Araf51 with improved transglycosylation activities: One engineered biocatalyst for one specific acceptor

Article abstract of DOI:10.1016/j.carres.2014.10.031

A random mutagenesis of the arabinofuranosyl hydrolase Araf51 has been run in order to have access to efficient biocatalysts for the synthesis of alkyl arabinofuranosides. The mutants were selected on their ability to catalyze the transglycosylation reaction of p-nitrophenyl α-l-arabinofuranoside (pNP-Araf) used as a donor and various aliphatic alcohols as acceptors. This screening strategy underlined 5 interesting clones, each one corresponding to one acceptor. They appeared to be much more efficient in the transglycosylation reaction compared to the wild type enzyme whereas no self-condensation or hydrolysis products could be detected. Moreover, the high specificity of the mutants toward the alcohols for which they have been selected validates the screening process. Sequence analysis of the mutated enzymes revealed that, despite their location far from the active site, the mutations affect significantly the kinetics properties as well as the substrate affinity of these mutants toward the alcohol acceptors in the transglycosylation reaction.

Full text of DOI:10.1016/j.carres.2014.10.031

Carbohydrate Research 402 (2015) 50–55
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Araf51 with improved transglycosylation activities: one engineered
biocatalyst for one specific acceptor
a,b,
Alizé Pennec a,b, Richard Daniellou a,b, , Pascal Loyer b,c, Caroline Nugier-Chauvin a,b, , Vincent Ferrières
⇑
⇑
a Ecole Nationale Supérieure de Chimie de Rennes, CNRS, UMR 6226, 11 Allée de Beaulieu, CS 50837, 35708 Rennes Cedex 7, France
b Université européenne de Bretagne, France
c INSERM UMR 991, Hôpital Pontchaillou, 35033 Rennes Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 8 July 2014
Received in revised form 10 October 2014
Accepted 31 October 2014
Available online 8 November 2014
A random mutagenesis of the arabinofuranosyl hydrolase Araf51 has been run in order to have access to
efficient biocatalysts for the synthesis of alkyl arabinofuranosides. The mutants were selected on their
ability to catalyze the transglycosylation reaction of p-nitrophenyl a-L-arabinofuranoside (pNP-Araf) used
as a donor and various aliphatic alcohols as acceptors. This screening strategy underlined 5 interesting
clones, each one corresponding to one acceptor. They appeared to be much more efficient in the transgly-
cosylation reaction compared to the wild type enzyme whereas no self-condensation or hydrolysis prod-
ucts could be detected. Moreover, the high specificity of the mutants toward the alcohols for which they
have been selected validates the screening process. Sequence analysis of the mutated enzymes revealed
that, despite their location far from the active site, the mutations affect significantly the kinetics proper-
ties as well as the substrate affinity of these mutants toward the alcohol acceptors in the transglycosy-
lation reaction.
Keywords:
Furanosides
Transglycosylation
Random mutagenesis
Furanosidase
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
preparation of non-natural furanosyl-containing conjugates (for a
review, see Ref. 11). Indeed, glycosidases are good candidates in
One of the current needs in chemical research is the develop-
ment of greener and more eco-friendly reactions. The glycosylation
is an example thereof, being one of the most important and diffi-
cult reactions for biomolecule synthesis. Indeed, it still remains a
challenge in chemistry requiring multiple reaction steps and the
use of various protecting groups. In Nature there are multiple
enzymes as glycosyltransferases, polysaccharide lyases, and glyco-
sidases able to assemble or cleave easily the glycosidic bond with
high regioselectivity.1 Nowadays, biotechnological approaches are
recognized as powerful tools for the synthesis of glycoconjugates
and improving biocatalyst efficiency seems to be one of the most
promising approaches.2–7 Moreover it is noteworthy that the
chemical synthesis of arabinofuranosides with a high degree of ste-
reocontrol and without any pyranosidic forms, suffers from a lack
of generality. For instance the synthesis of alkyl arabinofuranosides
is generally obtained from the key intermediates per-O-protected
furanoses.8–10 However, biocatalyzed furanosylation has been
recently developed and its versatility was further extended to the
comparison to other enzymes, thanks to their stability, ease of pro-
duction, and low-cost of substrates which can be bio-sources. A
recent example of directed evolution was applied to the b-glycosi-
dase of Thermus thermophilus in order to increase its ability to syn-
thesize oligopyranosides versus its hydrolytic activity.12–14 To
achieve these results, Koné et al. developed a high-throughput dig-
ital imaging screening methodology to detect transglycosidase
mutants in Escherichia coli cells.14
Herein, we adapted and completed this kind of screening to
improve arabinofuranosyl transglycosylation with the aim to apply
it to a wide range of aliphatic alcohols (Scheme 1). As a model reac-
tion, we chose to synthesize alkyl
a-L-arabinofuranosides which
are simple molecules with diverse properties and potential uses.
