Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Carbohydr Res. Author manuscript; available in PMC 2016 June 26.
Published in final edited form as:
PMCID: PMC4481193

Comparing substrate specificity of two UDP-sugar Pyrophosphorylases and Efficient One-pot Enzymatic Synthesis of UDP-GlcA and UDP-GalA

Yuxi Guo, Junqiang Fang, Tiehai Li, Xu Li, Cheng Ma, Xuan Wang, Peng G. Wang,* and Lei Li*


Uridine 5’-diphosphate-glucuronic acid (UDP-GlcA) and UDP-galacturonic acid (UDP-GalA), the unique carboxylic acid-formed sugar nucleotides, are key precursors involved in the biosynthesis of numerous cell components. Limited availability of those components has been hindering the development of efficient ways towards facile synthesis of bioactive glycans such as glycosaminoglycans. In current study, we biochemically characterized two UDP-sugar pyrophosphorylases from Arabidopsis thaliana (AtUSP) and Bifidobacterium infantis ATCC15697 (BiUSP), and compared their activities towards a panel of sugar-1-phosphates and derivatives. Both enzymes showed significant pyrophosphorylation activities towards GlcA-1-phosphate, and AtUSP also exhibited comparable activity towards GalA-1-phosphate. By combining with monosaccharide-1-phosphate kinases, we have developed an efficient and facile one-pot three-enzyme approach to quickly obtain hundreds milligrams of UDP-GlcA and UDP-GalA.

Keywords: UDP-GalA, UDP-GlcA, UDP-sugar pyrophosphorylase, Synthesis

1. Introduction

Hexuronic acids (HexAs), a class of sugar acids with both carbonyl and carboxylic acid functional groups, are the oxidation products of the corresponding monosaccharides. Glucuronic acid (GlcA) is one of the most commonly found HexAs in the plant and animal kingdoms.1 It is also prevalent in carbohydrate chains of proteoglycans.2 Galacturonic acid (GalA) is a major sugar residue of plant pectic polysaccharides and a minor component of certain plant arabinogalactan proteins.3 GalA is also found in various cell surface polysaccharides of different gram-negative bacteria,4,5 as well as capsular polysaccharide of the gram-positive human pathogens, such as type I Streptococcus pneumonae.6

Because of the biological significance and structure complexity, glycans containing HexAs have attracted considerable attention in past decades.79 One of the major bottlenecks in basic research and application of these components is the limited and cost-effective availability of uridine 5’-diphosphate glucuronic acid (UDP-GlcA) and uridine 5’-diphosphate galacturonic acid (UDP-GalA), the active forms of corresponding monosaccharides. UDP-GlcA and UDP-GalA are essential glycosyl donors for glycosyltransferases in the biosynthesis of various cell components including glycosaminoglycans, some N- and O-linked glycans, glycosphingolipids, pectins, etc.10,11 Recently, several chemical or coupled enzymatic approaches have been developed for the synthesis of UDP-GlcA and UDP-GalA.12,13 However, low yield, tedious steps and strict condition requirements have restricted the application of these approaches in large scale sugar nucleotide preparation.14 Therefore, a simple, cost-effective and efficient approach for the preparation of UDP-GlcA and UDP-GalA is highly desirable for subsequent glycosyltransferase-catalyzed reactions.

Enzymatic synthesis, which mimics the biosynthesis pathway, is considered more attractive with advantages of high conversion rate and mild reaction conditions. As shown in Figure 1, the de novo biosynthesis of UDP-GlcA initiates from glucose-6-phosphate (Glc-6-P), an intermediate of the glycolytic pathway. Three enzymatic catalyzed steps were involved, including the equilibrium conversion of Glc-6-P and Glc-1-P, the pyrophosphorylation step to yield UDP-Glc, and the oxidoreduction step catalyzed by UDP-Glc 6’-oxidoreductase. The complication of the pathway, the similarities between the final product and intermediates, as well as the consuming of expensive NAD+ or NADP+ have highly limited the application of such a pathway in large scale UDP-GlcA preparation. The de novo biosynthetic pathway of UDP-GalA is even more complicated (Fig. 1).

Figure 1
De novo and salvage biosynthetic pathway of UDP-GlcA and UDP-GalA. A), Glucuronokinase (GlcAK); B) UDP-sugar pyrophosphorylase (USP); C) UDP-glucose 6-dehydrogenase; D) Phosphoglucomutase; E) UTP: glucose-1-phosphate uridylyltransferase (GalU); F) UDP-Glc ...

