Role of hydrophobic residues in the aglycone binding subsite of a GH39 β-xylosidase in alkyl xylosides synthesis

Article abstract of DOI:10.1016/j.molcatb.2013.06.001

The GH39 β-xylosidase from Bacillus halodurans (BhXyl39) was previously reported to catalyze the synthesis of alkyl xylosides from donors such as pNP β-d-xylopyranoside or xylobiose and from aliphatic alcohols acting as acceptors with chain length inferior to five carbons. In the present study, the role played by aromatic residues present in the aglycone binding subsite of BhXyl39 in the hydrolysis and transglycosylation reactions of the enzyme was investigated. In this way, site-directed mutagenesis was carried out in order to highlight the role of three targeted hydrophobic residues F116, F167 and Y284. These residues were replaced by alanine to decrease the steric hindrance or were mutated into serine to evaluate the impact of the presence of a polar residue into the aglycone binding subsite of BhXyl39. Taking into account kinetic parameters and yields of transglycosylation, the function of each mutated residue in the catalytic mechanism was studied. Results concerning transglycosylation reactions in the presence of pentan-1-ol and octan-1-ol indicated that yields of transglycosylation were impacted both by the position and the nature of the mutated residues. These results were consistent with molecular docking performed with both acceptors which notably confirms that among the three targeted residues, F116 represents the most interesting one for mutagenesis to increase the transglycosylation reactions in presence of long chain alcohols.

Full text of DOI:10.1016/j.molcatb.2013.06.001

Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Role of hydrophobic residues in the aglycone binding subsite of a
GH39 ␤-xylosidase in alkyl xylosides synthesis
Marjorie Ochsa,b, Nicolas Belloyc,d, Manuel Dauchezc,d, Murielle Muzarde,
Richard Plantier-Royone, Caroline Rémonda,b,∗
a Université de Reims Champagne-Ardenne, UMR614 Fractionnement des AgroRessources et Environnement, 51100 Reims, France
b INRA, UMR614 Fractionnement des AgroRessources et Environnement, 51100 Reims, France
c FRE CNRS 3481 Matrice Extracellulaire et Dynamique Cellulaire, UFR Sciences Exactes et Naturelles, Université de Reims Champagne-Ardenne, 51687
Reims Cedex 2, France
d Plateforme de Modélisation Moléculaire Multi-échelle, UFR Sciences Exactes et Naturelles, Université de Reims Champagne-Ardenne, 51687 Reims Cedex
2, France
e Institut de Chimie Moléculaire de Reims ICMR, CNRS UMR 7312, Université de Reims Champagne-Ardenne, 51687 Reims Cedex 2, France
a r t i c l e i n f o
a b s t r a c t
Article history:
The GH39 ␤-xylosidase from Bacillus halodurans (BhXyl39) was previously reported to catalyze the syn-
thesis of alkyl xylosides from donors such as pNP ␤-d-xylopyranoside or xylobiose and from aliphatic
alcohols acting as acceptors with chain length inferior to five carbons. In the present study, the role
played by aromatic residues present in the aglycone binding subsite of BhXyl39 in the hydrolysis and
transglycosylation reactions of the enzyme was investigated. In this way, site-directed mutagenesis was
carried out in order to highlight the role of three targeted hydrophobic residues F116, F167 and Y284.
These residues were replaced by alanine to decrease the steric hindrance or were mutated into serine
to evaluate the impact of the presence of a polar residue into the aglycone binding subsite of BhXyl39.
Taking into account kinetic parameters and yields of transglycosylation, the function of each mutated
residue in the catalytic mechanism was studied. Results concerning transglycosylation reactions in the
presence of pentan-1-ol and octan-1-ol indicated that yields of transglycosylation were impacted both
by the position and the nature of the mutated residues. These results were consistent with molecular
docking performed with both acceptors which notably confirms that among the three targeted residues,
F116 represents the most interesting one for mutagenesis to increase the transglycosylation reactions in
presence of long chain alcohols.
Received 12 March 2013
Received in revised form 23 May 2013
Accepted 3 June 2013
Available online 11 June 2013
Keywords:
␤-Xylosidase
Transglycosylation reaction
Alkyl xylosides
Hydrophobic residues
Aglycone binding subsite
Molecular docking
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
and the xylo-oligosaccharide acts as the donor. The transglycosyla-
tion ability of ␤-xylosidases or xylanases represents an interesting
Depending on their amino acid sequence similarities, glycoside
hydrolases (GH) are classified into different families by the CAZy
database (http://www.cazy.org) [1]. The catalytic mechanisms of
these enzymes hydrolyzing glycosidic linkages lead either to the
retention or to the inversion of the configuration of the anomeric
carbon [2].