Depending on the nature of the alkyl chain, they find applications
as immunostimulating agents or antiparasitic drugs.15–17 More
generally, the alkyl pyranosidic counterparts act as chemical build-
ing blocks for further derivatization and industrial preparation of
alkyl polyglycosides.18,19
In this context, we performed random mutagenesis on the
a-L
-arabinofuranosidase (Araf51) from Clostridium thermocellum
in an attempt to improve the transglycosylation activities for the
synthesis of alkyl -arabinofuranosides. For the selection of
suitable variants, a screening in two steps has been developed.
⇑
Corresponding authors. Tel.: +33 (0)2 23238066.
E-mail address: caroline.nugier@ensc-rennes.fr (C. Nugier-Chauvin).
Present address: Institut de Chimie Organique et Analytique (ICOA), UMR CNRS
7311, Université d’Orléans, BP 6759, Rue de Chartres, 45067 Orléans Cedex 2, France.
a-L
http://dx.doi.org/10.1016/j.carres.2014.10.031
0008-6215/Ó 2014 Elsevier Ltd. All rights reserved.

A. Pennec et al. / Carbohydrate Research 402 (2015) 50–55
51
O
O
H
O
1
O
H
t
O
H
O
O
H
O
H
O
H
O
H
O
O
2
ra
f51
w
H
O
A
O
H
O
NP
Op
H
O
H
O
H
O
O
O
H
O
3
O
O
H
u an s o ra
M t
f A f51
H
O
t
H
O
H
O
H
O
H
O
O
O
4
O
H
H
O
-
ra
f
NP A
p
H
O
H
O
O
H
O
O
5
O
H
H
O
H
O
O
H
O
O
6
H
O
O
H
O
O
H
O
H
O
O
7
O
O
O
H
O
H
O
H
O
8
H
O
O
O
H
O
H
O
H
O
9
Scheme 1. Enzymatic synthesis of alkyl
a
-
L-arabinofuranosides.
The initial one, directly on plates, allowed colony selection of
mutants with good hydrolytic activity on an Araf-containing chro-
mogenic substrate. Indeed in order to maintain the physical quality
of the agar gels and cell viability, the presence of alcohols as accep-
tors was excluded for this first screening step. The second one
assessed the transglycosylation activities toward the convenient
alcohol acceptor of the previous selected enzymes by 96-well spec-
trophotometry. In this approach, one of the main objectives was to
obtain mutants with high selectivity toward nucleophilic alcohols:
one specific mutant for one specific acceptor.
of 5-bromo-indolyl a-L
-arabinofuranosyl.25 This molecule is hydro-
lyzed by the Araf51 releasing a blue air-oxidized di-indolyl com-
pound. In this way, the colonies staining in blue expressed
enzymes successfully cloned which still recognize the arabinofur-
anosyl moiety as donor. Ideally, mutants that tend to favor trans-
glycosylation were supposed to perform slow aglycon release in
the absence of a suitable acceptor and therefore display a pale blue
staining. Nevertheless, a low expression level of some mutated
enzymes could also result in a pale coloration. Also, the selection
of all the blue colonies was performed and the transglycosylation
properties of the resulting enzymes assayed.
This first step allowed us to select more than 120 colonies for
further expression of the mutated enzymes. The corresponding cell
free extracts were heated at 70 °C to remove most of the E. coli
derived proteins and also confirmed the thermostability of the
selected mutants. In the second step of the screening procedure,
each enzyme was tested in the presence of pNP-Araf and an alcohol
in order to evaluate their transglycosylation ability (T) over hydro-
lysis potential (H) (Scheme 2). UV-measurements were performed
during the initial time where the deglycosylation of the glycosyl-
enzyme intermediate is the rate-limiting step. If the transglycosy-
lation is favored, the initial speed rate in these conditions should
be higher than the one during the hydrolysis. The initial reaction
rates were determined thanks to the release of p-nitrophenol
(405 nm). For each cell free extract, two types of reaction were per-
formed, in presence and absence of alcohol acceptor. The alcohol
volume was made up to 25% (v/v) to limit the presence of water,
and so to favor as much as possible the target transglycosylation.
DMSO was added to the reaction mixture as an optimized solubil-
ity factor for aliphatic alcohol. Reactions were performed at pH 8 to
ensure maintaining the phenolate form of the leaving group and
thus proper UV–visible monitoring. A range of linear alcohols were
tested from methanol to n-hexanol as well as i-propanol, allylic
alcohol, and solketal.
2. Results and discussion
2.1. Identification of mutants with improved transglycosylation
abilities
The thermostable arabinofuranosyl hydrolase, Araf51, naturally
removes
a-(1 ? 2), a-(1 ? 3), and a-(1 ? 5)-linked L-arabinofur-
anosyl units from arabinans and arabinoxylans. To the best of
our knowledge, this enzyme belongs to the only arabinofuranosi-
dase family (GH51) showing the ability to catalyze transglycosyla-
tion reactions. In addition, Araf51 has been described several times
for its substrate versatility as it recognizes arabinofuranosides as
well as galactofuranosides.17,20–23 Preliminary experiments
showed the ability of the Araf51 wild type (wt) to transglycosylate
pNP-Araf to small aliphatic alcohol in order to obtain alkyl a-L-ara-
binofuranosides.24 However, the activity strongly decreased when
acceptors contained more than 3 carbons in their chain. With the
aim to improve the transglycosylation activity toward aliphatic
alcohols, mutants were thus created by random mutagenesis and
selected from a simple screening procedure.
