On the phosphorylase activity of GH3 enzymes: A β-N-acetylglucosaminidase from Herbaspirillum seropedicae SmR1 and a glucosidase from Saccharopolyspora erythraea

Article abstract of DOI:10.1016/j.carres.2016.09.015

A phosphorolytic activity has been reported for beta-N-acetylglucosaminidases from glycoside hydrolase family 3 (GH3) giving an interesting explanation for an unusual histidine as catalytic acid/base residue and suggesting that members from this family may be phosphorylases [J. Biol. Chem. 2015, 290, 4887]. Here, we describe the characterization of Hsero1941, a GH3 beta-N-acetylglucosaminidase from the endophytic nitrogen-fixing bacterium Herbaspirillum seropedicae SmR1. The enzyme has significantly higher activity against pNP-beta-D-GlcNAcp (K_m?=?0.24?mM, k_cat?=?1.2 s^?1, k_cat/K_m?=?5.0?mM^?1s^?1) than pNP-beta-D-Glcp (K_m?=?33?mM, k_cat?=?3.3?×?10^?3s^?1, k_cat/K_m?=?9?×?10^?4?mM^?1s^?1). The presence of phosphate failed to significantly modify the kinetic parameters of the reaction. The enzyme showed a broad aglycone site specificity, being able to hydrolyze sugar phosphates beta-D-GlcNAc 1P and beta-D-Glc 1P, albeit at a fraction of the rate of hydrolysis of aryl glycosides. GH3 beta-glucosidase EryBI, that does not have a histidine as the general acid/base residue, also hydrolyzed beta-D-Glc 1P, at comparable rates to Hsero1941. These data indicate that Hsero1941 functions primarily as a hydrolase and that phosphorolytic activity is likely adventitious. The prevalence of histidine as a general acid/base residue is not predictive, nor correlative, with GH3 beta-N-acetylglucosaminidases having phosphorolytic activity.

Full text of DOI:10.1016/j.carres.2016.09.015

Carbohydrate Research 435 (2016) 106e112
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
On the phosphorylase activity of GH3 enzymes: A b-N-
acetylglucosaminidase from Herbaspirillum seropedicae SmR1 and a
glucosidase from Saccharopolyspora erythraea
,
, c, *
Diogo R.B. Ducatti a b, Madison A. Carroll c, David L. Jakeman a
a College of Pharmacy, Dalhousie University, 5968 College Street, P.O. Box 15000, Halifax, Nova Scotia B3H 4R2, Canada
Departamento de Bioquímica e Biologia Molecular, Universidade Federal do Parana, Centro Politecnico, CEP 81-531-990, P.O. Box 19046, Curitiba, Parana,
b
ꢀ
ꢀ
ꢀ
Brazil
c Department of Chemistry, Dalhousie University, 6274 Coburg Road, P.O. Box 15000, Halifax, Nova Scotia B3H 4R2, Canada
a r t i c l e i n f o
a b s t r a c t
Article history:
A phosphorolytic activity has been reported for beta-N-acetylglucosaminidases from glycoside hydrolase
family 3 (GH3) giving an interesting explanation for an unusual histidine as catalytic acid/base residue
and suggesting that members from this family may be phosphorylases [J. Biol. Chem. 2015, 290, 4887].
Here, we describe the characterization of Hsero1941, a GH3 beta-N-acetylglucosaminidase from the
endophytic nitrogen-fixing bacterium Herbaspirillum seropedicae SmR1. The enzyme has significantly
higher activity against pNP-beta-D-GlcNAcp (Km ¼ 0.24 mM, kcat ¼ 1.2 sꢀ1, kcat/Km ¼ 5.0 mMꢀ1sꢀ1) than
Received 2 August 2016
Received in revised form
21 September 2016
Accepted 21 September 2016
Available online 23 September 2016
pNP-beta-D-Glcp (Km ¼ 33 mM, kcat ¼ 3.3 ꢁ 10ꢀ3
s
ꢀ1, kcat/Km ¼ 9 ꢁ 10ꢀ4 mMꢀ1sꢀ1). The presence of
Chemical compounds studied in this article:
phosphate failed to significantly modify the kinetic parameters of the reaction. The enzyme showed a
broad aglycone site specificity, being able to hydrolyze sugar phosphates beta-D-GlcNAc 1P and beta-D-
Glc 1P, albeit at a fraction of the rate of hydrolysis of aryl glycosides. GH3 beta-glucosidase EryBI, that
does not have a histidine as the general acid/base residue, also hydrolyzed beta-D-Glc 1P, at comparable
rates to Hsero1941. These data indicate that Hsero1941 functions primarily as a hydrolase and that
phosphorolytic activity is likely adventitious. The prevalence of histidine as a general acid/base residue is
not predictive, nor correlative, with GH3 beta-N-acetylglucosaminidases having phosphorolytic activity.
© 2016 Elsevier Ltd. All rights reserved.