Endo-xylanases (EC 3.2.1.8) are glycoside hydrolases that cat-
alyze the hydrolysis of xylans into xylose and xylo-oligomers
which could be further hydrolyzed into xylose by ␤-xylosidases
(EC 3.2.1.37) [3]. In the presence of alcohols, ␤-xylosidases can also
catalyze the synthesis of alkyl xylosides by transglycosylation or
reverse hydrolysis [4]. In this case, the alcohol acts as the acceptor
route for the valorization of pentoses coming from agricultural by-
products in a context of biorefineries development [5–8]. Indeed,
alkyl pentosides have exhibited interesting surfactants properties
for detergent applications [8,9]. However, the transxylosylation
yields depend on the alcohol chain length as low yields were
obtained with long chain alcohols [5,7,10]. Alkyl chain length of
alkyl glycosides impacts their physico-chemical properties such as
critical micellar concentration that decreases with increasing alkyl
chain length [11].
The active site of glycoside hydrolases is composed of subsites
that bind the donor (glycone subsite) and others that bind the
acceptor (aglycone subsite). These subsites allow the correct posi-
tioning of the substrate [12]. It has been pointed out that aromatic
residues in the aglycone binding subsite are implicated in the trans-
glycosylation reaction catalyzed by chitinases and cyclodextrine
glycotransferases as their replacement by non aromatic residues
∗
Corresponding author. Tel.: +33 326 773 652; fax: +33 326 773 599.
E-mail address: caroline.remond@univ-reims.fr (C. Rémond).
1381-1177/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.molcatb.2013.06.001

22
M. Ochs et al. / Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
Table 1
Cells were broken by two passages in a French press apparatus at
Sequence of the primers used for the mutagenesis.
16,000 psi. The solution was then centrifuged (20 min, 10 000 × g)
in order to eliminate the cellular fragments. The supernatant
obtained was filtrated through a 0.45 ␮m membrane and charged
on IMAC resin for purification. The resin was rinsed three times
by sodium phosphate/NaCl buffer pH 7.4 containing imidazole
(20 mM) and then the BhXyl39 and the mutants were eluted
three times by sodium phosphate/NaCl buffer pH 7.4 containing
imidazole (250 mM). The purity of each mutant and wild-type
BhXyl39 was analyzed by SDS-PAGE (12%, w/v). The concentra-
tion of purified enzyme was determined by the Beer–Lambert
method, measuring the absorbance at 280 nm and using ε as
Mutation
Sequence (5ꢀ→3ꢀ)
F116A
F116S
F167A
F167S
Y284A
Y284S
ggcttctggggaacaaacgattgccgattggcaaggg
ggcttctggggaacaaacgattagcgattggcaaggg
ggaatgagccgaatttaatcaacgcttggcagcatgccgataa
ggaatgagccgaatttaatcaacagttggcagcatgccgataa
ttacggagtataatacatccgccagtccgattaacccggttc
ttacggagtataatacatccaccagtccgattaacccggttc
decreases or even suppresses the transglycosylation [13–15]. In
these studies, the acceptor was a glycosidic compound. However,
the role of aromatic residues in the aglycone binding subsite when
linear alcohols act as acceptors during the synthesis of alkyl glyco-
sides remains unknown.
1.30505 mL mg−1 cm−1
.
2.3. Circular dichroism (CD) analysis
The aim of this work consists in evaluating the role of aro-
matic residues in the aglycone binding pocket of a ␤-xylosidase
from Bacillus halodurans belonging to the GH39 family. This enzyme
has exhibited good transxylosylation abilities in the presence
of pNP ␤-d-xylopyranoside or xylotriose as acceptor and donor
[16] for oligosaccharides synthesis. In the presence of pNP ␤-d-
xylopyranoside and primary alcohols with chain lengths up to 5
carbons, alkyl xylosides were obtained in satisfying amounts [7].