Firstly, a library of mutated Araf51 genes has been generated by
error prone PCR and cloned into pCRÒ 2.1-TOPOÒ vector. After cell
transformations, E. coli BL21 DE3 clones were spread on LB plates
for the screening. This latter consisted in a first selection of positive
colonies overexpressing an active Araf51. During this step, the
colonies were platted on LB media containing IPTG and 1.5 mM
To easily reveal interesting enzymes, the initial speed rates
obtained (Vi) were normalized with the one corresponding to the
wild type. Then, a ratio R was calculated correlating directly to
the increase of activity in presence of the acceptor (Eq. 1).

52
A. Pennec et al. / Carbohydrate Research 402 (2015) 50–55
NO2
OH
O
OH
H
O
Acid/Base
HO
O
OH
2
H
=
O
O
H
ROH
O
O
NO2
NO2
HO
H
O
O
O
O
R
OH
OH
ROH
T
=
HO
O
O
HO
Alcohol
OH
OH
O
O
OR
O
OH
HO
O
Nucleophile
OH
,
,
n-,
)
R=
CH CH
3-(
CH2-CH=CH2
CH(CH3)2
2
n=
O
0-5
Scheme 2. Araf51-catalyzed transglycosylation with pNP-Araf as donor and alcohols as acceptors.
ViHþ1
ViH
n-pentanol, and n-hexanol (entries 3, 4, 7, 8, 9, respectively). In com-
parison to the native enzyme, the engineered biocatalysts were all
more promising in terms of reaction time and conversion in trans-
glycosylation, with improvement from +13% for M20 (n-propanol,
entry 3) to +154% for M57 (n-hexanol, entry 9). As an example, time
course evolution between the wild type biocatalyst and the M20
mutant is given in Figure 1. Unfortunately, the proposed protocol
did not afford improved biocatalyst for solketal and allyl alcohol.
Nevertheless, under the transglycosylation conditions performed
with the selected mutants, hydrolysis of the resulted arabinofurano-
sides was not observed, even over 120 min and after total consump-
tion of the substrate. Finally, no other transglycosylation by-
products (autocondensation) were observed, so that the targeted
a-L-arabinofuranosides were specifically formed, confirming the
diastereoselectivity of the mutants according to the glycosidic bond,
whatever the nature of the biocatalyst.
In order to evaluate the selectivity toward the aliphatic acceptor
thanks to the mutated biocatalysts, transglycosylation reactions
were performed with other structurally close alcohols (Table 2).
It appeared that the mutants showed significantly better transgly-
cosylation ability for the alcohol for which they were selected than
for other acceptors. For instance, M20 mutant was particularly effi-
cient and selective for n-propanol.
R ¼
ð1Þ
with ViH+T corresponds to the initial rate in presence of alcohol, and
ViH to the initial rate without alcohol.
For each alcohol tested, around twenty mutants showed posi-
tive activation with R value above 1. In order to detect these
mutants with potentially better transglycosylation than the wild
type, analytic time course reactions were performed and conver-
sions (Table 1) determined by NMR spectroscopy: the protons cor-
responding to the pNP group are up-field shifted when released,
and the appearance of a new anomeric proton from the product
was clearly observed (see Supplementary materials). On this basis,
we focused our attention on five mutants. Then preparative reac-
tions were performed, and the transglycosylation products were
isolated by column chromatography (Table 1). The alkyl resulting
a-L
-arabinofuranosides were characterized by 1H and 13C NMR
analysis. These data were in accordance with previously recorded
results for compounds 1–5, 926–28 and new furanosides 6–8 were
fully described in the experimental part.
Data from Table 1 firstly showed that moderate (i-propanol, solk-
etal, n-butanol, and n-hexanol) toexcellent (methanol, ethanol) con-
versions in transglycosylated derivatives were generally obtained
using the wild type biocatalyst but long reaction times were
required (up to 2 h). Secondly, it is important to note that no auto-
condensation product has been observed, probably due to the high
concentration of alcohol acceptor that favored the desired transgly-
cosylation. Subsequently, the mutation procedure of Araf51 wt
enabled the isolation of 5 mutants with significant improved
transglycosylation activities for n-propanol, i-propanol, n-butanol,
2.2. Kinetic characterization and mutation identification of the
selected enzymes
Finally, the kinetic parameters of the different mutants were
determined under hydrolysis conditions of pNP-Araf at 25 °C and
Table 1
Conversions and yields by transglycosylation using Araf51 wt and selected mutants for different alcohol acceptors
Entry
Acceptor
Araf51 wt
Conversiona (%)
Mutant
Improvementc
Time (min)
Conversiona (%)
Time (min)
Yieldb (%)
1
2
3
4
5
6
7
8
9
MeOH
EtOH
95
90
85
30
38
76
42
72
37
60
60
—
—
—
—
60
60
—
—
—
60
30
—
n-PrOH
i-PrOH
Solketal
AllOH
n-BuOH
n-PentOH
n-HexOH
120
120
120
120
50
96 (M20)
38 (M22)
—
+13%
+27%
—
—
—
92 (M12)
96 (M60)
94 (M57)
20
120
120
68
80
71
+119%
+33%
+154%
120
120
a
b
c
Conversion in transglycosylated product.
