A chemoenzymatic route to synthesize unnatural sugar nucleotides using a novel N-acetylglucosamine-1-phosphate pyrophosphorylase from Camphylobacter jejuni NCTC 11168

Article abstract of DOI:10.1016/j.bmcl.2013.06.003

A novel N-acetylglucosamine-1-phosphate pyrophosphorylase was identified from Campylobacter jejuni NCTC 11168. An unprecedented degree of substrate promiscuity has been revealed by systematic studies on its substrate specificities towards sugar-1-P and NTP. The yields of the synthetic reaction of seven kinds of sugar nucleotides catalyzed by the enzyme were up to 60%. In addition, the yields of the other nine were around 20%. With this enzyme, three novel sugar nucleotide analogs were synthesized on a preparative scale and well characterized.

Full text of DOI:10.1016/j.bmcl.2013.06.003

Bioorganic & Medicinal Chemistry Letters 23 (2013) 4303–4307
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A chemoenzymatic route to synthesize unnatural sugar nucleotides
using a novel N-acetylglucosamine-1-phosphate pyrophosphorylase
from Camphylobacter jejuni NCTC 11168
a,c
Junqiang Fang a, Mengyang Xue a,b, Guofeng Gu a, Xian-wei Liu a, , Peng George Wang
⇑
a National Glycoengineering Research Center, Shandong University, Jinan, Shandong 250100, People’s Republic of China
b The State Key Laboratory of Microbial Technology and School of Life Sciences, Shandong University, Jinan, Shandong 250100, People’s Republic of China
c Department of Chemistry, Georgia State University, PO Box 4098, Atlanta, GA 30302-4098, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel N-acetylglucosamine-1-phosphate pyrophosphorylase was identified from Campylobacter jejuni
NCTC 11168. An unprecedented degree of substrate promiscuity has been revealed by systematic studies
on its substrate specificities towards sugar-1-P and NTP. The yields of the synthetic reaction of seven
kinds of sugar nucleotides catalyzed by the enzyme were up to 60%. In addition, the yields of the other
nine were around 20%. With this enzyme, three novel sugar nucleotide analogs were synthesized on a
preparative scale and well characterized.
Received 19 March 2013
Revised 30 May 2013
Accepted 1 June 2013
Available online 11 June 2013
Keywords:
N-Acetylglucosamine-1-phosphate
pyrophosphorylase
Ó 2013 Elsevier Ltd. All rights reserved.
Substrate specificity
UDP-GlcNAc
Unnatural sugar nucleotide
Chemoenzymatic synthesis
Unnatural sugar nucleotides, analogs of common sugar nucleo-
tides equipped with functional groups or its derivatives, are
indispensable materials for deciphering structure–function rela-
tionships in carbohydrate-associated pathways and discovering
carbohydrate-based drugs. In addition to presenting nine common
sugar nucleotides identified in mammalian cells, several unnatural
sugar nucleotides have been discovered from different organisms,
development of chemical and/or enzymatic methods to derivatize
UDP-GlcNAc with diverse moieties has attracted much attention,
however, for enzymatic methods, the major bottleneck is the lim-
ited availability of functional enzymes which have broad substrate
specificity. N-acetylglucosamine-1-phosphate pyrophosphorylase
(GlmU) is a cytoplasmic enzyme involved in prokaryotic biosynthe-
sis pathway and an attractive target for antibiotic drug discovery.10
Our previous work on Escherichia coli K12 GlmU (EcGlmU) demon-
strated that EcGlmU could be used to prepare UDP-GlcNAc deriva-
tives.11 Unfortunately, the relatively strict substrate specificity and
low yields of products have restricted the application of this
enzyme.
