Báo cáo khoa học: Crystal structure of a glycoside hydrolase family 6 enzyme, CcCel6C, a cellulase constitutively produced by Coprinopsis cinerea pot

Crystal structure of a glycoside hydrolase family 6
enzyme, CcCel6C, a cellulase constitutively produced by
Coprinopsis cinerea
Yuan Liu
1
, Makoto Yoshida
1
, Yuma Kurakata
2
, Takatsugu Miyazaki
2
, Kiyohiko Igarashi
3
,
Masahiro Samejima
3
, Kiyoharu Fukuda
1
, Atsushi Nishikawa
2
and Takashi Tonozuka
2
1 Department of Environmental and Natural Resource Science, Tokyo University of Agriculture and Technology, Japan
2 Department of Applied Biological Science, Tokyo University of Agriculture and Technology, Japan
3 Department of Biomaterial Sciences, Graduate School of Agricultural and Life Sciences, University of Tokyo, Japan
Introduction
Cellulose, a linear polymer made up of glucose units
linked by b-1,4-glucosidic linkages, is the predominant
structural component of plant cell walls and is the
most abundant biomass resource on Earth. Cellulases
hydrolyse the b-1,4-glucosidic bonds of cellulose chains

and are traditionally classified as endoglucanases (EC
3.2.1.4) or cellobiohydrolases (EC 3.2.1.91) based on
their activity profiles. Endoglucanases randomly cleave
the internal b-1,4-glucosidic bond of cellulose, whereas
cellobiohydrolases preferentially act on the end of the
chain and progressively cleave off cellobiose as the
main product [1–3].
Cellulases belonging to the glycoside hydrolase family
6 (GH6) are known as major cellulolytic enzymes pro-
duced by filamentous fungi. For example, GH6 cellulases
from the best studied cellulolytic organism, ascomycete
Hypocrea jecorina (formerly known as Trichoderma ree-
sei), make up 12–20% of total extracellular protein when
the fungus grows in cellulolytic culture [4]. Therefore,
GH6 enzymes have been considered attractive enzymes
for industrial application, such as biomass conversion.
The CAZy database ( [5] broadly
categorizes the GH6 enzymes into cellobiohydrolase-
type and endoglucanase-type enzymes. In 1990, the first
Keywords
basidiomycete; cellobiohydrolase; cellulase
induction; endoglucanase; glycoside
hydrolase family 6
Correspondence
T. Tonozuka, Department of Applied
Biological Science, Tokyo University of
Agriculture and Technology, 3-5-8
Saiwai-cho, Fuchu, Tokyo 183-8509, Japan
Fax: +81 42 367 5705
Tel: +81 42 367 5702

E-mail:
(Received 18 November 2009, revised 4
January 2010, accepted 14 January 2010)
doi:10.1111/j.1742-4658.2010.07582.x
The basidiomycete Coprinopsis cinerea produces the glycoside hydrolase
family 6 enzyme CcCel6C at low and constitutive levels. CcCel6C exhibits
unusual cellobiohydrolase activity; it hydrolyses carboxymethyl cellulose,
which is a poor substrate for typical cellobiohydrolases. Here, we deter-
mined the crystal structures of CcCel6C unbound and in complex with
p-nitrophenyl b-d-cellotrioside and cellobiose. CcCel6C consists of a dis-
torted seven-stranded b ⁄ a barrel and has an enclosed tunnel, which is
observed in other cellobiohydrolases from ascomecetes Hypocrea jecorina
(HjeCel6A) and Humicola insolens (HinCel6A). In HjeCel6A and
HinCel6A, ligand binding produces a conformational change that narrows
this tunnel. In contrast, the tunnel remains wide in CcCel6C and the con-
formational change appears to be less favourable than in HjeCel6A and
HinCel6A. The ligand binding cleft for subsite )3 of CcCel6C is also wide
and is rather similar to that of endoglucanase. These results suggest that
the open tunnel and the wide cleft are suitable for the hydrolysis of carb-
oxymethyl cellulose.
Abbreviations
CcCel6C, Coprinopsis cinerea Cel6C; GH, glycoside hydrolase family; (Glc)
2
-S-(Glc)
2
, methylcellobiosyl-4-thio-b-cellobioside; HinCel6A,
Humicola insolens Cel6A; HinCel6B, Humicola insolens Cel6B; HjeCel6A, Hypocrea jecorina Cel6A; PcCel7A, Phanerochaete chrysosporium,
Cel7A; pNPG2, p-nitrophenyl b-
D-cellobioside; pNPG3, p-nitrophenyl b-D-cellotrioside.
1532 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS

crystal structure of a cellulase was reported; it was
a catalytic domain of Hypocrea jecorina Cel6A
(HjeCel6A, formerly designated cellobiohydrolase II),
a GH6 cellobiohydrolase from an ascomycete [6]. In
that same decade, the crystal structure of the catalytic
domain in HinCel6A, another GH6 cellobiohydrolase
from ascomycete Humicola insolens, was determined
[7]. The catalytic domains of HjeCel6A and HinCel6A
consist of a distorted seven-stranded b⁄ a barrel; the
striking feature is that they have active sites enclosed
by N-terminal and C-terminal loops that form a
tunnel. The enclosed active sites trap the cellulose
chain in the tunnel and delay enzyme–substrate
dissociation, which promotes the cleavage of several
sequential substrate bonds [8,9]. In contrast and
despite displaying higher sequence similarity to
HjeCel6A and HinCel6A, the structure of a fungal
endoglucanase, Humicola insolens Cel6B (HinCel6B),
shows active sites in a cleft formed by a C-terminal
loop deletion coupled with the peeling open of an
N-terminal loop [10].
Many reports are available on the crystal structures
of the GH6 enzymes from ascomycetes, but no crystal
structure of the basidiomycete-derived GH6 enzyme
has yet been determined. Recently, we cloned five
genes encoding GH6 enzymes from a basidiomycete
Coprinopsis cinerea (formerly known as Coprinus cine-
reus), and the enzymes have been designated CcCel6A,
CcCel6B, CcCel6C, CcCel6D and CcCel6E [11]. The
amino acid sequences corresponding to the active site