4-nitrophenyl 2-acetamido-2-deoxy-
glucopyranoside (PubChem CID:102416)
4-nitrophenyl -D-glucopyranoside
(PubChem CID:92930)
-D-glucopyranosyl phosphate (PubChem
CID:122250)
2,4-Dinitrophenyl 2-fluoro-2-deoxy-beta-D-
glucopyranoside (PubChem CID:445227)
b-D-
b
b
2-acetamido-2-deoxy-a-D-glucopyranosyl
fluoride (PubChem CID:53906163)
Keywords:
Glycoside hydrolase
Glycosidase
GH3 N-acetylglucosaminidases
Phospho-sugar
Herbaspirillum seropedicae SmR1
Chemical compounds
1. Introduction
also for the synthesis of important targets in glycobiology [4].
Glycosidases catalyze the cleavage of glycosidic bonds, producing
Glycosidases are enzymes involved in carbohydrate metabolism
with broad distribution in all living organisms. These enzymes have
attracted considerable attention due to their application as protein
targets for drug design [1] and catalysis [2], being utilized not only
as catalysts for the depolymerization of complex glycans [3] but
an acetal or hemiacetal product through primarily either a retain-
ing or inverting mechanism. A great number of glycosidases show a
classical mechanism involving a general acid/base catalysis with
Asp or Glu acting as catalytic residues [5], although elegant and
alternative strategies have also been reported [6]. According to the
amino acid sequence similarity, these enzymes are classified into
135 families in the carbohydrate-active enzymes database (CAZy)
[7]. One of the important members in this classification system is
glycoside hydrolase family 3 (GH3). This family harbors exo-acting
* Corresponding author. College of Pharmacy, Dalhousie University, 5968 College
Street, P.O. Box 15000, Halifax, Nova Scotia B3H 4R2, Canada.
E-mail address: david.jakeman@dal.ca (D.L. Jakeman).
http://dx.doi.org/10.1016/j.carres.2016.09.015
0008-6215/© 2016 Elsevier Ltd. All rights reserved.

D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
107
enzymes found in plants, fungi and bacteria that are responsible for
assimilation or modification of glycosides and recycling of cell wall
components. These enzymes show broad substrate specificity
We analyzed the enzyme activity using twelve commercial
glycosides (Table S1). The assay indicated a higher activity for pNP-
b
-D-GlcNAcp than pNP-
hydrolysis against pNP-
b
b
-D-GalNAcp, together with slight levels of
-D-Galp and pNP- -D-Glcp. These data
against
b-D-glucosides (EC 3.2.1.21),
b
-D-xylosides (EC 3.2.1.37),
a
-
b
L-arabinosides (EC 3.2.1.55) and N-acetyl-
3.2.1.52) [8].
In particular, the N-acetylglucosaminidases from GH3 are
important catalysts in cell wall recycling in Gram-positive and
Gram-negative bacteria. In addition, they are responsible for the
production of metabolic intermediates involved in the regulation of
b
-D-glucosaminides (EC
were also confirmed by determination of steady state kinetic pa-
rameters in 50 mM Hepes buffer (Table 1). The kcat/Km ratio for the
GlcNAc glycoside was four orders of magnitude and 500 fold higher
than Glc and GalNAc, respectively, which indicated a significant
preference for the amino sugar with Glc configuration as substrate.
The kinetic parameters for pNP-b-D-GalNAcp and pNP-b-D-Galp
b
-lactamase expression [9,10]. Therefore, these glycosidases are
were not determined due to the low solubility and slow hydrolytic
rate, respectively. The pH effect over Hsero1941 activity was
investigated using pNP-b-D-GlcNAcp as substrate over a pH range
of 4.0e9.0 (Fig. S3). The graph showed a wide bell-shaped curve
with optimum activity at pH 7.1, which suggests two ionizable
amino acid residues in the catalytic site.
important candidates for design of inhibitors in biomedicinal
chemistry [11,12]. In the retaining mechanism proposed for the
catalysis, a conserved aspartate acts as a nucleophile to produce a
glycosyl-enzyme intermediate, while a histidine residue serves as
the acid/base residue [13e15]. Recently, a phosphorolytic activity
has been reported for the b-N-acetylglucosaminidase produced by
Cellulomonas fimi, giving an interesting explanation for the unusual
histidine in the catalytic site of GH3 members [16]. The presence of
the imidazole ring instead of carboxyl groups could avoid charge
repulsion with phosphate, making it possible that the enzyme
works as a phosphorylase. This discovery has raised the possibility
that all GH3 N-acetylglucosaminidases may not be hydrolases and
additionally suggesting bGlcNAc 1P as a new metabolic interme-
diate in bacterial cell wall metabolism. This phosphate sugar would
require at least one enzyme to connect it to the metabolic pathway,
2.2. Investigation of phosphorolytic activity of Hsero1941
In order to investigate any potential phosphorolytic activity for
Hsero1941, we determined the enzyme activity at different phos-
phate concentrations (Table 1). This activity is represented by
glycosyl phosphorylases, enzymes that catalyze the cleavage of a
glycosidic bond through the transfer of a glycosyl unit to inorganic
phosphate producing a glycosyl phosphate product [24]. The Km
and kcat values were similar in all assayed conditions, although
consistent decreases in kcat/Km were observed, as phosphate con-
centrations increased. Together, this data suggests that phosphate
has no significant effect on enzyme activity. We analyzed by 1H
NMR spectroscopy the cleavage products obtained from incubation
and therefore,
candidates for this task [16].
