates
compared with BMCC, their dehydration/rehydration
before experimental studies, significantly different
64
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
hydrolysis using commercial enzymatic preparations, a
problem that can be alleviated by the addition of
xylanase [44,45].
Materials and methods
Expression and purification of Th_Cel7A
For hydrolysis of all three crystalline substrates
(BMCC, Avicel and Sigmacell 20), the presence of
CBM is beneficial and leads to higher specific activity
and conversion yields. Th_Cel7A, as a processive
enzyme, is ineffective in CMC hydrolysis, a functional
feature shared by Tr_Cel7A which is known to have
low CMCase activity [46,47]. The presence of CBM
has no effect on the hydrolysis of CMC (Fig. 2), and
the tunnel-like shape of processive Th_Cel7A makes it
rather ineffective for hydrolysis of this soluble and
substituted noncrystalline substrate.
Expression and purification of the wild-type Cel7A from
T. harzianum IOC 3844 was carried out following the proto-
col described previously [49]. In short, the cells were cul-
tured for 48 h at 25 °C, then used to inoculate a 7.5-L
vessel BioFlo 115 Fermenter (New Brunswick Scientific Co.,
Edison, NJ, USA) and cultured for a further 5 days at
28 °C, pH 5.0 and 200 rpm. The aeration rate was adjusted
to keep the dissolved O2 levels below 60% in the culture
medium. The enzyme was purified by chromatography
though columns containing DEAE–Sephadex (A-50; Sigma-
Aldrich, St. Louis, MO, USA) and phenyl sepharose (GE
Healthcare Biosciences, Pittsburgh, PA, USA). pH and tem-
perature optima for Th_Cel7A enzymatic activity are pH
5.0 and 50 °C, respectively [49]. Possible endoglucanase con-
tamination in the protein preparations was investigated
using activity assays against hydroxyethyl cellulose (Sigma)
[26], and no significant activity was detected. The purified
protein was dialyzed overnight against 50 mM Tris/HCl, pH
7.0, and concentrated to 8 mgÁmLÀ1 before crystallization.
The protein concentration was monitored by measuring the
absorbance at 280 nm.
In agreement with our kinetic analysis, the Th_Cel7A
catalytic rates are consistently higher than the reported
activities for Tr_Cel7A [15,32] using both glucoside
and/or lactoside substrates. This seems to be consistent
with the more open catalytic active cleft of the enzyme.
Te_Cel7A and Pc_Cel7D, both presenting shorter
active-site loops and thus a more open and accessible
binding site, have higher kcat values compared with
Tr_Cel7A, particularly for short-chain oligosaccharides
[20,21]. A more open substrate-binding site also seems
to be positively correlated with the decreased processiv-
ity and increased endoglucanase activity of the enzyme
[20,48]. This allows us to speculate that Th_Cel7A will
probably have lower processivity but higher additional
endoglucanase activity compared with Tr_Cel7A. How-
ever, its endoglucanase activity will probably be smaller
than that of Pc_Cel7D [48].
Enzyme activity assays
One per cent (weight by volume) suspensions of the sub-
strates Avicel® PH101 (Sigma), Sigmacell 20 (Sigma) and a
solution of CMC (Sigma) were prepared according to the
manufacturer’s instructions and used for analysis of enzyme
activity on insoluble and soluble substrates, respectively.
BMCC was prepared from commercially available Aceto-
bacter xylinum cellulose, as described previously [50].
To
date,
Melanocarpus albomyces
Cel7B
(Ma_Cel7B) is the only enzyme that was structurally
characterized in its native form in the complex with ol-
igosaccharides [22]. (The Tr_Cel7A–cellooligosaccha-
ride complexes were determined using catalytically
deficient mutants [19]). The Ma_Cel7B is strongly
inhibited by the product of the enzymatic reaction, cel-
lobiose (Kd = 6 mM), which probably was an impor-
tant factor for success in obtaining the enzyme crystal
structures in the complex with oligosaccharides [22].
The Th_Cel7A enzyme shows very weak inhibition by
the end product of the reaction, cellobiose. Unfortu-
nately, all our attempts to solve its native X-ray struc-
ture in complex with oligosaccharides failed and
production of the catalytically deficient mutants was
impeded by the necessity to produce recombinant
enzymes expressed in its soluble form using an appro-
priate expression system. Although the molecular
mechanism underlying the low inhibition of Th_Cel7A
by cellobiose still remains elusive, this feature of the
enzyme might be important for its applications in
industrial settings.
