Probing the roles of conserved residues in uridyltransferase domain of Escherichia coli K12 GlmU by site-directed mutagenesis

Article abstract of DOI:10.1016/j.carres.2015.05.007

N-Acetylglucosamine-1-phosphate uridyltransferase (GlmU) is a bifunctional enzyme that catalyzes both acetyltransfer and uridyltransfer reactions in the prokaryotic UDP-GlcNAc biosynthesis pathway. Our previous study demonstrated that the uridyltransferase domain of GlmU (tGlmU) exhibited a flexible substrate specificity, which could be further applied in unnatural sugar nucleotides preparation. However, the structural basis of tolerating variant substrates is still not clear. Herein, we further investigated the roles of several highly conserved amino acid residues involved in substrate binding and recognition by structure- and sequence-guided site-directed mutagenesis. Out of total 16 mutants designed, tGlmU Q76E mutant which had a novel catalytic activity to convert CTP and GlcNAc-1P into unnatural sugar nucleotide CDP-GlcNAc was identified. Furthermore, tGlmU Y103F and N169R mutants were also investigated to have enhanced uridyltransferase activities compared with wide-type tGlmU.

Full text of DOI:10.1016/j.carres.2015.05.007

Carbohydrate Research 413 (2015) 70e74
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Probing the roles of conserved residues in uridyltransferase domain
of Escherichia coli K12 GlmU by site-directed mutagenesis
a
,
b
,
y
, Xuan Fu a , Yunpeng Liu , Xian-wei Liu , Lin Wang ,
,
y
b
,
c
a
d
Shuaishuai Wang
a,
*
a, c
Junqiang Fang , Peng George Wang
National Glycoengineering Research Center, Shandong University, Jinan, Shandong 250100, People's Republic of China
The State Key Laboratory of Microbial Technology and School of Life Sciences, Shandong University, Jinan, Shandong 250100, People's Republic of China
Department of Chemistry, Georgia State University, P.O. Box 4098, Atlanta, GA 30302-4098, USA
a
b
c
d
Department of Food Science and Engineering, Nanshan University, Yantai, Shandong 265713, People's Republic of China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 April 2015
Received in revised form
N-Acetylglucosamine-1-phosphate uridyltransferase (GlmU) is a bifunctional enzyme that catalyzes both
acetyltransfer and uridyltransfer reactions in the prokaryotic UDP-GlcNAc biosynthesis pathway. Our
previous study demonstrated that the uridyltransferase domain of GlmU (tGlmU) exhibited a flexible
substrate specificity, which could be further applied in unnatural sugar nucleotides preparation. How-
ever, the structural basis of tolerating variant substrates is still not clear. Herein, we further investigated
the roles of several highly conserved amino acid residues involved in substrate binding and recognition
by structure- and sequence-guided site-directed mutagenesis. Out of total 16 mutants designed, tGlmU
Q76E mutant which had a novel catalytic activity to convert CTP and GlcNAc-1P into unnatural sugar
nucleotide CDP-GlcNAc was identified. Furthermore, tGlmU Y103F and N169R mutants were also
investigated to have enhanced uridyltransferase activities compared with wide-type tGlmU.
© 2015 Elsevier Ltd. All rights reserved.
18 May 2015
Accepted 21 May 2015
Available online 2 June 2015
Keywords:
Uridyltransferase
UDP-GlcNAc
Substrate specificity
Site-directed mutagenesis
1
. Introduction
2.3.1.157) and uridyltransferase (EC 2.7.7.23) activities, which cat-
alyzes the formation of UDP-GlcNAc from glucosamine-1-P, acetyl-
0
Uridine 5 -diphosphate N-acetylglucosamine (UDP-GlcNAc), an
important cytoplasmic amino sugar nucleotide and the activated
form of GlcNAc, is the key precursor in enzymatic glycosylation by
Leloir-type glycosyltransferases to create incredible diversity of
CoA and UTP in prokaryotic UDP-GlcNAc de novo biosynthetic
10
pathway. The bacterial GlmU has been well studied from Escher-
10e12
ichia coli, Haemophilus influenza, and Streptococcus pneumonia.