Although several GH3 -N-acetylglucosaminidases have been
b-phosphoglucomutases have been suggested as
b
characterized [14,16e20], phosphorolysis activity was not consid-
ered and consequently this activity has not been monitored. Thus,
we decided to characterize a putative GH3 N-acetyl glucosamini-
dase (Hsero1941) from Herbaspirillum seropedicae SmR1, in order to
investigate its phosphorolytic activity. This organism is an endo-
phytic nitrogen-fixing bacterium that can be found in symbiosis
with important commercial crops, such as corn, rice, sugar-cane
and wheat and for this reason raises a biotechnological interest
[21]. Hsero1941 is a good candidate for this study because ac-
cording to UniProt database [22], H. seropedicae SmR1 codifies only
putative phosphoglucomutases (EC 5.4.2.2, EC 5.4.2.8, EC 5.4.2.10)
of Hsero1941 and pNP-b-D-GlcNAcp 1 in the presence of phosphate
(Fig. 1). We only observed the anomeric signals corresponding to
GlcNAc 2 at 5.19 (J1,2 ¼ 3.4 Hz, H1a) and 4.70 ppm (J1,2 ¼ 7.8 Hz, H1b).
By monitoring the hydrolysis by Hsero1941 of pNP-b-D-GlcNAcp in
the NMR tube, the beta anomer was observed initially, and upon
mutarotation, the alpha isomer was subsequently observed
(Fig. S4). These data showed that Hsero1941 has a hydrolytic ac-
tivity mediated by a double displacement retaining mechanism,
which is consistent with the formation of a glycosyl-enzyme in-
termediate [13].
with specificity for
a
-sugar phosphates. Further, the bacterial plant
We decided to investigate whether phosphate, or other anions,
could act as nucleophiles in order to cleave the covalently bound
intermediate. To this end, Hsero1941 was incubated with DNP2FGlc
(4), which functions as a covalent inhibitor, trapping the nucleo-
philic amino acid residue to produce a stable glycosyl-enzyme in-
termediate [25] (Fig. 2). The inhibition was not complete after 16 h
of incubation, since approximately 20% remaining activity was
observed. After removal of the excess fluoro glycoside inhibitor, the
enzyme was incubated with different anions and the residual ac-
tivity was evaluated. All anions tested promoted reactivation, with
azide (2 M) showing the highest rate of reactivation, almost 2-fold
greater than the inhibited enzyme (Fig. 2).
colonization seems to involve bacterial envelope alterations [23],
therefore, characterizing enzymes involved in this process will
provide insight into this symbiotic relationship.
2. Results
2.1. Gene cloning and hydrolytic enzymatic activity of Hsero1941
H. seropedicae SmR1 has 23 genes codifying glycoside hydro-
lases in its genome, according to CAZy analysis, and two of them
are classified as enzymes from family 3 (GH3). The Hsero1941
gene codifies a putative N-acetyl-
amino acids (M.W. ¼ 36.2 kDa). Sequence alignment with crys-
tallized -N-acetylglucosaminidases from GH3 produced by Gram
b-D-glucosaminidase with 342
2.3. Hydrolytic activity against glycosyl phosphates catalyzed by
Hsero1941, EryBI and DesR
b
negative and Gram positive bacteria showed high similarity,
including the conserved consensus motif containing the catalytic
acid/base histidine (H186) residue (Fig. S1) [13]. The potentially
nucleophilic aspartate (D255) residue was also identified.
Therefore, we amplified Hsero1941 from genomic DNA by PCR
and cloned it into a pET24b vector. Sequence analysis and partial
digestion using restriction endonucleases confirmed the glyco-
sidase gene insertion into the plasmid. The enzyme was
expressed with a C-terminal His6-tag to facilitate the purification
by affinity chromatography (Fig. S2).
To determine if Hsero1941 was able to hydrolyze a glycosyl
phosphate linkage, we incubated commercial
bGlc 1P (5) with
Hsero1941, and two known -glucosidases from GH3: DesR [26]
b
and EryBI [27]. The reaction products were analyzed by 1H NMR
spectroscopy (Fig. 3). At all three enzyme concentrations evaluated
for Hsero1941, signals corresponding to the anomeric protons of Glc
(6) at 5.19 and 4.60 ppm were observed indicating that Hsero1941
catalyzed the hydrolysis of 5. However, EryBI also showed compa-
rable activity at the highest concentration evaluated,

108
D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
Table 1
Kinetic parameters for substrate hydrolysis by Hsero1941.