enclosing loops of cellobiohydrolases have been
observed in all five enzymes. In the evolutionary tree,
however, four of the enzymes, CcCel6B–6E, have
mapped to a region distant from CcCel6A. There are
high sequence identities of CcCel6A–HjeCel6A (48%)
and CcCel6A–HinCel6A (52%), including an N-termi-
nal cellulose binding domain. In contrast, CcCel6B–6E
fall into a region closer to the endoglucanase
HinCel6B in the evolutionary tree, and no cellulose
binding domain is found in the four enzymes. For
example, the sequence identities of CcCel6C–HjeCel6A
and CcCel6C–HinCel6A are 36 and 39%, respectively,
whereas that of CcCel6C–HinCel6B is 43%. Transcript
analysis showed that the presence of cellobiose
strongly induced transcription of the CcCel6A gene,
but weakly induced transcription of the CcCel6B, -6D
and -6E genes. Interestingly, the transcript level of
CcCel6C was not influenced by either glucose or
cellobiose. When the enzymatic properties were
investigated, CcCel6B and CcCel6C exhibited cellobio-
hydrolase activity, but the enzymes hydrolysed carb-
oxymethyl cellulose, which is a poor substrate for
typical GH6 cellobiohydrolases [12]. These results indi-
cate that the physiological function and the substrate
binding mechanism of CcCel6C are expected to be dif-
ferent from those of known cellobiohydrolases. Here,
we present the crystal structure of CcCel6C. To our
knowledge, this is the first report of the crystal struc-
ture of a basidiomycete GH6 enzyme.
Results and Discussion

Overall structures of CcCel6C
The crystal structures of unliganded CcCel6C and the
enzyme–substrate complexes of CcCel6C–p-nitrophenyl
b-d-cellotrioside (pNPG3) and CcCel6C–cellobiose
were determined at 1.6, 1.4 and 1.2 A
˚
resolutions,
respectively (Table 1). The crystal belongs to the space
group P1, which contains one molecule in an
asymmetric unit. In Ramachandran plots, 95.7% (unli-
ganded CcCel6C), 95.7% (CcCel6C–pNPG3) and
95.4% (CcCel6C–cellobiose) of residues were shown to
be in favoured regions, and no residues were identified
as outliers, as calculated by the molprobity server
[13]. The electron density (2F
o
–F
c
) maps contoured at
1r show continuous density for almost all main chain
atoms except for the first  12 N-terminal residues
and the last  12 C-terminal segments containing the
His-tag sequence. The overall structure of CcCel6C
alone is shown in Fig. 1A. Like the fungal GH6 cello-
biohydrolases [6,7], CcCel6C consists of a seven-
stranded b ⁄ a barrel fold. a-Helices and b-strands are
numbered as a
1
–a
8

and b
0
–b
VII
, respectively, as shown
in Fig. 2, based on the numbering scheme for
HinCel6A [7].
Structural homology was researched using the dali
server [14] and CcCel6C was found to most resemble
the fungal GH6 enzymes: HinCel6A (cellobiohydro-
lase; Z score, 54.2) [7], HjeCel6A (cellobiohydrolase;
Z score 54.1) [6], HinCel6B (endoglucanase; Z score,
51.1) [10] and bacterial GH6 enzymes (e.g. 1UOZ [15]
and 1TML [16]; Z score  30). The Ca backbone of
unliganded CcCel6C was superposed with those of
cellobiohydrolases HjeCel6A, HinCel6A and endoglu-
canase HinCel6B using the program superpose in the
ccp4 suite [17]. The results indicated that the folds of
CcCel6C are almost identical to not only the cellobio-
hydrolases, but also to the endoglucanase (Fig. 1B).
The rmsd values are 1.16 A
˚
(CcCel6C–HinCel6A,
1BVW), 1.14 A
˚
(CcCel6C–HjeCel6A, 1QK0 chain A)
and 1.30 A
˚
(CcCel6C–HinCel6B, 1DYS chain A) for
main chain atoms.

The significant feature in cellobiohydrolases
HjeCel6A and HinCel6A is their active site located
inside an enclosed tunnel [6–10,18]; CcCel6C has a
Y. Liu et al. Structure of C. cinerea CcCel6C
FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS 1533
homologous tunnel and its conformation is very
similar in the liganded and unliganded enzymes
(Fig. 1C). Two loops, loop-1 and loop-2, forming this
tunnel are identified between b
II
and a
4
and between
b
VII
and a
8
’ (Fig. 2). Loop-1 and loop-2 contain
disulfide bridges of Cys103–Cys164 and Cys298–
Cys348, respectively, like those seen in HjeCel6A and
HinCel6A. Although the entire backbones of
CcCel6C, HjeCel6A and HinCel6A are essentially
identical, the two loops of the three enzymes are not
exactly superposed (Fig. 1B). In HinCel6A [7], a mag-
nesium ion forms a hexa-coordinated geometry
involved in the crystal contacts and the same magne-
sium-mediated geometry is found in the unliganded
CcCel6C, CcCel6C–pNPG3 and CcCel6C–cellobiose.
Here, this magnesium ion is located close to Asp109
in loop-1. Another hexa-coordinated magnesium ion,

which is observed near Glu33, is found only in
CcCel6C–cellobiose.
Ligand-bound structures
Comparing the ligand-bound structures of HjeCel6A
and HinCel6A with CcCel6C–pNPG3 and CcCel6C–
cellobiose enabled us to label the subsites of CcCel6C.
In CcCel6C–pNPG3, electron density for two ligand
molecules was seen in the active site (Fig. 3A). The
molecule bound to subsites )3to)1 was modelled as
p-nitrophenyl b-d-cellobioside (pNPG2, not pNPG3)
(average B = 25.4 A
˚
2
; Figs 3A, 4A), and a glucose
unit at the nonreducing end of pNPG3 was not identi-
fied in the difference Fourier map. The other molecule
bound to subsites +2 to +4 was also modelled as
pNPG2 (Figs 3A, 4B), but the map is better resolved
at the lower contoured level (average B = 40.8 A
˚
2
).
Weak electron density is present at subsite +1, but we
could not place the models. In CcCel6C–cellobiose,
electron density for two ligand molecules, a cellobiose
molecule bound to subsites +1 and +2 and a glucose
Table 1. Data collection and refinement statistics.
Unliganded Cellobiose pNPG3
Data collection
Beamline PF-AR NW12A PF-AR NW12A PF-AR NW12A