The N-terminal uridyltransferase domain of GlmU belongs to the
nucleoside triphosphate transferase superfamily and shares
1
glycoconjugates, including cell wall peptidoglycans, lipopolysac-
2
3
2 4
-G-X-G-T-X-M-X
-P-K.13
charides, glycosaminoglycans, enterobacterial common anti-
signature sequence motif of L-X
Till now, a series of GlmU crystal structures have been
gens, chitin5 and protein N-and O-linked glycans.
4
6,7
Although
1
1,12
UDP-GlcNAc has been widely used for many years, much still re-
mains unknown about its functional mechanisms and synthetic
pathway. Chemical synthesis of UDP-GlcNAc and its derivatives
have been developed, but the low yield and fastidious process
hinder its application.8 In contrast, enzymatic synthesis, which
mimics the biosynthetic pathway is considered more attractive
with the advantages of high region-/stereo-specificity and low cost.
N-Acetylglucosamine-1-phosphate uridyltransferase (GlmU) is
solved.
demonstrates that the overall structure of the uridyltransferase
domain is that of a seven-stranded mixed -helix surrounded by six
-sheets and a two-stranded -helix sitting atop the larger mixed
This domain can be divided into two substrate-binding
The reported crystal structure of GlmU (1HV9, PDB)
a
b
a
,9
14,15
sheet.
regions. The region that binds to the uridine of UTP, consists of
Ala12-Gly14, Ala18, Lys25, Gln76-Thr82, and Asp105 structural
segments and forms a solvent-exposed pocket. The region that
a
key bifunctional enzyme with both acetyltransferase (EC
binds to GlcNAc moiety, includes strands
b5-b7 and interacts pri-
marily with the GlcNAc moiety of GlcNAc-1-P.
We previously cloned the uridyltransferase domain of E. coli K12
GlmU, tGlmU, which had a smaller molecular weight and higher
soluble expression level, remained comparable catalytic activity
*
Corresponding author. Tel.: þ86 531 88366078; fax: þ86 531 88363002.
E-mail address: fangjunqiang@sdu.edu.cn (J. Fang).
Same contribution to this paper.
y
http://dx.doi.org/10.1016/j.carres.2015.05.007
008-6215/© 2015 Elsevier Ltd. All rights reserved.
0

S. Wang et al. / Carbohydrate Research 413 (2015) 70e74
71
with the full-length GlmU.16 This truncated tGlmU could also
recognize several nucleoside triphosphates (NTPs) to form the
corresponding sugar nucleotides in lower efficiencies. Although the
GlmU crystal structure has been available, the roles of highly
conserved residues in substrate-binding regions are still unclear,
leading to some uncertainty concerning residues involved in sub-
strate binding and enzyme reaction. In order to probe the roles of
these highly conserved residues, structure- and sequence-guided
mutagenesis studies were carried out to broad substrate speci-
ficity and to improve catalytic efficiency for the formation of sugar
nucleotides. Among mutants obtained by substituting the highly
conserved residues in substrate-binding regions, tGlmU Q76E
mutant with altered catalytic activity toward the formation of un-
natural sugar nucleotide CDP-GlcNAc was identified. Two tGlmU
mutants, Y103F and N169R with significantly enhanced uridyl-
transferase activities were also identified. The structural and
biochemical characterization of the uridyltransferase active sites
and catalytic mechanism described herein deepened our under-
standing about mechanisms of GlmU reactions.
order to determine the necessity of this highly conserved residue
involved in substrate recognition and catalytic activity, mutants
(Q76E, Q76A, Q76P) were constructed and analyzed (Fig. 2).
To our delight, a unique catalytic activity of tGlmU Q76E mutant
0
was identified. We observed that cytidine 5 -triphosphate (CTP),
which had an amide group instead of a carbonyl group on the base,
could be accepted and converted to unnatural sugar nucleotide
CDP-GlcNAc with a yield of 53.1%. However, the uridyltransferase
activity of Q76E mutant decreased dramatically, as about half of the
activity of the wild-type tGlmU. Hydrogen bonds, which formed
between hydrogen atoms and electronegative hydrogen bond ac-
ceptors, usually are relatively weak interactions but nonetheless are
crucial for enzymatic catalytic reactions. Apparently, the substitu-
tion of Gln76 to Glu residue disrupted the initial hydrogen-bonded
interaction between residue Gln76 and exocyclic oxygen O4 of the
uracil base, leading to poor conversion efficiency to form UDP-
GlcNAc. Interestingly, the exocyclic amide group of unnatural sub-
strate CTP, which acted as the hydrogen bond acceptor, could form a
nascent hydrogen bond with side chain of the substituted Glu
residue, thus facilitating CTP to enter to the uridyltransferase
pocket and endowing Q76E mutant a novel activity to form un-
natural sugar nucleotide CDP-GlcNAc. We further measured that
the optimal pH slightly changed from 8.0 to 7.5 followed by the
substitution of Gln76 residue to Glu (Fig. 3).