a
Substrate
Phosphate (mM)a
Km (mM)a
kcat (sꢀ1
)
k )
cat/Km (mMꢀ1sꢀ1
pNP-
pNP-
pNP-
pNP-
pNP-
pNP-
pNP-
b
b
b
b
b
b
b
-D-GlcNAcp
-D-GlcNAcp
-D-GlcNAcp
-D-GlcNAcp
-D-GlcNAcp
-D-Glcp
0
0.24 0.03
0.29 0.01
0.30 0.02
0.27 0.02
0.33 0.04
1.2 0.10
1.4 0.03
1.2 0.02
1.1 0.01
1.0 0.10
3.3 ꢁ 10ꢀ3 3 ꢁ 10ꢀ4
NDb
5.0
4.9
4.1
3.9
25
50
100
200
0
3.0
33
4
9 ꢁ 10ꢀ4
0.010
-D-GalNAcp
0
NDb
a
The reactions were performed in 50 mM Hepes pH 7.1 in absence or presence of phosphate. Results are expressed as mean standard error of the mean (SEM).
The parameters were not determined due to the linear relationship between Vo and So.
b
1H NMR spectroscopy (Fig. 4). Two major signals were observed in
the spectrum at 5.62 (J1,2 ¼ 2.7 and J1,F ¼ 52.8 Hz) and 5.19
(J1,2 ¼ 3.4 Hz, H1a) ppm and were attributed to the anomeric pro-
tons of substrate aFGlcNAc (7) and its hydrolytic byproduct GlcNAc
(2), respectively. However, a small signal was also noted at 4.93
(J ¼ 8.3 Hz) ppm, similar to the chemical shift reported in the
literature for the anomeric proton of bGlcNAc 1P (3) [16]. The cor-
relation found at 4.93/1.85 ppm in the 1H-31P HMBC spectrum
confirmed the formation of 3 in the reaction. Surprisingly, the
anomeric signal corresponding to the glycosyl phosphate was also
observed in the control reaction, indicating that the formation of
this glycoside was not catalyzed by the mutants, but the intrinsic
reactivity between 7 and phosphate under the reaction conditions.
WaterLOGSY NMR experiments were performed in order to
investigate whether
sidase and its mutants (Fig. S6). Although the mutants were not
able to catalyze the synthesis of GlcNAc 1P (3), the spectra indi-
aFGlcNAc (7) could bind the wild-type glyco-
b
cated that the glycosyl fluoride bound the mutants and wild-type
enzyme by the change of phase of signals at 5.62, 3.57 and
2.03 ppm, corresponding to H1, H4 and the N-acetyl of 7, respec-
tively. When the reaction between 7 and phosphate was performed
in the presence of Hsero1941 (Fig. 4), the production of
(3) was not detected, which was attributed to the hydrolysis of this
product by the wild type -N-acetylglucosaminidase.
Intrigued by the non-enzyme catalyzed nucleophilic substitu-
tion between FGlcNAc 7 and phosphate, we performed a reaction
in 50 mM Hepes buffer containing 2 M azide, a well-known good
nucleophile (Fig. 7S). The N3GlcNAc 8 was produced as major
bGlcNAc 1P
b
a
b
product after 48 h, as identified by 1H NMR analysis [29]. When the
reaction was carried out in the presence of Hsero1941, GlcNAc 2 was
the only product detected, indicating that 8 was hydrolyzed by the
wild type enzyme after its formation in the reaction mixture. The
hydrolytic activity of glycosidases against azido sugars has been
reported in the literature [30].
Fig. 1. Activity of Hsero1941 against pNP-b-D-GlcNAcp (1) at different phosphate
concentrations.
demonstrating that it was also able to hydrolyze bGlc 1P. In order to
evaluate if the hydrolysis of 5 could be a non-specific, or contam-
inating, activity, Hsero1941 was incubated with pNP-phosphate.
However, no hydrolytic activity was observed. Together, these data
indicated that both Hsero1941 and EryBI are able to hydrolyze 5 at
comparable rates.
3. Discussion
Hsero1941 is a glycosidase with higher specificity for glycoside
substrates of GlcNAc than GalNAc. This result is consistent with b-
N-acetylglucosaminidases from GH3, which have a preference for
amino sugars with Glc instead of Gal configuration [17,18].
Hsero1941 showed activity against Glc, however, according to the
kcat/Km values the preference for glycosides of Glc was less than that
observed for the GH3 phosphorylase from C. fimi [16,19]. Although
the anion rescue results have showed that phosphate was able to
react with the glycosyl-enzyme intermediate, the revival of cata-
lytic activity was similar to other anions tested. In addition, the
presence of phosphate in the reaction had no significant effect on
the enzyme activity. Those results, together with the identification
of the reaction product by 1H NMR, clearly demonstrated that
Hsero1941 is a glycoside hydrolase with a retaining mechanism.