To gain insight into chemoenzymatic synthesis of unnatural su-
gar nucleotides, especially those UDP-GlcNAc derivatives, we focus
on Campylobacter jejuni glycosylation system12–15 to find a feasible
and effective approach to synthesize unnatural sugar nucleotides
which can greatly benefit synthetic, biological and medicinal
chemistry. Herein, we provide a chemoenzymatic method to syn-
thesize multiple unnatural sugar nucleotides using a novel N-ace-
tylglucosamine-1-phosphate pyrophosphorylase (CjGlmU) from C.
jejuni NCTC 11168 (Scheme 1).
for example, UDP-2-acetamindo-2,6-dideoxy-
ulose,1 UDP-4-keto-6-deoxyl-GlcNAc,2 UDP-2,4-diacetamido-Bac,
ADP-Glc,3 dTDP-Glc, dTDP-4-keto- -rhamnose.4
-rhamnose, dTDP-
a-D-xylo-hexos-4-
L
L
Recent studies show that those unnatural sugar nucleotides are
considered to have functional impact on biological activity, selec-
tivity and pharmacokinetic properties of glycoconjugates.
Uridine 50-diphosphate N-acetylglucosamine (UDP-GlcNAc) is a
ubiquitous and essential cytoplasmic amino sugar nucleotide, plays
an important role in the biosynthetic pathway of peptidoglycan,5,6
the core and lipid A moieties of the lipopolysaccharides,7 entero-
bacterial common antigen, some O antigens of Gram-negative
bacteria and the teichoic acid of Gram-positive bacteria. Chemical
and/or enzymatic approaches have been thoroughly studied to
synthesize UDP-GlcNAc.8,9 Given the potential roles of unnatural
sugar nucleotides and the functional roles of UDP-GlcNAc, the
According to our previous work on Escherichia coli K12
GlmU,11,16 we cloned and overexpressed CjGlmU. The catalytic
activity of CjGlmU was detected under standard condition in a total
⇑
Corresponding author. Tel.: +86 531 88366078; fax: +86 531 88363002.
E-mail address: xianweiliu@sdu.edu.cn (X. Liu).
100 lL solution system containing 20 mM Tris–HCl (pH 7.5), 5 mM
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.06.003

4304
J. Fang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4303–4307
MgCl2, 5 mM GlcNAc-1-P, 5 mM UTP, 1 U/mL yeast inorganic pyro-
phosphatase and 10 L purified protein (about 6 g). Yeast inor-
l
l
ganic pyrophosphatase could drive the reaction forward by
2Pi
pyrophosphatase
PPi
NTP
O
O
R
R
O
O
O
P
O-
R'
O
O-
O
P
O-
O
P
O-
O
O
Cj GlmU
X
HO
Scheme 1. Enzymatic synthesis of unnatural sugar nucleotides using CjGlmU from
sugar-1-Ps and NTPs.
Figure 1. Divalent metal-ion preference.
Table 1
Yields of NDP-GlcNAc analogs synthesized by CjGlmU
Entry
R1
R2
(d)NTP substrate
UTP
Product
Exact mass
607.1
Reaction yielda (%)
97.9
1
–OH
UDP-GIcNAC
2
–H
dUTP
dUDP-GIcNAC
591.1
94.2
3
4
5
6
7
8
–H
dTTP
CTP
dTDP-GIcNAC
CDP-GIcNAC
605.1
606.1
590.1
614.1
630.1
614.1
97.7
64.6
3.6
–OH
–H
dCTP
dm5CTP
ATP
dCDP-GIcNAC
dm5CDP-GIcNAC
ADP-GIcNAC
dADP-GIcNAC
–H
N/Ab
3.3
–OH
–H
dATP
N/Ab
9
–H
dm6ATP
GTP
dm6ADP-GIcNAC
GDP-GIcNAC
628.1
646.1
10.8
10
–OH
N/Ab
a
Yield from profiles of capillary electrophoresis.
No product detected by MS or capillary electrophoresis.
b

J. Fang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4303–4307
4305
consuming the byproduct pyrophosphate. The reaction was per-
formed at 37 °C for 2 h and quenched by boiling at 100 °C for
15 min. Capillary electrophoresis (CE) was used to detect the
formation of sugar nucleotide. Divalent metal ion preference was
also characterized for pyrophosphorylase reaction catalyzed by
CjGlmU. The result (Fig. 1) illustrated that CjGlmU requires a
divalent metal ion as a cofactor for substrate binding and catalysis.