Avicel® and Sigmacel 20 have a degree of polymerization
(DP) between 150 and 500, and are mostly composed of
crystalline cellulose (85–95%), with minor fractions of
amorphous cellulose and up to 5% of xylose, mannose,
galactose and arabinose. The main difference between Avi-
cel® and Sigmacel 20 is their crystallinity, which is approxi-
mately 92% for Avicel and 85% for Sigmacell 20 [33,51].
Soluble ionic-substituted noncrystalline CMC is often
used for determining the activity of endoglucanases, which
randomly cleave intramolecular b-1,4-glucosidic bonds,
resulting in a dramatic reduction in the DP and the specific
viscosity of CMC [33]. CMC is rather a poor substrate for
cellobiohydrolases because owing to the steric confinement
imposed by the active-site tunnel, only nonsubstituted glu-
cose units are accessible for enzymatic hydrolysis [19,33,47].
Therefore, an exoglucanase will trim off the unsubstituted
residues until further progress along the CMC is prevented
by the carboxymethyl groups. Finally, BMCC has a low
DP (100–150) and higher crystallinity, accessibility and pur-
ity compared with Avicel and Sigmacel 20 [33]. This highly
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
65

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
ꢁ
crystalline cellulose, with a simple morphology, has been
used as a model substrate to study the mechanism of crys-
tal erosion [42] and cellobiohydrolase processivity [48]. The
crystallinity index of BMCC, measured using X-ray diffrac-
tion, was reported to be 95.2% [51].
1.67 A resolution, was obtained from optimized three-
dimensional crystals grown in 25% poly(ethylene glycol)
4000 and 5 mM CaCl2. The diffraction data were processed
using the program HKL2000 [54]. The crystal structures of
Cel7A CCD (residues 18–449) were determined by molecular
replacement with the program Phaser [55] using the CCD of
Cel7A from T. reesei (1CEL) as template (82% sequence
identity over 494 amino acids of the CCD). The 2Fo À Fc
map calculated from the unique solution of the top rotation
and translation searches showed a clear and contiguous elec-
tron density for the protein and active-site molecules not
included in the search model. Model building and refinement
was carried out using the program Coot [56]. Cycles of
restrained refinement with the program Refmac5 [57] were
carried out initially using overall temperature factors and
then by using isotropic temperature factors. One cycle of
simulated annealing was applied using the program Phenix
[58]. Water molecules were added using the program Arp/
wArp [59], and the stereochemical quality of the protein
complex was validated using the program MolProbity [60].
Statistics of the refined structures are presented in Table 2.
The visualization and all representations of the structures
were carried out using the program PyMol (The PyMOL
Activity assays for both full-length Th_Cel7A and its
CCD were carried out in 50 mM citrate buffer at the opti-
mum pH and temperature of the enzyme (pH 5.0 at 50 °C)
[49]. The reaction mixture of 15 lM of either full-length
Th_Cel7A or its CCD was incubated for 1 h for both the
soluble and the insoluble substrates. The conversion yields
were measured for 24 h under the same conditions described
above. No agitation was applied and to avoid inhibition of
Th_Cel7A by the cellobiose over time, 10 lM of Aspergillus
niger b-glucosidase (Novozymes SP188; Novo Nordisk A/S,
Bagsvaerd, Denmark) was added to the reaction mixtures
[38]. To ensure that Th_Cel7A was totally active during our
experiments, we monitored its activity during 24 h under
the same experimental conditions. No changes in the specific
activity were observed (data not shown). The quantification
of released reducing sugars was performed using the 3,5-
dinitrosalicylic acid [15]. All enzyme activities were
expressed in International Units [i.e. one unit of activity cor-
responded to the quantity of enzyme hydrolyzing 1 lmol of
substrate or releasing 1 lmol of reducing sugars (in glucose
equivalents) per minute].
€
Molecular Graphics System, Version 1.2r3pre, Schrodinger,
LLC; Cambridge, MA, USA).