Isolated yield.
Was obtained as 100 ꢀ (Conversion mutant ꢁ Conversion wt)/Conversion wt.

A. Pennec et al. / Carbohydrate Research 402 (2015) 50–55
53
within the active site nor close to it. Nevertheless, concerning the
biocatalyzed synthesis of the target alkyl arabinofuranosides, we
expected that the alkyl chains could fit the entry of the common
GH51 (b/
a
)8 barrel domain located at the extremity of the ꢁ1 sub-
site. Despite the location of the mutated sites, far from the active
site, these mutations seemed to significantly affect the kinetics
properties with regard to substrate affinity and transglycosylation
reaction activity.
Figure 1. Time course of transglycosylation into propyl
(black) from pNP-Araf (grey) by (a) Araf51 wt and (b) M20.
a-L-arabinofuranoside
3. Conclusion
Thanks to the coupling of random mutagenesis method con-
ducted on the arabinofuranosyl hydrolase Araf51 and an efficient
screening, we have obtained five mutants able to catalyze the
transfer of an arabinofuranosyl entity to various aliphatic alcohols.
In comparison to the wild type enzyme, these new biocatalysts
improved the transglycosylation conversions up to 96%. Addition-
ally, the procedure used in the study yielded new biocatalysts with
increased transglycosylation abilities and selectivity closely
depending on the nature of the acceptor for which they were
screened. This strategy was efficient with the aim of synthesizing
Table 2
Selectivity of the mutated enzymes (% of transglycosylation conversion)
Entry Mutant EtOH n-PrOH i-PrOH n-BuOH n-PentOH n-HexOH
1
2
3
4
5
M20
M22
M12
M60
M57
20
96
26
19
4
38
10
25
92
65
35
96
65
79
94
valuable chemical building blocks such as alkyl a-L-arabinofurano-
sides. Advantageous mutations in these rationally designed
enzymes appeared to be distant from the active site. Further direc-
ted evolution experiments targeted on these observed mutations
are now required in order to study the properties and catalytic
abilities of the generated enzymes and to accurately understand
the involvement of each modified site.
Table 3
Determination of Araf51 wt and selected mutants kinetic parameters for the
hydrolysis reaction of pNP-Araf (pH 8, 50 mM of Tris HCl buffer, at 25 °C)
Enzyme
Km (mM)
kcat (sꢁ1
)
kcat/Km
Araf51 wt
M20
M22
M12
M60
0.34 0.08
0.07 0.02
0.07 0.01
0.26 0.10
0.17 0.06
0.16 0.04
143
5
420
1900
298
360
694
133 6.4
20.9 1.2
93.8 12
118 8.7
73.6 4.7
4. Experimental
M57
460
4.1. Expression and purification of Araf51 and mutant
derivatives
Plasmid pET28a (+) (Novagen) containing Araf51 wild type from
Clostridium thermocellum29 and kanamycin resistance genes was
provided by Prof G. Davies, University of York, U. K. Plasmid
pCRÒ2.1-TOPOÒ (3.9 kb) from Invitrogen contains encoding
mutated enzymes genes as well as ampicillin and kanamycin resis-
tance genes. These plasmids were under the control of T7 pro-
moter. The enzymes were produced in Escherichia coli BL21 DE3
cells (Invitrogen) cultured in LB (Lysogeny broth) plates containing
0.1 mM of the corresponding selective agent at 37 °C. 10 mL of LB
medium supplemented with antibiotic were inoculated with one
single positive colony E. coli BL21 (DE3) and incubated overnight
at 37 °C, 250 rpm. That was subsequently used to prepare 1% inoc-
ulum (200 mL LB + antibiotic, 250 rpm, 37 °C). Cells were grown to
mid-exponential phase [Absorbance, A550: 0.7] at which point i-
Table 4
Identified mutations
Enzyme
Mutations
M20
M22
M12
M60
M57
R51G, V105L, T220S, S268T, A348S
P392S
K115T
V105L, D134G, T220S, S268T, A348S, E405D
V105L, T220S, S268T
pH 8 (50 mM Tris HCl buffer). Whereas Araf 51 is a thermophilic
enzyme with Topt around 82 °C, the parameters were determined
at 25 °C for practical reasons, as previously reported.29 The data
presented in Table 3 compare the values obtained with the Araf51
wt to the ones calculated for the selected enzymes. Most of the
new biocatalysts showed lower Km by comparison with the wild
type, especially for M20 and M22. Furthermore, kcat values for
the mutants were also lower than those for the wild type Araf51.