At this stage, we explored whether
bGlcNAc 1P (3) could be
synthesized from a fluoride glycosyl donor and phosphate, in a
similar strategy to the formation of glycoside bonds between two
sugars catalyzed by glycosynthases [28]. Thus, three nucleophile
mutants (Hsero1941-D255G, Hsero1941-D255A and Hsero1941-
D255S) were constructed by site-directed mutagenesis and over-
expressed in E. coli (Fig. S5). The mutants were inactive against
pNP-
ported for nucleophile mutants of
from GH3 [15]. The reactions for the production of
from FGlcNAc (7) were performed in 500 mM PBS and analyzed by
b
-D-GlcNAcp, which was consistent with data previously re-
-N-acetylglucosaminidases
GlcNAc 1P (3)
b
b
a

D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
109
Fig. 2. Anion reactivation of covalently inhibited 2FGlc-Hsero1941. (A) Scheme showing the enzyme inhibition using DNP2FGlc (4) as a covalent inhibitor. (B) Relative activity of
inhibited 2FGlc-Hsero1941 against pNP- -D-GlcNAcp after incubation with several anions.
b
Fig. 3. 1H NMR spectra showing hydrolysis of
(E) 5 mM 5, EryBI 10 M. (F) 5 mM 5, DesR 15
b
-D-Glc 1P 5. (A) Schematic hydrolysis of 5. (B) 10 mM 5, Hsero1941 70
mM. (C) 5 mM 5, Hsero1941 10 mM. (D) 10 mM 5, EryBI 70 mM.
m
m
M. (G) 5 mM 5, without enzyme.

110
D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
Fig. 4. Chemical synthesis and enzymatic hydrolysis of
bGlcNAc 1P 3. (A) Synthesis of 3 from
a
FGlcNAc 7. 1H NMR spectra showing products in reactions performed with 15 mM 7
and 30 mM of enzyme: (B) Hsero1941; (C) Hsero1941 D255S; (D) Hsero1941 D255A; (E) Hsero1941 D255G; (F) control reaction without enzyme.
It is worth noting that at lower enzyme concentrations,
Hsero1941 was able to hydrolyze the charged substrate Glc 1P 5,
while known GH3 -glucosidases DesR and EryBI were inactive
The capacity of Hsero1941 to hydrolyze azido-, phospho- and
pNP- glycosides indicated a broad aglycone site specificity, which
may be associated with the recognition in vivo of different substrates
b
b
against this substrate. Although the hydrolytic activity for EryBI was
observed at higher concentrations, the substrate consumption for
either enzyme was not complete after 4 days of incubation. It has
been proposed that some glycosidases containing His or Tyr as cat-
alytic residues instead of the usually found Glu or Asp, could accept
charged substrates due to a smaller Coulombic repulsion within the
catalytic site [16,31]. Indeed, DesR and EryBI show Glu/Asp and Asp/
Asp dyads, respectively, in the catalytic site [26,27], while Hsero1941
presents a His and Asp, which might explain the affinity difference
containing GlcNAc. This result is consistent with b-N-acetylgluco-
saminidases from GH3, which are involved in bacterial cell wall
recycling [9]. In addition, a recent study has been reported indicating
that NagZ from Neisseria gonorrhoeae is involved in biofilm meta-
bolism, suggesting a biological function for the enzymes from this
family beyond that of peptidoglycan recycling [35]. However, it is not
clear whether the ability of Hsero1941 to cleave the phosphorylated
sugar substrates, especially
bGlcNAc 1P is important in bacterial
metabolism, since the GlcNAc 1P motif is rare in the nature, being
b
for
be the only structural requirement to rationalize the catalysis of
charged phospho-substrates by glycosidases. For instance, an
b
Glc 1P (5). However, the identity of catalytic residues could not
found only in some bacterial glycolipids [36]. Another possibility to
consider is that glycosyl phosphate hydrolysis remains a promiscu-
ous activity detected in vitro, which could be associated with the
capacity to recognize phosphate groups by members of this family.
a-
retaining sucrose phosphorylase from GH13 has Glu and Asp as
catalytic residues that do not hamper the nucleophilic phosphate
attack on the glycosyl-enzyme intermediate [32]. Furthermore, the
first mannosidases from GH130 [33], a family known to contain
invertingmannoside phosphorylases have been characterized. These
glycoside hydrolases did not show conserved basic amino acid resi-
dues responsible for binding phosphate, and for this reason, a water
molecule acts as nucleophile. Therefore, additional features in the
The capacity of GH3 glucosidase EryBI to hydrolyze bGlc 1P indicates
that the presence of a His acid/base catalyst is not a prerequisite for
the recognition of charged substrates within the GH3 family.
In conclusion, Hsero1941 is a retaining beta-N-acetylglucosami-
nidase from GH3. The enzyme showed a broad aglycone site speci-
ficity, being able to recognize aryl glycosides andglycosyl phosphates
as substrates for hydrolysis. These data indicate that all GH3 family
members have hydrolytic activity, whilst only some of the family
members have phosphorolytic activity, and that the presence of a
histidine general acid/base is not predictive nor correlative with the
observed phosphorolytic activity. Further studies are needed to
determine whether the hydrolytic activity against glycosyl phos-
phates imparts physiological importance in bacterial metabolism.
tertiary structure of GH3 b-N-acetylglucosaminidases may also be
responsible for the interaction of phospho-compounds and phos-
phate in the catalytic site. Further 3D studies could be useful to
explain the presence of hydrolases and phosphorylases in this family.