However, aliphatic alcohols displaying longer alkyl chains were
unsuitable as acceptors. Various experimental conditions were
tested such as vigorous stirring or addition of co-solvents (tert-
butanol, DMSO) in order to alleviate the presence of the biphasic
system which could hinder the enzyme efficiency. The transglyco-
sylation rates were not improved in these cases and we assumed
that the aromatic residues present in the −1 subsite of the enzyme
could affect the arrival and/or stabilization of the acceptor into the
aglycone binding subsite. In this context, mutations of these aro-
matic amino acids into alanine or serine were investigated in order
to understand the effect of steric hindrance and polarity in the agly-
cone binding site during the synthesis of pentyl and octyl xylosides
catalyzed by the ␤-xylosidase.
Wild-type and mutant enzymes were prepared for CD analysis
by dialysis against potassium phosphate buffer pH 6.4. CD spec-
tra were recorded at 20 ◦C using a Jasco J-810 spectrometer (Jasco,
Tokyo, Japan), collecting data from 200 to 300 nm, at 50 nm min−1
with a 1 nm bandwidth using a 10 mm quartz cuvette contain-
ing the wild-type and mutant enzymes. Each spectrum shown is
the average of three individual scans. Collected CD data were cor-
rected for absorbance by the buffer and expressed in function of
the concentration of the enzyme.
2.4. Enzyme assays and kinetics
The standard activity assay used pNP ␤-d-Xylp as substrate.
Release of p-nitrophenol (pNP) was measured by continuous mon-
itoring at 401 nm. Reactions were performed in 50 mM Tris–HCl,
pH 7.5 and the substrate concentration was 5 mM. Assays were
performed at 45 ◦C for 10 min. Activity was determined by using
a standard calibration curve for pNP (ε = 12 514 M−1 cm−1). One
unit of ␤-xylosidase activity was defined as the amount of enzyme
releasing 1 ␮mol of pNP per minute.
Kinetic parameters Km and kcat were determined using vari-
able substrate concentrations (1–75 mM) in the standard assay.
Data were calculated from triplicate assays. Kinetic parameters
were derived from Michaelis–Menten representations using the
SigmaPlot 2000 software (version 6.1 with Enzyme Kinetics module
1.0; SPSS Science, Paris, France).
2. Experimental
2.1. Site-directed mutagenesis
The wild-type xylosidase (BhXyl39) was produced from the con-
struction pET28b-BH1068 encoding GH39 xylosidase which was
obtained in a previous work after amplification of the ORF BH1068
from B. halodurans C-125 [16]. The cloning strategy allowed the
insertion of a C-terminal His-tag [16]. Site-directed mutagenesis
experiments were performed by using the kit “QuickChange site-
directed mutagenesis” (Stratagene, La Jolla, CA, USA) using the
construction pET28b-BH1068 encoding GH39 xylosidase [16]. For
each single mutant, a sense/antisense primer was designed. The
primers corresponding to the sense strand are listed in Table 1 and
the codon harboring the mutation is underlined. The incorporation
of mutations was verified by DNA sequencing.
2.5. Transglycosylation reactions in the presence of aliphatic
alcohols
The synthesis of pentyl and octyl xyloside by the wild-type
enzyme and the mutant enzymes were carried out in water. The
reaction medium was composed by the donor, pNP ␤-d-Xylp
(5 mM), the acceptor, pentan-1-ol or octan-1-ol (5 and 10%, v/v,
respectively), the enzyme (0.1 IU/mL) and an internal standard for
HPLC analysis, hexyl glucoside (3 mM). The reactions (1–1.5 mL)
were carried out in closed glass vessels with a magnetic stirring
and were incubated in a thermostated oil bath for 1 h at 45 ◦C
under vigorous agitation (1000 r.p.m.). The reactions were stopped
by incubating the reaction mixtures during 10 min at 100 ◦C, then
were evaporated using a speed vac concentrator and resuspended
in DMSO (90 ␮L), centrifuged (10 min, 2000 × g) in order to pellet
the residues and filtered on 0.2 ␮m (PTFE membrane).
The quantification of alkyl xylosides was performed by HPLC
using a RP-C18 column (Nucleodur 100-5 C18 ec, 250 mm × 4 mm,
Macherey Nagel). Standard alkyl xylosides were purified as pre-
viously described [8]. Products were eluted at 0.6 mL/min with a
mobile phase composed of an acetonitrile:water mixture (20:80
for pentyl xylosides and 40:60 for octyl xylosides). The detection
of eluates was performed with a dynamic light scattering detector
2.2. Expression and purification of BhXyl39 and its mutants
Escherichia coli BL21star (DE3) cells were transformed by heat
shock with the pET28b-BH1068 construction that was or not
mutated as described above. Transformed bacteria were cultured
(37 ◦C, 150 r.p.m.) in 1.6 L of LB medium containing kanamycin
(30 ␮g/mL). When the optical density at 600 nm reached 0.6, the
production of enzymes was induced (20 ◦C, overnight, 150 r.p.m.)
by 0.2 mM of isopropyl thio-␤-d-galactoside. The induced bacteria
were harvested by centrifugation (15 min, 6000 × g), resuspended
in 40 mL of a sodium phosphate/NaCl buffer pH 7.4 and freezed.