On a structural point of view, the crystal structure of the wild
type enzyme has been previously solved (PDB file: 2C7F) and
showed that two essential catalytic residues are required for the
hydrolysis reaction were assigned as the acid/base residue
Glu173 and the nucleophile Glu292. Moreover, structural informa-
tion was also obtained from the co-crystallized of the inactive
mutant E173A form of Araf51 with natural arabinotriose and ara-
binoxylobiose.29 These studies suggest little specificity for the
exact orientation of the aglycone in the +1 subsite due to flexible
conformation adopted by Trp178 which stacks against the agly-
cone sugar unit. The sequence analysis of the mutants from this
study revealed one to six mutations (Table 4, see also Supplemen-
tary data). It clearly appeared that the mutations were neither
propyl b-D-thiogalactopyranoside (IPTG) was added to a final con-
centration of 1.0 mM and the cultures were incubated for 14 h at
37 °C. After centrifugation (20 min, 4000 rpm, 4 °C), the cells were
resuspended in 50 mM Tris HCl buffer pH 8 and sonicated
(3 ꢀ 10 s) for another centrifugation (30 min, 20000 rpm, 4 °C),
the supernatant was heated at 70 °C for 15 min to remove a major
amount of proteins and centrifuged again at 20000 rpm for 30 min.
Protein concentrations were determined by the Bradford
method.30 The purity of His-tagged Araf51 was confirmed by
SDS-PAGE after Ni-NTA agarose affinity chromatography (Novagen,
USA).
4.2. Determination of Araf51 kinetic parameters
The kinetic parameters Km and kcat of Araf51 were quantified in
triplicate by incubation of the enzyme (10
different concentrations of p-nitrophenyl
l
L of ꢂ1 mg/mL) with
a-L-arabinofuranoside

54
A. Pennec et al. / Carbohydrate Research 402 (2015) 50–55
(6 to 0.02 mM) in 50 mM Tris HCl buffer pH 8 at 25 °C. The release
of p-nitrophenol was measured at 405 nm during 5 min (Micro-
plates Spectrophotometer PowerWave XS/XS2, BioTek). The kinetic
parameters were determined by calculations using GOSA software
with repeatability.
buffer with 160
(a final concentration of 0.017 mg/mL is required) to the solution,
and finally completed to a final volume of 800 L and maintained
at 50 °C during 3 h. Aliquots (100 L) of the enzymatic reaction
lL of DMSO. The enzyme preparation was added
l
l
mixture were withdrawn at several times and directly freezed with
liquid nitrogen. After complete lyophilization, samples were solu-
4.3. Random mutagenesis
bilized in 500 lL of CD3OD to enable the analysis by NMR. Trans-
glycosylation activities using pNP-Araf as glycosyl donor were
determined by 1H NMR. By monitoring the decrease of the pNP-
Araf signal (aromatic protons d = 8.21 ppm and/or anomeric proton
d = 5.66 ppm) and the release of p-nitrophenol (aromatic protons
d = 8.12 ppm) corresponding to both the hydrolysis and the trans-
glycosylation of the donor, the residual starting material can easily
be quantified. The transglycosylation products were visualized by
the emergence of the anomeric proton signal of the furanoside
and/or the signal of the alkyl group protons. By reporting the rela-
tion between the proton signals, the resulting conversions were
determined.
Random mutagenesis was performed by GeneMorph II Random
Mutagenesis kit (Stratagene) using mutagenic PCR. The open read-
ing frame encoding Araf51 was amplified using the primers:
- forward T7 Promoter TACGACTCACTATAGGGGAA and reverse
T7 Promoter GCTAGTTATTGCTCAGCGGT (Strategy 1);
- forward T7 Promoter TACGACTCACTATAGGGGAA and reverse
T7 Promoter GTGAGTCGTATTAATTTCGCGGT (Strategy 2).
For low mutation number (mutation frequency 0–4.5 muta-
tions/kb), 500 ng of the initial target DNA was mixed with 250 ng
of each primer, 1
800 M each), 5
of Mutazyme II DNA polymerase (2.5 U/
l
L of 40 mM dNTP mix (final concentration of
4.6. General procedure for the synthesis of alkyl
arabinofuranosides in a preparative scale
l
l
L of 10ꢀ Mutazyme II reaction buffer and 1
l
l
L
L
lL) completed to 50
with H2O. The reaction was thermocycled as follows: one hot start
cycle (95 °C, 2 min) then 10 cycles: first the denaturing step (95 °C,
30 s), the hybridization step (60 °C, 30 s) and the elongation step
performed for 1 min/kb (72 °C, 7 min or 10 min); and finally one
cycle at 72 °C for 10 min. Only for the second strategy, mutagenesis
PCR products were directly cloned into a plasmid vector using the
TOPO TA CloningÒ Invitrogen protocol. The fresh PCR product
Experiments were performed in 50 mM Tris HCl buffer, pH 8 at
50 °C in the presence of pNP-Araf (30 mg, 0.11 mM), 1.4 mL of alco-
hol and 1.12 mL of DMSO in a total reaction volume of 5.6 mL.