Space group P1 P1 P1
Cell dimensions
a (A
˚
) 44.0 44.2 43.9
b (A
˚
) 45.1 45.4 45.2
c (A
˚
) 48.9 49.1 49.0
a (°) 77.8 77.6 77.6
b (°) 87.3 86.9 86.8
c (°) 68.8 68.6 68.8
Resolution range (A
˚
) 50–1.60 (1.66–1.60)
a
50–1.20 (1.24–1.20)
a
50–1.40 (1.45–1.40)
a
Measured reflections 85 577 197 765 125 859
Unique reflections 43 370 100 613 63 990
Completeness (%) 95.8 (94.0)
a
92.9 (89.4)
a
94.8 (90.6)
a

I ⁄ r(I) 33.2 (9.2)
a
29.4 (3.7)
a
25.7 (3.3)
a
R
merge
0.025 (0.090)
a
0.036 (0.184)
a
0.064 (0.175)
a
Refinement statistics
R
work
0.141 0.148 0.163
R
free
0.165 0.169 0.189
rmsd
Bond lengths (A
˚
) 0.008 0.008 0.008
Bond angles (°) 1.12 1.20 1.22
Number of atoms
Protein 2894 2959 2940
Ligand – 35 64
Magnesium 1 2 1

Water 469 579 494
Average B (A
˚
2
)
Protein 15.7 12.9 12.5
Ligand – 23.9 33.1
Magnesium 32.7 18.1 15.3
Water 30.1 25.7 25.8
a
The values for the highest resolution shells are given in parentheses.
Structure of C. cinerea CcCel6C Y. Liu et al.
1534 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS
molecule bound to subsite )2, were identified
(Fig. 3B). The cellobiose molecule gives a clear elec-
tron density map (average B = 23.0 A
˚
2
), whereas part
of the density for the glucose molecule is not visible
(average B = 25.8 A
˚
2
). A similar single glucose mole-
cule has been found in subsite )2 of the HinCel6A–
cellobiose complex, but it remains unclear whether the
glucose molecule is part of cellobiose or from contami-
nation in the commercial cellobiose preparation [18].
The structures of CcCel6C–pNPG3, CcCel6C–cello-
biose, HinCel6A–cellobiose [18] and HjeCel6A–methyl-

cellobiosyl-4-thio-b-cellobioside [(Glc)
2
-S-(Glc)
2
] [8]
were superposed to depict the characteristics of the
ligand binding site of CcCel6C (Fig. 3C). The glucose
units in subsites )2, +1, +2 and +3 (abbreviated as
Glc )2, +1, +2 and +3, respectively) overlaid well,
whereas the aromatic ring of pNPG3 in subsite )1 was
at a position markedly different from that of Glc )1.
Study of the HjeCel6A–(Glc)
2
-S-(Glc)
2
complex has
shown that Glc )1 adopts a distorted conformation,
and many hydrogen bonds between HjeCel6A and Glc
)1 appear to stabilize this energetically unfavoured
conformation [8]. The p-nitrophenyl group of pNPG3,
however, is not able to form similar hydrogen bonds
with CcCel6C, resulting in the different position in the
active site.
Although controversy exists concerning the active
site residues of GH6 enzymes [19], two Asp residues
are suggested to be catalytic [7]. The sequence align-
ment of HjeCel6A, HinCel6A and CcCel6C (Fig. 2)
indicated that Asp150 and Asp334 of CcCel6C are
the potential catalytic residues and could act as a
proton donor and a base, respectively. Another aspar-

tic acid residue (Asp175 of HjeCel6A) has been pro-
posed to contribute to the electrostatic stabilization
of the partial positive charge in the transition state
R343
D109
D150
R343
D109
D150
A
B
C
Fig. 1. Overall structures of CcCel6C.
(A) Stereoview of CcCel6C–cellobiose
shown as a ribbon model. a-Helices,
b-strands and disulfide bridges are indicated
in blue, orange and green, respectively.
Cellobiose and glucose molecules bound to
the active site are shown in red. Side chains
of Asp109, Asp150 and Arg343 are shown
in pink. (B) Stereoview of the Ca backbone
of CcCel6C–cellobiose (red), which is super-
posed on those of HjeCel6A–(Glc)
2
-S-(Glc)
2
(yellow; PDB id, 1QK2), HinCel6A–cellobiose
(cyan; PDB id, 2BVW) and the endoglucan-
ase HinCel6B (gray; PDB id, 1DYS). The
ligands bound to the active site are

illustrated as stick models. (C) Comparison
of the Ca backbones of unliganded CcCel6C
(blue), CcCel6C–pNPG3 (green) and
CcCel6C–cellobiose (red).
Y. Liu et al. Structure of C. cinerea CcCel6C
FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS 1535
[20], and a homologous residue in CcCel6C is proba-
bly Asp102. Two distinct conformations for the cata-
lytic acid have been observed in HjeCel6A and
HinCel6A, and both are proposed to be important in
the catalysis [7,20]. The F
o
–F
c
omit map shows that
these two conformations are present for Asp150 in
CcCel6C–cellobiose (Fig. S1A), but in CcCel6C–
pNPG3, one of the conformations is not seen, proba-
bly due to steric hindrance with the p-nitrophenyl
group of pNPG3.
To probe the interaction between CcCel6C and the
ligands, CcCel6C–pNPG3 and CcCel6C–cellobiose
were analysed using the program ligplot [21], and
taken together, 21 amino acid residues appear to par-
ticipate in ligand binding. Table 2 lists these residues,
plus the three conserved aspartic acid residues pro-
posed to be involved in catalysis. The amino acid resi-
dues in subsites )2 to +4 are highly conserved among
CcCel6C, HjeCel6A, HinCel6A and HinCel6B. Four
key tryptophan residues involved in substrate stacking