M. Ochs et al. / Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
23
(PL-ELS 1000; Polymer Laboratories). The amounts of xylosides
produced were corrected by using the internal standard concen-
tration detected. Results are expressed as mean values obtained
from triplicate reactions which allow calculating standard errors.
2.6. Sequence alignment
The work on proteic and nucleotidic sequences was performed
using online tools provided by EMBL-EBI (http://srs.ebi.ac.uk), par-
ticularly the tool clustalw2 for multiple sequences alignment.
2.7. Molecular docking
Molecular docking was based on the structure of the xylosi-
dase GH39 from Geobacillus stearothermophilus (PDB code: 2BFG).
Structures (n-pentanol, n-octanol and 2BFG) were built and/or
prepared for docking with the Schrödinger 2012 suite (Protein
Preparation Wizard and LigPrep module) [17]. The grid for the 2BFG
xylosyl-enzyme intermediate was calculated with a scaling of 0.8
for charges and Van der Waals radii in the binding pocket. This
strategy was used after trying to reproduce experimental confor-
mations of xylose in 2BS9 native GH39 from G. stearothermophilus
enzyme (data not shown). Docking was performed using Glide,
from Schrödinger suite [17], in XP precision in wild-type enzyme in
order to assess critical roles for amino acids surrounding the agly-
cone binding pocket. For n-octanol, docking was performed after
generating an ensemble (50) of low-energy conformations with
the Macromodel Systematic Torsional Sampling tool. These con-
formations were then docked in 2BFG enzyme aglycone binding
site. Suitable poses were those compatible with transglycosylation
reaction, i.e. with alcohol function properly oriented toward the
anomeric carbon of xylose. Analysis and pictures were done in Mae-
stro environment [17]. For the molecular modeling part, residue
numbering is based on G. stearothermophilus protein.
Fig. 1. Circular dichroism spectra of the wild-type BhXyl39 and of the mutants.
absorption spectra obtained. The spectra corresponding to the
enzymes harboring the mutations F116S, F167A, F167S, Y284A and
Y284S are similar to those obtained with the wild-type enzyme,
in particularly for wavelengths between 220 and 230 nm, corre-
sponding to ␣-helix. The spectrum of the mutant F116A has its
maximum absorbance slightly moved between 230 and 240 nm.
In a previous study with a ␤-glucosidase isolated in Zea mays for
which a phenylalanine was replaced by an alanine residue into the
aglycone subsite, authors obtained similar absorption spectra for
the reference enzyme and the mutant [19]. This result indicates
that the secondary structure, particularly ␣-helix, is not modified
by the mutation. Consequently, the differences of catalytic proper-
ties that could be observed for the mutants are directly related to
the mutations and could not be attributed to a modification of the
secondary structure of the proteins.
3. Results and discussion
3.1. Mutation strategy
3.3. Effect of mutations on hydrolytic abilities of the mutants
The structure of the BhXyl39 is currently not known.
Nevertheless, the structure of a GH39 ␤-xylosidase from G.
stearothermophilus showing 66% of identity with BhXyl39 (deter-
mined by sequence alignment) was described and the aglycone
binding subsite was studied [18]. The aglycone binding pocket of
GH39 ␤-xylosidase from G. stearothermophilus is defined by seven
residues among which four are aromatic residues. A sequence align-
ment of the ten known ␤-xylosidases belonging to the family 39 of
glycoside hydrolases (http://www.cazy.org/) showed that three of
these aromatic residues are highly conserved (see supplementary
data). These residues F116, F167 and Y284 for BhXyl39 were thus
chosen for site-directed mutagenesis. Mutagenesis experiments
were conducted in order to replace these three residues by ala-
nine or serine residues. Our objective was to decrease the steric
hindrance into the aglycone subsite by replacing aromatic residues
by alanine but also to study the impact of the presence of a polar
residue into this subsite by introducing a serine residue.
The kinetic parameters for hydrolysis of pNP ␤-d-Xylp were
calculated using the Michaelis–Menten model for the wild-type
BhXyl39 and for the mutants.