Araf51 wild type or mutated was added to a final concentration
of 0.017 mg/mL and the reaction was performed at 50 °C during
the optimal duration determined at the analytical scale experi-
ment. Then the mixture was concentrated under reduced pressure
and the reaction products were separated by column chromatogra-
phy on silica gel (AcOEt/AcOH/H2O, 50:1:1) affording the desired
product.
(2
TA CloningÒ Invitrogen kit: 1
vector and H2O was added up to a final volume of 6
l
L) was mixed with the different reagents provided in the TOPO
L salt solution, 1
L pCRÒ2.1-TOPOÒ
L. The reac-
l
l
l
tion was incubated for 5 min at room temperature (22–23 °C).
4.7. Purification and analysis of synthetic alkyl
arabinofuranosides
4.4. Screening of mutants
The blue colonies were selected and cultivated in 5 mL LB media
containing 0.1 mM kanamycin LB media. The enzymes were puri-
fied as previously described and analyzed for their transglycosyla-
tion activities. As previously, protein concentration was estimated
by Bradford method. Enzyme assays were performed to compare
transglycosylation activities (presence of the alcohol acceptor) to
hydrolytic activities (absence of the alcohol). Mutants and Araf51
wt enzymes were incubated in pH 8 Tris HCl 50 mM buffer with
20 mM pNP-Araf, 20% (v/v) DMSO, with or without 25% (v/v) alco-
Thin layer chromatography (TLC) analysis was conducted on E.
Merck 60 F254 Silica Gel non activated plates and compounds were
visualized by UV (254 nm), or a 5% solution of H2SO4 solution in
EtOH containing orcinol, followed by heating. For column chroma-
tography, Si 60 (40–63 lm) Silica Gel was used. NMR spectra were
recorded on a Bruker spectrometer ARX (400 MHz for 1H and
100 MHz for 13C). Chemicals shifts are expressed in d units
(ppm). Chemical shifts are calculated in Hertz and pattern abbrevi-
ations are as follows: s (singlet), d (doublet), t (triplet), q (quartet),
m (multiplet), and dd (double doublet). The HRMS were measured
with a MicrO-Tof-Q 2 (Bruker) equipped with electrospray and
APCI ionization sources, two quadrupole-orthogonal accelerators
and reflection time-of-flight analyzer, or a Q-tof 2 (Waters)
equipped with electrospray ionization source, two quadrupole-
orthogonal accelerators and reflection time-of-flight analyzer.
hol for a final volume of 140 lL at 50 °C. The final concentration in
enzyme necessarily reached 0.017 mg/mL, thus the collected vol-
ume having to be adapted to each attempt following the determi-
nation of the initial concentration by the Bradford method. The
release of p-nitrophenol was measured at 405 nm during 5 min
(Microplate Spectrophotometer Powerwave XS/XS2, Biotek) and
data evaluated with Gen5 Data Analysis Software (Biotek). The
initial activities of the enzyme and mutated enzymes were deter-
mined using the monitoring of UV curves of the enzymatic assays.
This enabled to compare the slope between Araf51 wt and one of
the mutants with or without the presence of alcohol acceptors,
and highlighted the mutants of interest. Indeed, the mutated
enzymes presenting a higher slope than the one of the Araf51 wt,
in presence of alcohol, showed higher reaction activations,
meaning that transglycosylation was preferred.
4.8. n-Pentyl -arabinofuranoside (6)
a-
L
n-Pentyl -L-arabinofuranoside 6 was obtained according to the
a
described general procedure for transglycosylation by incubation
of n-pentanol in the presence of the Araf51, and was isolated in
66% yield (16.1 mg) after purification by column chromatography.
1H NMR (CD3OD): d 4.84 (d, 1H, J1,2 = 2.0 Hz, H-1), 3.94 (dd, 1H,
J2,3 = 4.0 Hz, H-2), 3.92–3.89 (m, 1H, H-4), 3.82 (dd, 1H,
J3,4 = 6.4 Hz, H-3), 3.73 (dd, 1H, J4,5 = 3.2 Hz, J5a,5b = 12.0 Hz, H-5a),
3.71 (dt, 1H, JCH2,CH2 = 10.0 Hz, JCH2,CH2 = 6.4 Hz, OCH2CH2), 3.62
(dd, 1H, J4,5b = 5.2 Hz, H-5b), 3.41 (dt, 1H, JCH2,CH2 = 9.6 Hz,
JCH2,CH2 = 6.8 Hz, OCH2CH2), 1.61–1.57 (m, 2H, OCH2CH2), 1.36–
1.34 (m, 4H, CH2CH2CH3), 0.92 (t, 3H, JCH2,CH3 = 7.2 Hz, CH3). 13C
NMR (CD3OD): d 109.4 (C-1), 85.1 (C-4), 83.6 (C-2), 78.7 (C-3),
4.5. Analytical scale of transglycosylation reactions, NMR
kinetics
Enzymatic reactions were run from 20 mM pNP-Araf (4.3 mg)
and 200 lL of alcohol acceptor incubated in pH 8 Tris HCl 50 mM

A. Pennec et al. / Carbohydrate Research 402 (2015) 50–55
55
68.8 (OCH2CH2), 63.0 (C-5), 30.4 (OCH2CH2), 29.5 (CH2CH2CH3),
References
23.5 (CH2CH3), 14.4 (CH3). HRMS (ESI+): m/z calcd for C10H20O5Na
[M+Na]+ m/z 243.1208; found, m/z 243.1676.