The nucleophile mutants were not able to synthesize
bGlcNAc
1P (3) employing FGlcNAc (7) as donor. We did not observe gly-
a
cosynthase activity using a broad library of glycosides and alcohols
as acceptors (data not shown). The mutation of a glycosidase into a
glycosynthase has been reported as an interesting approach in
glycobiology to produce enzymes able to synthesize a variety of
glycosidic structures [34]. However, not all glycosidases can pro-
duce active glycosynthases [26], indeed, EryBI remains the only
active glycosynthase from GH3 [27].
4. Experimental
4.1. Cloning and mutagenesis of Hsero1941
The gene encoding the glycosidase was amplified from genomic
DNA of H. seropedicae SmR1 using high-fidelity polymerase and

D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
111
primers (Supplemental Table S2) containing NdeI and HindIII re-
striction sites. The gene was digested and ligated into a pET24b
vector to express the protein with a C-terminal His6-tag. Site-
directed mutagenesis was performed using QuickChange Light-
ning kit (Agilent Technologies), according to the manufacturer's
instructions. Primers were designed in order to introduce a silent
mutation together with the exchange of target amino acid
(Supplemental Table S2). The silent mutation added a new re-
striction site to assist in the identification of cells harboring the
mutated vector.
the activity and comparing with the activity level before the
centrifugation step. Afterwards, inhibited enzyme aliquots (final
concentration ¼ 1.37
Hepes buffer pH 7.1 and different anions added (azide, phosphate,
fluoride and formate, final volume ¼ 100 L). The reactions (n ¼ 6)
were incubated 19 h at room temperature, and enzyme activity was
estimated measuring initial velocity after addition of 100 L pNP-b-
mM) were added into wells containing 50 mM
m
m
D-GlcNAcp (3.6 mM). Results were expressed as mean relative to
the activity of non-inhibited wild type enzyme.
4.6. Hydrolysis of
bGlc 1P
4.2. Expression of Hsero1941 and its mutants
To a solution containing
b
Glc 1P in 50 mM Hepes/phosphate pH
The recombinant enzymes were expressed in E. coli BL21 (DE3)
cells according to previously reported procedures [26]. Briefly, the
cells were grown (37 ꢂC, 250 rpm) in Luria-Bertani media con-
7.1, the enzyme was added. The resulting mixture was incubated at
37 ꢂC, for 4 days, lyophilized and analyzed by 1H NMR spectroscopy.
taining kanamycin (50 mM/mL) until an O.D (600 nm) of 0.7e0.9
4.7. Chemical synthesis and enzymatic hydrolysis of
and N3GlcNAc
bGlcNAc 1P
was reached. Afterwards, IPTG was added to a final concentration of
1 mM, and the cultures incubated (24 h, 17 ꢂC, 250 rpm) and sub-
sequently centrifuged (4000g). The cell pellet was resuspended
with lysis buffer, stirred (30 min, 0 ꢂC), sonicated (5 ꢁ 5 s) and then
centrifuged (13,000 rpm, 4 ꢂC). The supernatant was loaded onto a
HisTrap HP column (5 mL) and proteins were eluted using a step-
wise imidazole gradient (25e250 mM). The collected tubes were
analyzed by SDS-PAGE, pooled, concentrated using centrifugal filter
devices (cut-off 10,000 Da) and desalted on a Sephadex G-25 col-
umn (10 mL) using 50 mM Hepes buffer pH 7.1 as eluent. The
protein concentrations were calculated measuring the solution
absorbance at 280 nm (ε280 ¼ 29,700 Mꢀ1cmꢀ1).
b
Wild type and mutants (150
160 mM FGlcNAc (see supplemental data for the synthesis) in
water, 2 M PBS pH 7.2 and 4 M NaN3 in 50 mM Hepes pH 7.1 stock
solutions were prepared. To a solution containing 15 mM FGlcNAc
M) were added. The
mM) in 50 mM Hepes pH 7.1,
a
a
in 500 mM PBS or 2 M NaN3, enzymes (30
m
resulting mixture was incubated at 37 ꢂC, for 2 days, lyophilized,
and analyzed by 1H NMR spectroscopy.
4.8. NMR analysis
NMR spectra were acquired on a Bruker AV 500 or 300 spec-
trometer, operating at 500 or 300 MHz for 1H. The chemical shifts
4.3. Steady state kinetic analysis
were measured relative to internal acetone (
d
¼ 2.208 ppm for 1H).
Continuous kinetic assays were performed in 96-well plates
using a Molecular Devices Spectromax 384 UV spectrophotometer,
monitoring the release of pNP at 400 nm (ε400 ¼ 7280 Mꢀ1cmꢀ1).
Initial velocities were calculated in Softmax Pro V4.6. Assays were
The data were analyzed using Bruker Topspin 3.5 software.