interactions are fully conserved in CcCel6C as Trp61
(subsite )2), Trp297 (+1), Trp198 (+2) and Trp201
(+4) (Fig. 2) as previously described [9,22]. A tyrosine
residue critical for the distortion of Glc )1 (Tyr169 in
HjeCel6A) [23–25] is identified in CcCel6C as Tyr86.
The enclosed tunnel
The cellobiohydrolases have been characterized by the
enclosed tunnel as described above. A conspicuous
feature of CcCel6C is that the enclosed tunnel is wider
than those of HjeCel6A and HinCel6A (Fig. 5A–C),
and loop-1 and loop-2 of CcCel6C revealed a more
open structure than those of HjeCel6A and HinCel6A.
The conformational changes in the two loops of Hje-
Cel6A and HinCel6A have been reported; binding of
the ligands such as (Glc)
2
-S-(Glc)
2
or cellobiose results
in a narrowing of the tunnel, and the additional empty
space is not seen in the vicinity of the ligands [8,18]. In
CcCel6C, however, the two loops of the unliganded
CcCel6C, CcCel6C–pNPG3 and CcCel6C–cellobiose
are superposed well (Fig. 1C), and the rmsd values for
all backbone atoms are 0.109 A
˚
(between unliganded
CcCel6C and CcCel6C–pNPG3) and 0.159 A
˚
(between

unliganded CcCel6C and CcCel6C–cellobiose). The
possibility that the two loops are trapped in the open
Fig. 2. Comparison of amino acid
sequences of CcCel6C and related
enzymes. The sequences were aligned
using the
CLUSTALW2 server, and manual
adjustment was carried out based on the
comparison of the crystal structures. The
numbering of amino acid residues and
secondary structures (a
1
–a
8
and b
0
–b
VII
) are
given. Residues listed in Table 2 are printed
with a red ⁄ pink (subsites )3to)1) or
blue ⁄ cyan (subsites +1 to +4) background.
The two loops, loop-1 and loop-2, are under-
lined. Other symbols: arrow, Asp150 and
Arg343; asterisk, three conserved aspartic
acid and four conserved tryptophan
described in the text; dashed line,
disulfide bridge.
Structure of C. cinerea CcCel6C Y. Liu et al.
1536 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS

conformations by crystal packing could not be
excluded, as the crystals of the complex structures were
obtained by soaking with pNPG3 or cellobiose.
In HjeCel6A–(Glc)
2
-S-(Glc)
2
, however, the conforma-
tional changes in the tunnel-forming loops have been
observed by addition of the ligand after the crystals
had reached full size [8]. The most significant move-
ment of HinCel6A occurs at residues Ala183 to
Gly188 [18], and the sequence of HjeCel6A ⁄ HinCel6A,
A-L ⁄ A-A-S-N-G, is composed of amino acids with rel-
atively small side chains. The corresponding region of
CcCel6C (residues 105–110) is A-K-A-S-D-G, which
contains a bulky lysine residue. It appears that the
conformational change in the two loops of CcCel6C is
less favourable and the tunnel still has an open space
near the binding sites of cellobiose or p NPG3.
Loop-1 contacts with loop-2 mainly via an interac-
tion between Asp109 and Arg343. In unliganded
CcCel6C, the electron density map for Asp109 is clear,
but the 2F
o
–F
c
map for Arg343 is better resolved at
the lower contoured level of 0.8r, and two hydrogen
bonds, Asp109 OD2-Arg343 NE and Ser108 O-Arg343

NH1 could form directly between loop-1 and loop-2 in
this model. As for Asp109 in both CcCel6C–pNPG3
and CcCel6C–cellobiose and Arg343 in CcCel6C–
pNPG3, the F
o
–F
c
omit maps show that there are at
least two different conformations (Fig. S1B). This
observation suggests that the enclosed tunnel of
CcCel6C is not completely ‘enclosed’, although the
ligand molecules are unable to pass through the open-
ing between loop-1 and loop-2.
From subsites )2 to +4, only one serine residue,
Ser236, is not conserved in the other fungal cellobiohy-
drolases (Table 2), and the corresponding residue of
HjeCel6A and HinCel6A is alanine (Ala304 and
Ala309, respectively). As described in the previous
section, the tunnel-forming loops of HinCel6A are
changed to adopt the closed conformation when the
ligands bind to the active site. As a result, in
HinCel6A–cellobiose, a serine residue in loop-1,
A
B
Fig. 4. Schematic drawing of the amino acid residues interacting
with the ligands observed at subsites ) 3to)1 (A) and +2 to +4
(B). Symbols: open circle, oxygen atom; closed circle, carbon atom;
gray circle, nitrogen atom; dashed line, hydrogen bond. The
residues involved in hydrophobic interactions are illustrated.
–3 –2 –1 +1 +2 +3 +4

K30
S26
D334
W61
S236
W297
W201
Y96
D102
D150
W198
A
B
C
Fig. 3. Comparison of the ligands bound to the active site. (A) The
pNPG3 F
o
–F
c
electron density maps at the 2.0 r contoured level.
The subsite numbers are labelled from )3 to +4. (B) The cellobiose
F
o
–F
c
electron density maps at the 2.0 r contoured level. (C) Over-
lays of the ligands in CcCel6C–pNPG3 (green), CcCel6C–cellobiose
(red), HjeCel6A–(Glc)
2
-S-(Glc)