1. Lombard, V.; Golaconda Ramulu, H.; Drula, E.; Coutinho, P. M.; Henrissat, B.
Nucleic Acids Res. 2014, 42, D490–D495.
2. Armstrong, Z.; Withers, S. G. Biopolymers 2013, 99, 666–674.
3. Cobucci-Ponzano, B.; Moracci, M. Nat. Prod. Rep. 2012, 29, 697–709.
4. Rempel, Brian P.; Mahuran, Don J.; Withers, Stephen G. Angew. Chem., Int. Ed.
2011, 50, 10381–10383.
4.9. n-Hexyl -arabinofuranoside (7)
a
-
L
n-Hexyl -L-arabinofuranoside 7 was obtained according to the
a
5. Kadokawa, J.-I. Chem. Rev. 2011, 111, 4308–4345.
described general procedure for transglycosylation by incubating
n-hexanol in the presence of the Araf51 wt. It was isolated in
71% yield (18.5 mg) after purification by column chromatography.
1H NMR (CD3OD): d 4.84 (d, 1H, J1,2 = 2.0 Hz, H-1), 3.94 (dd, 1H,
J2,3 = 4.0 Hz, H-2), 3.92–3.90 (m, 1H, H-4), 3.82 (dd, 1H,
J3,4 = 6.4 Hz, H-3), 3.74 (dd, 1H, J4,5 = 3.2 Hz, J5a,5b = 12.0 Hz, H-5a),
3.71 (dt, 1H, JCH2,CH2 = 6.8 Hz, JCH2,CH2 = 9.6 Hz, OCH2CH2), 3.62
(dd, 1H, J4,5b = 5.6 Hz, H-5b), 3.41 (dt, 1H, JCH2,CH2 = 9.6 Hz,
JCH2,CH2 = 6.4 Hz, OCH2CH2), 1.60–1.57 (m, 2H, OCH2CH2),
1.35–1.32 (m, 6H, CH2CH2CH2CH3), 0.91 (t, 3H, JCH2,CH3 = 6.8 Hz,
CH3). 13C NMR (CD3OD): d 109.4 (C-1), 85.1 (C-4), 83.6 (C-2), 78.7
(C-3), 68.8 (OCH2CH2), 63.0 (C-5), 32.8 (OCH2CH2), 30.7 (CH2CH2
CH2CH3), 26.9 (CH2CH2CH3), 23.7 (CH2CH3), 14.4 (CH3). HRMS
(ESI+): calcd for C11H22O5Na [M+Na]+ m/z 257.1365; found, m/z
257.1915.
6. Hancock, S. M.; Rich, J. R.; Caines, M. E. C.; Strynadka, N. C. J.; Withers, S. G. Nat.
Chem. Biol. 2009, 508–514.
7. Faijes, M.; Planas, A. Carbohydr. Res. 2007, 342, 1581–1594.
8. Dureau, R.; Legentil, L.; Daniellou, R.; Ferrières, V. J. Org. Chem. 2012, 77, 1301–
1307.
9. Lowary, T. L. Curr. Opin. Chem. Biol. 2003, 7, 749–756.
10. Sanchez, S.; Bamhaoud, T.; Prandi, J. Eur. J. Org. Chem. 2002, 3864–3873.
11. Ferrières, V.; Nugier-Chauvin, C.; Legentil, L.; Tranchimand, S. In Carbohydrate
Chemistry. Chemical and Biological Approaches; Rauter, A. P., Lindhorst, T.,
Queneau, Y., Eds.; The Royal Society of Chemistry (RSC editions), 2014; Vol. 40,
pp 401–417.
12. Teze, D.; Hendrickx, J.; Czjzek, M.; Ropartz, D.; Sanejouand, Y.-H.; Tran, V.;
Tellier, C.; Dion, M. Protein Eng. Des. Sel. 2014, 27, 13–19.
13. Feng, H.-Y.; Drone, J.; Hoffmann, L.; Tran, V.; Tellier, C.; Rabiller, C.; Dion, M. J.
Biol. Chem. 2005, 280, 37088–37097.
14. Koné, F. M. T.; Le Béchec, M.; Sine, J.-P.; Dion, M.; Tellier, C. Protein Eng. Des. Sel.
2009, 22, 37–44.
