Mutational analysis in the glycone binding pocket of Dalbergia cochinchinensis β-glucosidase to increase catalytic efficiency toward mannosides

Article abstract of DOI:10.1016/j.carres.2012.10.018

Dalcochinase and Abg are glycoside hydrolase family 1 β-glucosidases from Dalbergia cochinchinensis Pierre and Agrobacterium sp., respectively, with 35% sequence identity. However, Abg shows much higher catalytic efficiencies toward a broad range of glyco

Full text of DOI:10.1016/j.carres.2012.10.018

Carbohydrate Research 373 (2013) 35–41
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Mutational analysis in the glycone binding pocket of Dalbergia
cochinchinensis b-glucosidase to increase catalytic efficiency toward
mannosides
Khakhanang Ratananikom a, Khuanjarat Choengpanya b, Nusra Tongtubtim a, Theppanya Charoenrat c,
Stephen G. Withers d, Prachumporn T. Kongsaeree a,b,e,
⇑
a Department of Biochemistry, Faculty of Science, Kasetsart University, 50 Paholyothin Road, Ladyaw, Chatuchak, Bangkok 10900, Thailand
b Multidisciplinary Graduate Program in Genetic Engineering, Graduate School, Kasetsart University, Bangkok 10900, Thailand
c Department of Biotechnology, Faculty of Science and Technology, Thammasat University Rangsit Campus, Pathumthani 12120, Thailand
d Department of Chemistry, Faculty of Science, University of British Columbia, Vancouver, B.C., Canada V6T 1Z1
e Center for Advanced Studies in Tropical Natural Resources, NRU-KU, Kasetsart University, Bangkok 10900, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history:
Dalcochinase and Abg are glycoside hydrolase family 1 b-glucosidases from Dalbergia cochinchinensis
Pierre and Agrobacterium sp., respectively, with 35% sequence identity. However, Abg shows much higher
catalytic efficiencies toward a broad range of glycone substrates than dalcochinase does, possibly due to
the difference in amino acid residues around their glycone binding pockets. Site-directed mutagenesis
was used to replace the amino acid residues of dalcochinase with the corresponding residues of Abg, gen-
erating three single mutants, F196H, S251V, and M369E, as well as the corresponding three double
mutants and one triple mutant. Among these, the F196H mutant showed increases in catalytic efficiency
toward almost all glycoside substrates tested, with the most improved catalytic efficiency being a 3-fold
Received 5 September 2012
Received in revised form 17 October 2012
Accepted 19 October 2012
Available online 27 October 2012
Keywords:
Abg
Catalytic efficiency
Dalcochinase
b-Glucosidase
Protein engineering
increase for hydrolysis of p-nitrophenyl b-D-mannoside, suggesting a preferred polar residue at this
position and consistent with the presence of histidine at this position in two other GH1 glycosidases from
barley and rice that prefer b-mannosides. In addition, the M369E mutation resulted in a small increase in
catalytic efficiency for cleavage of p-nitrophenyl b-D-galactoside. By contrast, the multiple mutants were
up to 8-fold less efficient than the recombinant wild-type dalcochinase, and displayed primarily
antagonistic interactions between these residues. Thus, differences in catalytic efficiency between dalco-
chinase and Abg are therefore not primarily due to differences in the residues that directly contact the
substrate, but derive largely from contributions from more remote residues and the overall architecture
of the active site.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
and glucosylceramide in humans. They have been classified into
Carbohydrate Active EnZymes (CAZy) glycoside hydrolase (GH)
b-Glucosidases (3.2.1.21) are a heterogeneous group of enzymes
catalyzing the hydrolytic removal of b- -glucose from the non-
reducing end of b- -glucosides and b- -gluco-oligosaccharides.
They are involved in various physiological processes, such as host
defense mechanisms and growth control in plants, cellulose degra-
dation in bacteria and fungi, and hydrolysis of glucosylsphingosine
families GH1, GH3, GH5, GH9, and GH30, based on amino acid se-
quence and structural similarity.1–3 The GH1 family contains sev-
eral activities, including b-glucosidases, b-mannosidases, b-
fucosidases, b-galactosidases, and thioglucosidases. These enzymes
D
D
D
have an (b/a) barrel structure with the conserved glutamate res-
8
idues in the TL/FNEP and I/VTENG motifs, at the carboxyl-terminal
ends of b-stands 4 and 7, acting as the catalytic acid/base and
nucleophile, respectively, in the double displacement mecha-
nism.4,5 Apart from sharing overall structures, GH1 b-glucosidases
from different sources also have similar functional properties,
including optimal pH between 5 and 6 and molecular mass of
about 55–65 kDa.2
Abbreviations: GH, glycoside hydrolase; LFER, linear free energy relationships;
pNP-
side; pNP-b-
glucoside; pNP-b-
b- -xyloside.
a
-
L
-Ara, p-nitrophenyl
-Gal, p-nitrophenyl b-
-Man, p-nitrophenyl b-
a
-
L
-arabinoside; pNP-b-
-galactoside; pNP-b-
-mannoside; pNP-b-D
D
-Fuc, p-nitrophenyl b-
-Glc, p-nitrophenyl b-
-Xyl, p-nitrophenyl
D-fuco-
D
D
D
D-
D
D
D
Dalcochinase is a GH1 b-glucosidase from Dalbergia cochinchine-
⇑
Corresponding author. Tel.: +66 2 562 5555x2026; fax: +66 2 561 4627.