2
(yellow; PDB id, 1QJW) and Hin-
Cel6A–cellobiose (cyan; PDB id, 2BVW). Some critical residues
described in the text are indicated in black.
Y. Liu et al. Structure of C. cinerea CcCel6C
FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS 1537
Ser186, can directly form hydrogen bonds with the
ligand [18]. The conformational change of HjeCel6A–
(Glc)
2
-S-(Glc)
2
has been reported to be more compli-
cated, and four states (most closed, more open, even
more open, and most open) of the loop have been
identified. The complex of wild-type HjeCel6A–(Glc)
2
-
S-(Glc)
2
(PDB id, 1QK2) has been observed in the
‘more open’ form and the corresponding serine residue,
Ser181, does not interact with (Glc)
2
-S-(Glc)
2
. The
Y169F mutant of HjeCel6A complexed with (Glc)
2
-S-

(Glc)
2
(PDB id, 1QJW), on the other hand, adopts the
‘most closed’ form, and Ser181 is pointed into the )1
site and OG atom of the serine residue hydrogen
bonds with O5 of Glc )1, O4 of Glc )1, and O2 of
Glc )2 [8]. It is not easy to interpret the role of the
Ser181 ⁄ 186 residue during the catalysis, but they
appear to stabilize the distorted conformation of Glc
)1. However, the significant conformational changes
of the two loops of CcCel6C were not observed
(Fig. 1C) and in CcCel6C–cellobiose, Ser108, the posi-
tion equivalent to Ser181 ⁄ Ser186, does not directly
hydrogen bond with the ligand. In the endoglucanases,
the corresponding residue of Ser236 is found to be ser-
ine (Ser221, HinCel6B; Ser189, Thermobifida fusca
Cel6A) and in the complex of Thermobifida fusca,
Cel6A with (Glc)
2
-S-(Glc)
2
, Ser189 hydrogen bonds
with O6 of the distorted glucose unit Glc )1 [25]. The
role of Ser236 in CcCel6C is probably similar to that
of Ser189 in Thermobifida fusca Cel6A.
Subsite )3
In contrast to the high similarity of subsites from )2
to +4, two amino acid residues involved in subsite )3
of CcCel6C, Ser26 and Lys30, are strikingly differ-
ent from those of HjeCel6A (Tyr103 and Glu107,

respectively) and HinCel6A (Tyr104 and Glu108,
Table 2. Amino acid residues interacting with the ligands or poten-
tially involved in the catalysis, and the corresponding residues of
HjeCel6A, HinCel6A and HinCel6B. pNP, p -nitrophenyl group.
Closest
Glc ⁄ pNP
a
CcCel6C HjeCel6A HinCel6A HinCel6B
Glc )3 S26 (Y103)
b
(Y104)
b
(D16)
b
K30 (E107)
b
(E108)
b
K20
E332 E399 E403 E314
G361 G428 G432 G328
Glc )2 W61 W135 W137 W52
R101 R174 R179 R91
K328 K395 K399 K310
P329 P396 P400 P311
pNP )1 Y96 Y169 Y174 Y86
S236 (A304)
b
(A309)
b

S221
D102 D175 D180 D92
D334 D401 D405 D316
Glc +1 D150 D221 D226 D139
N237 N305 N310 N222
W297 W367 W371 W282
Glc +2 N158 N229 N234 (G147)
b
H195 H266 H271 (N183)
b
G197 G268 G273 G185
W198 W269 W274 W186
W294 W364 W368 W279
G295 G365 G369 G280
Glc +3 T157 T228 T233 T146
N204 N257 N280 (K192)
b
pNP +4 W201 W272 W277 W189
a
The closest Glc ⁄ pNP is determined based on the cartoon gener-
ated using the program
LIGPLOT.
b
Amino acid residues that are not
identical to those of CcCel6C are given in parentheses.
ABC
DEF
Fig. 5. Surface models of CcCel6C and
related enzymes. (A–C) Overall structures of
CcCel6C (A), HinCel6A (B) and the endoglu-

canase HinCel6B (C). (D–F) Close-up views
in the vicinity of subsite )3 of CcCel6C (D),
HinCel6A (E) and HinCel6B (F). To generate
the models, the structure of
CcCel6C–pNPG3 was superposed on those
of HinCel6A (PDB id, 2BVW) and HinCel6B
(PDB id, 1DYS), and pNPG2 was placed on
the models. Glc )3 is labeled as )3.
Residues forming protruding knobs at the
entrance of the cleft are shown in yellow
and red, and the distances between them
(A
˚
) are indicated. Ser26 and Lys30 of
CcCel6C and the corresponding residues are
indicated in cyan and pink, respectively.
Structure of C. cinerea CcCel6C Y. Liu et al.
1538 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS
respectively). For the cellobiohydrolases, subsite )3is
typically presumed unnecessary to produce cellobiose,
but the studies of HinCel6A have revealed no substan-
tial evidence to completely negate a )3 subsite [18] and
subsites from )4 to +4 of HinCel6A are proposed [9].
In CcCel6C, Glc )3ofpNPG3 not only forms mul-
tiple hydrogen bonds with Ser26, Lys30 and Trp61
through water molecules, but also makes hydrophobic
contacts with Pro329, Glu332 and Gly361 (Fig. 5D).
However, modelling pNPG3 in its CcCel6C-bound
conformation with either HjeCel6A or HinCel6A, Glc
)3 causes steric conflict with a tyrosine residue

(HjeCel6A Tyr103 and HinCel6A Tyr104) (Fig. 5E).
The cleft for subsite )3 of CcCel6C is also apparently
wider than those of HjeCel6A or HinCel6A (Fig. 5A,
B). Asp64 and Glu362 of CcCel6C form protruding
knobs at the entrance of the cleft, and the distance
between atom OD2 of Asp64 and atom OE2 of
Glu362 is 9.8 A
˚
(Fig. 5D). Similar knobs, which are
formed by Arg140 and Gln433, are observed at the
entrance of the cleft of HinCel6A, but the distance
between atom NH1 of Arg140 and atom NE2 of
Gln433 is only 6.9 A
˚
(Fig. 5E). These observations
indicate that the accessibility of subsite )3 of CcCel6C
is less restricted than those of HjeCel6A and Hin-
Cel6A. The cleft for subsite )3 of CcCel6C is rather
similar to that of the endoglucanase HinCel6B
(Fig. 5C, F). The residue of HinCel6B equivalent to
Ser26 of CcCel6C is identified as Asp16, an amino acid
residue with a relatively small side chain and, thus, no
steric conflict is found if the similar placement is tested
for the endoglucanase HinCel6B. The width of the
entrance of the cleft for subsite )3 (ND2 of Asn55-SD
of Met329) is 11.6 A
˚
, which is similar to that of
CcCel6C (Fig. 5F).
Implications for enzymatic activity