Acknowledgment
performed in 50 mM Hepes buffer, pH 7.1 containing 0.3
mM
We thank the Natural Sciences and Engineering Research
Council of Canada and CNPq e Conselho Nacional de Desenvolvi-
Hsero1941 and commercial pNP- -D-GlcNAcp or pNP- -D-Glcp as
b
b
substrates ranging from 0.15 to 4.8 mM and 10e100 mM, respec-
tively. For the reactions performed in the presence of phosphate,
Hepes buffer was supplemented with PBS pH 7.2 (25, 50, 100 and
200 mM). Kinetic parameters were obtained from at least 2
different assays (n ¼ 2 for each point) by fitting initial velocities
data to the Michaelis-Menten equation using Grafit 5.0 (Erithacus
software). Results were expressed as mean standard error of the
mean (SEM).
ꢀ
mento Científico e Tecnologico e Brasil (scholarship 205131/2014-
0) for funding. The authors are thankful to Prof. Marcelo M. dos
Santos for the cultivation of H. seropedicae SmR1 and extraction of
the genomic DNA.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.carres.2016.09.015.
4.4. Enzymatic activity in PBS buffer
References
To
a solution containing pNP-b-D-GlcNAcp and PBS, the
Hsero1941 was added. The resulting mixture was incubated at
[1] L. Bohlmann, G.D. Tredwell, X. Yu, C.-W. Chang, T. Haselhorst, M. Winger,
J.C. Dyason, R.J. Thomson, J. Tiralongo, I.R. Beacham, H. Blanchard,
M.V. Itzstein, Nat. Chem. Biol. 11 (2015) 955e957.
37 ꢂC, for 2 h, centrifuged using Nanosep 10 K
U filter devices (cut-
off 10,000 Da), lyophilized and analyzed by 1H NMR spectroscopy.
ꢀ
[2] B. Bissaro, P. Monsan, R. Faure, M.J. O'Donohue, Biochem. J. 467 (2015) 17e35.
[3] C.A. Santos, L.M. Zanphorlin, A. Crucello, C.C.C. Tonoli, R. Ruller, M.A.C. Horta,
M.T. Murakami, A.P. Souza, Biotechnol. Biofuels 9 (2016) 71.
[4] B. Cobucci-Ponzano, A. Strazzulli, M. Rossi, M. Moracci, Adv. Synth. Catal. 353
(2011) 2284e2300.
[5] D.J. Vocadlo, G.J. Davies, Curr. Opin. Chem. Biol. 12 (2008) 539e555.
[6] S.A.K. Jongkees, S.G. Withers, Acc. Chem. Res. 47 (2014) 226e235.
[7] V. Lombard, H.G. Ramulu, E. Drula, P.M. Coutinho, B. Henrissat, Nucleic Acids
Res. 42 (2014) D490eD495.
[8] D. Faure, Appl. Environ. Microb. 68 (2002) 1485e1490.
[9] J.W. Johnson, J.F. Fisher, S. Mobashery, Ann. N. Y. Acad. Sci. 1277 (2013)
54e75.
[10] K.A. Stubbs, M. Balcewich, B.L. Mark, D.J. Vocadlo, J. Biol. Chem. 282 (2007)
21382e21391.
4.5. Reactivation with anions
Hsero1941 (67
porting Information for the synthesis) (50 mM) in 50 mM Hepes
buffer pH 7.1 (200
L final volume) at 37 ꢂC. The enzyme activity
was monitored measuring the initial velocity as described above,
employing pNP- -D-GlcNAcp as substrate and using wild type
enzyme, incubated without inhibitor, as a control. After 16 h, the
inhibitor was removed by centrifugation using Nanosep 3 K filter
devices (cut-off 3000 Da) and washing with 50 mM Hepes buffer
(4 ꢁ 400 L). The enzyme concentration was estimated measuring
mM) was incubated with DNP2FGlc (see Sup-
m
b
U
[11] M.D. Balcewich, K.A. Stubbs, Y. He, T.W. James, G.J. Davies, D.J. Vocadlo,
B.L. Mark, Protein Sci. 18 (2009) 1541e1551.
m

112
D.R.B. Ducatti et al. / Carbohydrate Research 435 (2016) 106e112
[12] M. Mondon, S. Hur, G. Vadlamani, P. Rodrigues, P. Tsybina, A. Oliver, B.L. Mark,
L.M. Chueire, M.F. Nicolas, L. Pereira-Ferrari, J.L. Silva, N.M. Gioppo,
V.P. Margarido, M.A. Menck-Soares, F.G. Pinto, C. Simao Rde, E.K. Takahashi,
M.G. Yates, E.M. Souza, PLoS Genet. 7 (2011) e1002064.
ꢀ
D.J. Vocadlo, Y. Bleriot, Chem. Commun. 49 (2013) 10983e10985.
[13] D.J. Vocadlo, C. Mayer, S. He, S.G. Withers, Biochemistry 39 (2000) 117e126.
[14] S. Litzinger, S. Fischer, P. Polzer, K. Diederichs, W. Welte, C. Mayer, J. Biol.
Chem. 285 (2010) 35675e35684.
[15] J.-P. Bacik, G.E. Whitworth, K.A. Stubbs, D.J. Vocadlo, B.L. Mark, Chem. Biol. 19
(2012) 1471e1482.
[16] S.S. Macdonald, M. Blaukopf, S.G. Withers, J. Biol. Chem. 290 (2015)
4887e4895.