E-mail address: fscippt@ku.ac.th (P.T. Kongsaeree).
sis Pierre (Thai rosewood).6 Its natural substrate has been identi-
0008-6215/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.carres.2012.10.018

36
K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
fied as an isoflavonoid 12-dihydroamorphigenin-80-O b-
side, also known as dalcochinin-80-O-b- -glucoside.7 The pH opti-
mum and an apparent molecular mass were reported as 5.5 and
ꢀ330 kDa, respectively, and it was predicted to comprise 4–6 sub-
units of 66 kDa.8 Amino acid sequence alignment has placed the
catalytic acid/base and nucleophile residues at E182 and E396,
respectively, comparable to other members of GH1.6 Even though
D-gluco-
in the active site pockets (including both glycone and aglycone
subsites). Together these results allow us to better understand
the roles of specific amino acid residues around the glycone pocket
of dalcochinase and Abg, and provide information for future
protein engineering of GH1 b-glucosidases with an improved cata-
lytic efficiency.
D
dalcochinase is defined as a b-
p-nitrophenyl b- -fucoside (pNP-b-
ciency than p-nitrophenyl b- -glucoside (pNP-b-
specificity for the substituents at C-4 and C-6 positions is relatively
lax since it hydrolyzes p-nitrophenyl b- -galactoside (pNP-b-
Gal), p-nitrophenyl -arabinoside (pNP- -Ara), and p-nitro-
phenyl b- -xyloside (pNP-b- -Xyl), albeit with 8- to 80-fold lower
efficiency than the glucoside.8
D
-glucosidase, it actually hydrolyzes
-Fuc) with 5-fold greater effi-
-Glc). Indeed, its
2. Experimental procedures
D
D
D
D
2.1. Strains, plasmids, and chemicals
D
D-
Escherichia coli strain DH5
a and Pichia pastoris strain Y11430
a-
L
a-L
(Invitrogen, Carlsbad, CA, USA) were used for plasmid propagation
and protein expression, respectively. The recombinant plasmid
pPICZ-His8-trncTRBG that harbors the coding sequence of N-termi-
nally truncated form of dalcochinase (starting from the residue
D
D
Abg is another GH1 b-glucosidase from Agrobacterium sp. It
comprises a homodimer with 51 kDa subunits.9,10 The acid/base
and nucleophile residues were identified as E170 and E358, respec-
tively, by site-directed mutagenesis and labeling with the mecha-
V14 of the mature sequence) following the
a mating factor propep-
tide and 8 histidine residues, was used for the expression of the re-
combinant wild-type dalcochinase.26 Pfu DNA polymerase and
DpnI were purchased from Promega (Madison, WI, USA). All p-
nitrophenyl glycoside substrates were purchased from Sigma (St.
Louis, MO, USA).
nism-based inactivator.11,12 Abg also cleaves
a
range of p-
-Glc being the best substrate
and others cleaved about 2- to 360-fold slower. Abg also hydro-
nitrophenyl glycosides, with pNP-b-
D
lyzes p-nitrophenyl b-D-mannoside (pNP-b-D-Man) very slowly,
with a kcat/Km value of 6 mMꢁ1 sꢁ1, which is unusual among b-glu-
cosidases since hydrogen bond interactions between the C2-hydro-
xyl group and the residues in the active site pocket of the enzyme
contribute directly to transition state stabilization.13–15
2.2. Sequence alignment, homology modeling, and molecular
docking
The sequences of dalcochinase, Abg, and other GH1 b-glycosi-
dases were aligned using ClustalW 2.0.12.27 The homology model
of dalcochinase was generated by MODELLER 9v4 program,28 using
the structure of cyanogenic b-glucosidase from Trifolium repens L.
(white clover) (PDB code 1CBG) as a template.16 The overall struc-
ture of the model was checked by the PROCHECK, ProSA, Verify-3D,
and WHATIF programs.29–31 The active site was defined as 15 Å
around the pseudo-atom which was generated at the center of cat-
alytic residues, E182 and E396, of dalcochinase.
The structure of gluconolactone (PDB code LGC) was docked
into the active site of modeled dalcochinase using AutoDock ver-
sion 4.2. The coordinate files of both protein and ligand required
for docking calculation were prepared by AutoDockTools. The
non-polar hydrogens were deleted by the program, and the partial
charges were merged to the carbon atoms. The sugar substrates
were treated as rigid, and the rotatable bonds were set automati-
cally by the program. The protein portion was set as flexible recep-
tor by assigning the predicted catalytic residues (E182 and E396)
as flexible residues. AutoGrid was performed to pre-calculate the
grid map of interaction energy prior to docking. The grid size
was set at 60 ꢂ 60 ꢂ 60 with a grid point spacing of 0.258 Å at
the center of protein. A Lamarckian genetic algorithm was used
with a population size of 300, maximum number of energy evalu-
ations of 2,500,000, maximum number of generations of 25,000,
and uniform crossover mode. Other parameters were set as default.
The docked conformation was visualized using Accelrys DS Visual-
izer 3.0 (Accelrys Inc., San Diego, USA).
While sharing 35% sequence identity, Abg was 4- to 40-fold
more efficient than dalcochinase for the same p-nitrophenyl glyco-
side substrates. For pNP-b-
was greatest, the kcat of Abg was only 2-fold lower, but its Km was
70-fold lower than that of dalcochinase. For pNP-b- -Fuc, where
D-Glc, where the difference in efficiency
D
the difference in efficiency was the smallest, their kcat values are
very close, but the Km of Abg was 5-fold lower than that of dalco-
chinase. These parameters suggest that the differences in catalytic
efficiencies between these two enzymes are dominated by the gly-
cosylation step (reflected by the values of kcat/Km) over the degly-
cosylation step (reflected by the values of kcat), possibly reflecting
differences in the amino acid residues that form direct and/or indi-
rect interactions with the glycone moieties in their respective ac-
tive site pockets.