2
. Results and discussion
2.1. Crystal structure-based design of tGlmU mutants
Based on the sequence alignment of homologous proteins and
Beside the Q76E mutant, we designed other two mutants Q76A
and Q76P to fully demonstrate the role of this key residue. Q76A
mutant showed that the uridyltransferase activity decreased
dramatically (40% activity of the wide-type tGlmU), presumably
resulting from the weakened hydrogen bond interaction caused by
mutagenesis. In contrast, the substitution of Gln76 with Pro residue
resulted in a slightly enhancement to UTP recognition and catalytic
activity (nearly 1.1 fold compared with the wide-type tGlmU).
crystal structure data, several highly conserved amino acid resi-
dues, which involved in the interactions with UDP-GlcNAc through
hydrogen bonds were identified (Fig. 1). In order to probe the roles
of these highly conserved residues in catalytic activity, we
substituted residue Gln76 in the uridine-binding region and Thr82,
Glu154, Asn169 in the GlcNAc-binding region, respectively. Two
other residues, Tyr103 and Asp105, which located nearby the uri-
dyltransferase active pocket, were also investigated.
All the mutants designed were transformed into E. coli
BL21(DE3) strains and subsequently overexpressed under standard
conditions as previously described for tGlmU. The soluble protein
expression levels of wide-type tGlmU or mutants were approxi-
mately 40 mg/L.
2.3. Roles of highly conserved residues in GlcNAc-binding region
16
GlmU crystal structure indicates that the GlcNAc moiety is
involved in extensive hydrogen bond network formed between its
hydroxyl groups and protein atoms of the active site cavity, either
directly or water-mediated. The GlcNAc-binding region consisting
of Thr82, Tyr103, Thr137-Gln140, Glu154, and Asn169-Glu171
structural segments arranges to recognize the GlcNAc in GlcNAc-1-
P and structural rearrangements required to complete the uridyl-
transferase reaction. The N-acetyl moiety is recognized by two
2.2. Roles of highly conserved residue Gln76
GlmU structure data demonstrates that UTP substrate is
involved in extensive hydrogen bond networks formed between its
functional groups and protein atoms of the active site cavity, either
directly or water-mediated. The uracil base is reinforced by two
hydrogen bonds between its exocyclic N3, O4 ring atoms and Gln76
side chain Oε1, Nε2, respectively. This key residue was proposed to
play important role in overall catalytic activity of the enzyme. In
0
hydrogen bonds between its N2 and Glu154, as well as between its
0
O7 and Thr82, respectively. In order to probe the necessity of these
highly conserved residues involved in substrate recognition and
1
1
20%
00%
UTP
CTP
ATP
dATP
8
6
4
2
0%
0%
0%
0%
0%
Q76E
Q76A
Q76P
tGlmU
Fig. 1. Schematic drawing of hydrogen bonds between UDP-GlcNAc and residues in the
uridyltransferase active site of E. coli K12 GlmU.
Fig. 2. Comparative uridyltransferase activity of tGlmU Gln76 mutants. The activity for
wide-type tGlmU was set to 100%.

72
S. Wang et al. / Carbohydrate Research 413 (2015) 70e74
positive residue Lys (E154K) thoroughly impaired its activity, pre-
sumably resulting from the slightly secondary conformation
changes caused by mutation. These results indicated that both the
hydrogen bonds formed by residue Glu154 and its negative charged
side chain were necessary to provide benefit in accommodating
GlcNAc-1-P as substrate for uridyltransferase reaction.
