COMMUNICATION
Enzymatic synthesis of UDP-GlcNAc/UDP-GalNAc analogs using
N-acetylglucosamine 1-phosphate uridyltransferase (GlmU)w
a
b
a
Wanyi Guan,z Li Cai,z Junqiang Fang,z Baolin Wub and Peng George Wang*b
Received (in College Park, MD, USA) 26th August 2009, Accepted 17th September 2009
First published as an Advance Article on the web 20th October 2009
DOI: 10.1039/b917573c
Reports the generation of a library composed of UDP-GlcNAc/
UDP-GalNAc and investigates the substrate specificity of
Escherichia coli GlcNAc-1-P uridyltransferase GlmU.
constructed by GlmU using sugar-1-P and UTP. In addition,
yeast inorganic pyrophosphatase was added to drive the
GlmU reaction forward by degrading the byproduct PPi.
All sugar nucleotides produced here were purified by anion
exchange chromatography followed by desalting with size
exclusion chromatography.
In glycobiology research, analogs of sugars and glyco-
conjugates equipped with functional groups that enable facile
detection or further derivation are indispensable materials for
deciphering structure–function relationships in carbohydrate-
associated pathways1 and discovering carbohydrate-based
drugs.2 These analogs can be produced chemically or
enzymatically. Of the available enzymatic methods, glycosyl-
transferase-catalyzed oligosaccharide synthesis is generally
preferred as the reactions are regio- and stereoselective and
also eliminate the typical multiple protection/deprotection
steps required for chemical synthesis. As a result, access
to commercially unavailable unnatural sugar nucleotide
substrates of Leloir-type glycosyltransferases is of considerable
interest.
Out of seventeen sugar-1-P compounds tested, eleven were
accepted by GlmU (Table 1).9 GlcNAc-1-P, its 4-epimer
(GalNAc-1-P) and the 4-deoxy version were converted to
sugar nucleotides at comparable levels, indicating the 4-hydroxyl
group may not be necessary for enzyme recognition. In
contrast, 4-azido GalNAc-1-P failed in the reaction, presumably
resulting from the bulkiness of the azido group. The tolerance
for N-acyl modifications was different for GlcNAc-1-P and
GalNAc-1-P analogs. For the GlcNAc type, compounds with
relatively small N-acyl groups (GlcNAz-1-P and GlcNPr-1-P)
were good substrates, leading to generation of compounds 2
and 3. Reaction efficiency decreased for compound 4,
however, with its bulky acyl group, while compound 5 was
non-isolable and only detectable by MS, demonstrating poor
acceptance for bulky substitutions. Interestingly, none of the
N-acyl modified GalNAc-1-P analogs were accepted by
GlmU. The changes at both the N-acyl and C-4 positions
may enlarge these sugar-1-P compounds, thus causing them to
fail to enter the uridyltransferase pocket.10
N-acetylglucosamine (GlcNAc) and N-acetylgalactosamine
(GalNAc) are ubiquitous amino sugar components of glyco-
conjugates. GlcNAc is a component of peptidoglycans,3 while
both amino sugars play key roles in different glycosamino-
glycans.4 Moreover, they are prevalent in the core structures of
glycans in glycoproteins,5 and glycolipids,6 affecting cell–cell
interactions during various metabolic processes. Replacement
of GlcNAc/GalNAc residues by GlcNAc/GalNAc analogs in
a polysaccharide or glycoconjugate would thus be a good approach
to understand the mechanism of GlcNAc/GalNAc-related
pathways. GlmU, a GlcNAc-1-P uridyltransferase (pyrophos-
phorylase) from Escherichia coli,7 was accordingly employed
to obtain unnatural uridine 50-diphosphate-GlcNAc/uridine
50-diphosphate-GalNAc (UDP-GlcNAc/UDP-GalNAc) sugar
donors. This enabled generation of a library composed of
UDP-GlcNAc/UDP-GalNAc analogs and determination of
the substrate specificity of GlmU towards GlcNAc-1-P,
GalNAc-1-P and their corresponding analogs.
In contrast, the tolerance for 6-modified GlcNAc-1-P and
GalNAc-1-P analogs was quite similar. The yields of smaller
6-deoxy UDP-sugars were comparable to that of UDP-GlcNAc
and decreased dramatically for 6-azido compounds bearing a
relatively large substituent. The 3-epimer of GlcNAc-1-P was
also converted to the corresponding sugar donor but with
lower yield than GlcNAc-1-P.
To probe whether these UDP-sugar analogs are prospective
sugar donors for glycosyltransferases, 6-deoxy UDP-GlcNAc
(6) and Lac-1-OBn were chosen as donor and acceptor respectively
for LgtA (EC 2.4.1.56), which is a N-acetylglucosaminyl-
transferase. A trisaccharide with a terminal 6-deoxy-GlcNAc
GlcNAc-1-P/GalNAc-1-P and their analogs used as GlmU
substrates were prepared chemoenzymatically as previously
described.8 The corresponding sugar nucleotides were
was produced (Scheme 1). In addition,
a disaccharide
(6-deoxy-GlcNAcb1-3GalpNO2Ph) was detected by MS and
TLC (Fig. S2 and S3w) when p-nitrophenyl a-D-galactopyrano-
side was used as another acceptor.
a National Glycoengineering Research Center and The State Key
Laboratory of Microbial Technology, Shandong University, Jinan,
Shandong 250100, People’s Republic of China
In summary, we have reported the enzymatic synthesis of a
set of UDP-GlcNAc/UDP-GalNAc analogs. The substrate
specificity of the pyrophosphorylase GlmU towards sugar-1-P
compounds was also examined. The enzyme showed relaxed
tolerance for modifications at N-acyl, C-3, C-4, and C-6
positions, with a preference for small substituent groups.
The yields were low to moderate (10–65%) and some sugar-1-Ps
failed to generate the corresponding UDP-sugars due to the
b Departments of Chemistry and Biochemistry,
The Ohio State University, 484 W. 12th Ave., Columbus,
OH 43210, USA. E-mail: wang.892@osu.edu;
Fax: +1-614-688-3106; Tel: +1-614-292-9884
w Electronic supplementary information (ESI) available: Full experimental
procedures as well as full spectroscopic data for all new compounds. See
DOI: 10.1039/b917573c
z Contributed equally to this work.
ꢀc
This journal is The Royal Society of Chemistry 2009
6976 | Chem. Commun., 2009, 6976–6978
Guan, Wanyi