X-ray crystallography and site-directed mutagenesis studies of
various b-glucosidases showed that the residues involved in agly-
cone binding are variable in GH1 b-glucosidases, whereas those
interacting with the glycone moiety are mostly conserved.16–24
However, there are examples of unique residues in the glycone
binding pocket that are important for determining glycone
specificity. Exchanges of active site residues N206D in Pyrococcus
furiosus b-glucosidase and D206N in Pyrococcus horikoshii
b-mannosidase increased their preference for mannoside and
glucoside substrates, respectively.25 Also, the W433C substitution
in Sulfolobus solfataricus b-glycosidase resulted in 24-fold increase
in its preference for mannoside over galactoside substrates.18,44
Here, the amino acid residues that might be responsible for the
difference in catalytic efficiencies between dalcochinase and Abg
were probed by site-directed mutagenesis to generate single, dou-
ble, and triple dalcochinase mutants. Their kinetic analysis enabled
an evaluation of the interaction strengths between a pair of muta-
tions in multiple mutants, and the active site similarity between
the wild-type and mutant enzymes via linear free energy relation-
ships (LFER). Our results highlighted the preference for histidine at
position 196 instead of phenylalanine for hydrolysis of a b-manno-
side substrate. However, most mutations displayed antagonistic
interactions, and none could generate an active site that resembled
the active site architecture of Abg. So, substrate specificity is likely
determined by specific residues as well as the overall interactions
2.3. Construction of dalcochinase mutants
The dalcochinase mutants were made via site-directed muta-
genesis to replace amino acid residues at the positions F196,
S251, and M369 of dalcochinase (the numbers indicate their posi-
tions in the reported sequence of dalcochinase6) with the corre-
sponding residues of Abg, to generate three single mutants
F196H, S251V, and M369E, three double mutants F196H/S251V,
F196H/M369E, and S251V/M369E, and one triple mutant F196H/
S251V/M369E. The specific primers for each mutant were designed
based on the dalcochinase cDNA sequence (Genbank accession

K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
37
AF163097).6 The sequences of the sense mutagenic primers of
F196H, S251V, and M369E were 50-GGGTATGCATACGG-
stant that corresponds to (kcat/Km) [E]0. Since [E]0, which was the
concentration of enzyme used in the reaction, was known, the va-
lue of kcat/Km could be easily obtained.
TATGCATGCACCAGGTCGATGTTCTCC-30,
50-CATCAGAAAGGTAC-
AATAGGCATTGTTTTGCACGTAGTTTGGG-30, and 50-GGTCCAGT-
GACTCCCTCAGGATGGGAATGCATTTATCCAAAAGG-30, respectively
(mutation sites are underlined). Sequences of the antisense muta-
genic primers are the reverse complements of the sequences
shown above. The recombinant plasmid pPICZ-His8-trncTRBG
was used as a template for generating single mutations. The plas-
mids harboring single and double mutations were used as a tem-
plate for generating double and triple mutations, respectively.
Site-directed mutagenesis reactions were performed with 3 units
Pfu DNA polymerase according to the method published previ-
ously.32 The reactions were incubated with 10 units DpnI at 37 °C
overnight to remove the DNA template. The DpnI-treated DNA
was transformed into competent E. coli by electroporation. The
2.6. Calculation of activation free energy changes
The difference in the activation free energy (DDGà) for the gly-
cosylation step of hydrolysis was calculated by the following
equation:36
DDGz ¼ ꢁRTln½ðkcat=KmÞ =ðkcat=KmÞ ꢃ
ð2Þ
1
2
where R is the gas constant (8.314 Jmolꢁ1), T is the absolute temper-
ature (303 K), and kcat/Km is rate constant of the hydrolysis of the
same substrate by two different enzymes.
3. Results and discussion
transformants were selected on LB-agar plates containing 25 lg/
mL zeocin at 37 °C. The plasmids containing correctly mutated dal-
cochinase sequences were checked by DNA sequencing. Subse-
quently, the mutant plasmids were linearized with SacI,
transformed into P. pastoris by electroporation, and selected on
3.1. Homology modeling, molecular docking, and sequence
alignment
Since the 3-dimensional structures of both dalcochinase and
Abg are lacking, despite our efforts over several years, the homol-
ogy model of dalcochinase was generated (Supplementary Fig. 1)
using the structure of white clover b-glucosidase as a template
since it shows the highest sequence similarity (59%).16 The selected
YPDS plates with 100
Invitrogen.