Residue Asn169, which locates on the b-sheet secondary struc-
ture and forms two hydrogen bonds with GlcNAc motif, acts as a
structural switch to drive the uridyltransferase active site into the
closed conformation upon binding of UDP-GlcNAc. Substitution of
Asn169 to a simply hydrophobic Ala residue (N169A) decreased the
uridyltransferase activity (55% activity of the wide-type tGlmU),
presumably resulting from the unstable substrate binding caused
by the weakened hydrogen-bonded interactions. Altering Asn169
to Asp (N169D) or Gln (N169Q) did not affect the catalytic activities.
Remarkably, N169R mutant, which was substituted by a positive
charged Arg residue, showed an enhanced uridyltransferase activ-
ity (nearly 1.4 fold of the wide-type tGlmU). The secondary struc-
ture predication indicated that the N169R mutant caused a slightly
secondary structure changes, thus facilitating GlcNAc-1-P to enter
the active pocket through the additional interaction with N-acetyl
arm of GlcNAc moiety (Fig. 5).
Fig. 3. Effects of pH on wild-type tGlmU and tGlmU Q76E mutant activities.
catalytic activity, residues Thr82, Glu154, and Asn169 were inves-
tigated by site-directed mutagenesis (Fig. 4).
Residue Thr82 locates on the a-helix nearby the uridyltransfer-
ase pocket and is proposed to be an important residue for recog-
nition of the N-acetyl arm of GlcNAc. Therefore, the polar uncharged
residue Thr82 was constructed to Ser (T82S), Gln (T82Q), and Gly
2
.4. Roles of highly conserved residues near the uridyltransferase
active pocket
(
T82G). Activity assays using GlcNAc-1-P as sugar-1-P donor
showed that T82S, T82Q and T82G mutants impaired the uridyl-
transferase activities (21e35% activity of the wide-type tGlmU).
Moreover, none of these three mutants could tolerance C2 modified
sugar-1-P derivatives (Glc-1-P, Glcz-1-P) as substrate to form the
corresponding sugar nucleotides. These results indicated that sugar-
Tyr103-Asp105 segment, which locates at the floor of the uri-
dyltransferase active pocket, also involves in interactions with UTP
1
-P stabilizing effect by Thr82 hydrogen-bonded interaction was
more significant when GlcNAc-1-P, instead of Glc-1-P or Glcz-1-P,
was used as the substrate. Substitution to larger residues Gln, Ser, or
to the simple residue Gly did not provide assistance in stabilizing
the interaction between sugar-1-P and the enzyme.
The functionally important residue Glu154 is highly conserved
in most related bacterial sugar nucleotide uridyltransferases. In
tGlmU, Glu154 locates on the b-turn secondary structure and forms
two hydrogen bonds with the N-acetyl arm and C3 hydroxyl group
of GlcNAc moiety, respectively. Altering residue Glu154 with Asp
(
E154D) or with a hydrophobic residue Leu (E154L) showed that the
uridyltransferase activities decreased drastically (nearly 20% ac-
tivity of the wide-type tGlmU). Substitution of Glu154 with a
1
1
1
40%
20%
00%
8
6
4
2
0%
0%
0%
0%
0%
Q
A
9
D
Q
R
A
03
Y1
F
A
5
4
L
9
9
9
3
6
6
N1
6
N1
6
N1
0
Y1
0
D1
WT
T82G T82
T82S
E154D E154K E15 N1
Fig. 4. Comparative uridyltransferase activity of tGlmU mutants constructed in
Fig. 5. Predication result of secondary structure changes caused by tGlmU N169R
GlcNAc-binding region. The activity for wide-type tGlmU was set to 100%.
mutant.

S. Wang et al. / Carbohydrate Research 413 (2015) 70e74
73
and GlcNAc-1-P substrates. Previous studies indicated that this
segment had significantly conformation changing during the
enzyme reaction. In order to demonstrate the role of this segment,
three mutants were constructed. Activity assay results indicated that
two mutants occurred on Tyr103 residue (Y103A and Y103F)
exhibited enhanced catalytic activities to form UDP-GlcNAc, espe-
cially Y103F, which had a 1.2 fold activity of the wide-type tGlmU.
Based on structure data, we presumed that it was the hydrophobic
characteristic of Tyr103 residue that affected the enzymatic reaction.