No kinetic parameters could be established for the mutants
Y284S and Y284A because no steady-state was obtained when
the maximal rate in function of the substrate concentration was
represented (data not shown), indicating that the suppression of
the tyrosine at position 284 dramatically hinders the hydrolytic
function of the enzyme. In the case of the ␤-xylosidase from G.
stearothermophilus, it has been supposed that the corresponding
tyrosine could form stacking interactions with the pNP group from
the substrate [18]. Moreover, it has been pointed out that the aro-
matic residues in the aglycone subsites +1 and +2 from cellulases
do not appear to participate in the tight binding to the ligand, but
rather in structuring the active site tunnel appropriately for ideal
positioning of the ligand for catalysis [20]. Hence, we can assume
that the tyrosine in position 284 allows the right positioning of the
substrate pNP ␤-d-xylopyranoside required for the formation of the
glycosyl-enzyme intermediate.
The kinetic parameters calculated for the wild-type BhXyl39
and for the mutants F116A, F116S, F167A and F167S are described
in Table 2. The results indicate that the mutation considerably
affects the catalytic efficiency as the values of kcat/Km decrease
from 690 mM−1 s−1 for BhXyl39 to 40 and 44 mM−1 s−1 for F116A
and F116S mutants and to 5 and 6 mM−1 s−1 for F167A and F167S
mutants. This diminution is mainly due to the decrease of the cat-
alytic constant kcat, 14-fold and 57-fold, respectively, for F116S and
The six mutated enzymes (F116A, F116S, F167A, F167S, Y284A,
Y284S) as well as the wild-type BhXyl39, were successfully pro-
duced in E. coli and purified by immobilized metal ion affinity
chromatography (IMAC) with purity yield at or above 90% in all
cases (data not shown).
3.2. Secondary structure
Circular dichroism measurements were conducted on the wild-
type BhXyl39 and on the mutant enzymes. Fig. 1 represents the

24
M. Ochs et al. / Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
Table 2
Kinetic parameters of the wild-type BhXyl39 and of the mutants for the hydrolysis
of pNP ␤-d-Xylp at pH 7 and at 45 ◦C (values were calculated using the SigmaPlot
2000 software).
Enzyme
kcat (s−1
)
Km (mM)
kcat/Km (mM−1 s−1
)
Wild-type BhXyl39
F116S
F116A
F167S
F167A
9769 432
712 18
654 21
14.14 1.81
16.06 1.15
16.25 1.42
36.45 3.19
24.70 2.74
690.88 119.22
44.36 4.32
40.26 4.83
5.08 0.66
185
171
8
8
6.91 1.09
F167A. On the contrary, it seems that these mutations affect to a
lesser extent the Michaelis constant, as the Km value is slightly mod-
ified for F116A and F116S mutants compared to BhXyl39. Km value
is increased 1.7-fold and 2.6-fold for F167A and F167S mutants,
respectively. Other authors also observed a diminution of the cat-
alytic constant associated with an affinity constant less or not
affected by the mutation of aromatic residues in the aglycone sub-
sites. For instance, the mutation into alanine of a tyrosine at subsite
+2 from a xylanase led to a 100-fold decrease of the catalytic con-
stant whereas the affinity constant was not affected [21]. Other
authors studying the effect of the mutation into alanine of a phen-
ylalanine located in the aglycone subsite of a ␤-glucosidase also
showed that the effect of the mutation varies with the substrate,
Indeed the affinity constant for the pNP ␤-d-glucopyranoside was
divided by a factor 10 but not modified in case of 4-methyl umbel-
liferyl glucopyranoside and the catalytic constant decreased about
40-fold for 4-methyl umbelliferyl glucopyranoside and 100-fold for
pNP glucopyranoside [19].
Owing to our results, it seems that, for each position mutated,
the suppression of the aromatic character of the residue is the
main factor of decrease of hydrolysis activity. The nature of the
non-aromatic residue, polar or not, which replaces the aromatic
one has less influence on hydrolysis. Indeed, the kinetic parame-
ters are close for enzymes harboring the single mutation in alanine
or serine at the same position. Thus, mutants F116S and F116A
have catalytic constants of 712 and 654 s−1, respectively, and their
Michaelis constants are also very close (about 16 mM) and not mod-
ified compared to the wild-type BhXyl39 (14 mM). In the same way,
the kinetic parameters between the mutants F167A and F167S are
very similar.