The kinetic assays were carried out in 50 mM sodium
acetate buffer (pH 5.0) containing 700 nM enzyme and
pNPC and cNPL (Sigma) as soluble substrates at 12 differ-
ent final concentrations (from 0 to 15 mM). The reaction
was started by the addition of pNPC or cNPL at 50 °C for
10 min and stopped with 1.0 M sodium carbonate. The
amount of p- or c-nitrophenol released was measured at
k = 405 nm (spectrophotometer U-2001; Hitachi) and
quantified according to the respective standard curves. All
experiments were carried out in triplicate.
MD simulations
The initial coordinates for MD simulations of the CCD of
T. harzianum and T. reesei were taken from the obtained
crystallographic structure and from the PDB (PDB entry
8CEL), respectively. Hydrogen atoms were added and spe-
cial attention was paid to the ionizable residues, whose pro-
tonation states were estimated using the server H++ at pH 6
[61,62]. All substrate molecules and glucosamine groups were
removed. The structures were solvated using Packmol [63] by
insertion of water molecules within a rectangular box such
The effects of Cel7A inhibition were analyzed, as described
above, using pNPC as the substrate and different concentra-
tions of cellobiose (0–40 mM) at pH 5.0 and 50 °C. The
kinetic parameters were obtained from a Michaelis–Menten
plot V0 versus pNPC or cNPL by fitting to a rectangular
hyperbolic curve. Both linear and nonlinear fits were carried
out using GRAPHPAD PRISM version 5.04 for Windows (Graph-
Pad Software, La Jolla, CA, USA; www.graphpad.com).
ꢁ
that the solvent layer was at least 16 A thick. Sodium and
chloride ions were added to a total concentration of 0.10 M
in order to render the systems electrically neutral. The con-
structed systems consisted of approximately 60 000 atoms.
The simulations were performed in the ensemble NpT at
ꢀ
1 atm and 310 K using the Langevin/Nose-Hoover barostat
and the Langevin thermostat to control pressure and tem-
perature, respectively. The CHARMM force field [64] was
used for protein and the TIP3P model was used for water
molecules [65]. The van der Waals interactions were trun-
X-ray structure determination
ꢁ
cated at a cut-off radius of 12 A and electrostatics were
Collection of X-ray data was carried out on beamline MX2
at the Synchrotron Light National Laboratory (LNLS,
Campinas, Brazil) [52]. The first X-ray diffraction data set
evaluated using the particle mesh Ewald method [66]. All
bonds involving apolar hydrogen atoms were constrained
at their equilibrium lengths using the SHAKE algorithm
[67]. The time step for the numerical integration of the
equations of motion was 2.0 fs.
ꢁ
was collected at 2.9 A resolution from a needle-shaped crys-
tal obtained by the hanging-drop vapor-diffusion technique
in the presence of 25% poly(ethylene glycol) 4000 grown
using 35% poly(ethylene glycol) 4000 as a precipitant at pH
7.0 [53] (data not shown). The second data set, collected at
The following steps were performed to prepare the
systems for the production runs: step 1, 1000 steps of
66
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
conjugated gradient (CG) energy minimization followed by
a 200-ps MD run with all protein atoms harmonically
restrained; step 2, 1000 steps of CG followed by a 200-ps
MD with only the protein alpha carbons harmonically
restrained; and step 3, 600 ps MD without any restraints.
After these initial steps, 60-ns production trajectories were
generated and considered for analyses. For each system,
three independent simulations were performed. All simula-
tions were performed using NAMD [68].
7 Ahamed A & Vermette P (2008) Culture-based
strategies to enhance cellulase enzyme production from
Trichoderma reesei RUT-C30 in bioreactor culture
conditions. Biochem Eng J 40, 399–407.
8 Kleywegt GJ, Zou JY, Divne C, Davies GJ, Sinning I,
Stahlberg J, Reinikainen T, Srisodsuk M, Teeri TT &
Jones TA (1997) The crystal structure of the catalytic
core domain of endoglucanase I from Trichoderma
reesei at 3.6 angstrom resolution, and a comparison
with related enzymes. J Mol Biol 272, 383–397.
9 Coutinho PM & Henrissat B (1999) Carbohydrate-active
enzymes: an integrated database approach. In Recent
Advances in Carbohydrate Bioengineering (Gilbert HJ,
Davies GJ, Henrissat B & Svensson B, eds), pp. 3–12.
The Royal Society of Chemistry, Cambridge, UK.