In contrast, salvage biosynthetic pathways of sugar nucleotides, usually involving only two reactions (sugar-1-phosphate kinase catalyzed formation of sugar-1-P from monosaccharide and ATP, and NDP-sugar pyrophosphorylase catalyzed formation of NDP-sugar from sugar -1-P and NTP) were found much more practical in large scale synthesis of sugar nucleotides. Using recombinant enzymes from salvage pathways, we have successfully developed efficient approaches for 100 mg scale synthesis of UDP-Gal, UDP-GlcNAc, UDP-GalNAc, GDP-Fuc, GDP-Man and derivatives.1522 Recently, a salvage biosynthetic pathway of UDP-GlcA had been discovered in Arabidopsis thaliana (Fig. 1).23 In such a pathway, GlcA is first catalyzed by a glucuronokinase (AtGlcAK) to form GlcA-1-P in the presence of ATP and Mg2+. Subsequently, a UDP-sugar pyrophosphorylase (AtUSP) catalyzes the conversion of GlcA-1-P into UDP-GlcA in the presence of UTP and Mg2+.24,25 A galacturonokinase (AtGalKA) was also found in A. thaliana23 implies that a salvage biosynthetic way of UDP-GalA may also exist.

In this study, we cloned and characterized a UDP-sugar pyrophosphorylase from Bifidobacterium infantis ATCC15697 (BiUSP) which exhibited significant activity towards GlcA-1-P. In addition, we codon optimized the gene encoding AtUSP, and performed heterogeneous overexpression. AtUSP showed significant activities towards both GlcA-1-P and GalA-1-P. Furthermore, substrate specificity study exhibited that both USPs were promiscuous towards a panel of sugar-1-phosphates. Finally, by combing sugar-1-P kinases with USPs as well as pyrophosphatase, we developed an efficient one-pot three-enzyme approach to prepare hundreds of milligrams of UDP-GlcA and UDP-GalA.

2. Results and discussion

2.1 Gene cloning, expression, and purification of AtGlcAK, AtUSP, and BiUSP

Genes encoding AtGlcAK and AtUSP from A. thaliana were codon optimized (see DNA sequences in supporting information) and synthesized to achieve better heterologous expression in E. coli. The synthesized genes were cloned into pQE80L, a tightly controlled E. coli protein expression vector with a T5 promoter. Given that inductions were performed at a low temperature (16 °C) and with a low concentration of IPTG (0.2 mM), AtGlcAK and AtUSP were expressed almost entirely in soluble form. As shown in Figure 2, after one-step Ni-NTA affinity chromatography, both enzymes with N-terminal His6-tag were purified to 90% with high yields (90 mg for AtGlcAK, 42 mg AtUSP per L of LB cultures). The apparent molecular weights of the enzymes on SDS-PAGE (Fig. 2, Lane 1 & 2) are consistent with theoretical values.

Figure 2
SDS-PAGE analysis of purified enzymes. Lanes: 1, AtGlcAK; 2, AtUSP; 3, BiGalK; 4, BiUSP.

A bacterial UDP-sugar pyrophosphorylase gene, Blon_1169, was cloned from B. infantis ATCC 15697. The enzyme (BiUSP) encoded by this gene is highly identical (98%) to a previously reported UDP-sugar pyrophosphorylase, BLUSP.26 BiUSP was expressed and purified with a C-terminal His6-tag followed the abovementioned methods. The yield of pure BiUSP from 1 L of cultures is 43 mg. An approximate molecular weight of 60 KDa was observed on SDS-PAGE (Fig. 2, Lane 4), consistent with theoretical value (58 KDa).

2.2 Pyrophosphorylase activities of AtUSP and BiUSP

Capillary Electrophoresis (CE) was used to analyze the formation of UDP-sugars. Retention times of 21.7, 18.2, 16.6 min were observed on CE profiles for UTP, UDP-GlcA and UDP-GalA standards (Fig 3A, B & C), respectively. The results shows that BiUSP could catalyze the formation of UDP-GlcA from GlcA-1-P and UTP efficiently (Fig. 3D), whereas the formation of UDP-GalA was not clearly observed (CE profile not shown), indicating that the C-4 hydroxyl group may be necessary for BiUSP recognition. This is opposite with results yielded form activity assays towards Glc-1-P and Gal1-P, where both substrate can be well accepted (Table 2). One possible explanation could be the bulky carboxylic acid group on C-6 position altered the configuration of C-4 hydroxyl group, makes GalA not tolerable by BiUSP.