Fig. 2. Synthesis of pentyl xyloside (a) and octyl xyloside (b) catalyzed by the wild-
type BhXyl39 and the six mutants.
The results obtained for the synthesis of pentyl xylosides are
shown in Fig. 2A. It appears that the mutation of phenylala-
nine at position 116 has a huge impact on the synthesis abilities
of the enzyme as the amount of pentyl xyloside produced was
0.88 0.04 mM and 0.24 0.04 mM for F116A and F116S. Owing
to the kinetic parameters described above, the lowest amount of
pentyl xyloside synthesized could be explained by an inefficient
deglycosylation step during the catalysis. Indeed, docking poses
of pentanol (Fig. 3) show that it fits perfectly into the aglycone
subsite of 2BFG enzyme with F115, S284, P285, D324 and M325
3.4. Effect of mutations on transglycosylation capacities
In a previous study, we observed that the presence of long chain
alcohols (>5 carbons chain) as acceptors hinders the transglycosyla-
tion yields of alkyl xylosides catalyzed by the wild-type BhXyl39 [7].
We also pointed out in another work that the enzymatic synthesis
of octyl oligoxylosides performed by xylanases could be improved
by the addition of a co-solvent such as tert-butanol and by vigorous
stirring, ensuring a good contact between the glycosyl enzyme and
the acceptor poorly soluble in water [8]. Thus, transglycosylation
reactions were performed under vigorous stirring and the addition
of a co-solvent in the presence of octanol was tested. In our prelim-
inary experiments, the yields obtained for octyl xyloside synthesis
was 10-fold lower in the presence of 10% co-solvent. Addition of
higher concentrations of octanol and/or of tert-butanol in order to
promote better interactions between the enzyme and the accep-
tor did not improve the yields of octyl xyloside produced (data not
presented). Consequently, we assumed that these differences in the
presence of pentanol or octanol are directly related to the enzyme
transglycosylation ability and not to the reaction media. In the con-
ditions of study, the amount of product obtained for the synthesis
catalyzed by the wild-type BhXyl39 was 3.80 0.28 mM for pentyl
xyloside whereas the synthesis of octyl xyloside was 20-fold lesser
as 186 20 ␮M of product was synthesized.
˚
at less than 4 A of the methyl group of pentanol. These results
indicate that the steric hindrance due to the presence of the cor-
responding phenylalanine side chain in BhXyl39 (F116) allows an
appropriate positioning and stabilization of short alkyl chain alco-
hols such as pentanol. In case of mutations of the phenylalanine
into smallest residues, the increase of degree of freedom for posi-
tioning the acceptor in the aglycone pocket and the disruption
of hydrophobic interactions should be unsuitable for transgly-
cosylation. Nevertheless, the results obtained for the synthesis
of octyl xylosides, represented in Fig. 2B, are slightly different.
Indeed, when the phenylalanine in position 116 was mutated
into a serine residue, the transglycosylation yields decreased
(167 11 ␮M) compared to BhXyl39 (186 20 ␮M) but signifi-
cantly raised up to 325 38 ␮M in case of mutation into alanine.
Docking poses of octanol in 2BFG enzyme (Fig. 3) evidence a cur-
vature of the last 3 carbons at the end of aglycone binding pocket
with methylene groups 7 and 8 oriented toward the F115 cycle.

M. Ochs et al. / Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
25
Regarding the amounts of octyl xyloside produced, the mutations
at this position affect slightly the synthetic ability for the mutants
F167A and F167S, which, respectively, produced 207 12 ␮M and
249 20 ␮M of octyl xyloside. Furthermore, the distances mea-
sured between pentanol/octanol and F166 in the structure are
˚
between 3.7 and 4.7 A but concerns in both cases the methylene
groups 1–3 confirming that the phenylalanine at position 167 has
less influence in the arrival and/or stabilization of acceptors than
phenylalanine at position 116.