lg/mL zeocin, following the protocols from
2.4. Expression and purification of enzymes in P. pastoris
model showed the root mean square deviation of the Ca atoms of
0.39 Å with respect to the template, and satisfied all criteria as
evaluated by PROCHECK, ProSA, Verify-3D, and WHATIF programs
(Supplementary Table 1). The slot-like binding pocket of dalcochin-
ase was approximately 22 Å long and 8 Å wide (calculated from
The wild-type and mutant forms of dalcochinase were ex-
pressed in a 2-liter Biostat B fermenter (B. Braun Biotech Interna-
tional, Germany), or in
a shake-flask system as described
previously.26,33 Then, enzymes were purified from the culture
media by hydrophobic interaction chromatography followed by
immobilized metal-ion affinity chromatography.26
distances between P347-C
Nd2 and W368-Cd1, respectively), with the catalytic amino acid
residues located about 5.5 Å apart (between E182-O 1 and E396-
1) at the bottom of the binding pocket. In order to predict the
c and E455-Oe2, and between N189-
e
Protein concentration determination was performed according
to the Bradford method by using the BioRad Protein Assay Reagent
Kit (Bio-Rad, Hercules, CA, USA) and bovine serum albumin as a
standard. SDS–PAGE was performed in a discontinuous system
with 4.5% separating and 7.5% resolving gel.34 Western blot analy-
sis and chemiluminescent detection were done with mouse mono-
clonal antibody against natural dalcochinase (a gift from Professor
Watchara Kasinrerk, Chiangmai University, Thailand), horseradish
peroxidase-conjugated rabbit polyclonal antibody against mouse
immunoglobulin (Dako, Glostrup, Denmark), and ECL Plus Western
Blotting Detection reagents (GE Healthcare, Buckinghamshire, UK).
To remove conjugated oligosaccharides, the glycosylated proteins
were treated with endoglycosidase Hf (New England BioLabs,
USA) for 2 h at 37 °C under denaturing conditions according to
manufacturer’s instructions.
Oe
interactions between the enzyme and the glycone substrate, gluc-
onolactone, which mimics the proposed 4H3 half chair transition-
state conformation of the glucose substrate,44 was docked into
the active site pocket of the dalcochinase model (Fig. 1). Glucono-
lactone appeared to form hydrogen bonds with residues Q36,
N181, E182, E452, and W453 of dalcochinase, and the distances be-
tween the docked gluconolactone and the amino acid residues in
the glycone binding pocket of dalcochinase model as predicted
by molecular docking are summarized in Supplementary Table 2.
From the docked position of gluconolactone, 17 residues were
predicted to be located in the glycone binding pocket of dalcochin-
ase (Fig. 1). Among these, 14 residues are similar between dalco-
chinase and Abg, namely Q36, R90, H136, W137, N181, E182,
N323, Y325, W368, E396, W445, E452, W453, and F461 in dalco-
chinase, corresponding to Q24, R81, H125, W126, N169, E170,
N296, Y298, W331, E358, W404, E411, W412, and F420 in Abg,
respectively (Supplementary Fig. 2). In many cases, their corre-
sponding positions in other GH1 b-glucosidases have been re-
ported to make direct contacts with the glucosyl moiety in the -1
subsite. The two catalytic acid/base and nucleophile residues are
universally conserved in all GH1 enzymes as expected,2 with the
exception of myrosinases.37 The residues Q39 and E451 in Spodop-
tera frugiperda b-glycosidase, corresponding to Q36 and E452 in
dalcochinase, respectively, were found to interact with the hydro-
xyl groups at C3, C4, and C6, of the glycone moiety of the sub-
strate.19 The residues R77 and N206 in Pyrococcus furiosus b-
glucosidase, corresponding to R90 and N181 in dalcochinase,
respectively, were shown to interact with the catalytic nucleophile
and the equatorial C2-hydroxyl group of the non-reducing end su-
gar.25 The residues H142 and W457 in Zea mays (maize) b-glucosi-
dase, corresponding to H136 and W445 in dalcochinase,
2.5. Kinetic measurements
The catalytic efficiencies (kcat/Km) of the wild-type and mutant
forms of dalcochinase for hydrolysis of various pNP-glycoside sub-
strates were determined from progress curves at low substrate
concentrations (less than 1/5 Km of wild-type dalcochinase toward
the same substrate) in 0.1 M sodium acetate, pH 5.0 at 30 °C. The p-
nitrophenol released was monitored by following the absorbance
at 360 nm until substrate depletion was observed. When the sub-
strate concentration is low, the rate of reaction (
v
) is related to
the substrate concentration by the equation:35
m
¼ ðkcat=KmÞ½Eꢃ0½Sꢃ
ð1Þ
The change in absorbance with respect to time was fitted to
first-order rate equation using the program GraFit 5.0 (Erithacus
Software Limited, Horley, UK) to yield pseudo-first-order rate con-

38
K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
1–10 l
mol minꢁ1 mgꢁ1, respectively. All dalcochinase mutants
exhibited a broad band of about 66 kDa on SDS–PAGE, which is
similar to the recombinant wild-type dalcochinase (Supplemen-
tary Fig. 3) but different from a sharp protein band at 60 kDa of
natural dalcochinase.26,32 The broad protein bands could be sharp-
ened by treatment with endoglycosidase Hf (Supplementary Fig. 4),
and thus are likely due to greater glycosylation in yeast than in
plants.39 Detection of these bands with mouse monoclonal anti-
body against natural dalcochinase in a Western blot confirmed
their identity as dalcochinase proteins (Supplementary Fig. 3).
3.3. Kinetic study of dalcochinase mutants
The catalytic efficiencies of the recombinant wild-type and mu-
tant forms of dalcochinase for the hydrolysis of pNP-glycoside sub-
strates were determined by substrate-depletion assays (Table 1).