The overall divalent metal ion preference of CjGlmU was
Mg2+ > Zn2+ > Co2+ > Ni2+ > Mn2+ > Fe2+ > Ga2+ > Cu2+. As a control,
reaction with EDTA did not yield any products, illustrated that
divalent metal ion would be an absolutely necessary cofactor in
CjGlmU reaction.
ToinvestigatethesubstratespecificitytowardsNTPs, GlcNAc-1-P
was used as the sugar-1-P substrate. Out of ten nucleotide triphos-
phate (NTP) compounds tested, seven were accepted by CjGlmU
(Table 1). The natural substrate uridine 50-triphosphate (entry 1,
UTP) could convert to its corresponding product UDP-GlcNAc almost
completely. dUTP (entry 2) with small changes on ribose 20-hydroxyl
group and dTTP (entry 3)with an additional methyl group on the
nucleic acid base were good substrates as well, with reaction yields
of 94.2% and 97.7%, respectively. Remarkably, cytidine 50-triphos-
phate (entry 4, CTP), which has a NH2 group instead of a carbonyl
group on the base, can also be accepted and converted to a novel su-
gar nucleotide analog CDP-GlcNAc with a yield of 64.6%. However,
furthermodificationonthepentose(entry5, dCTP)orthebase(entry
6, dm5CTP) of the nucleotide resulted in unfavourable substrate for
the enzyme. As expected, nucleotide triphosphates bearing bulky
bases, such as ATP and GTP as well as their analogs (entry 7–10),
were not suitable for CjGlmU pyrophosphorylase reactions, except
entry 7 and 9 with poor yields. The purine ring of adensine or guano-
sine may exists steric hindrance and weaken binding affinity
between the nucleotide and the active site residues of CjGlmU. The
yields of NTP-GlcNAcs showed in Table 1 indicated a substrate
tolerance of CjGlmU for NTPs in the following order: UTP > dTTP > -
dUTP > CTP ꢀ dm6ATP > dCTP > ATP, suggesting that CjGlmU
prefers pyrimidine nucleotide than purine nucleotide.
Table 2
Yields of unnatural sugar nucleotides synthesized by CjGlmU
Entry
Sugar-1-P
(d)NTP substrate
UTP
Product
UDP-Glc
Exact mass
566.1
Reaction yielda (%)
60.4
11
12
13
14
dUTP
dTTP
UTP
dUDP-Glc
dTDP-Glc
UDP-Man
550.1
564.1
566.1
20.8
0.6
N/Ab
15
16
UTP
UTP
UDP-GlcNPr
UDP-GlcNBu
621.1
635.1
79.9
70.6
17
18
19
dUTP
dTTP
UTP
dUDP-GlcNPr
GTDP-GalNPr
UDP-GalNAc
605.1
619.1
607.1
21.6
2.9
12.7
20
21
dUTP
dTTP
dUDP-GalNAc
GTDP-GalNAc
591.1
605.1
0.8
0.5
a
Yield from profiles of capillary electrophoresis.
No product detected by MS or capillary electrophoresis.
b

4306
J. Fang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4303–4307
We further investigated the substrate tolerance of CjGlmU for
corresponding consensus motif of G-L-G-T-R-I-K-S-Q-K-P-K as a
signature of pyrophosphorylase family was also existed in CjGlmU
protein. Therefore, similar to its orthologs, CjGlmU might recognize
and bind to sugar-1-P through specific hydrogen bonds between
the hydroxyl groups of the glucosyl moiety and the residues in
the active sites.
Further investigation of CjGlmU substrate promiscuity was per-
formed with analogs for both substrates. The results showed the
CjGlmU activities rank of forward reactions as UDP-Glc-
NAc > dTDP-GlcNAc > dUDP-GlcNAc > UDP-GlcNPr > UDP-GlcNBu
> CDP-GlcNAc > UDP-Glc > dUDP-GlcNPr > dUDP-Glc > UDP-Gal-
NAc-1-P > dm6ADP-GlcNAc > dCDP-GlcNAc > ADP-GlcNAc > dTDP-
GlNPr > dTDP-Glc > dUDP-GalNAc > dTDP-GalNAc. Compared with
our previous data from EcGlmU,11 CjGlmU has an unprecedented
degree of substrate promiscuity towards sugar-1-P and NTP, which
illustrated a feasible route to chemoenzymatic synthesis of unnat-
ural sugar nucleotides (Table 3).