Figure 3
CE Profiles of compound standards, and BiUSP/AtUSP catalyzed reactions. (A) 0.5 mM UTP; (B) 0.5 mM UDP-GlcA; (C) 0.5 mM of UDP-GalA; (D) BiUSP catalyzed generation of UDP-GlcA; (E) AtUSP catalyzed generation of UDP-GlcA; (F) AtUSP catalyzed generation ...
Table 2
Substrate specificity of AtUSP and BiUSP towards sugar-1-phosphates. Specific activity of AtUSP towards GlcA-1-P (12.4 µmol min−1 mg−1) was set as 100. ND, not detected.

On the other hand, AtUSP could efficiently (68% and 65% towards GlcA-1-P and GalA-1-P) catalyze the formation of both UDP-GlcA and UDP-GalA (Fig. 3E & F). The activity of BiUSP and AtUSP towards GlcA-1-P are comparable (Fig. 3D & E). It is worth to note that AtUSP can be stored for up to 6 months at −20 °C without significant activity loss, whereas BiUSP showed dramatically activity decrease after 6 months storage under the same condition.

Freshly prepared AtUSP or BiUSP was applied in time course experiment and kinetics studies. As shown in Figure 4, time course results implies that for AtUSP, the formation rate of UDP-GalA is slightly lower than that of UDP-GlcA. In kinetics studies, to ensure initial reaction rates, enzyme dilutions were selected so that less than 15% of substrate conversion was achieved within 5 min for each reaction. Kinetic parameters were calculated according to Lineweaver–Burk equation (Table 1). The Km value of AtUSP towards GlcA-1-P and GalA-1-P are 2.36 × 10−3 µM and 5.50 × 10−3 µM, indicates that AtUSP has a higher affinity for GlcA-1-P that GalA-1-P. The kcat/Km values of AtUSP towards GlcA-1-P is 1.8 times as that of GalA-1-P, implies that AtUSP is more efficient for the pyrophosphorylation of GlcA-1-P. The kcat/Km values of AtUSP (4.66 × 103 s−1 µM−1) and BiUSP (4.72 × 103 s−1 µM−1) towards GlcA-1-P are similar, indicating they are equally efficient in the synthesis of UDP-GlcA. The kinetics of reverse reactions (hydrolysis of UDP-GlcA and UDP-GalA) (Table 1) revealed that AtUSP and BiUSP are equally active towards UDP-GlcA, whereas only AtUSP can catalyze the hydrolysis of UDP-GalA. Such results are consistent with that of forward reactions (formation of UPD-GlcA and UDP-GalA).

Figure 4
Time course for UDP-GlcA and UDP-GalA reactions. Reactions contains 1 mM UTP, 40 ng of At USP or BiUSP, and: 1, AtUSP and 0.05 mM GlcA-1-P; 2, AtUSP and 0.8 mM GlcA-1-P; 3, AtUSP and 0.05 mM GalA-1-P; 4, AtUSP and 0.8 mM GalA-1-P; 5, BiUSP and 0.05 mM ...
Table 1
Kinetic parameters of AtUSP and BiUSP towards GlcA-1-P, GalA-1-P, UDP-GlcA and UDP-GalA.

2.3 Substrate specificity study of AtUSP and BiUSP towards sugar-1-phosphates

It was previously reported that AtUSP possessed relaxed substrate specificity towards sugar-1-phosphate. For example, it could well accept Gal-1-P, Glc-1-P, 4-deoxy-4-azido-Gal-1-P, d-Fuc-1-P, l-Ara-1-P, and Xyl-1-P.21,24,25 To further investigate the substrate specificities of AtUSP and BiUSP, the specific activities towards a panel of sugar-1-phosphates were measured. Not surprisingly, BiUSP exhibited similar substrate specificity as BLUSP26 towards sugar-1-phosphates (Table 2). In addition, both AtUSP and BiUSP showed broad substrate specificities, but towards partially different sugar-1-phosphates. For example, AtUSP are highly active towards GlcA-1-P, GalA-1-P, Glc-1-P, Gal-1-P, dFuc-1-P, 2-deoxy-Glc-1-P, 2-deoxy-Gal-1-P, whereas BiUSP are highly active towards GlcA-1-P, Glc-1-P, Gal-1-P, Man-1-P, 2-deoxy-Glc-1-P, and 2-deoxy-2-amino-Glc-1-P. Such information are of great importance in applying the USPs in large scale synthesis of UDP-sugar and derivatives.