These kcat/Km values of each substrate were then used to calculate
the differences in the activation free energy for the formation of
glycosyl-enzyme intermediates (DDGà) as a result of each mutation
(Fig. 2). Comparison of catalytic efficiencies, thus active site speci-
ficities, between two enzymes for a range of substrates is best
achieved through the LFER plot, which is a plot of log kcat/Km of each
substrate for one enzyme versus the same parameter for the other
enzyme (Supplementary Fig. 5). Such a plot essentially directly
compares the free energy of activation for each enzyme/substrate
pair, thus constitutes a linear free energy relationship.36 In such a
plot, the correlation coefficient and slope provide measures of the
similarities of the two active sites, with values of 1 in each case
indicating extremely high similarity or identity. The absolute
efficiencies of the two enzymes are not captured in these num-
bers—but rather in relative kcat/Km values for a single substrate.
The LFER plot between the two wild-type enzymes showed a slope
of 1.0 and a correlation coefficient of 0.88, revealing the high simi-
larity of the two active sites (Table 2 and Supplementary Fig. 5).
Among the mutants created and kinetically characterized, the
only one with substantially improved kinetic parameters is the
Fl96H mutant, which has essentially identical catalytic efficiency
Figure 1. The homology model of wild-type dalcochinase containing a docked
structure of gluconolactone. The model is shown as viewed from the exterior of the
enzyme, looking into the active site pocket. The main chain conformation is
omitted. The catalytic acid/base and nucleophile at E182 and E396, respectively, are
shown as orange stick models. The 12 other residues that are likely located in the
glycone binding pocket of dalcochinase and are similar between dalcochinase and
Abg, are shown as green stick models, while the three residues in dalcochinase that
are targeted for site-directed mutagenesis (F196, S251, and M369) are shown as
pink stick models. The docked gluconolactone is shown as a yellow ball-and-stick
model. The picture was generated with PyMOL version 1.3 (DeLano Scientific, Palo
Alto, CA, USA). (For interpretation of the references to colour in this figure legend,
the reader is referred to the web version of this article.)
respectively, were shown to make hydrogen bond interactions
with the glycone moiety in its active site pocket.17 The residue
W433 in Sulfolobus solfataricus b-glycosidase, corresponding to
W453 in dalcochinase, was shown to form hydrogen bonds to
the C3-hydroxyl group of the sugar substrate.18
to that of the wild-type enzyme for pNP-b-D-Glc, but has efficien-
While the conservation of amino acid residues that make up the
glycone binding pockets of various GH1 b-glucosidases is indica-
tive of their specificity for the b-glucoside substrate, the slight vari-
ations in surrounding residues may subtly affect their interactions
with the glycone moiety. In this study, 3 out of 17 residues that
were likely located in the glycone binding pocket of dalcochinase,
were different between dalcochinase and Abg, namely F196, S251,
and M369 in dalcochinase, corresponding to H184, V224, and E332
in Abg, respectively (Fig. 1 and Supplementary Fig. 2). In the dock-
ing model, the residue F196 points toward the C2-hydroxyl group
of the docked gluconolactone. The residue S251, while appearing to
be far from the docked structure, corresponds to the residue V241
of Oryza sativa L. (rice) Os3BGlu7, that is in direct contact with the
sugar moiety in the -1 subsite.38 The residue M369 is close to the
C6-hydroxyl group of the docked gluconolactone. Therefore, these
amino acid residues in dalcochinase were replaced with the corre-
sponding residues of Abg by site-directed mutagenesis to see if
these made the enzyme more Abg-like and resulted in activity in-
creases. In total, three single mutants F196H, S251V, and M369E,
three double mutants F196H/S251V, F196H/M369E, and S251V/
M369E, and one triple mutant F196H/S251V/M369E, were
generated.
cies of up to 3-fold higher for pNP-b- -Man and approximately 2-
D
fold higher for most other substrates (Table 1). This result fits well
with the location of its corresponding position F205 in the active
site of Z. mays (maize) b-glucosidase ZmGlu1, near the 2-position
of its natural substrate DIMBOA- b-D
-glucoside.17,40 Additionally,
this position corresponds to H195 and H193 in Hordeum vulgare
L. (barley) HvBII and rice Os7BGlu26, respectively, both of which
showed preferences for pNP-b-D-Man over pNP-b-D-Glc, but are
N190 and D191 in rice Os3BGlu7 and Os3BGlu8, respectively,
which are primarily b-glucosidases.38 Presumably, improved
hydrogen bonding or Van der Waals interactions at the transition
state as a result of the F196H substitution bring about this im-
proved behavior. In comparison to the previous study, the substitu-
tion of aspartate at the conserved residue N206 in Pyrococcus
furiosus b-glucosidase did not increase its efficiency toward pNP-
b-D
-Man (49.8 and 3.1 sꢁ1 mMꢁ1 for the wild-type enzyme and
the N206D mutant, respectively), but its preference for manno-
side:glucoside increased 18-fold (from 0.7:100 to 13:100 for the
wild-type enzyme and the N206D mutant, respectively).25 In this
study, the substitution of histidine at the non-conserved residue
F196 in dalcochinase not only increased its efficiency toward
pNP-b-D
-Man (0.06 and 0.17 sꢁ1 mMꢁ1 for the wild-type enzyme
and the F196H mutant, respectively), but also improve its prefer-
ence for mannoside:glucoside about 3-fold (from 0.5:100 to
1.3:100 for the wild-type enzyme and the F196H mutant, respec-
tively). The difference may lie in the fact that the N206D mutation
in P. furiosus b-glucosidase resulted in about 300-fold reduction in
3.2. Production of dalcochinase mutants
All dalcochinase mutants were produced in P. pastoris, and
purified using two chromatographic steps. Purification yields
and specific activities of mutants range from 11% to 60% and
efficiency for hydrolysis of pNP-b-D-Glc but only 16-fold reduction

K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
39
Table 1
Catalytic efficiencies for hydrolysis of pNP-glycosides by the recombinant wild-type and mutant forms of dalcochinase
Enzyme
kcat/Km (sꢁ1 mMꢁ1
pNP-b-
)
pNP-b-
D
-Glc
pNP-b-
65.0 0.1
74
18.7 0.1
51.9 0.1
45.6 0.3
42.7 0.4
D
-Fuc
pNP-a-
L
-Ara
D-Gal
pNP-b-
D
-Xyl
pNP-b-D-Man
Wild-type
F196H
S251V
M369E
F196H/S251V
F196H/M369E
S251V/M369E
F196H/S251V/M369E
13.772 0.003
13.59 0.02
3.277 0.003
7.52 0.01
9.64 0.01
7.56 0.01
6.5 0.1
0.844 0.001
1.63 0.01
0.3898 0.0002
0.6340 0.0002
0.1822 0.0001
0.500 0.002
0.260 0.001
0.4615 0.0003
0.1214 0.0001
0.2260 0.0004
0.1483 0.0002
0.080 0.001
0.0628 0.0002
0.1731 0.0002
N.D.a
0.0558 0.0002
0.0114 0.0001
0.0274 0.0001
0.01386 0.00002
0.00786 0.00003
5
0.376 0.001
0.5000 0.0003
0.4079 0.0003
0.845 0.001
0.526 0.001
0.3178 0.0003
0.2225 0.0003
0.282 0.003
35
1
0.2339 0.0003
0.1059 0.0004
0.117 0.001
0.1059 0.0001
5.19 0.01
30.4 0.2
a
N.D., no absorbance change was detected during overnight incubation.