The structure of CcCel6C contains the enclosed tunnel
around its active site, indicating that the enzyme has
cellobiohydrolase activity. Indeed, our previous study
showed that CcCel6C hydrolysed phosphoric acid-
swollen cellulose with the release of cellobiose as a
main product [12]. However, the enzyme lacks a cellu-
lose binding domain, which is necessary to hydrolyse
crystalline cellulose, and most GH6 cellobiohydrolases
have this domain. In addition, we have reported that
the transcript level of CcCel6C was very low at the
active growth stage in the cellulose-degrading culture,
and almost the same transcript level was detected at
the active growth stage in the glucose culture. The
transcript level also did not change when the mycelia
were transferred to a medium containing glucose,
cellobiose or no carbon source [11]. These findings sug-
gest that the physiological role of CcCel6C does not
involve the degradation of crystalline cellulose.
Cellulose is insoluble in water; for the enzyme to
recognize it, the insoluble cellulose must be converted
into soluble saccharides, such as cellobiose and cellool-
igosaccharides. In the past several decades, it has been
assumed that low and constitutive levels of cellulases
react with cellulose to produce a soluble molecule that
enters the cell and induces transcription of cellulase
genes [26,27]. Considering the results of our biochemi-
cal and transcript analyses, CcCel6C probably pro-
duces a small amount of cellobiose when cellulose is
present. Similar activity has been reported in a GH7
enzyme produced by basidiomycete Phanerochae-

te chrysosporium, PcCel7A [28–30]. This enzyme shows
the amino acid sequence corresponding to an active
site tunnel also shown in GH7 cellobiohydrolases, but
like CcCel6C, lacks a cellulose binding domain. A low
level of PcCel7A transcripts was observed in the cul-
ture containing cellulose, whereas the transcripts were
detected in glucose culture. In a homology modelling
analysis, the enzyme was expected to have endo-type
activity [31]. These enzymatic and transcriptional prop-
erties are very similar to those of CcCel6C and, thus,
PcCel7A might also produce an inducer of cellulase
genes. Recently, it was also reported that basidiomy-
cete Coniophora puteana has GH6 and GH7 enzymes
without a cellulose binding domain [32]. Therefore, the
existence of cellobiohydrolases lacking a cellulose
binding domain might characterize soluble cellulose
degradation specifically in basidiomycetes because this
type of enzyme was not found in ascomycetes. Many
reports have shown that cellobiohydrolases are
typically active on crystalline cellulose [1]. The fungal
cellobiohydrolases, HjeCel6A and HinCel6A, have the
cellulose binding domain [33,34]. Specific conforma-
tional changes have been observed in the two active
site enclosing loops of HjeCel6A and HinCel6A that
seem to be critical in hydrolysing the crystalline
cellulose [8,18]. CcCel6A, whose amino acid sequence
is highly similar to those of HjeCel6A and HinCel6A,
probably has similar activity on amorphous cellulose
in Coprinopsis cinerea.
The structures determined here indicate that

CcCel6C has an enclosed tunnel similar to that of
HjeCel6A and HinCel6A. The tunnel is, however,
wider and more open than these fungal cellobiohydro-
lases, and virtually no conformational change in the
two loops of CcCel6C is induced. The ligand binding
cleft of CcCel6C is also wider due to the absence of
the bulky tyrosine residue in subsite )3 (Fig. 5), and
the structures of subsites )1 and )3 of CcCel6C resem-
Y. Liu et al. Structure of C. cinerea CcCel6C
FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS 1539
ble those of the endoglucanases HinCel6B rather than
HjeCel6A and HinCel6A. We have reported that
CcCel6C hydrolyses the chemically modified cellulose
derivative, carboxymethyl cellulose, whereas CcCel6A
does not [12]. The open tunnel and the wide cleft are
probably suitable for the hydrolysis of carboxymethyl
cellulose. Carboxymethyl cellulose and amorphous
cellulose have been reported to be substrates of most
endoglucanases, indicating that the enzyme activity is
mainly directed towards amorphous regions in the
cellulose molecule [1]. The results described above lead
us to conclude that the architecture of CcCel6C could
be suitable for hydrolysing amorphous cellulose to
serve cellobiose, an inducer for the expression of
CcCel6A in Coprinopsis cinerea.
Experimental procedures
Enzyme preparation and crystallization
The expression and purification of CcCel6C were carried out
as described previously [35]. Briefly, recombinant CcCel6C
fused with a His-tag was produced in Escherichia coli

BL21(DE3) cells and purified with Ni-NTA agarose (Qiagen,
Hilden, Germany). The enzyme was crystallized at 20 °C
using the hanging drop vapour diffusion method, where
1 lL CcCel6C (21.5 mgÆmL
)1
) was mixed with the same vol-
ume of well solution (100 mm Hepes ⁄ KOH pH 7.0, 30%
polyethylene glycol 8000, 150 mm magnesium acetate). The
obtained crystal was transferred to a cryo-solution of 40%
(w ⁄ v) polyethylene glycol 8000 in well solution and flash fro-
zen in a stream of nitrogen gas. The crystal of the complex
of pNPG3 (Seikagaku Corporation, Tokyo, Japan) or cello-
biose was obtained by soaking in the same well solution
(100 mm Hepes ⁄ KOH pH 7.0, 30% polyethylene glycol
8000, 150 mm magnesium acetate) containing 60 mm
pNPG3 for 2 h or 220 mm cellobiose for 5 min, and the
solution containing the ligand also acted as a cryoprotectant.
The diffraction data were collected at beamline PF-AR
NW12 (Photon Factory, Tsukuba, Japan). The data were
processed and scaled with the program hkl2000 [36]
(Table 1).
Model building and structure refinement
The structure of CcCel6C was solved by molecular replace-
ment with the program molrep in the ccp4 suite [17], and
a model of HinCel6A (PDB id, 1BVW) [7] was employed
as a probe model. The automated model building was per-
formed with the program arp ⁄ warp [37]. The refinement
was carried out using the program refmac in the ccp4
suite, and anisotropic refinement was applied for data bet-
ter than 1.2 A