2.4 Preparation and characterization of UDP-GlcA and UDP-GalA

Given the stability and efficiency, AtUSP was chosen for hundreds milligrams scale synthesis of UDP-HexA. As shown in Figure 5, a one-pot three enzyme approach was employed. In the case of UDP-GlcA preparation, the monosaccharide GlcA was first converted into GlcA-1-P catalyzed by AtGlcAK, the glucuronokinase from A. thaliana.23 The formation of UDP-GlcA was then achieved by AtUSP catalyzed reaction in the presence of UTP, and Mg2+. Since UDP-sugar pyrophosphorylases catalyzed relations are usually reversible,3 a pyrophosphatase from E. coli (EcPpA) was added to digest the diphosphate byproduct. The same as previously proved,20 addition of EcPpA greatly improved the final yield (207 mg pure UDP-GlcA in a yield of 89%) by driving the reaction forward. The synthesis of UDP-GalA was performed following the same route, except that AtGlcAK was substituted by BiGalK, a promiscuous galactokinase exhibiting phosphorylation activity towards GalA.27 The total yield of UDP-GalA after P2 purification was 70% (162 mg).

Figure 5
One-pot three-enzyme synthesis of UDP-GlcA and UDP-GalA. Enzymes used: AtGlcAK: Glucuronokinase from A. thaliana; BiGalK, Galactokinase from B. infantis ATCC15697; AtUSP, UDP-sugar pyrophosphorylase from A. thaliana; EcPpA, Pyrophosphatase from E. coli ...

2.4.1 Spectroscopic data of Uridine 5’-diphosphate glucuronic acid

HRMS (ESI): Calcd 580.0343 for C15H22N2O18P2; found: 579. 0256 [M-H+];

1H NMR (400 MHz, D2O): δ 3.51 (t, J = 9.6 Hz, 1 H), 3.57–3.60 (m,1 H), 3.74 (s,1 H), 3.79 (t, J = 9.6 Hz, 1H), 4.14 (d, J = 10.4 Hz, 1H), 4.20–4.29 (m,2 H), 4.37–4.38 (m,2 H), 5.62 (dd, J = 2.8 Hz, J = 7.2 Hz,1 H), 5.97–5.99 (m,2 H), 7.95 (d, J = 8.0 Hz, 1 H); 13C NMR (100 MHz, D2O): δ 58.33, 63.95 (JC-P= 20.4 Hz), 68.65, 70.37 (JC-P= 33.6 Hz), 70.85, 71.58, 72.02, 72.77, 82.27 (JC-P= 35.6 Hz), 87.27, 94.36 (JC-P= 26.0 Hz), 101.68, 140.56, 150.84, 165.23; 31P NMR (162 MHz, D2O): δ −12.06, −10.40;

2.4.2 Spectroscopic data of Uridine 5’-diphosphate galacturonic acid

HRMS (ESI): Calcd 580.0343 for C15H22N2O18P2; found: 579. 0258 [M-H+];

1H NMR (400 MHz, D2O): δ 3.83 (d, J = 10.4 Hz, 1 H), 3.98 (dd, J = 3.2 Hz, J = 10.4 Hz,1 H), 4.17–4.22 (m,2 H), 4.29–4.32 (m,2 H), 4.36–4.37 (m,2 H), 4.53 (s,1 H), 5.68–5.69 (m,1 H), 5.97–5.99 (m,2 H), 7.96 (d, J = 8.0 Hz, 1 H); 13C NMR (100 MHz, D2O): δ 58.31, 63.91 (JC-P= 16.0 Hz), 67.50 (JC-P= 29.2 Hz), 68.32, 68.63, 69.57, 71.70, 72.78, 82.28 (JC-P= 35.6 Hz), 87.25, 94.71 (JC-P= 20.4 Hz), 101.67, 140.57, 150.84, 165.23; 31P NMR (162 MHz, D2O): δ −11.90, −10.35; HRMS (ESI):