Figure 2. The differences in the activation free energy (DDGà) for the formation of the ESà complex between the mutant and the recombinant wild-type dalcochinase. Note
that the value of DDGà of the S251V mutant toward pNP-b-
D-Man is absent in this plot because the catalytic efficiency could not be determined.
Table 2
Parameters of LFER plots between the log kcat/Km for the recombinant wild-type and mutant forms of dalcochinase versus the corresponding value for Abg
LFER parameter
Dalcochinase forms
F196H/S251V F196H/M369E
wild-type
F196H
S251V
M369E
S251V/M369E
F196H/S251V/M369E
Slope
1.01
0.88
0.87
0.86
0.84
0.80
0.94
0.85
1.15
0.88
1.09
0.91
1.09
0.89
1.10
0.86
Correlation coefficient (R2)
in efficiency for hydrolysis of pNP-b-
mutation in dalcochinase, the catalytic efficiency for hydrolysis of
pNP-b- -Glc was not significantly affected while the efficiency for
hydrolysis of pNP-b- -Man increased 3-fold. So, it is possible that
D
-Man, whereas for the F196H
binding pockets that contribute to transition state stabilization,
possibly coupled with differences in dynamic behavior, geometry,
and electrostatic interactions of the rest of the enzyme.
D
D
the conserved residue N206 in P. furiosus b-glucosidase, corre-
sponding to N181 in dalcochinase, plays a dominant role in both
binding and catalysis of the glycoside substrate.
Ironically, however, although improving catalytic efficiencies,
the F196H mutation does not make the enzyme more Abg-like,
but rather worsens the correlation in the LFER—with a slope of
0.87 and a correlation coefficient of 0.86 (Table 2 and Supplemen-
tary Fig. 5). With the exception of the M369E mutation, which im-
3.4. Interaction between mutations
The double and triple mutants were created to test the interac-
tions among these 3 positions, following the theory proposed pre-
viously.41 The effects of multiple mutations were interpreted based
on the differences in the activation energy for the formation of the
glycosyl-enzyme intermediates (DDGà, from kcat/Km values) be-
tween the wild-type and mutant enzymes as shown in Fig. 2. In
proves catalytic efficiency for pNP-b-
D-Gal by 25% (while
this principle, the effects of the two single mutations (DDGà and
1
decreasing efficiencies for all other substrates), all other mutations
proved to be deleterious. This is reflected graphically in Fig. 2,
which shows changes in activation free energy for each sub-
strate/mutant pair. Therefore, the generally lower catalytic effi-
ciencies of dalcochinase than Abg for various glycoside substrates
do not derive only from differences in ‘‘first sphere’’ active site res-
idues located in the glycone binding pocket, but rather draw from
differences in more remote residues in both glycone and aglycone
DDGà2, where DDGà
>
DDGà2) were compared with the effect of
1
the double mutation (DDGà1+2) and the sum of the effects of the
two single mutations (DDGà DDGà2). Interactions of two muta-
+
1
tions can be classified into one of five categories, which are addi-
tion, partial addition, antagonism, synergism, and no effect.41,42
The analysis of the interactions between mutations performed
in this study is shown in Table 3 (and detailed analysis shown in
Supplementary Table 3). For most pairs of mutations, antagonistic

40
K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
Table 3
by the presence of unique amino acid residues, but also their con-
text in the overall active site architecture, including interactions
with remote residues, geometry, and electrostatic field of the rest
of the enzyme.