Concerning the mutation of the tyrosine in position 284, the
amounts of pentyl xyloside synthesized by the mutants Y284S and
Y284A (3.59 0.09 mM and 3.51 0.09 mM, respectively) were not
significantlydifferent in comparison with BhXyl39. On the contrary,
the mutation of the tyrosine at position 284 has an extensive influ-
ence on the transglycosylation with octanol as the yields of octyl
xyloside, compared to BhXyl39, dropped about 5-fold (37 20 ␮M)
when the reaction was catalyzed by the mutant Y284S. Further-
more, in case of mutation into alanine, the amounts of octyl xyloside
produced were so low that the concentrations measured by HPLC
were around the limit of detection. Consequently, the amount
shown in Fig. 2 corresponds to a mean of two values and no standard
error could be calculated. This dramatic reduction in the synthe-
sis of octyl xyloside associated with the absence of effect on the
synthesis of pentyl xyloside shows that the tyrosine at position
284 plays a crucial role for the arrival and/or the stabilization
of octanol. If we consider that octanol docking is less favorable
with F115 in wild-type enzyme, the Y283 could act as a block-
age upon the methylene groups 1–3 preventing the early release of
octanol from the aglycone binding pocket leading to a low amount
of octyl xyloside synthetized. Further analysis including molec-
ular dynamics and normal mode analysis could be conducted in
order to take into account the dynamic and breathing of the gly-
cone and aglycone subsites and their effects on the stability of the
complex.
Fig. 3. Best docking poses for n-pentanol and n-octanol in 2BFG xylosyl-enzyme
aglycone binding site. Electrostatic surface and residues targeted are represented.
F115, F166 and Y283 correspond to positions F116, F167 and Y284 in BhXyl39.
These results suppose that the mutation at position 116 could lead
to an enlargement of the aglycone pocket that could allow the
arrival of alcohols with a longer aliphatic chain such as octanol.
In addition, it seems that the presence of a polar group limits the
arrival of octanol by non-favorable interactions between the polar
hydroxyl group of serine and the hydrophobic aliphatic chain of
the acceptor.
4. Conclusion
The single mutation into alanine or serine of three aromatic
residues in the aglycone binding subsite of BhXyl39 did not pro-
mote important modifications of the protein folding showing that
the results obtained could be directly related to the mutation. The
kinetic parameters calculated for the mutants and compared to
the wild-type enzyme pointed out that F116, F167 and Y284 play
an important role in hydrolysis of pNP ␤-d-Xylp. Moreover, Y284
seems to play a crucial role by forming stacking interactions with
the pNP group of the substrate.
The study concerning the role of the aromatic residues in the
transglycosylation reactions in the presence of alcohols is more
complex to explain. Indeed, yields are impacted both by the
position and by the nature of the mutation. Nevertheless, the com-
parison of yields and kinetic parameters helped us to assess the
function of the mutated residues in the catalytic mechanism. Thus,
tyrosine 284 seems to play a crucial role for the affinity with the
donor molecule (pNP ␤-d-Xylp) and the arrival and/or the stabi-
lization of an acceptor with a long aliphatic chain such as octanol.
Results concerning the phenylalanine residues showed that both
phenylalanine 116 and 167 play an important role in the cat-
alytic mechanism as demonstrated by kinetic parameters results
obtained with pNP ␤-d-Xylp. Phenylalanine 116 could be involved
in the deglycosylation step and in the arrival or in the stabilization
of the acceptor alcohol.
The mutation of phenylalanine at position 167 impacts with
a lesser extent the synthetic abilities of the ␤-xylosidase (Fig. 2).
Indeed, the amount of pentyl xyloside produced by the mutant
F167S (3.52 0.05 mM) was not significantly different from the
amount synthesized by BhXyl39. In case of the mutant F167A,
pentyl xyloside production was low (2.52 0.16 mM). The results
concerning the effect of the suppression of phenylalanine residues
in the aglycone subsite on the transglycosylation reactions in the
presence of pentanol could be related to a previous work con-
cerning a ␤-glucosidase for which the effect of the introduction
of phenylalanine instead of valine or asparagine in the aglycone
subsite on transglucosylation was studied [22]. In the presence of
primary aliphatic alcohols with small chain lengths (≤4 carbons),
the transglucosylation was affected in the same proportions inde-
pendently of the residue mutated, as we observed for the mutation
of the phenylalanine of BhXyl39 at position 116 with pentanol as
acceptor where the decreases of the yields of pentyl xyloside pro-
duced were quite similar whether the phenylalanine is mutated
into a serine or an alanine. In the presence of a secondary alco-
hol, the transglucosylation with the ␤-glucosidase was increased in
case of a replacement of a phenylalanine located into the aglycone
subsite by an apolar residue but not in the case of its substitu-
tion by a polar residue [22]. These results are concordant with
the data obtained for the mutation of BhXyl39 at position 167.
Our results indicate that among the three mutated residues,
phenylalanine 116 represents the most interesting target residue
for mutagenesis to increase the transglycosylation reactions in the
presence of long chain alcohols.