[17] E. Chitlaru, S. Roseman, J. Biol. Chem. 271 (1996) 33433e33439.
[18] S. Yang, S. Song, Q. Yan, X. Fu, Z. Jiang, X. Yang, J. Agric. Food Chem. 62 (2014)
5181e5190.
[22] UniProt consortium, Nucleic Acids Res. 43 (2015) D204eD212.
[23] R.A. Monteiro, E. Balsanelli, R. Wassem, A.M. Marin, L.C.C. Brusamarello-San-
tos, M.A. Schmidt, M.Z. Tadra-Sfeir, V.C.S. Pankievicz, L.M. Cruz, L.S. Chubatsu,
F.O. Pedrosa, E.M. Souza, Plant Soil 356 (2012) 175e196.
[24] V. Puchart, Biotechnol. Adv. 33 (2015) 261e276.
[25] S.J. Williams, S.G. Withers, Carbohydr. Res. 327 (2000) 27e46.
[26] A. Sadeghi-Khomami, M.D. Lumsden, D.L. Jakeman, Chem. Biol. 15 (2008)
739e749.
[27] D.L. Jakeman, A. Sadeghi-Khomami, Biochemistry 50 (2011) 10359e10366.
[28] L.F. Mackenzie, Q. Wang, R.A.J. Warren, S.G. Withers, J. Am. Chem. Soc. 120
(1998) 5583e5584.
[19] C. Mayer, D.J. Vocadlo, M. Mah, K. Rupitz, D. Stoll, R.A.J. Warren, S.G. Withers,
FEBS J. 273 (2006) 2929e2941.
[20] S. Mine, Y. Kado, M. Watanabe, Y. Fukuda, Y. Abe, T. Ueda, Y. Kawarabayasi,
T. Inoue, K. Ishikawa, FEBS J. 281 (2014) 5092e5103.
[29] T. Tanaka, H. Nagai, M. Noguchi, A. Kobayashi, S.-I. Shoda, Chem. Commun.
(2009) 3378e3379.
ꢀ ꢀ ꢀ ꢀ
[30] P. Bojarova, L. Petraskova, E.E. Ferrandi, D. Monti, H. Pelantova, M. Kuzma,
[21] F.O. Pedrosa, R.A. Monteiro, R. Wassem, L.M. Cruz, R.A. Ayub, N.B. Colauto,
M.A. Fernandez, M.H. Fungaro, E.C. Grisard, M. Hungria, H.M. Madeira,
R.O. Nodari, C.A. Osaku, M.L. Petzl-Erler, H. Terenzi, L.G. Vieira, M.B. Steffens,
V.A. Weiss, L.F. Pereira, M.I. Almeida, L.R. Alves, A. Marin, L.M. Araujo,
E. Balsanelli, V.A. Baura, L.S. Chubatsu, H. Faoro, A. Favetti, G. Friedermann,
C. Glienke, S. Karp, V. Kava-Cordeiro, R.T. Raittz, H.J. Ramos, E.M. Ribeiro,
L.U. Rigo, S.N. Rocha, S. Schwab, A.G. Silva, E.M. Souza, M.Z. Tadra-Sfeir,
R.A. Torres, A.N. Dabul, M.A. Soares, L.S. Gasques, C.C. Gimenes, J.S. Valle,
R.R. Ciferri, L.C. Correa, N.K. Murace, J.A. Pamphile, E.V. Patussi, A.J. Prioli,
S.M. Prioli, C.L. Rocha, O.M. Arantes, M.C. Furlaneto, L.P. Godoy, C.E. Oliveira,
D. Satori, L.A. Vilas-Boas, M.A. Watanabe, B.P. Dambros, M.P. Guerra,
S.M. Mathioni, K.L. Santos, M. Steindel, J. Vernal, F.G. Barcellos, R.J. Campo,
ꢀ
ꢁ
P. Simerska, V. Kren, Adv. Synth. Catal. 349 (2007) 1514e1520.
[31] A.G. Watts, I. Damager, M.L. Amaya, A. Buschiazzo, P. Alzari, A.C. Frasch,
S.G. Withers, J. Am. Chem. Soc. 125 (2003) 7532e7533.
[32] A. Schwarz, L. Brecker, B. Nidetzky, Biochem. J. 403 (2007) 441e449.
ꢀ
ꢂ
[33] F. Cuskin, A. Basle, S. Ladevese, A.M. Day, H.J. Gilbert, G.J. Davies, G. Potocki-
ꢀ
ꢂ
Veroneese, E.C. Lowe, J. Biol. Chem. 290 (2015) 25023e25033.
[34] K. Yamamoto, B.G. Davis, Angew. Chem. Int. Ed. 51 (2012) 7449e7453.
[35] S.V. Bhoopalan, A. Piekarowicz, J.D. Lenz, J.P. Dillard, D. Stein, Sci. Rep. 6 (2016)
22372.
[36] N. Murazumi, S. Yamamori, Y. Araki, E. Ito, J. Biol. Chem. 254 (1979)
11791e11793.