Substitution to other hydrophobic residues might contribute to
substrate-enzyme interactions, thus facilitating GlcNAc-1-P and UTP
to enter the uridyltransferase pocket. In contrast, the tGlmU D105A
mutant, which affected the adjacent hydrophobic interactions,
impaired its activity (two folds lower than the wide-type tGlmU).
target for antibiotic drug discovery. Based on our previous study,
we investigated the functional roles of several highly conserved
residues in the uridyltransferase domain of E. coli GlmU. Among
mutants constructed in our experiment, tGlmU Q76E mutant
exhibited a novel catalytic activity to convert CTP and GlcNAc-1-P
into CDP-GlcNAc. Two other mutants, Y103F and N169R, showed
enhanced activities toward the formation of UDP-GlcNAc. This
work deepens our understanding of the reaction mechanisms of
GlmU. The results further illuminate the relationship of highly
conserved residues and enzyme reaction. Most importantly, site-
directed mutagenesis provides
a prototypic template for a
structure-function analysis of the catalytic domains of uridyl-
transferases, which will benefit us to assess the roles of sequence-
conserved residues in the catalytic mechanism.
2.5. The secondary structure changes in tGlmU mutants
4. Experimental section
In order to fully understand the functional roles of the tGlmU
4.1. Material
mutants, we predicated the secondary structures by SWISS-MODEL
Workspace to calculate conformation changes (Table 1). The
predication results demonstrated that all the mutants did not affect
the secondary structures where the mutated residues located.
However, substitution of the corresponding residues had effect on
the secondary structures formed by the amino acids adjacent and
far way from these mutated residues.
E. coli BL21(DE3) strain hosting the E. coli K12 tGlmU gene was
16
constructed in our Lab previously. Fast Multi-site-mutagenesis kit
was from Transgen Biotechnology Company. Sugar-1-Ps and NTPs
were from Sigma.
4
.2. Site-directed mutagenesis
2.6. Enzymatic synthesis of unnatural CDP-GlcNAc by coupling of
Site-directed mutagenesis was carried out using the Fast Multi-
Q76E mutant and NahK
site-mutagenesis kit according to the protocol from the manufac-
turer. The primers used were shown in Table 2.
In order to demonstrate the application of tGlmU Q76E mutant
for unnatural sugar nucleotide preparation, we performed a one-
pot de novo CDP-GlcNAc synthetic reaction. To facilitate the
biosynthesis reaction, the N-acetylhexosamine 1-kinase from Bifi-
dobacterium longum (NahK), which could convert ATP and GlcNAc
into GlcNAc-1-P, was combined with tGlmU Q76E mutant. The
purified CDP-GlcNAc was identified by ESI-MS and compared with
the CDP-GlcNAc produced by CjGlmU in our previous work.18
4
.3. Protein expression of wide-type tGlmU and mutants
All mutants constructed were further confirmed by sequencing
and subsequently transformed into E. coli BL21(DE3) strain for
protein expression. The E. coli cells were cultured in LuriaeBertani
medium supplemented with 100
17
mg/mL ampicillin. When OD600
reached 0.6e0.8, isopropyl-1-thio-
b-D-galactopyranoside was
added to a final concentration of 1 mM. The induced culture was
ꢀ
3
. Conclusion
further incubated at 28 C for 3 h with shaking at 200 rpm. Cells
were harvested by centrifugation at 12,000 rpm for 5 min. Protein
In summary, GlmU is a key cytoplasmic enzyme involved in
prokaryotic UDP-GlcNAc biosynthetic pathway and an attractive
purification was performed utilizing Ni-NTA resin (Amersham)
according to the manufacturer's instructions at 4 C. The purified
ꢀ
Table 1
The secondary structure changes caused by tGlmU mutants
Mutant
Predicted amino acid residues related to changes of secondary structurea
Q76A
T29(E/H), L30(E/H), L71(E/C), F93(C/H)
Q76E
Q76P
T29(E/H), L30(E/C), L71(E/C), Y103(E/C), E145(C/E)
K25(H/C), E145(C/E)
T82Q
T82G
T82S
K25(H/C), T29(E/H), L30(E/H), L71(E/C), E145(C/E)
T29(E/H), L30(E/H), L71(E/C), T82(C/H), Y103(E/C), E145(C/E), Q164(C/H)
T29(E/H), L30(E/H), L71(E/C), T82(C/H), E195(C/E)
T29(E/H), L30(E/H), F93(C/H), Y139(C/E), G140(C/E), E145(C/E)
K25(H/C), T29(E/H), L30(E/H), F93(C/H), Y139(C/E), G140(C/E)
T29(E/H), L30(E/H), L71(E/C)
L71(E/C), L108(C/E), I109(C/E), R163(H/C)
T29(E/H), L30(E/H), L71(E/C), F93(C/H)
T29(E/H), L30(E/H), L71(E/C), E145(C/E)
L71(E/C), F93(C/H)
Y103A
Y103F
D105A
E154D
E154L
E154K
N169A
N169D
N169R
N169Q
L71(E/C), L108(C/E), I109(C/E)
L71(E/C), F93(C/H), R163(H/C)
K25(H/C), L71(E/C), E145(C/E)
C represents coil structure.