3. Conclusion

In summary, we have biochemically characterized two highly efficient UDP-sugar pyrophosphorylases (AtUSP and BiUSP) for the purpose of large scale UDP-HexA preparation. Genes or expression conditions were optimized to obtain high yields of enzymes (40–90 mg per L culture). Both enzymes showed comparable activities towards GlcA-1-P, whereas only AtUSP exhibited activity towards GalA-1-P. Specific activity study revealed that both USPs possessed broad substrate specificity, but towards partially different sugar-1-phosphates. In addition, we have developed a one-pot three-enzyme approach for efficient and rapid preparation of UDP-GlcA and UDP-GalA starting from inexpensive monosaccharides and nucleotides. Using such an approach, 207 and 162 mgs of UDP-GlcA and UDP-GalA were prepared within one week.

4. Experimental section

4.1 Materials

All chemical reagents were purchased from Sigma unless otherwise noted. Ni-NTA agarose was purchased from Qiagen. Talose-1-P,20 d-Fucose-1-P,27 2-deoxy-Gal-1-P,27 2-deoxy-Glc-1-P20 were prepared as previously described. The preparation of 6-deoxy-6-azido-Gal-1-P will be reported subsequently. Codon optimized gene of AtGlcAK, AtUSP were synthesized by Genscript (Piscataway, NJ). Genomic DNA of B. infantis ATCC 15697 was purchase from ATCC. E. coli DH5α and BL21 (DE3), as well as vector pET22b were purchased from EMD Millipore (Bilerica, MA). Vector pQE80Lwas purchased from Qiagen (Germantown, MD). BiGalK expression E. coli strain was kept by our group.

4.2 Molecular Cloning, Protein Expression and Purification

The glucuronokinase gene (AtGlcAK) and UDP-sugar pyrophosphorylase gene (AtUSP) from A. thaliana was codon optimized and synthesized with restriction sites (SacI and HindIII) (see supporting information for optimized DNA sequences). The genes were then cleaved by SalI and HindIII, and inserted into pQE80L pretreated with corresponding restriction enzymes. The UDP-sugar pyrophosphorylase gene (BiUSP, Blon_1169) from B. infantis was amplified by PCR using B. infantis ATCC15697 genomic DNA as templates. Primers used are, Blon_1169F: GAGCATATGACCGAAATAAACGATAAGGCC and Blon_1169R: TATCTCGAGCACCCAATCGTCCGGTTCGAT. Yielded PCR fragments were digested with NdeI and HindIII and subsequently inserted into pET22b pretreated with corresponding enzymes. The recombinant plasmids were verified by DNA sequencing and transformed into E. coli BL21 (DE3) for protein expression.

E. coli BL21(DE3) cells harboring the above recombinant plasmids were grown in LB medium at 37 °C with shaking (20 rpm) till OD600 reached 0.6–0.8, followed by addition of isopropyl-β-d-thiogalactopyranoside (IPTG) to a final concentration of 0.2 mM. After 20 h induction at 16 °C the cells were harvested by brief centrifugation, and stored at −80 °C until used. Protein purification was performed utilizing a Ni-NTA resin (Qiagen) column according to the manufacture instructions. The purified proteins were desalted by PD-10 column (GE life science, USA) against 50 mM Tris-HCl (pH 8.0) and 20% of glycerol. Enzymes were kept at −20 °C until used.

4.3 Activity assays of AtUSP and BiUSP towards GlcA-1-P and GalA-1-P

A 50 µL reaction system in 20 mM Tris-HCl (pH 8.0) consisted of 0.8 mM GlcA-1-P or GalA-1-P, 1 mM UTP, 2 mM MgCl2, and varying amount of purified AtUSP or BiUSP. Reactions were allowed to proceed at 37 °C for 20 min, and quenched by the addition of equal volumes of cold ethanol. The formation of UDP-HexA and consumption of UTP was analyzed by Capillary Electrophoresis (Beckman Coulter) using the following conditions: 75 µm i.d. capillary (Beckman Coulter), 25 KV/140 µA, 5 s vacuum injection, 50 mM sodium tetraborate running buffer, pH 9.4., monitored at 254 nm.