10 Davies GJ, Wilson KS & Henrissat B (1997)
Nomenclature for sugar-binding subsites in glycosyl
hydrolases. Biochem J 321, 557–559.
Accession numbers
The coordinates and structure factors of Cel7A CCD from
T. harzianum have been deposited at the PDB (PDB codes
2YOK and R2YOKSF, respectively).
Acknowledgements
11 Lynd LR, Weimer PJ, van Zyl WH & Pretorius IS
(2002) Microbial cellulose utilization: fundamentals and
biotechnology (vol 66, pg 506, 2002). Microbiol Mol
Biol Rev 66, 739–739.
~
ꢂ
We thank the support from Fundaßcao de Amparo a
~
Pesquisa do Estado de Sao Paulo (FAPESP) via grants
2008/56255-9, 2009/54035-4 and 2010/08680-2; Conse-
ꢀ
lho Nacional de Desenvolvimento Cientıfico e Tec-
ꢀ
nologico (CNPq) via grants 490022/2009-0, 471834/
12 Fagerstam LG & Pettersson LG (1980) The 1.4-beta-glucan
cellobiohydrolases of Trichoderma reesei Qm-9414 – a new
type of cellulolytic synergism. FEBS Lett 119, 97–100.
13 Nummi M, Niku-Paavola ML, Lappalainen A, Enari
TM & Raunio V (1983) Cellobiohydrolase from
Trichoderma reesei. Biochem J 215, 677–683.
2009-2, 550985/2010-7 and 151951/2008-0 and Coorde-
~
ꢀ
naßcao de Aperfeicßoamento de Pessoal de Nıvel Supe-
rior (CAPESP).
14 Stals I, Sandra K, Geysens S, Contreras R, Van
Beeumen J & Claeyssens M (2004) Factors influencing
glycosylation of Trichoderma reesei cellulases. I:
postsecretorial changes of the O- and N-glycosylation
pattern of Cel7A. Glycobiology 14, 713–724.
References
1 Cao L, Bala G, Caldeira K, Nemani R & Ban-Weiss G
(2010) Importance of carbon dioxide physiological
forcing to future climate change. Proc Natl Acad Sci
USA 107, 9513–9518.
15 Tomme P, Vantilbeurgh H, Pettersson G, Vandamme 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, 575–581.
2 Idso SB (1989) Carbon Dioxide and Global Change:
Earth in Transition. IBR Press, Tempe, AZ (631 E.
Laguna Dr., Tempe 85282).
3 Koerner C & Bazzaz FA (1996) Carbon Dioxide,
Populations, and Communities. Academic Press, San Diego.
4 on Ossowski I, Stahlberg J, Koivula A, Piens K, Becker
D, Boer H, Harle R, Harris M, Divne C, Mahdi S
et al. (2003) Engineering the exo-loop of Trichoderma
reesei cellobiohydrolase, Ce17A. A comparison with
Phanerochaete chrysosporium Cel7D. J Mol Biol 333,
817–829.
16 Harrison MJ, Nouwens AS, Jardine DR, Zachara NE,
Gooley AA, Nevalainen H & Packer NH (1998)
Modified glycosylation of cellobiohydrolase I from a
high cellulase-producing mutant strain of Trichoderma
reesei. Eur J Biochem 256, 119–127.
17 Medve J, Lee D & Tjerneld F (1998) Ion-exchange
chromatographic purification and quantitative analysis
of Trichoderma reesei cellulases cellobiohydrolase I. II
and endoglucanase II by fast protein liquid
5 Gao JM, Weng HB, Zhu DH, Yuan MX, Guan FX &
Xi Y (2008) Production and characterization of
cellulolytic enzymes from the thermoacidophilic fungal
Aspergillus terreus M11 under solid-state cultivation of
corn stover. Bioresour Technol 99, 7623–7629.
6 Zhou J, Wang YH, Chu J, Zhuang YP, Zhang SL &
Yin P (2008) Identification and purification of the main
components of cellulases from a mutant strain of
Trichoderma viride T 100–14. Bioresour Technol 99,
6826–6833.
chromatography. J Chromatogr A 808, 153–165.
18 Teeri TT, Lehtovaara P, Kauppinen S, Salovuori I &
Knowles J (1987) Homologous domains in Trichoderma
reesei cellulolytic enzymes: gene sequence and
expression of cellobiohydrolase I. Gene 51, 43–52.