To demonstrate the application of the novel CjGlmU for synthe-
sis of unnatural sugar nucleotides, we preformed synthetic reac-
tions in multiple mg scale for CDP-GlcNAc, UDP-GlcNBu and
UDP-GlcNPr, respectively. The reaction system contained 5 mM
sugar-1-P and 5 mM NTP in a final volume of 10 mL. Bio-gel P2
gel filtration column (Bio-Rad) was used to separate the products
from the reaction mixture and the fractions containing unnatural
sugar nucleotides were pooled and concentrated. The isolated
unnatural sugar nucleotides were identified by ESI-MS or
LCMS-IT-TOF and NMR spectroscropy.19
sugar-1-P. Out of six sugar 1-phosphates (sugar-1-P) compounds
tested, five were accepted by CjGlmU (Table 2). Contrast to the
yield of 97.9% with the natural substrate of GlcNAc-1-P, reaction
yield with substrate analog in entry 11 (Glc-1-P) bearing a C2 hy-
droxyl group instead of N-acetyl group decreased dramatically.
Man-1-P, a C2 isomer, was not a suitable substrate (entry 14),
demonstrating poor acceptance for C2 position epimerization.
Those results indicated that the C2 N-acetyl group may play a sig-
nificant role in substrate recognition and reaction efficiency.
Remarkably, compounds with longer N-acyl groups (GlcNBu-1-P
and GlcNPr-1-P) were also excellent substrates, leading to generate
UDP-GlcNBu and GlcNPr at comparable level (entries 15 and 16).
However, reaction efficiency decreased for C4 isomer of
GlcNAc-1-P (entry 19, GalNAc-1-P), presumably resulting from
the steric hindrance of 4-hydroxyl group. The modifications at
N-acyl or C4 hydroxyl group may change the size of these
sugar-1-P compounds, thus decreasing their interactions with
CjGlmU and causing them to fail to enter the catalytic pocket.
Till now, the crystal structures of several prokaryotic GlmU pro-
teins have been resolved.17,18 Data suggested that pyrophosphory-
lase domain of these enzymes which adopts a globular structure
formed by the central b-sheet and flexible loops, shares similarity
with the nucleotide diphosphate sugar pyrophosphorylase family
with a strict conservation motif of G-X-G-T-(RS)-M-X4-P-K.17
A
N-Acetylglucosamine-1-phosphate pyrophosphorylase is
a
Table 3
cytoplasmic enzyme involved in prokaryotic biosynthesis path-
way and an attractive target for antibiotic drug discovery. Based
on our previous work, we cloned a novel CjGlmU from C. jejuni
NCTC 11168. An unprecedented degree of substrate promiscuity
has been revealed by systematic studies on CjGlmU substrate
specificities towards Sugar-1-P and NTP, providing a chemoenzy-
matic method to synthesize multiple unnatural sugar nucleo-
tides. Total sixteen unnatural sugar nucleotides could be
synthesized using this novel enzyme, the yields of the synthetic
reaction of seven kinds of sugar nucleotides were up to 60%, in
addition, the yields of the other nine were around 20%. With this
enzyme, three novel sugar nucleotide analogs were synthesized
on a preparative scale and well characterized. Most importantly,
this novel CjGlmU provides a prototypic template for structure–
function analysis of the catalytic domains of pyrophosphorylase
enzymes. Further mutagenesis, kinetic and structural studies will
be necessary to assess the roles of the functionally important
residues in the catalytic mechanism and the specificity of
CjGlmU.