E represents extened/beta.
H represents helix.
a
The secondary structure changes were predicted by SWISS-MODEL Workspace.

74
S. Wang et al. / Carbohydrate Research 413 (2015) 70e74
Table 2
4.5. tGlmU Q76E mutant catalyzed the formation of CDP-GlcNAc
Primers used for site-directed mutagenesis
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
Q76E forward
Q76E revise
Q76A forward
Q76A revise
Q76P forward
Q76P revise
T82G forward
T82G revise
T82Q forward
T82Q revise
T82S forward
T82S revise
Y103A forward
Y103A revise
Y103F forward
Y103F revise
D105A forward
D105A revise
E154D forward
E154D revise
E154L forward
E154L revise
E154K forward
E154K revise
N169R forward
N169R revise
N169D forward
N169D revise
N169Q forward
N169Q revise
N169A forward
N169A revise
5 aactgggtgcttgaggcagagcagctgggtacgggtcatgc3
The one-pot two-enzyme synthetic reaction was carried out
coupling of tGlmU Q76E mutant and NahK. The reaction system
contained 5 mM GlcNAc, 5 mM CTP and 5 mM ATP in a final volume
of 10 mL and incubated at 37 C for 48 h. Yeast inorganic pyro-
phosphatase, which could drive the reaction forward by consuming
0
5 agctgctctgcctcaagcacccagttaaggttgtcgtgtttg3
0
5 accttaactgggtgcttgcggcagagcagctg3
0
5 cgcaagcacccagttaaggttgtcgtctttc3
ꢀ
0
5 accttaactgggtgcttccggcagagcagctg3
0
5 cggaagcacccagttaaggttgtcgtctttc3
0
5 caggcagagcagctgggtgggggtcatgcaatgc3
the byproduct pyrophosphate, was also added to a final 1 U/mL. The
0
0
5 ccacccagctgctctgcctgaagcacccagttaagg3
ꢀ
reaction was quenched by boiling at 100 C for 10min followed by
0
5 caggcagagcagctgggtcagggtcatgcaatgc3
concentration at 13,000 rpm for 30 min. Bio-gel P2 gel filtration
column (Bio-Rad) was used to separate the product from the re-
action mixture and the fractions containing CDP-GlcNAc product
were pooled and lyophilized.
5 tgacccagctgctctgcctgaagcacccagttaagg3
0
5 caggcagagcagctgggtagcggtcatgcaatgc3
0
5 gctacccagctgctctgcctgaagcacccagttaagg3
0
5 atgaagacattttaatgctcgccggcgacgtgcc3
0
5 ggcgagcattaaaatgtcttcatcatcggcaa3
0
5 atgaagacattttaatgctcttcggcgacgtgcc3
0
5 gaagagcattaaaatgtcttcatcatcggcaa3
Acknowledgements
0
5 taatgctctacggcgccgtgccgctgatctctg3
0
5 ggcgccgtagagcattaaaatgtcttcatcat3
The authors acknowledge the support of National Basic Research
Program of China (973 Program, No. 2012CB822102), the Key Grant
Project of Chinese Ministry of Education (No. 313033), and Pro-
motive Research Fund for Excellent Young and Middle-Aged Sci-
entists of Shandong Province (No. BS2013SW023).
0
0
0
1
2
3
4
5
6
5 gcaaagttaccggcattgttgaccacaaagatgc3
5 gtcaacaatgccggtaactttgccgttttcacggg3
0
0
5 gcaaagttaccggcattgttttgcacaaagatgc3
5 aaaacaatgccggtaactttgccgttttcacggg3
0
5 gcaaagttaccggcattgttcagcacaaagatgc 3
0
5 gaacaatgccggtaactttgccgttttcacggg 3
0
5 gtcagattcaggagatccgcaccggcattctg3
0
5 cggatctcctgaatctgacgctgctcgtcgg3
Supplementary data
0
5 gtcagattcaggagatcgacaccggcattctg3
0
5 cgatctcctgaatctgacgctgctcgtcgg3
0
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.carres.2015.05.007.