19 Divne C, Stahlberg J, Reinikainen T, Ruohonen L,
Pettersson G, Knowles JKC, Teeri TT & Jones TA
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
67

Crystal structure of Cel7A from T. harzianum
L. C. Textor et al.
(1994) The three-dimensional crystal structure of the
catalytic core of cellobiohydrolase I from Trichoderma
reesei. Science 265, 524–528.
Vehmaanpera J & Koivula A (2008) Cloning, expression,
and characterization of novel thermostable family 7
cellobiohydrolases. Biotechnol Bioeng 101, 515–528.
30 Deshpande MV, Eriksson K-E & Gran Pettersson L
(1984) An assay for selective determination of exo-1,4,-
b-glucanases in a mixture of cellulolytic enzymes. Anal
Biochem 138, 481–487.
20 Munoz IG, Ubhayasekera W, Henriksson H, Szabo I,
Pettersson G, Johansson G, Mowbray SL & Stahlberg
J (2001) Family 7 cellobiohydrolases from
Phanerochaete chrysosporium: crystal structure of the
catalytic module of Cel7D (CBH58) at 1.32 A
resolution and homology models of the isozymes.
J Mol Biol 314, 1097–1111.
31 Vonhoff S, Piens K, Pipelier M, Braet C, Claeyssens M
& Vasella A (1999) Inhibition of cellobiohydrolases from
Trichoderma reesei. Synthesis and evaluation of some
glucose-, cellobiose-, and cellotriose-derived hydrox-
imolactams and imidazoles. Helv Chim Acta 82, 963–980.
32 Jager G, Wu Z, Garschhammer K, Engel P, Klement
T, Rinaldi R, Spiess AC & Buchs J (2010) Practical
screening of purified cellobiohydrolases and
21 Grassick A, Murray PG, Thompson R, Collins CM,
Byrnes L, Birrane G, Higgins TM & Tuohy MG (2004)
Three-dimensional structure of a thermostable native
cellobiohydrolase, CBH IB, and molecular characterization
of the cel7 gene from the filamentous fungus, Talaromyces
emersonii. Eur J Biochem 271, 4495–450622.
22 Parkkinen T, Koivula A & Rouvinen J (2008) Crystal
structures of Melanocarpus albomyces cellobiohydrolase
Cel7B in complex with cello-oligomers show high
flexibility in the substrate binding. Protein Sci 17, 1383–
1394.
endoglucanases with alpha-cellulose and specification of
hydrodynamics. Biotechnol Biofuels 3, 3–18.
33 Zhang YHP, Himmel ME & Mielenz JR (2006)
Outlook for cellulase improvement: screening and
selection strategies. Biotechnol Adv 24, 452–481.
34 Bragatto J, Segato F, Cota J, Mello DB, Oliveira MM,
Buckeridge MS, Squina FM & Driemeier C (2012)
Insights on how the activity of an endoglucanase is
affected by physical properties of insoluble celluloses.
J Phys Chem B 116, 6128–6136.
23 Kraulis PJ, Clore GM, Nilges M, Jones TA, Pettersson
G, Knowles J & Gronenborn AM (1989)
Determination of the 3-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–
7257.
35 Esteghlalian AR, Bilodeau M, Mansfield SD & Saddler
JN (2001) Do enzymatic hydrolyzability and simons’
stain reflect the changes in the accessibility of
lignocellulosic substrates to cellulase enzymes?
Biotechnol Prog 17, 1049–1054.
24 Divne C, Stahlberg J, Teeri TT & Jones TA (1998)
High-resolution crystal structures reveal how a cellulose
chain is bound in the 50 angstrom long tunnel of
cellobiohydrolase I from Trichoderma reesei. J Mol Biol
275, 309–325.
36 Arantes V & Saddler JN (2010) Access to cellulose
limits the efficiency of enzymatic hydrolysis: the role of
amorphogenesis. Biotechnol Biofuels 3, 4.
37 Arantes V & Saddler JN (2011) Cellulose accessibility
limits the effectiveness of minimum cellulase loading on
the efficient hydrolysis of pretreated lignocellulosic
substrates. Biotechnol Biofuels 4, 3.