Conversion ratio comparison between CjGlmU and EcGlmU
Sugar-1-P
NTPs
Yielda (%)
CjGlmUb EcGlmUb
OH
O
N
O
HO
HO
NH
O
NH
O
O
O
O
O
O P OH
OH
97.9
94.2
97.7
60.4
20.8
0.6
90.9
72.4
65.8
35.3
24.2
15.3
HO P O P O P O
O
OH OH OH
OH OH
O
NH
O
O
O
O
N
O
HO P O P O P O
OH OH OH
OH
H
O
N
H3C
NH
O
O
O
O
HO P O P O P O
O
OH OH OH
OH
H
OH
O
N
O
Acknowledgments
HO
HO
NH
O
OH
O
O
O
O
O P OH
OH
HO P O P O P O
The authors acknowledge the support of National Basic Re-
search Program of China (973 Program, No.2012CB822102), Na-
tional Natural Science Foundation of China (NNSFC,
No.31100588) and Shandong Excellent Young Scientist Research
Award Fund (SDEYSRAF, No.BS2010SW002). The authors thank
the genome of C. jejuni NCTC 11168 as a gift from Dr. Maojun
Zhang (Chinese Center for Disease Control and Prevention).
O
OH OH OH
OH OH
O
NH
O
O
O
O
N
O
HO P O P O P O
OH OH OH
OH
H
O
N
H3C
Supplementary data
NH
O
O
O
O
HO P O P O P O
Supplementary data (methods of CjGlmU protein expression,
the CE profiles and ESI-MS/LCMS-IT-TOF data for substrates and
products, NMR spectra for CDP-GlcNAc, UDPGlcNPr and UDP-
GlcNBu) associated with this article can be found, in the online ver-
sion, at http://dx.doi.org/10.1016/j.bmcl.2013.06.003.
O
OH OH OH
OH
Yield from profiles of capillary electrophoresis.
H
a
b
Enzyme concentration ꢁ1.5 mg/100
lL reaction system.

J. Fang et al. / Bioorg. Med. Chem. Lett. 23 (2013) 4303–4307
4307
17. Brown, K.; Pompeo, F.; Dixon, S.; Mengin-Lecreulx, D.; Cambillau, C.; Bourne, Y.
EMBO J. 1999, 18, 4096.
References and notes
18. Sulzenbacher, G.; Gal, L.; Peneff, C.; Fassy, F.; Bourne, Y. J. Biol. Chem. 2001, 276,
11844.
1. Guerry, P.; Ewing, C. P.; Schoenhofen, I. C.; Logan, S. M. J. Bacteriol. 2007, 189,
6731.
2. Vijayakumar, S.; Merkx-Jacques, A.; Ratnayake, D. B.; Gryski, I.; Obhi, R. K.;
Houle, S.; Dozois, C. M.; Creuzenet, C. J. Biol. Chem. 2006, 281, 27733.
3. Petreikov, M.; Shen, S.; Yeselson, Y.; Levin, I.; Bar, M.; Schaffer, A. A. Planta
2006, 224, 1465.
4. Giraud, M. F.; Leonard, G. A.; Field, R. A.; Berlind, C.; Naismith, J. H. Nat. Struct.
Biol. 2000, 7, 398.
5. Barbosa, M. D.; Yang, G.; Fang, J.; Kurilla, M. G.; Pompliano, D. L. Antimicrob.
Agents Chemother. 2002, 46, 943.
6. Ha, S.; Walker, D.; Shi, Y.; Walker, S. Protein Sci. 2000, 9, 1045.
7. Creuzenet, C.; Lam, J. S. Mol. Microbiol. 2001, 41, 1295.
8. Okuyama, K.; Hamamoto, T.; Ishige, K.; Takenouchi, K.; Noguchi, T. Biosci.
Biotechnol. Biochem. 2000, 64, 386.
9. Lazarevic, D.; Thiem, J. Carbohydr. Res. 2006, 341, 569.
10. Mengin-Lecreulx, D.; van Heijenoort, J. J. Bacteriol. 1993, 175, 6150.
11. Fang, J.; Guan, W.; Cai, L.; Gu, G.; Liu, X.; Wang, P. G. Bioorg. Med. Chem. Lett.
2009, 19, 6429.
12. Nachamkin, I.; Allos, B. M.; Ho, T. Clin. Microbiol. Rev. 1998, 11, 555.
13. Karlyshev, A. V.; Ketley, J. M.; Wren, B. W. FEMS Microbiol. Rev. 2005, 29, 377.
14. Schoenhofen, I. C.; Vinogradov, E.; Whitfield, D. M.; Brisson, J. R.; Logan, S. M.
Glycobiology 2009, 19, 715.