4.4 Determination of kinetic parameters

Kinetic parameters of both AtUSP and BiUSP towards Glc-1-P and Gal-1-P were measured in 50 µL reactions in 20 mM Tris-HCl (pH 8.0). The reaction mixtures contain varying concentrations of GlcA-1-P or GalA-1-P (0.05, 0.1, 0.2, 0.4, 0.6, 0.8 mM), 1 mM UTP, 2 mM MgCl2. Reactions were initiated by the addition of 40 ng of AtUSP or BiUSP, performed at 37 °C for 5 min, and quenched by addition of equal volumes of cold ethanol. The sugar-1-P substrate concentrations used in the time course studies were 0.05 and 0.8 mM. The reactions were quenched at 0, 2, 5, 10, 15, 30, and 60 min. Analysis were performed by CE as described above, and the average value of triplicate assays was employed to generate Line-weaver-Burk plot, and Km, Vmax, kcat values were calculated accordingly. Kinetic parameters of AtUSP and BiUSP towards UDP-GlcA and UDP-GalA were measured similarly, by replacing GlcA-1-P or GalA-1-P into UDP-GlcA or UDP-GalA, and replacing UTP into pyrophosphate.

4.5 Substrate specificity study of AtUSP and BiUSP towards various sugar-1-phosphates

The reaction mixtures (50 µL total volume) contain 0.8 mM of different sugar-1-phosphates (Table 2), 1 mM UTP and 2 mM MgCl2. Reactions were initiated by the addition of varying amounts (0.4–40 µg, to make sure the reactions convert no more than 25%) of AtUSP or BiUSP, performed at 37 °C for 10 min, and quenched by the addition of equal volumes of cold ethanol. Analysis were performed by CE as described above, and the average value of triplicate assays was employed to generate specific activities towards each sugar-1-phosphate.

4.6 One-pot three-enzyme synthesis of UDP-GlcA and UDP-GalA

Preparative-scale synthesis of UDP-GlcA and UDP-GalA were performed in a 20 mL reaction system containing 50 mM pH 8.0 Tris-HCl buffer, 20 mM GlcA or GalA, 22 mM ATP, 22 mM UTP, 20 mM MgCl2, and 0.1 to 2 mg of different enzymes. For the synthesis of UDP-GlcA, 0.6 mg of AtGlcAK, 0.4 mg of AtUSP, and 0.1 mg of EcPpA20 were employed; for the synthesis of UDP-GalA, 2 mg of BiGalK, 0.4 mg of AtUSP, and 0.1 mg of EcPpA were employed. The reaction mixtures were incubated at 37 °C for 4 to 12 hours with gentle shaking. The generation of UDP-sugars were monitored by TLC (Isopropanol: NH4OH: H2O = 7: 3: 2). After 90% of monosaccharides converted into UDP-sugar, reactions were quenched by boiling for 5 min, centrifuged to remove precipitants, and concentrated for separation. The separation of UDP-sugars from the reaction mixtures were performed by gel filtration (Bio-Gel P2, Bio-Rad), fractions containing products were pooled and lyophilized. Digestion of nucleotides by alkaline phosphatase and further P2 purification were applied to achieve minimum 95% purity. The final products were lyophilized and characterized by HR-MS and NMR.

UDP-sugar pyrophosphorylase (AtUSP) from Arabidopsis thaliana showed significant activity towards both GlcA-1-P and GalA-1-P

UDP-sugar pyrophosphorylase (BiUSP) from Bifidobacterium infantis showed significant activity towards GlcA-1-P

Both USPs showed broad substrate specificity towards various sugar-1-phosphates.

Both AtUSP an BiUSP were expressed and purified in high yields (over 40 mg per L);

A one-pot three-enzyme approach was applied efficiently for hundreds mgs scale of UDP-GlcA and UDP-GalA.

Supplementary Material



This work was supported by National Institutes of Health (Grant No.: R01 GM085267 & R01 AI083754 to P.G.W).


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Supplementary materials

Supplementary materials associated with this article can be found in the online version.