25 Barr BK, Hsieh YL, Ganem B & Wilson DB (1996)
Identification of two functionally different classes of
exocellulases. Biochemistry 35, 586–592.
26 Stahlberg J, Divne C, Koivula A, Piens K, Claeyssens
M, Teeri TT & Jones TA (1996) Activity studies and
crystal structures of catalytically deficient mutants of
cellobiohydrolase I from Trichoderma reesei. J Mol Biol
264, 337–349.
38 Jalak J & Vaeljamaee P (2010) Mechanism of initial
rapid rate retardation in cellobiohydrolase catalyzed
cellulose hydrolysis. Biotechnol Bioeng 106, 871–883.
ꢃ
€
€
39 Jalak J, Kurasin M, Teugjas H & Valjamae P (2012)
Endo-exo synergism in cellulose hydrolysis revisited.
J Biol Chem 287, 28802–28815.
27 Castro AM, Pedro KC, Cruz JC, Ferreira MC, Leite
SG & Pereira N Jr (2010) Trichoderma harzianum IOC-
4038: a promising strain for the production of a
cellulolytic complex with significant beta-glucosidase
activity from sugarcane bagasse cellulignin. Appl
Biochem Biotechnol 162, 2111–2122.
40 Atalla RH & VanderHart DL (1984) Native cellulose: a
composite of two distinct crystalline forms. Science 223,
283–285.
41 Wiley JH & Atalla RH (1987) Band assignments in the
Raman spectra of celluloses. Carbohydr Res 160, 113–129.
42 Hayashi N, Sugiyama J, Okano T & Ishihara M (1998)
The enzymatic susceptibility of cellulose microfibrils of
the algal-bacterial type and the cotton-ramie type.
Carbohydr Res 305, 261–269.
28 Claeyssens M, Van Tilbeurgh H, Tomme P, Wood TM
& McRae SI (1989) Fungal cellulase systems.
Comparison of the specificities of the cellobiohydrolases
isolated from Penicillium pinophilum and Trichoderma
reesei. Biochem J 261, 819–825.
43 Qing Q, Yang B & Wyman CE (2010) Xylooligomers
are strong inhibitors of cellulose hydrolysis by enzymes.
Bioresour Technol 101, 9624–9630.
29 Voutilainen SP, Puranen T, Siika-Aho M, Lappalainen
A, Alapuranen M, Kallio J, Hooman S, Viikari L,
68
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS

L. C. Textor et al.
Crystal structure of Cel7A from T. harzianum
44 Berlin A, Gilkes N, Kilburn D, Bura R, Markov A,
Skomarovsky A, Okunev O, Gusakov A, Maximenko
V, Gregg D et al. (2005) Evaluation of novel fungal
cellulase preparations for ability to hydrolyze softwood
substrates – evidence for the role of accessory enzymes.
Enzyme Microb Technol 37, 175–184.
56 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D Biol
Crystallogr 60, 2126–2132.
57 Skubak P, Murshudov GN & Pannu NS (2004) Direct
incorporation of experimental phase information in
model refinement. Acta Crystallogr D Biol Crystallogr
60, 2196–2201.
45 Varnai A, Siika-aho M & Viikari L (2010) Restriction
of the enzymatic hydrolysis of steam-pretreated spruce
by lignin and hemicellulose. Enzyme Microb Technol 46,
185–193.
58 Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis
IW, Echols N, Headd JJ, Hung LW, Kapral GJ,
Grosse-Kunstleve RW et al. (2010) PHENIX: a
comprehensive Python-based system for
46 Irwin DC, Spezio M, Walker LP & Wilson DB (1993)
Activity studies of 8 purified cellulases – specificity,
synergism, and binding domain effects. Biotechnol
Bioeng 42, 1002–1013.
macromolecular structure solution. Acta Crystallogr D
Biol Crystallogr 66, 213–221.
59 Perrakis A, Harkiolaki M, Wilson KS & Lamzin VS
(2001) ARP/wARP and molecular replacement. Acta
Crystallogr D Biol Crystallogr 57, 1445–1450.
60 Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral
GJ, Wang X, Murray LW, Arendall WB, Snoeyink J,
Richardson JS et al. (2007) MolProbity: all-atom
contacts and structure validation for proteins
and nucleic acids. Nucleic Acids Res 35, W375–
W383.