5 gtcagattcaggagatccagaccggcattctg3
0
5 ctggatctcctgaatctgacgctgctcgtcgg3
0
5 gtcagattcaggagatcgccaccggcattctg3
0
5 ggcgatctcctgaatctgacgctgctcgt3
References
1
. Barbosa MD, Yang G, Fang J, Kurilla MG, Pompliano DL. Antimicrob Agents
Chemother 2002;46:943e6.
fraction was dialyzed with a 10K centrifugal filter to remove
imidazole and buffer components. Protein concentration was
quantified using Bradford protein assay kit according to the pro-
tocol from the manufacturer. The wide-type tGlmU was also
expressed, purified, and quantified as the positive control. All the
2. Whitfield C, Trent MS. Annu Rev Biochem 2014;83:99e128.
3. Masuko S, Bera S, Green DE, Weiwer M, Liu J, DeAngelis PL, et al. J Org Chem
012;77:1449e56.
2
4. Ovchinnikova OG, Rozalski A, Liu B, Knirel YA. Biochemistry 2013;78:
798e817.
ꢀ
5. Zhang D, Wang PG, Qi Q. Biotechnol Lett 2007;29:1729e33.
proteins were stored at ꢁ20 C with 20% glycerol existed.
6. Niemann MC, Bartrina I, Ashikov A, Weber H, Novak O, Spichal L, et al. Proc Natl
Acad Sci U.S.A 2015;112:291e6.
7. Mohorko E, Glockshuber R, Aebi M. J Inherit Metab Dis 2011;34:869e78.
. Chang R, Vo TT, Finney NS. Carbohydr Res 2006;341:1998e2004.
4.4. Activity assays for mutant enzymes
8
9
. Zeng Y, Shabalin Y, Szumilo T, Pastuszak I, Drake RR, Elbein AD. Anal Biochem
Enzymatic reactions were carried out under standard conditions
1996;239:99e106.
in duplicates in a total 100
pH 7.5), 100 mM NaCl, 5 mM sugar-1-P, 5 mM NTPs, 5 mM MgCl
and 1 U/mL pyrophosphatase. The reactions were performed at
mL solution containing 100 mM TriseHCl
10. Mengin-Lecreulx D, van Heijenoort J. J Bacteriol 1993;175:6150e7.
11. Mochalkin I, Lightle S, Zhu Y, Ohren JF, Spessard C, Chirgadze NY, et al. Protein
Sci 2007;16:2657e66.
(
2
,
12. Kostrewa D, D'Arcy A, Takacs B, Kamber M. J Mol Biol 2001;305:279e89.
ꢀ
ꢀ
3
7 C for 60 min and quenched by boiling at 100 C for 5 min fol-
13. Han S, Forman MD, Loulakis P, Rosner MH, Xie Z, Wang H, et al. J Mol Biol
lowed by centrifugation at 13,000 rpm for 15 min. The consumption
of NTP and the formation of sugar nucleotide during reactions were
analyzed by capillary electrophoresis (CE).
2006;360:814e25.
1
1
4. Brown K, Pompeo F, Dixon S, Mengin-Lecreulx D, Cambillau C, Bourne Y. EMBO
J 1999;18:4096e107.
5. Olsen LR, Roderick SL. Biochemistry 2001;40:1913e21.
The products were characterized by a P/ACE MDQ Capillary
16. Fang J, Guan W, Cai L, Gu G, Liu X, Wang PG. Bioorg Med Chem Lett 2009;19:
6429e32.
Electrophoresis System. Electrophoresis was run in 75
40 cm to detector) bare fused silica capillary, under 25 kV with UV
detection at 254 nm.
mmꢂ50 cm
1
1
7. Nishimoto M, Kitaoka M. Appl Environ Microbiol 2007;73:6444e9.
8. Fang J, Xue M, Gu G, Liu XW, Wang PG. Bioorg Med Chem Lett 2013;23:
4303e7.
(