˚
resolution. Manual adjustment and rebuild-
ing of the model were carried out with the program coot
[38], and the models for the ligands were built from both
2F
o
–F
c
and F
o
–F
c
electron density maps. Solvent molecules
were introduced using the program arp ⁄ warp. Refinement
statistics are listed in Table 1. Superpositioning of CcCel6C
with other protein structures and calculation of the rmsd
values were carried out using the program superpose in the
ccp4 suite. Sequence identities were calculated using the
program clustalw2 on the ebi server (.a-
c.uk/Tools/clustalw2/) [39] with the default values. Figures
were generated using ligplot [21] and pymol (http://
www.pymol.org/). The coordinates and structure factors of
unliganded CcCel6C, CcCel6C–pNPG3 and CcCel6C–cello-
biose have been deposited in the Protein Data Bank under
the accession codes 3A64, 3ABX and 3A9B, respectively.
Acknowledgements
This research was supported, in part, by the Green
Biomass Research for Improvement of Local Energy
Self-sufficiency Program of the Ministry of Education,
Culture, Sports, Science and Technology of Japan.

This research was performed with the approval of the
Photon Factory Advisory Committee (2008G013),
the National Laboratory for High Energy Physics,
Tsukuba, Japan.
References
1 Baldrian P & Vala
´
skova
´
V (2008) Degradation of cellu-
lose by basidiomycetous fungi. FEMS Microbiol Rev
32, 501–521.
2 Bayer EA, Chanzy H, Lamed R & Shoham Y (1998)
Cellulose, cellulases and cellulosomes. Curr Opin Struct
Biol 8, 548–557.
3 Lynd LR, Weimer J, van Zyl WH & Pretorius IS (2002)
Microbial cellulose utilization: fundamentals and bio-
technology. Microbiol Mol Biol Rev 66, 506–577.
4 Rosgaard L, Pedersen S, Langston J, Akerhielm D,
Cherry JR & Meyer AS (2007) Evaluation of minimal
Trichoderma reesei cellulase mixtures on differently
pretreated barley straw substrates. Biotechnol Prog 23,
1270–1276.
5 Cantarel BL, Coutinho PM, Rancurel C, Bernard T,
Lombard V & Henrissat B (2009) The Carbohydrate-
Active EnZymes database (CAZy): an expert resource
for glycogenomics. Nucleic Acids Res 37 Database issue,
D233–D238.
6 Rouvinen J, Bergfors T, Teeri T, Knowles JK & Jones
TA (1990) Three-dimensional structure of cellobiohy-

drolase II from Trichoderma reesei. Science 249, 380–
386.
7 Varrot A, Hastrup S, Schu
¨
lein M & Davies GJ (1999)
Crystal structure of the catalytic core domain of the
Structure of C. cinerea CcCel6C Y. Liu et al.
1540 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS
family 6 cellobiohydrolase II, Cel6A, from Humicola
insolens, at 1.92 A
˚
resolution. Biochem J 337, 297–304.
8 Zou J, Kleywegt GJ, Sta
˚
hlberg J, Driguez H, Nerinckx
W, Claeyssens M, Koivula A, Teeri TT & Jones TA
(1999) Crystallographic evidence for substrate ring
distortion and protein conformational changes during
catalysis in cellobiohydrolase Ce16A from Trichoderma
reesei. Structure 7, 1035–1045.
9 Varrot A, Frandsen TP, von Ossowski I, Boyer V,
Cottaz S, Driguez H, Schu
¨
lein M & Davies GJ (2003)
Structural basis for ligand binding and processivity in
cellobiohydrolase Cel6A from Humicola insolens.
Structure 11, 855–864.
10 Davies GJ, Brzozowski AM, Dauter M, Varrot A &
Schu
¨

lein M (2000) Structure and function of Humicola
insolens family 6 cellulases: structure of the endoglucan-
ase, Cel6B, at 1.6 A
˚
resolution. Biochem J 348,
201–207.
11 Yoshida M, Sato K, Kaneko S & Fukuda K (2009)
Cloning and transcript analysis of multiple genes encod-
ing the glycoside hydrolase family 6 enzyme from
Coprinopsis cinerea. Biosci Biotechnol Biochem 73, 67–73.
12 Liu Y, Igarashi K, Kaneko S, Tonozuka T, Samejima
M, Fukuda K & Yoshida M (2009) Characterization of
glycoside hydrolase family 6 enzymes from Coprinopsis
cinerea. Biosci Biotechnol Biochem 73, 1432–1434.
13 Davis IW, Leaver-Fay A, Chen VB, Block JN, Kapral
GJ, Wang X, Murray LW, Arendall WB III, Snoeyink
J, Richardson JS et al. (2007) MolProbity: all-atom
contacts and structure validation for proteins and
nucleic acids. Nucleic Acids Res 35 Web Server issue,
W375–W383.
14 Holm L & Sander C (1993) Protein structure compari-
son by alignment of distance matrices. J Mol Biol 233,
123–138.
15 Varrot A, Leydier S, Pell G, Macdonald JM, Stick RV,
Henrissat B, Gilbert HJ & Davies GJ (2005) Mycobac-
terium tuberculosis strains possess functional cellulases.
J Biol Chem 280, 20181–20184.
16 Spezio M, Wilson DB & Karplus PA (1993) Crystal
structure of the catalytic domain of a thermophilic
endocellulase. Biochemistry 32, 9906–9916.