Analysis of the interactions between mutations performed in this study
Substrate
F196H/
S251V
F196H/
M396E
S251V/
M369E
F196H/
S251V/
M369E
pNP-b-
pNP-b-
pNP-a-L-Ara
D
D
-Glc
-Fuc
Antagonism Antagonism Antagonism Antagonism
Antagonism Synergism Antagonism Antagonism
Antagonism Antagonism Antagonism Synergism
Acknowledgments
The authors especially thank Watchara Kasinrerk for providing
the mouse monoclonal antibody against natural dalcochinase. The
project was financially supported by Grants from the Thailand Re-
search Fund and the Commission on Higher Education
(RMU5480004); the Commission on Higher Education; Kasetsart
University Research and Development Institute (V-T(D)22.50);
the Higher Education Research Promotion and National Research
University Project of Thailand; the Faculty of Science (ScRF-S2-
18/2551) and the Graduate School, Kasetsart University; and the
Natural Sciences and Engineering Research Council of Canada.
K.R. and K.C. are recipients of the Ph.D. scholarships from the Com-
mission on Higher Education, Thailand.
pNP-b-
D
-Gal
-Xyl
Antagonism Synergism
Antagonism Synergism
Antagonism Synergism
pNP-b-
D
Partial
Antagonism
addition
Antagonism Antagonism
pNP-b-
D
-Man
Antagonism Synergism
interactions, in which the effects of the double mutations were less
than the effects of the more damaging single mutation, were ob-
served, suggesting that these mutations represented opposing
structural effects on the same catalytic step such that the effect
of one mutation would partially rescue the damaging effect of
the other mutation. On the other hand, synergistic interactions,
in which the effects of the double mutations exceeded the sum
of the effects of the two single mutations, were observed in six
cases. Synergism can be caused by three conditions; (1) anti-coop-
erative interaction of residues that introduce strain into the transi-
tion state of the same rate-limiting step, (2) extensive unfolding of
enzyme as a result of the double mutations, or (3) noninteracting
residues that slow down the same non rate-limiting step such that
it becomes a rate-limiting step in the double mutant.41,42 The first
condition was unlikely as F196 and M369 were placed 12.2 Å apart
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.carres.2012.
10.018.
References
1. Henrissat, B.; Bairoch, A. Biochem. J. 1996, 316, 695–696.
2. Ketudat Cairns, J. R.; Esen, A. Cell. Mol. Life Sci. 2010, 67, 3389–3405.
3. www.cazy.org/Glycoside-Hydrolases.html.
(measured from F196-CZ and M369-Ce) in the dalcochinase model,
whereas interacting residues that contribute to a change in DDGà
should be located within 4 Å to be in Van der Waals contact.43
The second condition was also unlikely as the kinetic parameters
for these mutants are roughly comparable to those of the wild-type
enzyme. So, the third condition seemed the most likely explanation
for the synergistic interaction between these positions. Lastly, a
partially additive effect, in which the effect of the double
mutations was greater than the effect of the more damaging single
mutation alone but still lower than the sum of the effects of
the two single mutations, was observed in the hydrolysis of pNP-
4. Esen, A. b-Glucosidase: Overview. In b-Glucosidase Biochemistry and Molecular
Biology; Esen, A., Ed.; American Chemical Society: Washington, DC, 1992; pp 1–
14.
5. McCarter, J. D.; Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885–892.
6. Ketudat-Cairns, J. R.; Champattanachai, V.; Srisomsap, C.; Wittman-Liebold, B.;
Thiede, B.; Svasti, J. J. Biochem. 2000, 128, 999–1008.
7. Svasti, J.; Srisomsap, C.; Techasakul, S.; Surarit, R. Phytochemistry 1999, 50, 739–
743.
8. Srisomsap, C.; Svasti, J.; Surarit, R.; Champattanachai, V.; Sawangareetrakul, P.;
Boonpuan, K.; Subhasitanont, P.; Chokchaichamnankit, D. J. Biochem. 1996, 119,
585–590.
9. Wakarchuk, W. W.; Kilburn, D. G.; Miller, R. C., Jr; Warren, R. A. J. Mol. Gen.
Genet. 1986, 205, 146–152.
10. Day, A. G.; Withers, S. G. Biochem. Cell Biol. 1986, 64, 914–922.
11. Withers, S. G.; Rupitz, K.; Trimbur, D.; Warren, R. A. J. Biochemistry 1992, 31,
9979–9985.
12. Wang, Q.; Trimbur, D.; Graham, R.; Warren, R. A. J.; Withers, S. G. Biochemistry
1995, 34, 14554–14562.
13. Kempton, J. B.; Withers, S. G. Biochemistry 1992, 31, 9961–9969.
14. Rivera-Sagredo, A.; Canada, F. J.; Nieto, O.; Jimenez-Barbero, J.; Martin-Lomas,
M. Eur. J. Biochem. 1992, 209, 415–422.
b-D-Xyl by the S251V/M369E mutant. Partially additive interac-
tions could be due to (1) partially cooperative interactions between
residues in the same step, or (2) independent residues acting in the
consecutive non rate-limiting steps.41,42 Since these residues were
placed 14.16 Å apart (measured from S251-OG and M369-Ce) in
the dalcochinase model, only the second explanation is possible.
15. Namchuk, M. N.; Withers, S. G. Biochemistry 1995, 34, 16194–16202.
16. Barrett, T.; Suresh, C. G.; Tolley, S. P.; Dodson, E. J.; Hughes, M. A. Structure
1995, 3, 951–960.
4. Conclusion
17. Czjzek, M.; Cicek, M.; Zamboni, V.; Burmeister, W. P.; Bevan, D. R.; Henrissat,
B.; Esen, A. Biochem. J. 2001, 354, 37–46.
18. Corbett, K.; Fordham-Skelton, A. P.; Gatehouse, J. A.; Davis, B. G. FEBS Lett. 2001,
509, 355–360.