47 Penttila ME, Andre L, Lehtovaara P, Bailey MJ, Teeri
TT & Knowles JKC (1988) Efficient secretion of two
fungal cellobiohydrolases by Saccharomyces cerevisiae.
Gene 63, 103–112.
48 Kurasin M & Valjamae P (2011) Processivity of
cellobiohydrolases is limited by the substrate. J Biol
Chem 286, 169–177.
49 Colussi F, Serpa V, Delabona PS, Manzine LR,
Voltatodio ML, Alves R, Mello BL, Pereira N, Farinas
CS, Golubev AM et al. (2011) Purification, and
biochemical and biophysical characterization of
cellobiohydrolase I from Trichoderma harzianum IOC
3844. J Microbiol Biotechnol 21, 808–817.
61 Gordon JC, Myers JB, Folta T, Shoja V, Heath LS &
Onufriev A (2005) H++: a server for estimating pKas
and adding missing hydrogens to macromolecules.
Nucleic Acids Res 33, W368–W371.
62 Myers J, Grothaus G, Narayanan S & Onufriev A
(2006) A simple clustering algorithm can be accurate
enough for use in calculations of pKs in
50 Valjamae P, Sild V, Nutt A, Pettersson G & Johansson
G (1999) Acid hydrolosis of bacterial cellulose reveals
different modes of synergistic action between cellobio-
hydrolase I and endoglucanase I. Eur J Biochem 266,
327–334.
macromolecules. Proteins 63, 928–938.
63 Martinez L, Andrade R, Birgin EG & Martinez JM
(2009) PACKMOL: a package for building initial
configurations for molecular dynamics simulations.
J Comput Chem 30, 2157–2164.
51 Park S, Johnson DK, Ishizawa CI, Parilla PA & Davis
MF (2009) Measuring the crystallinity index of cellulose
by solid state C-13 nuclear magnetic resonance.
Cellulose 16, 641–647.
64 MacKerell AD, Bashford DM, Dunbrack RL,
Evanseck JD, Field MJ, Fischer S, Gao J, Guo H, Ha
S et al. (1998) All-atom empirical potential for
molecular modeling and dynamics studies of proteins.
J Phys Chem B 102, 3586–3616.
52 Guimaraes BG, Sanfelici L, Neuenschwander RT,
Rodrigues F, Grizolli WC, Raulik MA, Piton JR, Meyer
BC, Nascimento AS & Polikarpov I (2009) The MX2
macromolecular crystallography beamline: a wiggler
X-ray source at the LNLS. J Synchrotron Radiat 16,
69–75.
65 Jorgensen WL, Chandrasekhar J, Madura JD, Impey
RW & Klein ML (1983) Comparison of simple
potential functions for simulating liquid water. J Chem
Phys 79, 926–935.
53 Colussi F, Textor LC, Serpa V, Maeda RN, Pereira N
& Polikarpov I (2010) Purification, crystallization and
preliminary crystallographic analysis of the catalytic
domain of the extracellular cellulase CBHI from
Trichoderma harzianum. Acta Crystallogr Sect F Struct
Biol Cryst Commun 66, 1041–1044.
66 Darden T, York D & Pedersen L (1993) Particle mesh
Ewald: an N. log(N) method for Ewald sums in large
systems. J Chem Phys 98, 10089–10092.
67 Ryckaert JP, Ciccotti G & Berendsen HJC (1997)
Numerical integration of the Cartesian equations of
motion of a system with constraints: molecular
dynamics of n-alkanes. J Comput Chem 23, 327–341.
68 Phillips JC, Braun R, Wang W, Gumbart J,
54 Otwinowski Z & Minor W (1997) Processing of X-ray
diffraction data collected in oscillation mode. Macromol
Crystallogr Pt A 276, 307–326.
ꢀ
Tajkhorshid E, Villa E, Chipot C, Skeel RD, Kale L &
55 Storoni LC, McCoy AJ & Read RJ (2004) Likelihood-
enhanced fast rotation functions. Acta Crystallogr D
Biol Crystallogr 60, 432–438.
Schulten K (2005) Scalable molecular dynamics with
NAMD. J Comput Chem 26, 1781–1802.
FEBS Journal 280 (2013) 56–69 ª 2012 The Authors Journal compilation ª 2012 FEBS
69