15. Suleman, M.; Gangneux, J.-P.; Legentil, L.; Belaza, S.; Cabezas, Y.; Manuela, C.;
Dureau, R.; Sergent, O.; Burel, A.; Daligault, F.; Ferrières, V.; Robert-Gangneux,
F. Antimicrob. Agents Chemother. 2014, 58, 2156–2166.
16. Legentil, L.; Audic, J.-L.; Daniellou, R.; Nugier-Chauvin, C.; Ferrières, V.
Carbohydr. Res. 2011, 346, 1541–1545.
4.10. 3-O-a-L-arabinofuranosyl-O-isopropylidene-sn-glycerol
(8)
ˇ
17. Chlubnová, I.; Filipp, D.; Spiwok, V.; Dvoráková, H.; Daniellou, R.; Nugier-
Chauvin, C.; Králová, B.; Ferrières, V. Org. Biomol. Chem. 2010, 8, 2092–2102.
18. von Rybinski, W.; Hill, K. Angew. Chem., Int. Ed. 1998, 37, 1328–1345.
19. Goodby, J. W.; Haley, J. A.; MacKenzie, G.; Watson, M. A. J.; Plusquellec, D.;
Ferrières, V. J. Mater. Chem. 1995, 5, 2209–2220.
The desired furanoside 8 was obtained according to the general
procedure for transglycosylation by incubation of solketal in the
presence of the Araf51 wt. After chromatographic purification, it
was isolated in 36% yield (10.5 mg). 1H NMR (CD3OD): d 4.92 (d,
1H, J1,2 = 1.6 Hz, H-1), 4.31–4.27 (m, 1H, CH2CHCH2), 4.05 (ddd,
1H, J = 2.4 Hz, J = 6.8 Hz, J = 8.4 Hz, O-1CH2CHCH2O), 3.99 (dd, 1H,
J2,3 = 3.6 Hz, H-2), 3.92–3.90 (m, 1H, H-4), 3.84 (dd, 1H,
J3,4 = 6.0 Hz, H-3), 3.78–3.74 (m, 2H, O-1CH2CH-CH2, O-1CH2),
3.71 (dd, 1H, J4,5a = 5.2 Hz, J5a,5b = 12 Hz, H-5a), 3.64 (dd, 1H,
J4,5b = 5.6 Hz, H-5b), 3.50–3.47 (m, 1H, O-1CH2), 1.38 (s, 3H, CH3),
1.32 (s, 3H, CH3).13C NMR(CD3OD): d 109.7 (C-1), 85.7 (C-4), 83.1
(C-2), 78.8 (C-3), 75.9 (CH), 69.1 (O-1CH2CH), 67.6 (O-1CH2CHCH2),
63.0 (C-5), 27.0 (CH3), 25.6 (CH3). HRMS (ESI+): calcd for C11H20O7-
Na [M+Na]+ m/z 287.1107; found, m/z 287.0985.
ˇ
20. Chlubnová, I.; Králová, B.; Dvoráková, H.; Hošek, P.; Spiwok, V.; Filipp, D.;
Nugier-Chauvin, C.; Daniellou, R.; Ferrières, V. Org. Biomol. Chem. 2014, 12,
3080–3089.
21. Rémond, C.; Plantier-Royon, R.; Aubry, N.; Maes, E.; Bliard, C.; O’Donohue, M. J.
Carbohydr. Res. 2004, 339, 2019–2025.
22. Rémond, C.; Plantier-Royon, R.; Aubry, N.; O’Donohue, M. J. Carbohydr. Res.
2005, 340, 637–644.
23. Rémond, C.; Ferchichi, M.; Aubry, N.; Plantier-Royon, R.; Portella, C.;
O’Donohue, M. J. Tetrahedron Lett. 2002, 43, 9653–9655.
24. Pennec, A.; Chlubnova, I.; Daniellou, R.; Ferrières, V.; Nugier-Chauvin, C. U.S.
Provisional 2012, No 61/636883.
25. Berlin, W.; Sauer, B. Anal. Biochem. 1996, 243, 171–175.
26. Finch, P.; Iskander, G. M.; Siriwardena, A. H. Carbohydr. Res. 1991, 210, 319–
325.
27. Mizutani, K.; Kasai, R.; Nakamura, M.; Tanaka, O.; Matsuura, H. Carbohydr. Res.
1989, 185, 27–38.
28. Dutton, G. G. S.; Tanaka, Y.; Yates, K. Can. J. Chem. 1959, 37, 1955–1958.
29. Taylor, E. J.; Smith, N. L.; Turkenburg, J. P.; D’Souza, S.; Gilbert, H. J.; Davies, G. J.
Biochem. J. 2006, 395, 31–37.
Acknowledgments
30. Bradford, M. M. Anal. Biochem. 1976, 72, 248–254.
We are grateful to the Région Bretagne for a Grant to A.P. We
also thank Prof. G.J. Davies for providing the wild type and Jean-
Paul Guégan (ENSCR) for recording NMR spectra.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.carres.2014.10.
031.