17 Collaborative Computational Project (1994) The CCP4
suite: programs for protein crystallography. Acta
Crystallogr Sect D 50, 760–763.
18 Varrot A, Schu
¨
lein M & Davies GJ (1999) Structural
changes of the active site tunnel of Humicola insolens
cellobiohydrolase, Cel6A, upon oligosaccharide binding.
Biochemistry 38
, 8884–8891.
19 Vuong TV & Wilson DB (2009) The absence of an
identifiable single catalytic base residue in Thermobifida
fusca exocellulase Cel6B. FEBS J 276, 3837–3845.
20 Koivula A, Ruohonen L, Wohlfahrt G, Reinikainen T,
Teeri TT, Piens K, Claeyssens M, Weber M, Vasella A,
Becker D et al. (2002) The active site of cellobiohydro-
lase Cel6A from Trichoderma reesei: the roles of
aspartic acids D221 and D175. J Am Chem Soc 124 ,
10015–10024.
21 Wallace AC, Laskowski RA & Thornton JM (1995)
LIGPLOT: a program to generate schematic diagrams
of protein–ligand interactions. Protein Eng 8, 127–134.
22 Koivula A, Kinnari T, Harjunpa
¨
a
¨
V, Ruohonen L,
Teleman A, Drakenberg T, Rouvinen J, Jones TA &
Teeri TT (1998) Tryptophan 272: an essential determi-
nant of crystalline cellulose degradation by Trichoderma

reesei cellobiohydrolase Cel6A. FEBS Lett 429, 341–346.
23 Koivula A, Reinikainen T, Ruohonen L, Valkeaja
¨
rvi A,
Claeyssens M, Teleman O, Kleywegt GJ, Szardenings
M, Rouvinen J, Jones TA et al. (1996) The active site
of Trichoderma reesei cellobiohydrolase II: the role of
tyrosine 169. Protein Eng 9, 691–699.
24 Andre
´
G, Kanchanawong P, Palma R, Cho H, Deng X,
Irwin D, Himmel ME, Wilson DB & Brady JW (2003)
Computational and experimental studies of the catalytic
mechanism of Thermobifida fusca cellulase Cel6A (E2).
Protein Eng 16, 125–134.
25 Larsson AM, Bergfors T, Dultz E, Irwin DC, Roos A,
Driguez H, Wilson DB & Jones TA (2005) Crystal
structure of Thermobifida fusca endoglucanase Cel6A in
complex with substrate and inhibitor: the role of tyro-
sine Y73 in substrate ring distortion. Biochemistry 44,
12915–12922.
26 Sternberg D & Mandels GR (1979) Induction of cellulo-
lytic enzymes in Trichoderma reesei by sophorose.
J Bacteriol 139, 761–769.
27 Mach RL & Zeilinger S (2003) Regulation of gene
expression in industrial fungi: Trichoderma. Appl Micro-
biol Biotechnol 60, 515–522.
28 Covert SF, Bolduc J & Cullen D (1992) Genomic orga-
nization of a cellulase gene family in Phanerochaete
chrysosporium. Curr Genet 22, 407–413.

29 Covert SF, Vanden Wymelenberg A & Cullen D (1992)
Structure, organization, and transcription of a cellobio-
hydrolase gene cluster from Phanerochaete chrysospori-
um. Appl Environ Microbiol 58, 2168–2175.
30 Vanden Wymelenberg A, Covert S & Cullen D (1993)
Identification of the gene encoding the major cellobio-
hydrolase of the white rot fungus Phanerochaete chry-
sosporium. Appl Environ Microbiol 59, 3492–3494.
31 Mun
˜
oz IG, Ubhayasekera W, Henriksson H, Szabo
´
I,
Pettersson G, Johansson G, Mowbray SL & Sta
˚
hlberg
J (2001) Family 7 cellobiohydrolases from Phanerocha-
ete 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.
32 Kajisa T, Igarashi K & Samejima M (2009) The genes
encoding glycoside hydrolase family 6 and 7 cellulases
from the brown-rot fungus Coniophora puteana. J Wood
Sci 55, 376–380.
Y. Liu et al. Structure of C. cinerea CcCel6C
FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS 1541
33 Carrard G & Linder M (1999) Widely different off rates

of two closely related cellulose-binding domains from
Trichoderma reesei. Eur J Biochem 262, 637–643.
34 Shoseyov O, Shani Z & Levy I (2006) Carbohydrate
binding modules: biochemical properties and novel
applications. Microbiol Mol Biol Rev 70, 283–295.
35 Kurakata Y, Tonozuka T, Liu Y, Kaneko S, Nishikawa
A, Fukuda K & Yoshida M (2009) Heterologous
expression, crystallization and preliminary X-ray
characterization of CcCel6C, a glycoside hydrolase
family 6 enzyme from the basidiomycete Coprinopsis
cinerea. Acta Crystallogr Sect F 65, 140–143.
36 Otwinowski Z & Minor W (1997) Processing of X-ray
diffraction data collected in oscillation mode. Methods
Enzymol 276, 307–326.
37 Perrakis A, Morris R & Lamzin VS (1999) Automated
protein model building combined with iterative
structure refinement. Nat Struct Biol 6, 458–463.
38 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr Sect D
60, 2126–2132.
39 Larkin MA, Blackshields G, Brown NP, Chenna R,
McGettigan PA, McWilliam H, Valentin F, Wallace
IM, Wilm A, Lopez R et al. (2007) ClustalW and
ClustalX version 2.0. Bioinformatics 23, 2947–2948.
Supporting information
The following supplementary material is available:
Fig. S1. Stereoview of the F
o
–F
c

electron density maps
for residues observed in multiple conformations at the
2.0 r contoured level.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Structure of C. cinerea CcCel6C Y. Liu et al.
1542 FEBS Journal 277 (2010) 1532–1542 ª 2010 The Authors Journal compilation ª 2010 FEBS