19. Marana, S. R.; Jacobs-Lorena, M.; Terra, W. R.; Ferreira, C. Biochim. Biophys. Acta
2001, 1545, 41–52.
20. Verdoucq, L.; Morinière, J.; Bevan, D. R.; Esen, A.; Vasella, A.; Henrissat, B.;
Czjzek, M. J. Biol. Chem. 2004, 279, 31796–31803.
21. Isorna, P.; Polaina, J.; Latorre-García, L.; Cañada, F. J.; González, B.; Sanz-
Aparicio, J. J. Mol. Biol. 2007, 371, 1204–1218.
22. Nijikken, Y.; Tsukada, T.; Igarashi, K.; Samejima, M.; Wakagi, T.; Shoun, H.;
Fushinobu, S. FEBS Lett. 2007, 581, 1514–1520.
23. Chuenchor, W.; Pengthaisong, S.; Robinson, R. C.; Yuvaniyama, J.; Oonanant,
W.; Bevan, D. R.; Esen, A.; Chen, C.-J.; Opassiri, R.; Svasti, J.; Cairns, J. R. K. J. Mol.
Biol. 2008, 377, 1200–1215.
24. Tribolo, S.; Berrin, J.-G.; Kroon, P. A.; Czjzek, M.; Juge, N. J. Mol. Biol. 2007, 370,
964–975.
Mutations of the residues in the glycone binding pocket of dal-
cochinase to those of Abg showed that the F196H mutation could
improve the activity of the enzyme on some substrates, especially
the mannoside. This is consistent with the fact that two other GH1
glycosidases (barley HvBII and rice Os7Bglu26), which prefer b-
mannosides over b-glucosides as substrates, both naturally have
histidine at this position. Presumably, the histidine residue is able
to form an improved hydrogen bond with manno-configured sub-
strates at the transition state for the glycosylation step. Since
hydrogen bond interactions between the hydroxyl group at the
C-2 position and the residues in the active site pocket of the en-
zyme contribute directly to transition state stabilization, our result
has implications for the molecular mechanism for determining
glycone specificity in other GH1 b-glucosidases. However, other
mutations made the enzyme less Abg-like in its substrate
specificity. So, the behavior of the enzyme is not only determined
25. Kaper, T.; Van Heusden, H. H.; Van Loo, B.; Vasella, A.; Van der Oost, J.; De Vos,
W. M. Biochemistry 2002, 41, 4147–4155.
26. Toonkool, P.; Metheenukul, P.; Sujiwattanarat, P.; Paiboon, P.; Tongtubtim, N.;
Ketudat-Cairns, M.; Ketudat-Cairns, J.; Svasti, J. Protein Expr. Purif. 2006, 48,
195–204.

K. Ratananikom et al. / Carbohydrate Research 373 (2013) 35–41
41
27. Thompson, J. D.; Higgins, D. G.; Gibson, T. J. Nucleic Acids Res. 1994, 22, 4673–
4680.
36. Withers, S. G.; Rupitz, K. Biochemistry 1990, 29, 6405–6409.
37. Burmeister, W. P.; Cottaz, S.; Rollin, P.; Vasella, A.; Henrissat, B. J. Biol. Chem.
2000, 275, 39385–39393.
38. Kuntothom, T.; Luang, S.; Harvey, A. J.; Fincher, G. B.; Opassiri, R.; Hrmova, M.;
Ketudat Cairns, J. R. Arch. Biochem. Biophys. 2009, 491, 85–95.
39. Cereghino, J. L.; Cregg, J. M. FEM Microbiol. Rev. 2000, 24, 45–66.
40. Hill, A. D.; Reilly, P. J. Biopolymers 2008, 89, 1021–1031.
41. Mildvan, A. S.; Weber, D. J.; Kuliopulos, A. Arch. Biochem. Biophys. 1992, 294,
327–340.
28. Eswar, N.; Webb, B.; Marti-Renom, M. A.; Madhusudhan, M. S.; Eramian, D.;
Shen, M. Y.; Pieper, U.; Sali, A. Curr. Protoc. Protein Sci. 2007, 50, 2.9.1–2.9.31.
29. Luthy, R.; Bowie, J. U.; Eisenberg, D. Nature 1992, 356, 83–85.
30. Laskowski, R. A.; MacArthur, M. W.; Moss, D. S.; Thornton, J. M. J. Appl. Cryst.
1993, 26, 283–291.
31. Wiederstein, M.; Sippl, M. J. Nucleic Acids Res. 2007, 35, 407–410.
32. Kongsaeree, P. T.; Ratananikom, K.; Choengpanya, K.; Tongtubtim, N.;
Sujiwattanarat, P.; Porncharoennop, C.; Onpium, A.; Svasti, J. J. Mol. Catal. B:
Enzyme 2010, 67, 257–265.
42. Mildvan, A. S. Biochemistry 2004, 43, 14517–14520.
43. Wells, J. A. Biochemistry 1990, 29, 8509–8517.
33. Charoenrat, T.; Ketudat-Cairns, M.; Jahic, M.; Enfors, S.-O.; Veide, A. J.
Biotechnol. 2006, 122, 86–98.
44. Hancock, S. M.; Corbett, K.; Fordham-Skelton, A. P.; Gatehouse, J. A.; Davis, B. G.
ChemBioChem 2005, 6, 866–875.
34. Laemmli, U. K. Nature 1970, 227, 680–685.
35. Wicki, J.; Schloegl, J.; Tarling, C. A.; Withers, S. G. Biochemistry 2007, 46, 6996–
7005.