1. Banhegyi G, Braun L, Csala M, Puskas F, Mandl J. Free Radic Biol Med. 1997;23:793–803. [PubMed]
2. Selleck SB. Trends Genet. 2000;16:206–212. [PubMed]
3. Ohashi T, Cramer N, Ishimizu T, Hase S. Anal Biochem. 2006;352:182–187. [PubMed]
4. Roberts IS. Annu Rev Microbiol. 1996;50:285–315. [PubMed]
5. Knirel' Iu A, Kochetkov NK. Biokhimiia. 1994;59:1784–1851. [PubMed]
6. Stroop CJ, Xu Q, Retzlaff M, Abeygunawardana C, Bush CA. Carbohydr Res. 2002;337:335–344. [PubMed]
7. Xu Y, Cai C, Chandarajoti K, Hsieh PH, Li L, Pham TQ, Sparkenbaugh EM, Sheng J, Key NS, Pawlinski R, Harris EN, Linhardt RJ, Liu J. Nat Chem Biol. 2014;10:248–250. [PMC free article] [PubMed]
8. Xu Y, Masuko S, Takieddin M, Xu H, Liu R, Jing J, Mousa SA, Linhardt RJ, Liu J. Science. 2011;334:498–501. [PMC free article] [PubMed]
9. Leach FE, 3rd, Ly M, Laremore TN, Wolff JJ, Perlow J, Linhardt RJ, Amster IJ. J Am Soc Mass Spectrom. 2012;23:1488–1497. [PMC free article] [PubMed]
10. Freeze HH, Elbein AD. In: Essentials of Glycobiology. Varki A, et al., editors. 2009.
11. Ridley BL, O'Neill MA, Mohnen D. Phytochemistry. 2001;57:929–967. [PubMed]
12. Broach B, Gu X, Bar-Peled M. FEBS J. 2012;279:100–112. [PMC free article] [PubMed]
13. Kleczkowski LA, Decker D, Wilczynska M. Plant Physiol. 2011;156:3–10. [PubMed]
14. Tanaka H, Yoshimura Y, Jorgensen MR, Cuesta-Seijo JA, Hindsgaul O. Angew Chem Int Ed Engl. 2012;51:11531–11534. [PubMed]
15. Cai L, Guan W, Wang W, Zhao W, Kitaoka M, Shen J, O'Neil C, Wang PG. Bioorg Med Chem Lett. 2009;19:5433–5435. [PubMed]
16. Cai L, Guan W, Kitaoka M, Shen J, Xia C, Chen W, Wang PG. Chem Commun (Camb) 2009:2944–2946. [PubMed]
17. Yi W, Liu X, Li Y, Li J, Xia C, Zhou G, Zhang W, Zhao W, Chen X, Wang PG. Proc Natl Acad Sci U S A. 2009;106:4207–4212. [PubMed]
18. Cai L, Guan W, Chen W, Wang PG. J Org Chem. 2010;75:3492–3494. [PMC free article] [PubMed]
19. Zhao G, Guan W, Cai L, Wang PG. Nat Protoc. 2010;5:636–646. [PMC free article] [PubMed]
20. Li L, Liu Y, Wan Y, Li Y, Chen X, Zhao W, Wang PG. Org Lett. 2013;15:5528–5530. [PMC free article] [PubMed]
21. Liu J, Zou Y, Guan W, Zhai Y, Xue M, Jin L, Zhao X, Dong J, Wang W, Shen J, Wang PG, Chen M. Bioorg Med Chem Lett. 2013;23:3764–3768. [PubMed]
22. Zou Y, Xue M, Wang W, Cai L, Chen L, Liu J, Wang PG, Shen J, Chen M. Carbohydr Res. 2013;373:76–81. [PubMed]
23. Pieslinger AM, Hoepflinger MC, Tenhaken R. J Biol Chem. 2010;285:2902–2910. [PMC free article] [PubMed]
24. Kotake T, Hojo S, Yamaguchi D, Aohara T, Konishi T, Tsumuraya Y. Biosci Biotechnol Biochem. 2007;71:761–771. [PubMed]
25. Litterer LA, Schnurr JA, Plaisance KL, Storey KK, Gronwald JW, Somers DA. Plant Physiol Biochem. 2006;44:171–180. [PubMed]
26. Muthana MM, Qu J, Li Y, Zhang L, Yu H, Ding L, Malekan H, Chen X. Chem Commun (Camb) 2012;48:2728–2730. [PubMed]
27. Li L, Liu Y, Wang W, Cheng J, Zhao W, Wang P. Carbohydr Res. 2012;355:35–39. [PubMed]