26
M. Ochs et al. / Journal of Molecular Catalysis B: Enzymatic 96 (2013) 21–26
Acknowledgments
[7] M. Muzard, N. Aubry, R. Plantier-Royon, M. O’Donohue, C. Remond, J. Mol. Catal.
B: Enzym. 58 (2009) 1–5.
[8] M. Ochs, M. Muzard, R. Plantier-Royon, B. Estrine, C. Remond, Green Chem. 13
(2011) 2380–2388.
[9] E. Papadopoulou, A. Hatjiissaak, B. Estrine, S. Marinkovic, Eur. J. Wood Wood
Prod. 69 (2011) 579–585.
[10] H. Shinoyama, Y. Kamiyama, T. Yasui, Agric. Biol. Chem. 52 (1988) 2197–2202.
[11] W. von Rybinski, K. Hill, Angew. Chem. Int. Ed. 37 (1998) 1328–1345.
[12] D.B. Jordan, X.L. Li, Biochim. Biophys. Acta: Prot. Proteomics 1774 (2007)
1192–1198.
[13] Y. Lue, H. Yang, H. Hu, Y. Wang, Z. Rao, C. Jin, Glycoconj. J. 26 (2009) 525–534.
[14] B.A. van der Veen, H. Leemhuis, S. Kralj, J.C.M. Uitdehaag, B.W. Dijkstra, L.
Dijkhuizen, J. Biol. Chem. 276 (2001) 44557–44562.
This work was supported by the “CPER 2007-2013” (Pentoraf
program). The authors are grateful to the “Région Champagne-
Ardenne” for a doctoral fellowship (M. Ochs) and material funds.
We also thank Ms Aubry (UMR FARE, Reims, France) for technical
help. Circular dichroism analysis and molecular docking were per-
formed at the Multiscale Molecular Modeling Platform (P3M) from
University of Reims Champagne Ardenne (France).
[15] H. Zakariassen, M.C. Hansen, M. Joranli, V.G.H. Eijsink, M. Sorlie, Biochemistry
50 (2011) 5693–5703.
Appendix A. Supplementary data
[16] I. Smaali, C. Remond, M.J. O’Donohue, Appl. Microbiol. Biotechnol. 73 (2006)
582–590.
[17] Suite 2012: Schrödinger Suite 2011 Protein Preparation Wizard; Epik version
2.3; Impact version 5.8; Prime version 3.1; LigPrep, version 5.8; Maestro, ver-
sion 9.3, Schrödinger, LLC, New York, NY, 2012.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.molcatb.
2013.06.001.
[18] M. Czjzek, A. Ben David, T. Braman, G. Shoham, B. Henrissat, Y. Shoham, J. Mol.
Biol. 353 (2005) 838–846.
[19] R. Dopitova, P. Mazura, L. Janda, R. Chaloupkova, P. Jerabek, J. Damborsky, T.
Filipi, N.S. Kiran, B. Brzobohaty, FEBS J. 275 (2008) 6123–6135.
[20] C.M. Payne, Y. Bomble, C.B. Taylor, C. McCabe, M.E. Himmel, M.F. Crowley, G.T.
Beckham, J. Biol. Chem. 286 (2011) 41028–41035.
[21] A. Pollet, S. Lagaert, E. Eneyskaya, A. Kulminskaya, J.A. Delcour, C.M. Courtin,
Biochim. Biophys. Acta: Prot. Proteomics 1804 (2010) 977–985.
[22] P.T. Kongsaeree, K. Ratananikom, K. Choengpanya, N. Tongtubtim, P. Sujiwatta-
narat, C. Porncharoennop, A. Onpium, J. Svasti, J. Mol. Catal. B: Enzym. 67 (2010)
257–265.
References
[1] B. Henrissat, Biochem. J. 280 (1991) 309–316.
[2] S.G. Withers, Carbohydr. Polym. 44 (2001) 325–337.
[3] D. Shallom, Y. Shoham, Curr. Opin. Microbiol. 6 (2003) 219–228.
[4] P. Drouet, M. Zhang, M.D. Legoy, Biotechnol. Bioeng. 43 (1994) 1075–1080.
[5] M. Kurakake, T. Fujii, M. Yata, T. Okazaki, T. Komaki, Biochim. Biophys. Acta:
Gen. Sub. 1726 (2005) 272–279.
[6] G. Mamo, S. Kasture, R. Faryar, S. Hashim, R. Hatti-Kaul, Process Biochem. 45
(2010) 700–705.