15. Gundogdu, O.; Bentley, S. D.; Holden, M. T.; Parkhill, J.; Dorrell, N.; Wren, B. W.
BMC Genomics 2007, 8, 162.
19. Spectroscopic data of CDP-GlcNAc: 1H NMR (600 MHz, D2O): d 7.90 (d, 1H,
J = 7.6 Hz, H-600), 6.06 (d, 1H, J = 7.6 Hz, H-500), 5.90 (d, 1H, J = 4.2 Hz, H-10), 5.43
(dd, 1H, J = 2.8, 6.8 Hz, H-1), 4.18–4.28 (m, 4H), 4.08–4.14 (m, 1H), 3.61–3.90
(m, 5H), 3.46 (t, 1H, J = 9.6 Hz, H-4), 1.99 (s, 3H, –NHCOCH3); 13C NMR
(150 MHz, D2O): d 174.8, 165.5, 156.8, 141.7, 96.5, 94.5, 89.4, 82.7, 74.3, 73.1,
71.0, 69.5, 69.3, 64.7, 59.4, 53.7, 22.1. ESI-MS (negative ion) calcd for
C
17H28N4O16P2: 606.10; found: 605.4 [MꢂH]ꢂ, 627.4 [Mꢂ2H+Na]ꢂ;
spectroscopic data of UDP-GlcNPr: 1H NMR (400 MHz, D2O): d 7.90 (d, 1H,
J = 8.1 Hz, H-600), 5.93 (d, 1H, J = 3.3 Hz, H-10), 5.91 (d, 1H, J = 8.1 Hz, H-500), 5.46
(dd, 1H, J = 3.2, 7.1 Hz, H-1), 4.29–4.32 (m, 2H, H-20, H-30), 4.11–4.23 (m, 3H, H-
40, H-5a, b0), 3.73–3.97 (m, 5H, H-5, H-4, H-3, H-6a, b), 3.49 (t, 1H, J = 8.7 Hz),
2.30 (q, 2H, J = 7.7 Hz, CH3CH2CO), 1.07 (t, 2H, J = 7.7 Hz, CH3CH2CO); 13C NMR
(100 MHz, D2O): d 178.8, 166.2, 151.8, 141.7, 102.7, 94.6, 88.5, 83.3, 73.8, 73.1,
71.0, 69.7, 65.0, 62.6, 60.4, 53.6, 29.1, 9.5. ESI-MS (negative ion) calcd for
C
18H29N3O17P2: 621.10; found: 620.5 [MꢂH]ꢂ, 642.3 [Mꢂ2H+Na]ꢂ;
spectroscopic data of UDP-GlcNBu: 1H NMR (400 MHz, D2O): d 7.90 (d, 1H,
J = 8.1 Hz, H-600), 5.93 (d, 1H, J = 3.3 Hz, H-10), 5.92 (d, 1H, J = 8.1 Hz, H-500), 5.46
(dd, 1H, J = 3.4, 7.1 Hz, H-1), 4.30–4.32 (m, 2H, H-20, H-30), 4.13–4.23 (m, 3H, H-
40, H-5a, b0), 3.73–3.97 (m, 5H, H-5, H-4, H-3, H-6a, b), 3.49 (t, 1H, J = 8.7 Hz),
2.27 (t, 2H, J = 7.3 Hz, CH3CH2CH2CO), 1.56 (m, 2H, CH3CH2CH2CO), 0.86 (t, 2H,
J = 7.3 Hz, CH3CH2CH2CO); 13C NMR (100 MHz, D2O): d 177.8, 166.3, 151.9,
141.7, 102.8, 94.7, 88.5, 83.4, 73.8, 73.1, 71.0, 69.8, 65.0, 60.4, 59.6, 53.6, 37.7,
19.0, 12.9. ESI-MS (negative ion) calcd for C19H31N3O17P2: 635.11; found: 634.4
[MꢂH]ꢂ, 656.4 [Mꢂ2H+Na]ꢂ.
16. Guan, W.; Cai, L.; Fang, J.; Wu, B.; Wang, P. G. Chem. Commun. (Camb.) 2009,
6976.